1

Methods For Identifying Surgical Wound Infection

After Discharge From Hospital

Emily Petherick, Jane Dalton and Nicky Cullumª

Department of Health Sciences

University of York

ªCorresponding Author: Professor Nicky Cullum, Department of Health Sciences, Univeristy of York, Area 2, Seebohm Rowntree Building, Alcuin College, York YO10 5DD. Email: nac2@york.ac.uk

2

Table of Contents

Abstract .......................................................................................................4

Background .......................................................................................................8

Definition of Surgical Site Infection (SSI)................................................................................9

Aims ......................................................................................................11

Methods ..................................................................................................... 12

Literature Review..................................................................................................................12

National Audit ......................................................................................................................13

Results ..................................................................................................... 15

Literature Review..................................................................................................................15

Discussion .....................................................................................................29

Audit of current surveillance practice: Results.......................................................................30

Telephone Interviews with Survey Respondents....................................................................31

Other relevant research currently being undertaken or planned within the United Kingdom.36

Conclusions .....................................................................................................37

Question 1: What are the features of existing post discharge surveillance systems in terms of

coverage; denominator data; diagnostic criteria; other data collected; methods of data

collection?.............................................................................................................................37

Question 2: What is the current evidence for the validity, reliability and practicality of

different methods of post-discharge surveillance of SSIs? .....................................................38

Question 3: What research is needed to fill knowledge gaps?................................................39

Appendix A Search strategies used in the report..........................................44

Strategy no. 1. Surgical wound infection: prospective studies.................................................44

Search strategy no. 2. Surgical wound infection: measurement and validation studies............47

Search Strategy no. 3. Surgical wound infection: surveillance and monitoring studies............51

3

Appendix B Critical Appraisal of Included non-comparative PDS studies....

.....................................................................................................55

Appendix C Example of audit sent to Infection Control lead within Trusts .

.....................................................................................................90

Appendix D Telephone interview 1st contact................................................92

Appendix E Telephone interview 2nd contact...............................................93

APPENDIX F Individual components of wound definition systems not

previously described in Bruce et al (2001). ...........................................................................95

Appendix G Excluded Studies.......................................................................96

4

Acknowledgements

We would like to thank the following for their assistance in the production of this report:

Ms Lisa Mather, Centre for Reviews and Dissemination, University of York, for literature

searching; Mr Peter Moore, Consultant Surgeon, Northern Lincolnshire and Goole Hospitals

NHS Trust for his expert clinical advice; and Sandi Newby and Corinna Petre for their secretarial

assistance.

5

Abbreviations

ASEPSIS Additional treatment, Serous discharge, Erythema, Purulent exudate,

Separation of deep tissues, Isolation of bacteria, and duration of

inpatient Stay – a system for scoring wound infections

CABG Coronary artery bypass graft

CDC Centers for Disease Control

HPA Health Protection Agency

ICD International Classification of Disease

NINSS Nosocomial Infection National Surveillance Service (UK based)

NNIS National Nosocomial Infections Surveillance System (USA based)

NS Not Stated

PDS Post discharge surveillance

SSI Surgical site infection

6

Abstract

Background: Most surgical site infections are thought to occur after patients are discharged

from hospital, therefore accurate methods of detecting infection post-discharge are important in

order to monitor surgical performance and evaluate interventions aimed at reducing surgical site

infections.

Objectives: To establish

(i) the features of existing post discharge surveillance systems;

(ii) the current evidence for the validity, reliability and practicality of different

methods of post-discharge surveillance of SSIs;

(iii) what research is needed to fill knowledge gaps and

(iv) possible study sites where evaluations of methods of detection of SSI post-

discharge might be incorporated into current activity.

Methods: Update of previous systematic review and national audit of current surveillance

practices.

Results: Only 5 studies (3 new to this review) were identified which compared alternative

methods of post-discharge surveillance for surgical wound infection. These studies were unique

comparisons; there is no consensus as to the gold standard reference method of detection –

candidates include direct observation of patient wounds by health professionals and data capture

from primary care (including of routine data). A promising method for comparison would be

CDC-criteria driven patient self-diagnosis but an appropriate tool has yet to be developed. The

audit identified a wide range of surveillance programmes in operation in England, Scotland and

Wales. Some of these involve direct observation of patients’ wounds and some primary care

data capture. National SSI surveillance programmes currently do not require post-discharge

surveillance.

7

Conclusions: More research on methods to measure surgical site infection rates after hospital

discharge is needed. Preliminary work should develop an appropriate reference (gold standard)

system that is sufficiently accurate. This should then serve as a comparison to alternatives such

as patient self-diagnosis. It is likely that some of the centres currently undertaking PDS SSI

could host new research.

8

Background

Surgical site infections (SSIs) have been estimated to occur in up to 15% of elective surgical

patients and approximately 30% of patients whose surgical procedure was classed as

contaminated or “dirty”(Bruce et al, 2001). The proportion of SSIs which is preventable is

unknown however there are wide variations in infection rates and an international drive to

minimise them (Bruce et al, 2001). Accurate, standardised methods of defining and monitoring

SSIs are essential in order to detect and treat infections (treatment), compare the performance of

different surgical services (performance monitoring) and of different interventions (research). Some

investigators have viewed SSI surveillance itself as an intervention aimed at reducing SSI through

an audit and feedback mechanism.

Recent figures from Scotland completed as part of the Scottish Surveillance of Healthcare

Programme (SSHAIP) found that infections occurring after discharge from hospital accounted

for 65.3% (range: 26-89%) of all SSIs in those sites that include post-discharge surveillance

(Scottish Centre for Infection and Environmental Health, 2003).

Bruce et al (2001) completed an extremely helpful review of the measurement and monitoring of

surgical adverse events in 2001. This review (the search for which covered only the years 1993 –

1999) identified 41 different definitions of SSI and 13 different grading scales, of varying

complexity. The review emphasised the importance of post-discharge surveillance for SSI (PDS

SSI) since the mean duration of post-operative stay is reducing (including an increase surgical day

cases). It has been suggested that between 50% and 70% of SSIs occur after the patient is

discharged from hospital (Holtz & Wenzel, 1992). Ascertainment of SSI after discharge presents

its own methodological challenges however. Methods which have been studied include direct

observation of patients’ wounds by health care professionals in different settings; questionnaires

and diaries completed either by patients or health care professionals; telephone surveys to

9

patients. The interpretation of these data to date has presented challenges however which can be

summarised as:

• The lack of an agreed gold standard method of PDS SSI

• Paucity of studies comparing different methods in a consistent way (large variations in

definition of SSIs; staffing, setting and timings used in existing studies).

Definition of Surgical Site Infection (SSI)

The previous systematic review (Bruce et al, 2001) described the most current definition of

surgical wound infection as that produced by the Centers for Disease Control and Prevention

(CDC) in 1992. The CDC has since produced updated guidelines for the prevention of surgical

site infections, however the definitions provided of surgical site infections have not changed

(Mangram et al, 1999). Full details of this definition are provided in Box 1.

BOX 1 Definition of surgical site infection: CDC (Horan et al 1992).

For surveillance classification purposes, SSIs are divided into incisional SSIs and organ/space SSIs. Incisional space SSIs are further classified as involving only the skin and subcutaneous tissue (superficial incisional SSIs) or involving deep soft tissues (e.g. fascial and muscle layers) of the incision (deep incisional SSIs). Organ/space SSIs involve any part of the anatomy (organs or spaces) other than the incision opened or manipulated during the operative procedure.

Superficial incisional SSI Superficial incisional SSIs must meet the following criteria: infection occurs within 30 days after the operative procedure and involves only skin or subcutaneous tissue of the incision, and at least one of the following is present: • Purulent drainage from the superficial incision. • Organisms isolated from an aseptically obtained culture of fluid or tissue from the superficial incision. • At least one of the following signs or symptoms of infection: pain or tenderness, localised swelling, redness or heat and superficial incision is deliberately opened by a surgeon, unless culture of incision is negative. • Diagnosis of superficial incisional SSI by the surgeon or attending physician. The following are not reported as superficial incisional SSIs: • stitch abscess (minimal inflammation and discharge confined to the points of suture penetration) • infection of an episiotomy or a neonate’s circumcision site* (episiotomy and circumcision are not considered NNIS operative procedures) • infected burn wound* • incisional SSI that extends into the fascial and muscle layers (see deep incisional SSI).

Deep incisional SSI Deep incisional SSIs must meet the following criteria: infection occurs within 30 days after the operative procedure if no implant† is left in place or within 1 year if an implant is left in place and the infection appears to be related to the operative procedure and infection involves deep soft tissues (e.g. fascial and muscle layers) of the incision and at least one of the following is present: • Purulent drainage from the deep incision but not from the organ/space component of the surgical site. • A deep incision spontaneously dehisces or is deliberately opened by a surgeon when the patient has at least one of the following signs or symptoms: fever (> 38°C), localised pain, or tenderness, unless culture of the incision is negative.

10

BOX 1 Definition of surgical site infection: CDC (Horan et al 1992). (Continued)

• An abscess or other evidence of infection involving the deep incision is found on direct examination, during re-operation, or by histo-pathological or radiological examination. • Diagnosis of a deep incisional SSI by a surgeon or attending physician. Organ/space SSI An organ/space SSI involves any part of the anatomy (e.g. organs or spaces), other than the incision, opened or manipulated during the operative procedure. Organ/space SSIs must meet the following criteria: infection occurs within 30 days after the operative procedure if no implant is left in place or within 1 year if an implant is left in place and the infection appears to be related to the operative procedure and infection involves any part of the anatomy (e.g. organs or spaces) other than the incision opened or manipulated during the operative procedures, and at least one of the following is present: • Purulent drainage from a drain that is placed through a stab wound‡ into the organ/space. • Organisms isolated from an aseptically obtained culture of fluid or tissue in the organ/space. • An abscess or other evidence of infection involving the organ/space on direct examination, during reoperation, or by histopathological or radiological examination. • Diagnosis of an organ/space SSI by a surgeon or attending physician. * Specific criteria are used for infected episiotomy and circumcision sites and for burn wounds. † An implant is defined as a non-human-derived implantable foreign body (e.g. prosthetic heart valve, non-human vascular graft, mechanical heart, hip prosthesis) that is permanently placed in a patient during operation. ‡ If the area around a stab wound becomes infected, it is not an SSI. It is considered a skin or soft tissue infection, depending on its depth.

Specific sites of organ/space surgical site infection • Arterial or venous infection • Breast abscess or mastitis • Disc space • Ear, mastoid • Endometritis • Other infections of the lower respiratory tract • Other infections of the urinary tract • Other male or female reproductive tract • Intra-abdominal, not specified elsewhere • Intracranial, brain or dural infections abscess • Joint or bursa • Mediastinitis • Meningitis or ventriculitis • Myocarditis or pericarditis • Oral cavity (mouth, tongue or gums) • Osteomyelitis • Endocarditis • Eye, other than conjunctivitis • Gastrointestinal tract • Spinal abscess without meningitis • Sinusitis • Upper respiratory tract, pharyngitis • Vaginal cuff

SSI involving more than one specific site Infection that involves both superficial and deep incision sites is classified as a deep incisional SSI . Occasionally an organ/space infection drains through the incision. Such infection generally does not involve re-operation and is considered a complication of the incision. It is therefore classified as a deep incisional SSI.

The CDC guidelines concluded that, with respect to post-discharge surveillance;

11

‘At this time no consensus exists on which post-discharge surveillance methods are the most

sensitive, specific and practical. Methods chosen will necessarily reflect the hospital’s unique

mix of operations, personnel resources, and data needs.’

The guidelines also examined the evidence for surgical site surveillance in the

outpatients department and concluded that;

‘At this time, no single detection method can be recommended. Available resources and

data needs will determine which method(s) should be used and which operations should be

monitored. Regardless of which detection method is used, it is recommended that the CDC

NNIS definitions of SSI be used without modification’.

We have taken as the starting point for this review (and in agreement with the NCCRM) the

CDC definition of surgical site infection and recommend that this should be used in all future

studies unless evidence arises to suggest the definition is unsatisfactory or that a more valid or

reliable definition is developed.

Aims

This brief review aims to answer the following questions:

1. What are the features of existing post discharge surveillance systems in terms of

coverage; denominator data; diagnostic criteria; other data collected; methods of data

collection?

2. What is the current evidence for the validity, reliability and practicality of different

methods of post-discharge surveillance of SSIs?

3. What research is needed to fill knowledge gaps?

4. Are there possible study sites where evaluations of methods of detection of SSI post-

discharge might be incorporated into current activity?

12

Methods

Existing surveillance systems were identified and described using a combination of literature

review and a national audit.

Literature Review

The literature search undertaken for the Bruce review was updated. The search strategy is

outlined in Appendix A and the databases searched are shown in Table 1.

Table 1 Electronic databases searched for this review

Database Period/Issue covered

Medline (OVID) 1999-02/2004

EMBASE 1999-03/2004

CINAHL 1999-03/2004

The Cochrane Library 1999-2004 Issue 1

In order to be eligible for inclusion, studies had to:

- Describe a method of post discharge surveillance for surgical site infection

OR

- Compare at least two methods of post discharge surveillance for surgical site infection

OR

- Describe an economic evaluation of post discharge surveillance for surgical site infection.

Papers were excluded if they met any of the following criteria;

13

- Paper gave no indication that post discharge follow up was performed

- Post-discharge follow up was carried out for reasons other than for surveillance of

surgical wound infection

- Appeared to be routine patient follow up post operation with no methodological detail

provided of how any post-discharge follow up was achieved.

National Audit

The answer to Question 1 was further supplemented by an audit of current post-discharge

surveillance across the UK. The audit involved two phases in an effort to gain a high response

rate. For the first phase, a brief audit form was sent to Infection Control Personnel within all

primary care trusts and hospital trusts within the UK National Health Service. The aim of this

national audit was to identify those Trusts that were undertaking any post-discharge surveillance

for surgical site infection. A database of infection control personnel was purchased from

Binley’s (http://www.binleys.com/default.asp) current to January 2004. Audit forms were sent

to multiple contacts within the Trusts, where the database listed more than one infection control

staff member. A sample of Trusts that identified themselves as undertaking surveillance

participated in a telephone interview in order to collect more detailed information.

A copy of the audit form sent to Infection Control teams can be found in Appendix C.

Telephone Interviews

Sample selection Of the 42 respondents who indicated that their Trust was currently carrying out post discharge

surveillance for surgical site infection, nine were selected in line with the following criteria:

14

(a) respondents had indicated their willingness to be contacted for further discussion

(only 20 out of 42);

(b) the Trust’s post discharge surgical site infection (SSI) surveillance appeared to extend

beyond the provision of routine follow up;

(c) to reflect a variety of methods of surveillance in a range of surgical specialities.

The resulting sample comprised a range of surgical specialities: general, gynaecology,

orthopaedics, neurosurgery and cardio-vascular. A range of surveillance methods was also

evident within the identified programmes. These included patient telephone surveys (n=4),

patient questionnaire (for example delivered at an outpatient visit) (n=3), GP/health professional

surveys (n=2), surgeon survey (n=1), direct observation (n=5) and others (n=2) in connection

with the use of hospital information systems.

Development of interview schedule

Broad categories of questions were identified with close reference to issues arising from the

systematic review and others contained within the research brief. The interview schedule was

constructed to cover the following:

(a) descriptions of the practice setting, population and surgical speciality

(b) methods of surveillance (including any previously tried and discarded)

(c) the operational definition of surgical site infection in use

(d) perceived practicality of methods currently employed

(e) any evaluation and linkage of data (see 2nd contact - Interview Schedule: Appendix E).

A draft interview schedule was prepared and a final version agreed by the investigators.

15

Procedure for contacting prospective interviewees

Prospective interviewees were contacted by telephone to arrange a convenient time for interview.

Their agreement was ascertained in relation to (a) the tape recording of the interview and, (b) the

identification of their Trust in the write up of results (see Telephone interview – 1st contact

sheet: Appendix D). A copy of the interview schedule was faxed or e-mailed to the prospective

interviewee in advance of the interview date where possible.

Nine Trusts were initially contacted to arrange an interview. Seven responded and

representatives from six were interviewed (one cancelled the appointment and was unable to re-

arrange a suitable time).

Results

Literature Review

Our literature search overlapped with the Bruce et al (2001) search by one year (1998). The

search yielded a total of 3, 548 article titles and/or abstracts. Two abstractors assessed these

articles independently applying the inclusion and exclusion criteria previously stated. Any

disagreements were agreed by consensus.

Full text versions of research reports were obtained if they potentially met any one of the criteria.

A total of 130 papers were ordered for full text assessment. Of these 74 met the eligibility criteria

and were abstracted in full. Of those that did not meet inclusion criteria 13 were narrative

reviews, 26 did not perform post discharge surveillance in the study, 3 were editorials, 9

contained information that had been published in duplicate, 1 study was systematic review that

was not limited to studies of post-discharge surveillance (references were checked for relevant

articles although none were located) and 1 study was unable to be translated (sufficient details

confirmed that this was not a study comparing methods of post discharge surveillance), 2 were

16

unable to be retrieved and in 1 study the authors were unable to separate the results for the post-

discharge period that was included in the study. Full references of the excluded studies are

included in Appendix G.

A study assigned as Level I evidence is considered the most rigorous and least susceptible to

bias, while a study deemed to contain Level IV evidence is considered the least rigorous and

most susceptible to bias (see Table 2). Systematic reviews of Level I studies of diagnostic tests

were also considered Level I evidence.

Table 2 Levels of evidence for diagnostic test studies (primary research)

Level of Evidence Criteria

I Independent blind comparison of an appropriate spectrum of consecutive patients, all of

who have undergone both the diagnostic and the reference standard

II Independent, blind or objective comparison but in a set of non-consecutive patients, or

confined to a narrow spectrum of study individuals (or both), all of whom have undergone

both the diagnostic test and the reference standard.

III Independent blind comparison of an appropriate spectrum but the reference standard was

not applied to all patients

IV Any of: reference standard was not applied blinded or not applied independently.

No reference standard was applied (case series).

The QUADAS tool (Whiting et al, 2003) was used to assess the validity of the studies comparing

different methods of post discharge surveillance methods.

A total of 75 papers were found that referred to post discharge surveillance for surgical site

infections. Of these, 3 papers not included in the previous review, reported research that

compared different surveillance methods. The remaining 72 papers described a single

surveillance programme. No studies were located that attempted to validate or compare wound

17

classification systems, nor were any studies found that looked at the impact on patient outcomes

of having post discharge wound surveillance programme.

Methods of post-discharge surveillance for surgical site infections in the literature

In the previous review by Bruce et al (2001) a total of 82 studies were found that had some

mention of post-discharge surveillance. Of these the majority were single arm studies, with only

2 that looked at a comparison of different methods of post discharge surveillance.

This review update identified a total of 72 studies that described a single system of post

discharge surveillance for surgical site infection. These studies did not use a reference standard

and as such can be designated level IV studies (Table 2). All but one study looked at non-

objective testing methods such as patient or physician assessment of surgical wound status. One

study looked at blood testing for diagnosis of mediastinitis in a population of patients who had

undergone CABG. As in Bruce et al (2001), the majority of papers found monitoring systems

that undertook intermittent and selective rather than continuous or hospital wide surveillance.

These studies can be further differentiated into three categories based on the primary aim as

reported by the author. These being;

1) Studies undertaken to test a method of post-discharge surveillance without a

comparison group (11 studies identified).

2) Studies undertaken to report the effect of an intervention (surgical or other) and

report on the complications (SSIs or other) experienced by patients undergoing the

therapy. These studies may compare methods of an intervention but the method of

post-discharge surveillance is uniform between the two groups (i.e. a randomised

controlled trial comparing two surgical techniques). Or alternatively these studies

18

may have been undertaken to examine the prevalence of surgical site infections post

discharge within a particular ward or hospital using one method of post discharge

surveillance (45 studies).

3) Studies undertaken with an alternate aim or aim not specified (16 studies).

The majority of studies were undertaken in Western Europe and North America, with 9 studies

having been undertaken in the United Kingdom, in a wide variety of surgical specialties including

general, cardiac, orthopaedic, obstetrics amongst others.

All papers were abstracted irrespective of stated primary aim if it appeared that some attempt

had been made by the authors to describe their post-discharge surveillance programmes. The full

details of all abstracted single arm studies can be found in Appendix B.

This section gives an overview of the methods and duration of surveillance of the 72 studies

included.

The methods used to detect post-discharge SSI were:

• direct observation of the wound by health professional (n=31)

• telephone interviews with patients (n=17)

• patient questionnaire (n=13)

• other methods (n=21). Other methods used included review of operating logs to

examine surgical revisions, cards being given to patients to give to their general

practitioners to return to the hospital should a surgical site infection have been detected,

examination of hospital readmission data, review of pharmacy data and using a mixture

of methods amongst others.

19

• not stated (n=9)

• staff questionnaires (n=8).

It should be noted that more than one method was used in some studies however in these

studies the methods were not compared with each other.

Full details of each surveillance method, duration, study sample size and response rate can be

found in Appendix B.

The CDC definition was the most commonly used definition of surgical wound infection and

was applied in more than 38% (n=28) of studies included. Other definitions used included

authors’ own (n=8, 11% of studies) and other methods or methods unclear (n=10, 14% of

studies). No formal definition of surgical wound infection was provided in 26 (36%) of studies.

Appendix F presents a summary.

The duration of follow up within the post discharge surveillance programmes identified varied

from 3 days post-discharge to several years, with 30 days being the most common duration

(n=34, 50%). The use of a thirty-day follow up point was associated with use of the CDC

definition of surgical wound infection (30 days is stipulated by the CDC as the required duration

of follow up). In many cases the time at which follow up assessments took place was probably

chosen to reflect other outcomes of interest such as functional assessments but this was not

explicitly stated. Only one study stated the CDC criteria as the rationale for following patients up

for a year after they had received a prosthetic implant.

20

These wound infection criteria are provided in Appendix F for descriptive purposes as evidence

does not exist to determine the comparability or validity of these compared to other definitions

e.g., CDC. It should further be noted that these other definitions provide no guidance as to

optimum duration of follow up.

Studies comparing the validity of alternate methods of post-discharge surgical site

infection surveillance.

Bruce et al (2001) categorised the body of literature related to the accuracy and validity of

surveillance of surgical wound infections in four ways. Firstly, there are attempts to compare

different processes of case ascertainment, including case ascertainment by different health care

professionals. Second, these are assessments of patients’ own ability to self diagnose wound

infection, compared with health professional diagnosis. Third, there are validation reports of data

capture methods, in particular, manual versus automated data entry. Finally, there are studies of

the validity of systems and examination of the feasibility of using existing data-collections

systems (e.g. feasibility of antibiotic utilisation as a measure of surgical site infections).

In order for validation studies to be eligible for inclusion in this review, studies had to compare

alternative ascertainment techniques, preferably in consecutive patients. In addition patients had

to receive both methods of surveillance regardless of the results of either method.

Validation of case ascertainment

We regarded studies of case ascertainment as being those which compared an alternative method

of surveillance with a reference standard which involved a purposeful patient examination by a

health care professional with the goal of identification of infection. We found none of these

studies.

21

Validity of systems and existing data collection systems.

Only one study was found that compared the validity of using an existing data system to capture

patients with surgical site infections (Sands et al, 2003). This study did not focus exclusively on

post discharge surveillance.

In this study, Sands et al, (2003) included all patients who were members of the Harvard Pilgrim

Health Care (USA) who underwent CABG surgery between March 1, 1993 and June 30, 1997, if

their algorithm derived probability of postoperative wound infection was = 0.1 or greater were

entered into the study. This resulted in 388 eligible patients from a total of 1352 CABG patients.

The study compared two different surveillance methods; ascertainment of SSI by the capture of

data from existing systems (claims and pharmacy dispensing data, medical records) and

prospective hospital based surveillance using NNIS criteria. One hospital within this study also

used automatic screening of antibiotic use. The study attempted to capture SSIs that occurred

both during the inpatient and post-discharge phases. Patients whose probability of infection was

calculated to be 0.1 had their patient records reviewed by a trained abstractor to determine

whether they had experienced an SSI as defined by the criteria of the CDC National Nosocomial

Infections Surveillance (NNIS) system. No information was provided regarding how the

probability of infection was calculated.

Patients were then classified as infected, or not, according to hospital-based surveillance, claims

based surveillance or neither. The denominator used for the calculation of sensitivity and

specificity was the combined total of infected patients irrespective of method of diagnosis.

Surveillance based on health plan data identified approximately 50% more infections than did

hospital-based surveillance and more than twice the number of infections occurring post

discharge.

22

There are considerable methodological problems inherent in this study; not least the lack of gold

standard comparison and therefore no sense of the extent of misclassification by either system,

neither of which involved purposeful patient examination. Rather the authors have taken the

combined figure of positive results of infection irrespective of method of diagnosis to be the true

positive rate. Whilst this may have increased the reported sensitivity of the tests, the ratio of

difference between the two tests would remain the same. Sufficient detail was not available to

calculate the specificity of either of the methods.

Therefore whilst this study suggests that, in the USA, administrative data were able to detect

more patients with SSIs than hospital based surveillance, its applicability in the UK is low due to

the amount, type and quality of routine data capture in the UK health system. Furthermore, the

study is uninformative regarding post-discharge SSI detection.

Validation of patient self-diagnosis

Four studies were identified that looked at the validation of patient self-diagnosis as a method of

post discharge surveillance (Martini et al, 2000; Mitchell et al, 1999 Seaman & Lammers, 1991 and

Whitby et al, 2001). Two of these studies (Mitchell et al, 1999 and Seaman & Lammers, 1991)

were included in the Bruce et al (2001) review and were re-extracted for this review for

completeness.

In the first study Seaman & Lammers (1991) compared standardised patient interview by health

professional (nurse, nurse practitioner, physician’s assistant, physician, or medical interpreter)

with medical examination (physician, nurse practitioner, or physician’s assistant). Patients in this

study were those whose lacerations had been previously sutured in the emergency department

thus whilst these were not strictly surgical wounds we felt the information was likely to be

generalisable to surgical wounds.

23

The authors of this study calculated patients’ ability to detect infection. This study found that

patients were unable to recognise infections in their own wounds. Of the 21 wound infections

that were identified by health care professional assessment only 11 were detected by patients

themselves, equating to a false negative diagnostic rate of 48%. However the paper did not

detail the questions asked of the patients.

A second study Mitchell et al (1999) compared patient identification of wound infection using

postal questionnaire with a postal questionnaire to surgeons requesting information on the

wound status of the patient obtained at their postoperative review. Patients were chosen to

represent the major elective procedures performed by cardiothoracic, vascular, abdominal,

orthopaedic and gynaecological surgeons at the hospital and covered both clean and

contaminated classes of surgery. Infection was diagnosed and classified on the basis of modified

CDC criteria modified.

Response rates to the questionnaire ranged from 50% to 65% for patients and 42% to 61% for

surgeons. Forms relating to 641 wounds were returned by both the patients and the surgeon; the

total number of eligible wounds is not clear but since there were 1360 eligible patients there is

clearly a high proportion of missing data. Both patients and surgeons agreed that wound

infection was absent in 565 cases. Surgeons classified infection as present in 59 wounds whilst

patients classified infection as present in 74 wounds; the researchers on further investigation

then re-classified 23 wounds regarded as not infected by surgeons as infected – in other words

they rejected the gold standard in favour of patient assessment! Reasons for misclassification

were given as surgical wound assessment preceding the development of infection and patients

having reported their infection to someone other than their surgeon. False negative rates for

patient assessment were also very low. Overall the agreement between surgeon and patient

assessment was fair, with a Kappa of 0.73.

24

The results of this study would appear to suggest that patients may be able to self diagnose

wound infection with a good level of agreement with the surgeon’s routine post-operative

review. The interpretation of these results is slightly hampered by the relatively low response

rates from both the patients and surgeons, which may bias the estimates.

In a third study, Whitby et al (2001) analysed the validity of post-discharge self-diagnosis in

Brisbane, Australia. Patients were chosen from a variety of surgical specialties sampled to contain

a high proportion of patients undergoing ‘contaminated’ and ‘dirty’ procedures; specifically

cardiac, breast, laparotomy, upper gastrointestinal tract, lower gastrointestinal tract, hepatobiliary,

hernia repair, orthopaedics, urology, gynaecology and vascular surgery.

Self-diagnosis including recall of antibiotic prescribing by General Practitioners was compared

with diagnosis by one of two research nurses (the standard method of surveillance in that

hospital and regarded as the reference standard). These two methods were also compared with

diagnosis from wound photographs by a Surgeon and a Medical Microbiologist. One Infection

Control Nurse visited the patient and assessed for the presence of infection or no infection.

Patients were blinded as to the purpose of the post-discharge visits in an effort to eliminate the

potential bias that a weekly visit by the nurse may engender.

Nurse assessment was then compared with self-diagnosis that comprised a postal survey at four

and six weeks postoperatively (the second survey to establish inter-rater reliability the patients

being told it was to assist in rectifying computer problems associated with the initial recording of

questionnaire responses). A third questionnaire was then sent to patients who had replied to the

questionnaire sent at week 4 but not at week 6.

Patients were questioned in the survey as to whether they had noticed or remembered any of the

following events;

25

1. Presence or recall of yellow discharge (with the appearance of pus) alone or

2. a) Presence or recall of fever and

b) Redness or swelling and/or

3. Recall by the pa tient of antibiotics prescribed by a General Practitioner for

suspected SSI.

These criteria were also used for nurse diagnosis. The results of both the patient self diagnosis by

questionnaire and the gold standard of Infection Control Nurse diagnosis were compared with

independent assessment of the photographs the doctors and were then classified as being an SSI

or not. Medical diagnosis was based only on the presence of discharge and/or swelling as visible

from the photographs.

Insufficient details are provided as to judge how well these patients were representative of a full

spectrum of disease and risk.

The authors calculated the correlation of all methods with the gold standard of infection control

nurse diagnosis. Unfortunately correlation is inappropriately applied here since the correlation

co-efficient measures the strength of a relation between the two variables, not the agreement

between them and it is possibly to have poor agreement but a high correlation (Bland & Altman,

1986). In this case there was poor correlation between each of patient recall, surgeon diagnosis

and infection department physician/microbiologist with the gold standard of infection control

nurse assessment (correlation coefficients of r=0.37, 0.39 and 0.38 respectively). Correlation

between the surgeon and physician proved only moderate (r=0.50; 95% C.I. 0.41-0.62). The

greatest correlation with the gold standard was the patients’ recall of General Practitioner

antibiotic prescription (r=0.76; 95% C.I. 0.66-0.87). P values were not reported for the values of

correlation. Inter-rater reliability was not provided for the two research nurses that undertook

the evaluation that formed the gold standard of the study.

26

It is unclear whether having a nurse visit on weekly visit, even if intended only to judge the

cosmetic appearance of the wound, may have altered the perception of the patient with regards

to taking action that they may have felt was appropriate to treat the wound (i.e. would they be

more or less likely to visit their G.P. with concerns about the infection in a wound if they knew

there was a nurse coming to visit regardless).

All patients received the reference tests (i.e. questionnaires and surgeon/physician assessments)

regardless of the findings of the gold standard.

This study found that patients were unable to effectively self assess surgical site infections post

discharge (positive predictive value of 29% - in other words only 29% of people who assessed

themselves as having an infection actually had one). However the negative predictive value for

patient assessment was high (98% of patients who assessed themselves as not having an

infection did not have one). The results of this study suggest that patient recall of antibiotic

prescriptions by General Practitioner warrants further research to test whether it is a viable

alternative to visits by infection control nurses for the detection of surgical site infections that

occur post discharge. This requires further confirmation since correlational data only were

presented. Data presented in the publication were not sufficient to allow for the calculation of

the sensitivity and specificity of any of the methods compared to the gold standard which is the

information required to make an assessment of the validity of the methods being studied.

Furthermore both the positive and negative predictive values that are reported in this study are

highly dependant upon the prevalence of infection in the population and as such cannot be

generalised to other populations.

In the fourth study, Martini et al (2000) compared the results of a patient questionnaire with

outpatient clinic records for the surveillance of post discharge surgical wound infections. The

27

study participants were all patients (n=1664) who had undergone orthopaedic surgery over a 20

month period at a hospital in Tubingen, Germany.

Patients were sent a questionnaire 3 months post discharge asking them to report any wound

infection or inflammation that had occurred post surgery. The results of this survey were then

compared with the patient records in the outpatient clinic where patients had been routinely

followed up post discharge. In additional patients’ general practitioners were contacted to

confirm if a diagnosis of surgical site infection had been made. It was not stated at what time

point/s patients attended an outpatient clinic.

Of the 92% of patients who responded, 64 reported having had an infection or inflammation of

the wound. When this was compared with clinic and general practitioner data the number of

confirmed cases was 9, and in all of these cases the wound infection had been captured at the

outpatient clinic. No a priori definition of wound infection was described. It is unclear if all

practitioners were using the same criteria by which a judgement of wound infection could be

objectively and consistently made.

The authors of this study felt that results of this study showed that patients who were regularly

followed up in outpatient clinics did not require additional questionnaires to capture post

discharge wound infections, however it is not clear that routine outpatient follow up is a robust

enough method of detection to be regarded as a reference method.

No studies were found that compared a “gold standard” reference method with an existing data

collection system only although there was some component of this in one of the studies found

and described above.(Sands et al, 2003)

All studies were then assessed for then assessed for quality using the QUADAS tool. The result

of this assessment is shown below in Table 3.

28

Table 3 Assessment of comparative studies using the QUADAS tool

QUADAS Item Sands (2002)

Whitby (2002)

Seaman & Lammers (1991)

Mitchell (1999)

Martini (2000)

Was the spectrum of patients representative of the patient who will receive the test in practice?

No. CABG patients only.

No. Chose moderate to high risk patients only.

No. Only included patients with lacerations treated at A & E

Yes Unclear. Not enough information presented

Were the selection criteria clearly described?

Yes No No No No

Is the reference standard likely to correctly classify the target condition.

No. Relies on reporting; no observation.

Unclear* Yes Unclear No

Is the time period between the reference standard and index test short enough to be reasonably sure that the target condition did not change between the two tests?

Unclear Unclear No Unclear

Did the whole sample or a random selection of the sample, receive verification using a reference standard of diagnosis?

Yes Yes Yes No. Only those whose surgeon completed the questionnaire

Yes

Did patients receive the same reference standard regardless of the test result?

Yes Yes Yes Yes Yes

Was the reference standard independent of the index test (i.e. the index test did not form part of the reference standard)?

No Yes Yes Yes Yes

Was the execution of the index test described in sufficient detail to permit replication of the test?

Yes Yes Yes Unclear Yes

Was the execution of the reference test described in sufficient detail to permit replication of the test?

No Yes Yes Yes Yes

Were the index test results interpreted without knowledge of the results of the reference test?

Unclear Yes Yes Unclear Yes

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear Yes No Unclear Yes

Were the same clinical data available when test results were interpreted as would be available when the test is used in practice?

No. The information available in this standard would not be available in the UK.

Yes Yes Yes Yes

Were uninterpretable/ Intermediate test results reported?

Unclear Unclear Unclear Unclear Unclear

Were withdrawals from the study explained?

Unclear Unclear Yes Unclear Unclear

Overall, methodological weaknesses were observed in all 5 comparative studies and the quality of

reporting was also poor, as evidenced by the frequency of “unclear” responses to the QUADAS

questions.

29

Discussion

Only five studies comparing alternating methods of surveillance for SSI were located. In brief

these studies compared;

• ascertainment of SSI by the capture of data from existing systems (claims and pharmacy

dispensing data, medical records) with prospective hospital based surveillance using

NNIS criteria (Sands et al, 2003). This study found that data from existing administrative

systems were better able to detect surgical site infections than routinely collected data.

• Surgeon questionnaire-based assessment of patient wound status compared with patient

self-assessment of wound status by postal questionnaire (Mitchell et al, 1999). This study

found that there was fair agreement between surgeons and patients regarding the status

of their wounds however data were missing for the surgeon assessment in 50% of

patients.

• Patient self diagnosis by interview compared with health professional diagnosis and

found that patients were unable to diagnose infection or recognise signs of inflammation

producing false negative rates of 48% (Seaman & Lammers, 1991).

• Infection control nurse diagnosis compared with patient self-assessment, surgeon

diagnosis, infection department physician/microbiologist and patient recall of general

practitioner antibiotic prescription (Whitby et al, 2002). This study found that patients

were not able to adequately identify infected wounds (i.e., a high false negative rate).

Further results in this study are uninterpretable since analysed as correlation rather than

agreement, although it is noteworthy that correlation between methods was poor.

• Patient reported symptoms of infection with outpatient clinic physician diagnosis

(Martini et al, 2000). This study found that similar rates of infection were detected via

patient report and outpatient clinic follow up.

30

Two main issues arose from the comparative studies. Firstly variations in data collection

procedures and classification systems between countries limits the comparability and synthesis of

the post discharge surveillance data. As will be further described many studies do not provide

clear descriptions of the criteria by which a diagnosis of infection was made. Secondly, the

source of data affects the external validity of the study as information collected in one country

may not be readily available in another. Together these two factors limit the applicability of the

available evidence to the United Kingdom where vastly different systems are in place. In addition

methodological limitations are evident in these studies that make the conclusions drawn difficult

due to interpret in light of the limited internal and external validity that is apparent in these

studies.

Audit of current surveillance practice: Results

In total 361 Infection Control personnel were sent an audit form (Appendix C) and asked to

return the form irrespective of whether they were performing post discharge surveillance.

Overall 46% (n=146) of trusts returned the audit form (only one response was counted in the

numerator where multiple responses were received from single institutions). Of those trusts that

responded n=42, 29% reported performing some form of post-discharge surveillance and n=104

(71%) said they were not.

Orthopaedics was the specialty most commonly involved with obstetrics (post-caesarean section)

also being common (23%, n=10). In a small minority of trusts that reported performing PDS,

14% (n=6/42) reported doing so in all surgical areas. Other surgical procedures that were

followed in smaller numbers were vasectomies, craniotomies, CABG, large bowel surgery, breast

surgery, general surgery, hernia repair, vascular surgery and day surgery.

31

The most common methods of PDS reported in the audit was that of routine clinical follow up

45% (n=19/42) and direct observation of the wound in 41% (n=17/42) of positively responding

trusts. An even greater number of respondents 50% (n=21) reported using another method or a

combination of methods. These included; giving forms to patients or primary care providers to

return in 30 days; looking at hospital readmission data at 30 days post discharge; writing letters to

or telephoning General Practitioners to ask if the patient had any signs of infection post

hospitalisation or a combination of these and other methods. Other methods such as surgeon

surveys, patient telephone or postal questionnaires were less frequently undertaken. Consent to

pass on details of Trusts was only obtained from Trusts who participated in telephone

interviews.

Telephone Interviews with Survey Respondents

Nine Trusts were initially contacted to arrange an interview. Seven responded and six were

interviewed.

Summary of findings of telephone surveys of current practice of post-discharge within

the NHS

Interviews were conducted with representatives from the following NHS Trusts:

- Nottingham City Hospital

- University College London Hospitals Trust

- Bedford NHS Hospital Trust

- South Glasgow University Hospitals Trust

- Ayrshire and Arran Acute Hospitals Trust

32

- Poole Hospital NHS Trust

All post discharge surveillance (PDS) systems discussed below represent activity that was

considered to be beyond routine follow up. Four locations (Nottingham City Hospital NHS

Trust, University College London Hospitals NHS Trust, Bedford NHS Trust and South

Glasgow University Hospitals Trust) were operating a system in more than one surgical

speciality.

The surgical specialities involved in post-discharge surveillance amongst the five locations that

were available for interview were obstetrics, orthopaedics and neurosurgery. Methods of post

discharge surveillance included patient telephone survey (n=3), patient questionnaire completed

at outpatient visit (n=1), patient questionnaire sent to patients’ residence (n=1), GP/health

professional survey (n=2), direct observation (n=3) and others (hospital re-admissions data)

(n=2). Two locations (Ayrshire & Arran Acute Hospitals Trust and South Glasgow University

Hospitals Trust) were using more than one method in the same surgical speciality.

Centres using patient telephone survey as the method (or part method) of PDS

Systems in operation at Ayrshire & Arran Acute Hospitals NHS Trust and South Glasgow

University Hospitals Trust and University College London Hospital in London included the use

of a patient telephone survey as part of their PDS system. At Ayrshire & Arran, all caesarean

section patients received a telephone call from a midwife (at day 28 post surgery) to confirm

wound status. SSHAIP criteria (Scottish Surveillance of Healthcare Associated Infection

Program – issued by SCIEH) were used to classify SSI. The system, established since October

2002 in conjunction with direct observation (see below), was deemed to work well as part of an

established after-care pathway in this surgical speciality. No additional resource requirements

had been identified due to its perceived ease of incorporation within the community midwife

role. Data regarding patients with surgical site infection were fed back within the hospital (to

33

clinical staff, surgeons, anaesthetists, Infection Control Committee) and at national level (to

SCIEH). A previous telephone survey system (without community midwife involvement) had

been tested at this location and constituted the pilot for the current procedure.

At South Glasgow, all patients with primary hip/knee replacements were monitored using CDC

guidelines. Due to some patients often being discharged to outlying areas (with no routine

follow up appointment made at the treatment location), telephone contact was made by

surveillance nurses during the first year after discharge in order to assess wound status with the

patients. No costings had been undertaken for this part of the PDS procedure. This system was

accompanied by checks on routine hospital data (see below). Again, data regarding post

discharge infections were fed back both locally (to clinicians, multidisciplinary teams and allied

health care professionals) and nationally to SCIEH and the Pan-Celtic Collaboration.

At University College London hospital post-discharge surveillance programmes have been in

place since 2000. Telephone surveillance is undertaken over one 6 month period each year

capturing all patients who have undergone surgery in that institution within that time period. In

addition patients are also asked to complete a postal questionnaire. Patients are asked a series of

questions so that the infection control team are able to assess and classify the patients’ wounds

using a comparison of four methods; ASEPSIS, CDC, NNINS and the definition provided by

Cruse & Foord (1973). They have found that the range of systems is more informative than any

single definition of wound infection, however the questionnaires have not been compared with

direct observation of the wound. The questions that are asked relate to each patient’s condition

for the entire period post-discharge (i.e. not solely at the time of survey completion). The average

response rates for the two methods of surveillance have been estimated at 88 to 90%.

34

The cost of this programme has been estimated at approximately £90,000 per annum (the

respondents claim a large proportion of these costs are offset by cost savings due to the

surveillance programme).

Centres using patient questionnaire as the method (or part method) of PDS

The PDS system at Nottingham City Hospital NHS Trust consisted of a patient questionnaire

administered at outpatient visits following all orthopaedic surgical procedures. This system was

provided as an extension to the mandatory inpatient surveillance after orthopaedic surgery

(Department of the Communicable Disease Surveillance Centre at the Health Protection Agency

- April 2004) and data were fed back at national level.

Centres using GP/health professional survey as the method (or part method) of PDS

Two centres used reports from GPs or other primary health care professionals as the basis of

surveillance. At Bedford NHS Trust, infection control nurses surveyed GPs 2 to 2.5 years after

surgery, using NINSS guidelines (cases identified from the hospital register). GPs were asked for

a simple “Yes/No” answer as to whether any of their patients had been referred back to an

orthopaedic surgeon for an “important” infection (defined as “deep”, not “superficial”) during

that time period. A good response rate was reported from GP surgeries (generally over 85%).

The system was considered to be easy to set up and run, although costs to Primary Care were

unknown. Feedback of data was given at local level (within hospital).

A new system of PDS was established at Poole Hospital NHS Trust in January 2004 involving

follow up of all patients with repair of fractured neck of femur. A standard leaflet regarding

wound care was issued to patients upon discharge; at the same time GPs, district nurses and

practice nurses are advised (by letter) regarding the purpose of the proposed PDS system. They

were advised to report any wound problem occurring in patients during at least the first year after

surgery. Regular telephone contact with GPs was made (the timing and frequency of this was

35

reported to be variable and dependent upon resources) in order to identify whether the patient

has represented to the surgery. Given that the hospital would subsequently follow up the patient

(this time using modified CDC definitions of wound infection), the resource impact of this

procedure upon GPs was considered to be minimal. Despite this, the practicality of this system

was often hindered by low GP response rates. There was some internal (within hospital)

feedback of data, but no link to any national audit.

Centres using direct observation as the method (or part method) of PDS

Three of the five locations (Nottingham City Hospital NHS Trust, Ayrshire & Arran Acute

Hospitals Trust and Bedford NHS Trust) were using direct observational techniques during the

process of post discharge surveillance on all caesarean section patients. All systems were

conducted in the community setting, with one (Bedford Trust) maintaining a linkage of data

between hospital and community recording systems. Patient information on wound self-

assessment was given in one location (Ayrshire & Arran). This was delivered in the form of a

letter, which also contained instructions for the patient to contact their GP or midwife as and

when required. Established surgical wound classification systems were in operation across the

three sites (CDC/NINSS and SSHAIP). Surveillance after caesarean section was considered

practically feasible since regular home visits by midwives and health visitors is standard practice.

However, it was pointed out (by the representative of Nottingham Trust) that stronger data links

should be forged with primary, secondary and community care services in order to utilise

established communication channels more effectively. Although no formal economic evaluation

had been carried out at any of the locations, no significant extra costs were perceived for

community services. One location (Ayrshire & Arran) reported that feedback was directed at

national level (SCIEH), the other two reporting formally on an internal (within hospital) basis

only.

36

Centres using other methods for PDS South Glasgow University Hospitals Trust was using hospital re-admissions data for the

monitoring of surgical site infections in both orthopaedic and neurosurgery specialities. The

system in orthopaedics was used in conjunction with telephone follow up (reported above). The

system in neurosurgery was of a limited nature. The programme (which comprised checks on re-

admissions data) was reported to be a difficult procedure to administer, mainly due to the

dispersal of patients returning to regional care centres after discharge from South Glasgow. Such

a method of surveillance will only identify the most severe cases of infection and depends on

having adequate record flagging or linkage.

Other relevant research currently being undertaken or planned within the United

Kingdom

Complementary work is being currently or about to be performed in a number of relevant areas.

Risk factors for surgical wound infections

Dr Barney Reeves of the London School of Hygiene and Tropical Medicine is undertaking work

to investigate risk adjustment methods for surgical wound infection by (a) systematic reviews and

(b) statistical modelling using existing databases.

First, systematic reviews will be carried out on: (a) risk factors; (b) stratification by procedure

type; (c) evidence about the effect on risk-adjustment models of post-discharge surveillance.

Second, data contained in each database, methods of data collection and potential biases arising

will be described. Third, statistical modelling will be carried out: (a) uni-variable analyses of

individual risk factors and potential two-way interactions between risk factors; (b) multi-variable

modelling, initially without and subsequently with consideration of the data hierarchy (clustering

37

patients within surgeons and centres); (c) exploration of alternative definitions of SSI. This work

was due to be completed by March 2004.

Determining diagnostic characteristics without a gold standard

This work also commissioned by the National Co-ordinating Centre for Research Methodology

seeks to commission an analytical piece that will give guidance to researchers as to the best

methods to use in the absence of a gold standard. This is relevant to detection of surgical

infection in that there is no consensus as to what the precise components of a gold standard

method of detection are, and no objective measure of the presence of infection.

Clinical Practice Guideline on Surgical Site Infection: prevention and treatment

This work has been commissioned by the National Institute for Clinical Excellence and is being

undertaken at the National Nursing and Supportive Care Collaborating Centre (RCN Institute).

The guideline will be restricted to incisional procedures, in people at risk of SSI. Diagnosis,

classification and surveillance of infection are all within the scope of this guideline. Guideline

development was reported as being due to begin June/July 2004.

Conclusions

Question 1: What are the features of existing post discharge surveillance systems in

terms of coverage; denominator data; diagnostic criteria; other data collected; methods

of data collection?

There is great variation in the methods of PDS SSI and particularly in the sources of data used.

Multiple methods are commonly used including routine clinic follow up plus data from primary

care providers. Orthopaedics and obstetrics are the most frequently observed surgical

specialities undertaking PDS SSI. Importantly whilst there are na tional policies to capture SSIs

38

(for example the Health Protection Agency’s mandatory orthopaedic surgical site infection

surveillance which began on April 1st 2004), these schemes does not require post-discharge

surveillance and do not give guidance as to how this could be undertaken.

Question 2: What is the current evidence for the validity, reliability and practicality of

different methods of post-discharge surveillance of SSIs?

Only five studies have compared alternative methods of PDS SSI and none of these is a proper

study of diagnostic test accuracy comparing an alternative method with a gold standard or

reference method of surveillance. The five studies identified represent five unique comparisons

in varied and sometimes highly selected populations. The following methods have been studied

at least once in a comparison:

- capture of routine data (insurance claims, pharmacy data, records)

- prospective hospital based (including outpatient visits) surveillance (3 studies)

- patient interview by health care professional

- patient postal questionnaire (3 studies)

- surgeon questionnaire

- general practitioner questionnaire

- home based infection control nurse assessment

- doctors’ assessment using photographs

There is no uniformity of gold standard or reference method; in at least one study the surgeons’

assessment was over-ridden by the patient assessment. The diagnostic accuracy of routine

outpatient follow up as a method of surveillance has not been proven. Methods of routine data

capture which have performed well in the USA are not applicable to the UK.

39

The studies of patient self-diagnosis gave conflicting results (and used varied methods of aiding

self-diagnosis that were not usually described in sufficient detail to permit replication). The only

study which suggested patients’ assessments were accurate had so much missing data as to

completely undermine the findings.

The survey of national practice gives some impression of which methods are feasible and

practical. It is noteworthy that routine outpatient follow up is the most commonly used method

but probably what constitutes routine follow up varies between sites, and the research literature

does not provide evidence of its reliability as a method. The data suggests that at least in some

locations, patient telephone survey is feasible and this may be a good method of capturing data

from patients who do not attend for routine follow up. Some centres are routinely capturing

data from primary care – a method that shows promise in some of the research.

Question 3: What research is needed to fill knowledge gaps?

• Criteria to be used to define and classify surgical site infection

The international standard is the CDC definition (Mangram et al, 1999) and this should be used

as the basis for patient and professional assessment.

There are those who believe that alternative methods of wound assessment (e.g., ASEPSIS) have

benefits over the CDC criteria (particularly the fact that CDC criteria allow for physician or

surgeon impression of infection alone to be sufficient evidence) and comparison with an

alternative system could be undertaken within any surveillance study.

40

• Evaluating alternative methods of post-discharge surveillance

The crucial decision to be made relates to what the appropriate “gold standard” or reference

surveillance method will be for comparative purposes. The best candidate for this is not clear

but given the absence of data supporting alternatives it should probably be health professional

assessment at around 30 days post-op (timing dictated by CDC criteria). It appears from our

survey that there are centres where this currently occurs (Bedford for caesarean section patients

for example) and which might be candidates to host further comparative research. If health

professional assessment proves prohibitively expensive or impractical then use of primary care

data (antibiotic prescriptions, GP survey, primary care data capture) may be a valuable

alternative. Again Bedford NHS Trust use this method and report good response rates from

GPs whilst Poole NHS Trust complains of poor response rates.

The first step in any research to explore the use of patient self-diagnosis will be the development

and pilot of a suitable instrument for use by patients, and it should be fairly straightforward to

use CDC criteria as the basis for this.

• Can patients self-diagnose surgical wound infection at an acceptable level of

accuracy?

The identification of a sufficiently valid, reliable and cost-efficient method of post-discharge

surveillance for surgical site infection is essential given the need for, and expansion in, this

activity. It is probably fair to say that the reference standard method of surgical site surveillance

is currently the direct observation of the surgical site by a health professional for up to 30 days

for a standard operative procedure and up to a year where a prosthesis has been implanted.

However the precise components of this reference standard are impossible to define at this time.

To what extent does the type of health professional assessing the wound make a difference?

41

Could health care assistants be trained to undertake this role effectively (probably)? How

frequently should wounds be assessed in order to optimally capture clinically important infection

(no studies comparing different time periods have been identified)?

An alternative method to health professional assessment, for detecting surgical site infections

after discharge, should be developed and evaluated. Such a method would trade off sensitivity

and specificity in favour of efficiency. Unfortunately the existing evidence does not give clear

direction as to what the components of this new method should be but an important method to

evaluate would be CDC criteria-based patient self-assessment. The diagnostic validity of the new

method should then be compared with professional assessment (see desirable features of future

studies, below).

As regular home visits by health care professionals or health care assistants are expensive to

undertake, consideration could be given to researching self-diagnosis in patient groups already

receiving follow up after discharge. Women who have had a caesarean delivery; people who

have undergone coronary artery bypass grafting or hip surgery (it will be important that a range

of ages is represented) are potential groups in whom feasibility could be explored. An economic

evaluation should also be incorporated in order to weigh the costs and benefits of the new

method of surveillance.

• What is the most reliable method of detecting surgical site infection post discharge

using patients as informants?

It is not clear whether patient generated detection is best achieved via questionnaires or

telephone interviews at fixed time points. Telephone interviews are likely to give better response

rates but are far more resource intensive therefore the obvious candidate for initial comparison

42

with professional assessment would be patient completed postal questionnaire. Telephone

interviews would be an alternative only if response rates were inadequate via postal

questionnaire.

Desirable Features of Future Studies

a) The spectrum of patients chosen should be representative of those patients who will

receive the surveillance in practice and include a wide spectrum of underlying risk

and co-morbidities. The inclusion of a wide spectrum of patient means that the

clinical applicability (generalisability) is increased.

b) It is likely that future studies will use multiple raters and it is essential that all raters

are fully trained for their role; are using the same criteria for assessment; and that

inter-rater reliability (or agreement) is periodically measured within the study (and

reported).

c) The time period between the application of the reference standard and the

comparator method of surveillance should be sufficiently short to ensure stability in

the wound status. Furthermore there is a need for both tests to be undertaken at a

time period by which a surgical site infection should be reasonably expected to have

occurred (i.e. surveillance carried out one week post-discharge may be less likely to

capture surgical site infections than one carried out at 30 days).

d) Patients need to have both the new and the reference method of surveillance in order

to calculate the predictive validity of the new surveillance method. In addition both

methods need to be applied irrespective (and independently) of the results of either

test.

e) Description of both the reference and alternate method should be described in

sufficient detail to permit replication.

43

f) Interpretation of the data from alternative surveillance methods should occur

independently.

g) The clinical information that health professionals access when judging whether a

surgical wound infection has occurred during the study should be the same as is

available in routine clinical practice within the United Kingdom.

h) Any withdrawals from the study should be clearly explained so as to reduce any bias

that may occur due to systematic differences between patients who withdraw and

those who remain can be examined.

i) Future studies should integrate mechanisms for dealing with equivocal or ambiguous

assessments.

• Does post-discharge surveillance make a difference to patient outcomes?

The evidence currently suggests that audit and feedback has only modest effects on professional

practice (Grimshaw et al, 2004) therefore a simple evaluation of the impact of PDS SSI (e.g., in a

randomised controlled trial) does not seem sensible (since audit and feedback alone would not be

likely to change professional practice and impact on patient outcomes). Any new method of

PDS SSI should be used to measure outcomes of other interventions aimed at reducing surgical

site infection and as part of ongoing performance monitoring once an adequate method of risk

adjustment has been identified.

44

Appendix A Search strategies used in the report Strategy no. 1. Surgical wound infection: prospective studies

Medline search strategy 1 exp surgical wound infection/ 2 exp wound infection/ 3 (postoperative wound infection$ or post operative wound infection$).tw. 4 surgical wound infection$.tw. 5 wound infect$.tw. 6 (wound infect$) adj10 surgery.tw. 7 surgical site infection$.tw. 8 (nosocomial infection$) adj10 surg$.tw. 9 or/1-8 10 exp prospective studies/ 11 exp cohort studies/ 12 exp longitudinal studies/ 13 exp follow-up studies/ 14 prospective stud$.tw. 15 cohort stud$.tw. 16 longitudinal stud$.tw. 17 (follow-up stud$ or followup stud$).tw. 18 panel stud$.tw. 19 or/10-18 20 9 and 19 21 animal/ not (animal/ and human/) 22 20 not 21

45

EMBASE search strategy 1 exp surgical infection/ 2 exp wound infection/ 3 exp surgical wound/ 4 (postoperative wound infection$ or post operative wound infection$).tw. 5 Surgical wound infection$.tw. 6 wound infect$.tw. 7 (wound infect$) adj10 surgery.tw. 8 Surgical site infection$.tw. 9 (nosocomial infection$) adj10 surg$.tw. 10 or/1-9 11 exp prospective studies/ 12 exp cohort analysis/ 13 exp longitudinal studies/ 14 exp follow-up studies/ 15 prospective stud$.tw. 16 cohort stud$.tw. 17 longitudinal stud$.tw. 18 (follow-up stud$ or followup stud$).tw. 19 panel stud$.tw. 20 or/11-19 21 10 and 20 22 animal/ or nonhuman/ 23 human/ 24 22 not (22 and 23) 25 21 not 24

46

CINAHL search strategy 1 exp wound infection/ 2 exp surgical wound infection/ 3 (postoperative wound infection$ or post operative wound infection$).tw. 4 surgical wound infection$.tw. 5 wound infect$.tw. 6 (wound infect$) adj10 surgery.tw. 7 surgical site infection$.tw. 8 (nosocomial infection$) adj10 surg$.tw. 9 or/1-8 10 exp prospective studies/ 11 prospective stud$.tw. 12 cohort stud$.tw. 13 longitudinal stud$.tw. 14 (follow-up stud$ or followup stud$).tw. 15 panel stud$.tw. 16 or/10-15 17 9 and 16 Cochrane Library search strategy 1 wound-infection explode all trees (MeSH) 2 surgical-wound-infection explode all trees (MeSH) 3 postoperative wound infection* or post operative wound infection* 4 surgical wound infection* 5 wound infect* 6 (wound infect*) near surgery 7 surgical site infection* 8 (nosocomial infection*) near surg* 9 (#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8) 10 prospective-studies explode all trees (MeSH) 11 cohort-studies explode all trees (MeSH) 12 longitudinal-studies explode all trees (MeSH) 13 follow-up-studies explode all trees (MeSH) 14 prospective study or prospective studies 15 cohort study or cohort studies 16 longitudinal study or longitudinal studies 17 follow-up study or follow-up studies or followup study or followup studies 18 panel study or panel studies 19 (#10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18) 20 (#9 and #19)

47

Search strategy no. 2. Surgical wound infection: measurement and validation

studies

Medline search strategy 1 exp surgical wound infection/ 2 exp wound infec tion/ 3 (postoperative wound infection$ or post operative wound infection$).tw. 4 surgical wound infection$.tw. 5 wound infect$.tw. 6 (wound infect$) adj10 surgery.tw. 7 surgical site infection$.tw. 8 (nosocomial infection$) adj10 surg$.tw. 9 or/1-8 10 exp reproducibility of results/ 11 exp "Sensitivity and Specificity"/ 12 exp predictive value of tests/ 13 exp diagnostic errors/ 14 validity.tw. 15 reliability.tw. 16 sensitivity.tw. 17 specificity.tw. 18 reproducibility.tw. 19 test retest.tw. 20 measurement.tw. 21 grading system$.tw. 22 scoring system$.tw. 23 definition.tw. 24 or/10-23 25 9 and 24 26 wound scoring.tw. 27 wound grading.tw. 28 wound classification.tw. 29 or/26-28 30 25 or 29 31 animal/ not (animal/ and human/) 32 30 not 31

48

EMBASE search strategy 1 exp surgical infection/ exp wound infection/ exp surgical wound/ (postoperative wound infection$ or post operative wound infection$).tw. surgical wound infection$.tw. wound infect$.tw. (wound infect$) adj10 surgery.tw. surgical site infection$.tw. (nosocomial infection$) adj10 surg$.tw. or/1-9 exp reproducibility/ exp diagnostic accuracy/ exp measurement/ exp reliability/ exp observer variation/ exp accuracy/ validity.tw. reliability.tw. sensitivity.tw. 20. specificity.tw. 21. reproducibility.tw. 22. test retest.tw. 23. measurement.tw. 24. grading system$.tw. 25. scoring system$.tw. 26. definition.tw. 27. or/11-26 28. 10 and 27 29. wound scoring.tw. 30. wound grading.tw. 31. wound classification.tw. 32. or/29-31 33. 28 or 32 34. animal/ or nonhuman/ 35. human/ 36. 34 not (34 and 35) 37. 33 not 36

49

CINAHL search strategy 1 exp wound infection/ 2 exp surgical wound infection/ 3 (postoperative wound infection$ or post operative wound infection$).tw. 4 surgical wound infection$.tw. 5 wound infect$.tw. 6 (wound infect$) adj10 surgery.tw. 7 surgical site infection$.tw. 8 (nosocomial infection$) adj10 surg$.tw. 9 or/1-8 10 exp reproducibility of results/ 11 exp reliability and validity/ 12 exp test-retest reliability/ 13 exp clinical assessment tools/ 14 exp diagnostic errors/ 15 validity.tw. 16 reliability.tw. 17 sensitivity.tw. 18 specificity.tw. 19 reproducibility.tw. 20 test retest.tw. 21 measurement.tw. 22 grading system$.tw. 23 scoring system$.tw. 24 definition.tw. 25 or/10-24 26 9 and 25 27 wound scoring.tw. 28 wound grading.tw. 29 wound classification.tw. 30 or/27-29 31 26 or 30

50

Cochrane Library search strategy 1 wound-infection explode all trees (MeSH) 2 surgical-wound-infection explode all trees (MeSH) 3 postoperative wound infection* or post operative wound infection* 4 surgical wound infection* 5 wound infect* 6 (wound infect*) near surgery 7 surgical site infection* 8 (nosocomial infection*) near surg* 9 (#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8) 10 reproducibility of results explode all trees (MeSH) 11 diagnostic errors explode all trees (MeSH) 12 predictive value of tests explode all trees (MeSH) 13 sensitivity and specificity explode all trees (MeSH) 14 validity 15 reliability 16 sensitivity 17 specificity 18 reproducibility 19 test retest 20 measurement 21 grading system* 22 scoring system* 23 definition 24 (#10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22

or #23) 25 (#9 and #24) 26 (wound classification) 27 (wound scoring) 28 (wound grading) 29 (#26 or #27 or #28) 30 #25 or #29

51

Search Strategy no. 3. Surgical wound infection: surveillance and monitoring

studies

Medline search strategy 1 exp surgical wound infection/ 2 exp wound infection/ 3 (postoperative wound infection$ or post operative wound infection$).tw. 4 surgical wound infection$.tw. 5 wound infect$.tw. 6 (wound infect$) adj10 surgery.tw. 7 surgical site infection$.tw. 8 (nosocomial infection$) adj10 surg$.tw. 9 or/1-8 10 exp population surveillance/ 11 exp sentinel surveillance/ 12 exp data collection/ 13 exp registries/ 14 exp medic al records/ 15 exp nursing records/ 16 aftercare/ 17 exp patient discharge/ 18 exp outpatients/ 19 exp ambulatory surgical procedures/ 20 surveillance.tw. 21 monitoring.tw. 22 patient discharge.tw. 23 hospital discharge.tw. 24 ambulatory surgery.tw. 25 or/10-24 26 9 and 25 27 animal/ not (animal/ and human/) 28 26 not 27

52

EMBASE search strategy 1 exp surgical infection/ 2 exp wound infection/ 3 exp surgical wound/ 4 (postoperative wound infection$ or post operative wound infection$).tw. 5 surgical wound infection$.tw. 6 wound infect$.tw. 7 (wound infect$) adj10 surgery.tw. 8 surgical site infection$.tw. 9 (nosocomial infection$) adj10 surg$.tw. 10 or/1-9 11 exp hospital discharge/ 12 exp ambulatory surgery/ 13 exp health survey/ 14 exp registries/ 15 exp medical record/ 16 exp outpatient/ 17 ambulatory surgery.tw. 18 hospital discharge.tw. 19 patient discharge.tw. 20 surveillance.tw. 21 monitoring.tw. 22 or/11-21 23 10 and 22 24 animal/ or nonhuman/ 25 human/ 26 24 not (24 and 25) 27 23 not 26

53

CINAHL search strategy 1 exp surgical wound infection/ 2 exp wound infection/ 3 (postoperative wound infection$ or post operative wound infection$).tw. 4 surgical wound infection$.tw. 5 wound infect$.tw. 6 (wound infect$) adj10 surgery.tw. 7 surgical site infection$.tw. 8 (nosocomial infection$) adj10 surg$.tw. 9 or/1-8 10 exp disease surveillance/ 11 exp data collection/ 12 exp registries, disease/ 13 exp medical records/ 14 exp patient discharge/ 15 exp outpatients/ 16 exp ambulatory surgery/ 17 after care/ 18 surveillance.tw. 19 monitoring.tw. 20 patient discharge.tw. 21 hospital discharge.tw. 22 ambulatory surgery.tw. 23 or/10-22 24 9 and 23

54

Cochrane Library search strategy 1 wound-infection explode all trees (MeSH) 2 surgical-wound-infection explode all trees (MeSH) 3 postoperative wound infection* or post operative wound infection* 4 surgical wound infection* 5 wound infect* 6 (wound infect*) near surgery 7 surgical site infection* 8 (nosocomial infection*) near surg* 9 (#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8) 10 population-surveillance explode all trees (MeSH) 11 patient-discharge explode all trees (MeSH) 12 ambulatory-surgical-procedures explode all trees (MeSH) 13 sentinel-surveillance explode all trees (MeSH) 14 data-collection explode all trees (MeSH) 15 registries explode all trees (MeSH) 16 medical-records explode all trees (MeSH) 17 nursing-records explode all trees (MeSH) 18 aftercare explode all trees (MeSH) 19 outpatients explode all trees (MeSH) 20 surveillance 21 monitoring 22 patient discharge 23 hospital discharge 24 ambulatory surgery 25 (#10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22

or #23 or #24) 26 (#9 and #25)

55

Appendix B Critical Appraisal of Included non-comparative PDS studies Table B1 Descriptive characteristics of the study First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Amiel (1999) Israel 1033/1098 (94% to telephone survey, 87.6% attended the 7 day outpatient clinic follow-up)

Intermittent To evaluate the cosmetic outcomes and complications of the application of N-Butyl-2-cyanocrylate for the approximation of elective surgical incisions in a paediatric population.

Hospital Paediatrics HAB (Histoacryl Blue, N-Butyl-2-cyanoacrylate) tissue adhesive: Orchidopexy, inguinal hernia, umbilical hernia or hydrocele repair.

Andreasen (2002)

Denmark 180 operations (100%)

Intermittent To gain experience with continuous electronic registration of data regarding postoperative wound infection following heart surgery.

Hospital Cardiac All heart operations

56

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Arbman (2000) Sweden 80 patients (%followed up varied depending on the time point but ranged from 84% and 97%)

Intermittent The aim of this study was to compare closed (Ferguson) haemorrhoidectomy to open (Milligan-Morgan) haemorrhoidectomy regarding postoperative conditions, complications and long term results

Hospital Gastrointestinal Closed and open haemorrhoidectomy

Astagneau (2001)

France 38973 Intermittent To determine whether SSI is associated with severe or fatal outcome in surgical patients and what is its role, independent of other survival predictors.

Private and public hospitals

Various Various

Avato (2002) USA 1271/1324 (96%) Intermittent To assess the influence of post discharge infection surveillance on risk-adjusted SSI rates for CABG procedures

Tertiary care referral hospital

Cardiac CABG procedures that included both chest and leg incisions

57

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Beaujean (2002) Netherlands 515/576 (89%) Intermittent To compare the efficacy of two surveillance methods in the diagnosis of SSIs. CS (using complication surveillance criteria). NIS using CDC criteria.

Hospital General, trauma & vascular (emergency admissions included/outpatients excluded)

Various

Bhatia (2003) India 615 patients (% response not reported)

Intermittent To report postoperative wound infection in CABG surgery patients

Hospital Cardiac CABG

Bitzer (2000) Germany 502 patients(73% at 3 months)

Intermittent To determine the success of surgery for inguinal hernia from the patients perspective.

NS NS Inguinal hernia repair

Cadwallader (2001)

Australia 510 procedures (100%)

Intermittent Two compare two different methods of surgical site infections following orthopaedic surgery

Hospital Orthopaedics Hemi-arthoplasty or internal fixation of fractured neck of femur.

Callaghan (1999) USA 24 patients (50% follow up at final follow-up time not stated)

Intermittent To assess whether one stage revision surgery for infection after total hip arthroplasty in selected patients predictably eradicates the infection

NS Orthopaedics One-stage revision surgery of the infected hip.

58

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Canonico (2001) Italy 18 patients Intermittent To describe the preliminary results of a special method of wound closure in varicose vein surgery

NS NS Varicose vein surgery

Dagan (1999) Canada 455 patients (% followed up not reported)

Intermittent To evaluate whether changes have occurred at our centre in the rate of nosocomial infections and in the infectious organisms consequent to changes in policy and procedure as of 1987.

Hospital Cardiac surgery Cardiopulmonary bypass/no bypass

De Boer (2001) The Netherlands Number followed up post discharge not reported

NS To assess the relative importance of risk factors for surgical site infections following total hip and knee prosthesis in the Netherlands

Hospitals Orthopaedics Total hip and knee prostheses.

Delgado-Rodriguez (2001)

Spain 1444/1477 (97.8%) followed up

Intermittent To study postoperative infections in hospital and after discharge and to identify the risk factors for such infections

Hospital General surgery Various

59

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Derzie (2000) USA 331 patients NS To compare to two different wound closure techniques in gastric surgery for morbid obesity to see which is associated with lower rates of acute wound complications.

NS Gastric Gastric Bariatric operations

Eriksen (2003) Tanzania 396 operations on 388 patients (64% follow up)

Intermittent To identify the incidence of SSI and associated risk factors.

Hospital General surgery (orthopaedic, urological and gynaecological operations were not included)

Various

Ferraz (2001) Brazil 50 patients (100% follow up)

Intermittent To analyse the late clinical follow-up of patients operated on for Chagas megaseophogus with the Thal-Hatafuku procedure.

Hospital NS Thal-Hatafuku procedure.

Fields (1999) USA 50 surgeons (83-88% response to survey questionnaires)

Intermittent To evaluate a PDS system using surgeon questionnaires

Hospital General Various

Fowler(2003) USA 5500 patients (15.5% patients)

Intermittent To evaluate the utility of blood cultures in identifying patients with mediastinitis.

Hospital Cardiac CABG

60

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Friedman (2001) USA 555 TKA (total knee arthroplasty)

Intermittent To evaluate whether the additional use of electronic chart review is an effective and efficient means of identifying SSI

Hospital Orthopaedic TKA (Total knee arthroplasty)

Gerbitz (2000) Switzerland 106 patients Intermittent Retrospective

Unclear Hospital NS Hernia operations

Goldsborough (1999)

USA 547 patients (no response rate reported)

Unclear To determine the prevalence of postoperative leg wound complications after CABG at our institution and to determine factors that place patients at higher risk for these complications.

Hospital Cardiac CABG alone or in combination with another cardiac procedure.

Habib (2002) Saudi Arabia 754 caesarean sections

Intermittent To measure the rate of wound infection after caesarean section and to assess risk factors

Hospital Obstetrics Caesarean section

Heah (2000) UK 25 NS To evaluate the role of laparoscopic suture rectopexy without resection as a safe and effective treatment for full- thickness rectal prolapse.

Hospital Colorectal surgery Laparoscopic suture rectopexy

61

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Henderson (2001)

Australia 158/207 (76%) Intermittent To establish whether patient information during hospitalisation is sufficient to enable patient self-care upon discharge

Hospital General surgery Abdominal/colorectal

Henriksen (2003) Denmark 1090/1946 (56%) Intermittent To assess the incidence and severity of short term complications after breast augmentation

Private/public clinics of plastic surgery

Plastic surgery Breast augmentation

Higgins (1999) USA 450 patients Intermittent To test the hypothesis that prophylactic antibiotic treatment in elective laparoscopic cholecystectomy does not lower the already low infection rate associated with this procedure

Hospital NS Laparoscopic cholecystectomy

Hui (1999) USA 621/1412 (44%) Intermittent Prospective analysis of mechanisms of PGB (perforation of gallbladder) and postoperative infectious complications

Hospital NS Laparoscopic cholecystectomy

62

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Jamali (2001) Pakistan 252 patients Intermittent To compare the effect of theatre condition on postoperative infection rate.

Hospitals NS Clean surgery only

Jonkers (2003) Netherlands 1885 (100% patient response) 91% response from family physician

Intermittent To study the prevalence of sternal wound and donor site infections during hospitalisation and at 30/90 days post operatively

Hospital Cardiac Various

Kent (2001) Australia 1257/1291 (98.7% followed up)

Intermittent To examine the impact of effective post-discharge SSI follow up on the overall SSI rate

Hospital General surgery Joint replacement Gall bladder surgery Abdominal/gynaecological surgery

Killian (2001) USA 765 caesarean sections (30% response rate from physician survey)

Intermittent To identify risk factors associated with SSIs following caesarean sections

Hospital Obstetrics & Gynaecology Caesarean section

Krupova (2001) NS 117 patients (% followed up not reported)

NS NS Hospital Neurosurgery Various

Kumar (2001) UK 20 patients (100% follow up)

Intermittent To review the safety, practicality and cost-effectiveness of day surgery anterior cruciate ligament reconstruction

NS (presume hospital)

Orthopaedics Anterior cruciate ligament reconstruction

63

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Letrelliart (2001) France 2199/7540 (29%) Intermittent To study both surgical and nonsurgical nosocomial infections (NIs) seen by primary-care physicians (general practitioners [GPs]) in France.

General practice NS NS

Mahatharadol (2001)

Thailand 102 (100% follow up)

Intermittent To assess postoperative complications in patients receiving antibiotic prophylaxis in elective laparoscopic cholecystectomy

Hospital NS Elective laparoscopic cholecystectomy

Mahomed (2003) Canada/USA 75,051 patients Population-based study

Continuous To characterise rates and immediate postoperative outcomes of primary and revision total hip replacement

Medicare population

Orthopaedics Primary & revision total hip replacement

Marroni (2003) Italy 668 patients NS NS Hospital Various Vascular surgery Amputation Other cardiovascular surgery

McCowan (1999) Scotland 62 Intermittent Pilot study to establish means of assessing post discharge SSI

Hospital NS NS

Melling (2001) UK 421patients (% response unclear)

Intermittent within an RCT

To asssess whether warming patients before short duration, clean surgery would reduce infection rates.

Hospital Various Elective hernia repair, varicose vein surgery or breast surgery.

64

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Nelzen (2000) Sweden 149 operations on 138 patients

Intermittent To establish morbidity, patient satisfaction and long term results of SEPS

Hospital Vascular surgery Subfascial endoscopic perforator surgery (SEPS)

Noy (2002) Australia 247/277 (89%) Intermittent To evaluate a multi-method approach to PDS SSI

Hospital Obstetrics/Gynaecology Caesarean section

Oliveira (2002) Brazil 398/504 patients (79%)

Intermittent To determine the incidence of surgical wound infections in surgical patients and to compare the frequency of surgical wound infection diagnosed in hospital

Hospital Digestive tract surgery Not further specified.

Paajanen (2003) Finland 317/964 using Lichtenstein procedure

Intermittent To determine the effect of surgeon’s training level on long term results of groin hernia repair

Hospital NS Groin hernia repair using Lichtenstein technique

Perencevich (2003)

USA 89/4571 procedures (1.9% had SSI)

Intermittent To assess clinical outcome and resource utilisation in an 8 week postoperative period associated with PD SSI

Hospital Non-obstretric in or outpatient procedures not further specified.

Various

Pernice (2001) Italy 56 (100% follow up)

Intermittent To assess long term outcome of circular staple haemorrhoidectomy

Medical unit NS Circular staple haemorrhoidectomy

65

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Platt (2002) USA 149/536 (34%) that had sufficient documentation that could be retrieved.

Intermittent To determine if infection indicators were sufficiently consistent across health plans to allow comparison of hospital risks for infection after CABG

Hospitals Cardiac CABG

Porras-Hernandez (2003)

Mexico 428/530 (80.7% follow up)

Intermittent To quantify SSI rates and identify risk factors in a tertiary care paediatric hospital

Hospital General surgery, cardiovascular and neurosurgical

Various

Powell (2002) South Africa 731/912 (80.15%) patients

Intermittent To determine if cost effectiveness is sustained, whether there are hidden complications and whether there is a need for routine follow-up.

Hospital day unit Day case procedures confined to the inguinoscrotal region

Inguinal herniotomy, orchidopexy and circumcision.

Rammelt (2003) Germany 169/314 (54% follow up)

Intermittent Critical results of the medium- term results of open reduction and internal fixation (ORIF) of displaced intra-articular calcaneus fractures with a standardized protocol in a greater patient cohort.

Hospital NS Open reduction and internal fixation (ORIF) of displaced intra-articular calcaneus fractures

66

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Reid (2002) New Zealand 1904/1934 (98.5% follow up)

Intermittent To determine incidence of and risk factors for infections following clean surgery

Hospital General surgery Various

Reilly (2002) Scotland 2202 patients Intermittent To evaluate the effectiveness of a gold standard surveillance methodology for SSI

Hospital General surgery Clean, elective operations

Ridderstolpe (2001)

Sweden 3008/3026 (99%) Continuous from January 1996-September 1999

To detect the incidence of (and risk factors for) sternal wound infections, post discharge

Hospital Cardiac Various, including CABG and valve surgery

Rishi (2001) Libya 1295 Intermittent NS Hospital Various Various Rovera (2003) Italy 346 NS To compare the

incidence and types of infections occurring after 2 different VATS (Video-assisted thorascopic surgery) procedures

Centre for Thoracic Surgery (University of Insurbia, Varese, Italy)

Thoracic surgery Lung wedge resection. Biopsy of pleura/mediastinal mass.

Saeed (2001) UK 127/150 Intermittent To assess patient-based outcomes following radial artery harvesting for CABG

NS Cardiac Radial artery harvesting for CABG

Schneeberger (2002)

The Netherlands 89 operations (% followed up not stated)

Intermittent NS Hospital Orthopaedic NS

67

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical Specialty Procedure

Sewonou (2002)

France 5, 183 patients Intermittent To estimate the incidence of surgical-site infections in ambulatory surgery and to identify risk factors based on the surveillance network INCISO in 1999-2000.

Hospital Various Orthopaedic, gynaecologic/obstetrics, head and neck, skin and soft tissue surgery accounted for 83% of ambulatory procedures.

Sjol (2002) Denmark 855/1034 (82.7%) patients

Intermittent To assess the impact of surgical caps and masks on infection rates during cardiac catheterisation

Hospital Cardiac Cardiac catheterisation

Sorensen (2003)

Denmark 57/60 (95%) patients

Intermittent To test the hypothesis that short-term pre-operative cessation of smoking in colorectal surgery decreases the incidence of postoperative tissue and wound complications

Hospital Colorectal Open colonic or rectal operative procedure with formation of an enteric anatomises.

Stockley (2001) UK 667 patients (92.7% up)

Intermittent NS Hospital Various Total abdominal hysterectomy, inguinal hernia repair, large bowel surgery, vascular surgery.

68

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting

Surgical specialty Procedure

Taylor (2004) Scotland 3046 (93.3%) of eligible population 2665 (87.5%) complete follow-up data available.

Intermittent To determine the incidence of, and risk factors for, SSI after hernia repair

32 Scottish hospitals

NS Hernia repair

Tegnell (2002) Sweden 200 175/200 (87.5%) complete follow-up data available

Intermittent Describing the early development of wound infections after cardiac surgery to identify markers for upcoming infections

University Hospital

Cardiac CABG, valve surgery, or combined procedures.

Thibon (2002) France 3705 Of those 1525 (41.2%) had follow-up for greater than 30 days

Intermittent To take into account the proportion of patients lost to follow-up when calculating surgical-site infection rates.

27 surgical units within 11 institutions

Various Various inpatient procedures (oupatient procedures were not included in this study)

Vilar-Compte (2001)

Mexico 1556 surgeries. 1350 followed up at 30 days (87%)

Intermittent To evaluate the ambulatory surgical site infection rate and risk factors associated with SSI.

Tertiary-care cancer hospital

Ambulatory outpatient surgeries (80% clean surgeries)

Various

69

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Vrijland (2000) Netherlands 136/171 (80%) included in analysis + 6 patients lost to follow up, therefore n=130 (76%)

Intermittent To determine whether enterocutaneous fistulas occurred after repair of incisional hernias with polypropylene mesh

NS NS Incisional hernia repair

Whitney (2001) USA 24 patients (% response unclear)

Intermittent To explore clinical healing outcomes and complications in the two study groups.

NS NS Inguinal hernia repair

Yerdel (2001) Turkey 280 Intermittent (within an RCT)

To assess the value of a single-dose, intravenous, prophylactic ampicillin and sulbactum in the prevention of wound infections during open prosthetic inguinal hernia repair by a double blind, prospective randomised trial.

NS NS Inguinal hernia repair

Yokoe (2001) USA 100 patients (random sample)

Intermittent To identify postpartum infections and describe the epidemiology of these infections

Hospital Obstetrics & Gynaecology Caesarean/vaginal delivery

70

Table B1 Descriptive characteristics of the study (Continued) First author (year)

Location Sample size (% response)

Type of surveillance (continuous or intermittent)

Study aim Practice setting Surgical specialty Procedure

Zhu (2001) Hong Kong 180 patients Intermittent (within an RCT)

To compare the rates of postoperative infectious complications of neurosurgery with prophylactic ceftriaxone and ampicillin/sulbactum, a less expensive antibiotic.

Hospital Neurosurgery Craniotomies, burr holes, cerebral cyst shuntings, transphenodial surgeries, cranioplasties and laminectomies for intradural pathology.

NS-not stated

71

Table B2 Methods and definitions of post discharge surveillance First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Amiel (1999) Outpatient clinic visit and telephone survey

7 days and then at least 6 months postoperatively. Patients or physicians could also request additional follow-up if required but this was not recorded

NS NS NS NS

Andreasen (2002)

A registration chart given to the patient upon discharge to be returned upon contact with primary health care workers. Outpatient/telephone follow up

30 days postoperative Postoperative wound infection was defined as infection in skin and subcutaneous tissue. Postoperative deep wound infection was defined as infection in skin and subcutaneous tissue.

NS Yes. Authors own. Authors own includes clean, potentially contaminated, certainly contaminated and severely contaminated

Arbman (2000) Follow up was done by visit or telephone. Staff involved not reported

3 weeks, 6 weeks and after at least a year.

NS NS NS NS

Astagneau (2001)

Standardised form completed by hospital staff

30 days postoperative CDC (1992) NS NNIS risk index Altemeier classification (Altemeier et al 1992)

Avato (2002) Data collection checklist to record any infection since patient discharge and for use at outpatient visit. Microbiological reports reviewed periodically and at 30 days post operative

Periodically and 30 days post operative over a period of 27 months

CDC NNIS CDC NNIS risk index NS

72

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Beaujean (2002)

NIS study arm did not do PDS. PDS in “CS” Study arm only: Outpatient visit at least 30 days post discharge. Telephone and postal questionnaires also used.

30 days post discharge System developed by Association of Surgeons of the Netherlands that is not specific to surgical wound infection.

NS NS NS

Bhatia (2003) Outpatient visits 2 months post operatively

CDC NNIS (modified) NS NS NS

Bitzer(2000) Postal questionnaire Average of 3 and 14 months post operatively

NS NS NS NS

Cadwallader (2001)

Routine patient data and forms for surgeons to fill in when patients had a SSI that was detected during an outpatient visit

NS CDC (1992) NS NS CDC

Callaghan (1999)

NS NS although patients were followed up for an average of 9.1 years

NS NS NS NS

Canonico (2001)

Direction observation by healthcare worker

7 days postoperatively NS NS NS NS

Dagan (1999) Not specified 2 months post discharge

CDC (1988) NS Authors own based on the therapeutic intervention scoring system

NS

De Boer (2001)

Not specified, although reported to be according to a standardised protocol.

Follow up until 1 year after the operation

CDC Altmeier (1984) NNIS CDC

73

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used

Wound classification system

Delgado-Rodriguez (2001)

Review of emergency department forms and telephone questionnaire

30 days post discharge CDC NS SENIC/NNIS risk index

Clean, clean- contaminated, contaminated, dirty

Derzie (2000) NS 30 days post discharge CDC (1992) NS NS CDC Eriksen (2003)

Review of patient files at 30 days post operation. Outpatient clinic.

30 days post operation CDC (1999) NS NNIS CDC

Ferraz (2001) Outpatient follow-up Time period post operative not stated

NS NS NS NS

Fields (1999) Questionnaire mailed to surgeons 30 days postoperative. Responses returned by mail or communicated by telephone

At least 30 days post operative

NS NS NS Clean, clean-contaminated

Fowler (2003) Blood culture drawn Within 90 days of CABG

Mediastinitis was defined according to CDC definition

NS NS NS

Friedman (2001)

Traditional surveillance methods: microbiology reports & follow up of readmissions. Additionally, electronic chart review from clinic visits.

Electronic chart review at 2 weeks, 6 weeks 6 months, and 1 year

NNIS ASA score NNIS risk index Clean, clean-contaminated, contaminated, dirty

Gerbitz (2000)

Patient questionnaire. 3 months post operative.

NS NS NS NS

74

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Goldsborough (1999)

The nurse case managers were notified by the patient or by other health care providers (home health nurse or primary physician) of any wound complications.

Up to 30 days postoperatively, exact time unclear.

NS NS NS NS

Habib (2002) Inspection of wound at outpatient postnatal clinic

5-10 days post discharge

NS Part ASEPSIS (Unclear) NS NS

Heah (2000) Outpatient appointments or where necessary by telephone review

Unclear NS NS NS NS

Henderson (2001)

Telephone interview 1-2 weeks post discharge

NS NS NS NS

Henrikson (2003)

Standardised follow up sheets completed by surgeons

At each follow up visit, planned consultation or extra visit due to complications (No uniform time period)

NS NS NS NS

Higgins (1999) Examination 7 to 10 days after the procedure and were followed up for 30 days.

At 7,10 and 30 days post operation

Authors own NS ASA and the Culver (1991) wound infection risk classification

Infected wounds were classified as superficial surgical site, deep surgical site and distant.

Hui (1999) Mailed patient questionnaire & telephone interviews

6-86 months post operative. (Average 49 months)

NS NS NS NS

Jamali (2001) Direct observation of wound by health care provider.

48 hours, 4 days and at 4 weeks postoperatively

A wound with a positive culture was defined as infected.

NS NS NS

75

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Jonkers (2003)

Observation of wound by medical attendant and patient questionnaire administered at 2-6 week outpatient visit. At 90 days, fa mily physician asked to fill in questionnaire

2-6 weeks; 90 days CDC (1992) NS NS NS

Kent (2001) Surgeon questionnaire and three follow up telephone calls. Medical records consulted where necessary

Questionnaire 3-4 weeks post discharge. Telephone calls 1 month post survey

CDC (1992) NS NNIS NS

Killian (2001) Hospital readmission data and physician questionnaire, following NNIS system

30 days post operative CDC/NNIS NS NS NS

Krupova (2001)

NS NS NS NS NS NS

Kumar (2001) Physiotherapist home visits and an in clinic visit

Post-operative visits occurred on days 1,3 and 6 post surgery. Clinic visits were conducted at 2 and 6 weeks post operatively.

NS NS NS NS

76

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Letrilliart (2001)

Data collected (via teleinformatics) from GP regarding (amongst other items) patient consultations for post discharge infection

Within 30 days of discharge CDC (1988) NS NS NS

Mahatharadol (2001)

Data collected at outpatient clinic or by telephone contact

30 days post operative Authors own Unclear NS Authors own (Superficial or deep)

Mahomed (2003)

Data collected from Medicare claims submitted by hospitals, physicians and outpatient facilities

90 days postoperative ICD-9-CM/CPT codes

NS NS Only deep wound infections were captured by this study

Marroni (2003)

NS 30 days post operative and 1 year post operative if there was an implant

CDC (1999) NS CDC NNIS CDC

McCowan (1999)

Post-card patient questionnaire, telephone interviews, clinic review

NS NS NS NS NS

77

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Melling (2001) Outpatient clinic visit or visit to patients home and patient diary

At 2 and 6 weeks postoperatively Authors own ASEPSIS NS NS

Nelzen (2000) Outpatient review at 7-10 days post operative/surgeon review at 3-4 months/patient postal questionnaire to assess long-term complications

7-10 days; 3-4 months Long term (exact time period not stated)

NS NS NS NS

Noy (2002) Patient postal questionnaire/telephone follow up/physician consultation/pathological report/chart review

30 days post operative CDC (1999) NS NS NS

Oliviera (2002)

Outpatient clinic visits Within 30 days post operatively. CDC (1999) NS NNIS Unclear- possibly CDC

Paajanen (2003)

Patient postal questionnaire & clinical examination

Mean 3 years (range 1-6) NS NS NS NS

Perencevich (2003)

Automated medical record screening, pharmacy records, hospital readmission/emergency room claims patient postal questionnaire/telephone follow up/administrative databases

30 days or 8 weeks postoperative for questions regarding resource use.

CDC NNIS (1991) NS NS NS

78

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Pernice (2001)

Outpatient visit and telephone follow up

Outpatient visits at 1,2,4 weeks; telephone follow up 5-120 months (mean 33)

NS NS NS NS

Platt (2002) Review of medical records that contained indicator codes suggestive of postoperative SSI

Unclear, although utilised patient records that contained information on outpatient claims up to 30 days post operative

CDC (1992) NS NS NS

Porras-Hernandez (2003)

Outpatient clinic visits and full chart review at 30 days post operative. Also 6 month chart review where implant in place

30 days/6 months post operative CDC (1992) NS NNIS Altmeier et al (1984) and CDC (1992) (Note these two methods were not compared in this study).

Powell (2002) Outpatient clinic visit 2 weeks post surgery. NS Southampton wound assessment scale

ASA Southampton wound assessment scale

Rammelt (2003)

NS NS NS NS NS NS

Reid (2002) Telephone interview with patient. Clarification with GP or surgeon review of wound where required

30 days post discharge ACHC (1995) NS NS NS

Reilly (2002) Patients reviewed prospectively at home or at special wound surveillance clinics

30 days post operative As those used by the Second UK NPS 1992

Southampton wound assessment scale

Risk analysis conducted but no index mentioned

NS

79

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Ridderstolpe (2001)

Review of operating room logs to classify surgical revisions (CDC criteria). Patient telephone follow up, plus questionnaire. Further data collected at surgeon consultation + physician questionnaire

Telephone follow up at 2 weeks postoperative; questionnaire at 6 weeks. Data from surgeon consultation 2/6 months

CDC (1999) NS NS CDC

Rishi (2001) Visit to outpatient clinic within 28 of operation.

Within 28 days of operation, unclear that all patients were observed at the same time period post discharge

NS NS NS National Research Council classification of clean, clean contaminated, contaminated and dirty.

Risnes (2001) Wound inspection. 3 days and 6 weeks postoperatively A score of 4 or more on the Authors wound scoring system

Authors own NS NS

Rovera (2003)

NS 30 days post operatively CDC (1999) NS NS NS

Saeed (2001) Telephone survey of patients

Median time post surgery 8 months Range 0.2-15 months

Defined by author as clinical infection requiring antibiotic treatment.

NS NS NS

Schneeberger (2002)

Outpatient clinic visit. One year period postoperatively although exact dates that patients visited the outpatient clinic were not stated

CDC (1992) NS NNIS CDC

Sewonou (2002)

Outpatient clinic visit and assessment of readmission data from hospital records

30 days post operative CDC NS NNIS CDC

80

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Sjol (2002) Patient questionnaire About two months post operatively NS NS NS NS Sorensen (2003)

Outpatient clinic visit 30 days post operative NS NS NS NS

Stockley (2001)

Telephone contact of patients at home Subgroup nurse visit at home Patients also given a questionnaire to be given to health professional if a visit was made

25-35 days post operative, subgroup at 10 days Time visited at home not stated. Survey tio be returned after patient in contact with health provider.

NINSS NS NS NS

Taylor (2004) Telephone contact 10,20,30 days after surgery CDC (1992) Unclear Unclear NS Tegnell (2002)

Telephone contact or contact via mail and records of patients from healthcare visits 3 months post-operatively

3-4 months after surgery NS ASEPSIS (1986) NS NS

Thibon (2002) Examination of patient records

30 days post operative CDC (1992) NS NS NS

Vilar-Compte (2001)

Full review of microbiology reports and medical records

30 days post operatively CDC (1992) NS NS National Academy of Sciences/ National Research Council wound classification

81

Table B2 Methods and definitions of post discharge surveillance (Continued) First author (year)

Surveillance Method Time period when surveillance carried out

Definition of surgical wound infection given

Wound scoring system been used (reference)

SSI risk index used Wound classification system

Vrijland (2000)

Review of medical records from outpatient visits or questionnaire sent to GP

NS Authors own NS NS A mild wound infection was defined as redness surrounding the laparotomy wound. A moderate wound produced pus and a severe infection was defined as a non-healing defect of the wound extending into the subcutaneous tissue.

Whitney (2001)

Examination of patient records and a follow-up phone call.

30 days post discharge. ASEPSIS Method used by Wilson et al (1986) ASEPSIS

NS NS

Yerdel (2001) Follow up visits to the surgeon

At suture removal 7-9 days post operatively and then 4-6 weeks, 6 months, and then at 1 year

CDC (1992) NS NS CDC (1992)

Yokoe (2001) Automated Health Maintenance Organisation data, pharmacy dispensing data, administrative claims & full text ambulatory medical records. Patient questionnaire

30 days postpartum 6 weeks post discharge

CDC (1992) NS NS NS

Zhu (2001) NS 6 weeks post operatively CDC (1992) NS NS Unclear NS-not stated

82

Table B3 Personnel involved and intrinsic features of post discharge surveillance First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Amiel (1999) Data was collected at outpatient clinic visit and telephone surveys. Staff involved not reported.

NS Yes, but related to preventative steps of wound infection not surveillance.

NS NS NS

Andreasen (2002)

Doctor recorded the occurrence of PWI in medical records. Secretarial staff were also employed to retrieve records outside of county.

NS NS NS NS NS

Arbman (2000) NS NS NS NS NS NS Astagneau (2001)

Nursing staff, anaesthetists and surgeons completed a standardised form for each patient. Data was then entered into an EPI-INFO based programme

NS NS NS Before analysis all data was checked for missing or incorrect values using a logical EPI-INFO software programme

NS

Avato (2002) Operating room nurse/surgeon/nurse practitioner/infection control practitioner

Training on form completion given by infection control practitioners

NS Medical chart, microbiological data, prescription of antibiotics and data from post surgical visit

Medical charts reviewed by ICP

NS Some economic analysis p365

Beaujean (2002)

NS NS NS NS NS NS

Bhatia (2003) Outpatient clinic staff NS NS NS NS NS Bitzer (2000) Patient self report, no staff

required NS NS NS NS NS

83

Table B3 Personnel involved and intrinsic features of post discharge surveillance (Continued) First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Cadawallader (2001)

Surgeons and ward staff Collected data on infections occuring at outpatient clinic visits and if patients were readmitted with suspected SSIs.

NS NS NS NS NS Although a comparative study post-discharge methods not compared

Callaghan (1999)

NS NS NS NS NS NS

Canonico (2001)

NS NS NS NS NS NS

Dagan (1999) Infectious disease nurse NS NS NS NS NS De Boer (2001)

NS NS NS NS Checks of data were performed

NS

Delgado-Rodriguez (2001)

Review of all emergency department forms and telephone interview with patient. Staff not stated

NS NS In-hospital SSI data NS NS

Derzie (2000) NS NS NS NS NS NS Eriksen (2003) Hospital staff (not further

specified) data was collected from patient records

NS NS Patient files were marked so that medical personnel at other wards could notify the investigator if a patient included in the study was admitted into a different section of the hospital

NS NS

Ferraz (2001) Outpatient clinic Staff (exact staff not reported)

NS NS NS NS NS

84

Table B3 Personnel involved and intrinsic features of post discharge surveillance (Continued) First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Fields (1999) Clerical staff, Infection Control staff

Surgeons given CDC guidelines

NS NS NS NS Resource analysis conducted

Fowler (1999) Infection control practioners, clinical staff

NS NS Archived records of blood cultures and the database containing operative data for patients undergoing cardiac surgery

NS NS

Friedman (2001)

Researchers reviewed the electronic chart

NS NS NS NS NS

Goldsborough (1999)

Patients or healthcare providers (home health nurse or primary physician) notified nurse case manager of any wound infections.

NS NS NS NS NS

Habib (2002) Obstetrician NS NS NS NS NS Heah (2000) NS NS NS NS NS NS Henderson (2001)

Clinical nurse consultant/researchers/ Patient self report

NS Yes, this is the intervention

Unclear NS NS

Henrikson (2003)

Surgeon completed standardised forms

NS NS NS NS NS

Higgins (1999) Attending surgeon NS NS NS NS NS Hui (1999) PDS staff not stated. Patient

self report NS NS NS NS NS

Jamali (2001) Data was collected from patients at time of evaluation. Staff involved in this process were not reported

NS NS Cultures taken from wounds

NS NS

Jonkers (2003) Medical attendants, family physician, patient self report

NS NS Outpatient/family physician data

NS NS

85

Table B3 Personnel involved and intrinsic features of post discharge surveillance (Continued) First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Kent (2001) Nurses, anaesthetic, surgical staff, practice secretaries

Yes: definitions, guidelines, training handouts etc

NS Inpatient/outpatient SSI rates

NS Risk adjusted SSIrates were fed back to induvidual surgeons

Krupova (2001)

NS NS NS NS NS NS

Kumar (2001) Physiotherapists and day clinic staff

NS NS NS NS NS, although clinic staff did perform the post-operative assessment.

Cost analysis for day patient vs. inpatient

Killian (2001) Physicians. Patient records were checked.

NS NS NS NS NS

Letrilliart (2001)

GPs. Patient notes were checked.

Standard protocol used – refs 19, 20

NS NS GP/hospital physician telephone interviews

NS

Mahatharadol (2001)

NS NS NS NS NS NS

Mahomed (2003)

NS NS NS NS NS NS

Marroni (2003) NS NS NS NS NS NS McCowan (1999)

NS NS NS Authors refer to “triangulation of methods” p.71 suggests data was cross-validated

NS NS

Melling (2001) Single trained observer and from patient diary.

NS NS NS NS NS

Nelzen (2000) Trained nurses, surgeon NS NS NS NS NS

86

Table B3 Personnel involved and intrinsic features of post discharge surveillance (Continued) First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Noy (2002) NS NS NS Yes, via multi-method approach

NS SSI rates provided on monthly basis to surgeons

Economic analysis carried out. Examples of postal & telephone questionnaire tools given

Oliveira (2002) Data was collected from patient records collected from attendance at outpatient clinics. Staff involved not specified.

NS NS NS Checked regularly for completeness. Not specified by whom

NS

Paajanen (2003)

NS NS NS NS NS NS Questionnaire details given

Perencevich (2003)

Review of records judged to indicate a post discharge SSI by initial screening.

NS NS NNIS criteria with initial screening results for SSI p.196

NS NS Economic analysis of SSI (not PDS system)

Pernice (2001) NS NS NS NS NS NS Platt (2002) Health maintainence

organisations databases were searched for indicator codes. These were then confirmed as infected or not by a trained abstractor using the CDC classification of SSI

NS NS Unclear NS NS

Porra-Hernandez (2003)

Data retrieved from medical records. Staff involved not mentioned.

NS NS NS NS NS

Powell (2002) Surgeon NS Yes NS NS Surgeon that performed the operation performed the post discharge follow-up

87

Table B3 Personnel involved and intrinsic features of post discharge surveillance (Continued) First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Rammelt (2003)

NS NS NS NS NS NS

Reid (2002) Research nurse, medical/surgical staff

NS NS NS NS Monthly feedback to surgeons on number of clean wounds and number of infections

Discussion on wound definition systems p.342

Reilly (2002) Wound surveillance nurse collected data from case notes and daily visits were made to the patients.

Intervention part of a wider programme of change for evidence-based surgical practice

NS NS NS Feedback given to surgeons and nurses participating in the study

Ridderstolpe (2001)

Surgeons notes were reviewed. Both a physician and patient questionnaire were undertaken.

NS NS NS NS NS

Rishi (2001) Staff member not stated although data collected at the time of patient attendance at an outpatient clinic

NS NS NS NS NS

Risnes (2001) NS NS NS NS NS NS Rovera (2003) Unclear NS NS NS NS NS Saeed (2001) Patients were interviewed by

telephone. Staff undertaking the survey not reported.

NS NS NS NS NS

Schneeberger (2002)

Data was collected from patient record data collected at outpatient clinic visit. Exact staff involved not stated.

Yes. Training on ASEPSIS principles given twice yearly.

NS NS Performed by another national body.

Surgeons and assisting personnel were informed of the results of SSI’s

88

Table B3 Personnel involved and intrinsic features of post discharge surveillance (Continued) First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Sewonou (2002)

Surgical staff completed standardised forms for each patient which described SSI occurrence

* * * * *

Sjol (2002) Patient reported symptoms of infection as reported in patient questionnaire

* * * * *

Sorensen (2003)

Staff member not stated although data collected at the time of patient attendance at an outpatient clinic

NS NS NS NS NS

Stockley (2001)

Infection control audit nurse gave patient information pack whilst in hospital and gave form to be completed if treatment was required by health professional

NS Yes NS NS NS Methods changed over time in this study as such rates observed may have been affected by changes in methods.

Taylor (2004) G.P.s and other healthcare workers not defined

NS NS NS NS NS

Tegnell (2002) G.P.s and nursing staff NS - but yes on scoring system

NS NS NS NS

Thibon (2002) NS NS NS NS 5% of randomly selected files were verified compared to original patient files

NS

Whitney (2001)

Patient records from visits to hospital or community clinics made within 30 days of surgery

NS NS NS Interater reliability of ASEPSIS scoring between the three members of the research team.

NS

89

Table B3 Personnel involved and intrinsic features of post discharge surveillance (Continued) First author (year)

Staff and method of data collection

Training given to staff

Patient education given

Linkage of data Validation of data Feedback to users Others

Villar-Compte (2001)

Data was obtained from microbiology reports and medical records

NS NS NS NS NS

Vrijland (2000) NS NS NS NS NS NS Yerdel (2001) Patient records from follow

up visits NS NS NS NS Yes

Yokoe (2001) NS NS NS NS NS NS Zhu (2001) Unclear NS NS NS NS NS NS-not stated

90

Appendix C Example of audit sent to Infection Control lead within Trusts

Department of Health Sciences

Research Section, First Floor, Area 2 Seebohm Rowntree Building Heslington York YO10 5DD

Telephone (01904) 321362 Fax (01904) 321382

www.york.ac.uk/healthsciences

MONITORING OF SURGICAL WOUNDS FOR INFECTION, POST HOSPITAL DISCHARGE. (Please complete and return this questionnaire even if you are not monitoring surgical wounds for infection, post hospital discharge) SECTION 1 – Contact Details To be completed by the person responsible for infection control (or, where appropriate, the person responsible for monitoring of surgical wounds, post hospital discharge). If this has been sent to you incorrectly, please pass this form to the appropriate person within your Trust, or return it to Emily Petherick by fax: 01904 321382 with new contact details. Name: Position: Tel: Fax: Email: Name of Hospital Trust: NHS Region: SECTION 2 – Monitoring of surgical wound infection, post hospital discharge 1. Does your Trust currently monitor surgical patients for wound infection after discharge from hospital?

91

Yes � No � If you have answered no to question 1, please return this questionnaire by fax to Emily Petherick: 01904 321382 by Friday 14 th May 2004. We thank you for your co-operation. If you have answered yes, please complete questions 2 and 3. 2. Which patients (or types of surgery) are currently included in the monitoring system? ………………………………………………………………………………… ………………………………………………………………………………… ………………………………………………………………………………… ………………………………………………………………………………… ………………………………………………………………………………… 3. What methods are currently used in the monitoring system? (please tick all that apply): Routine clinic follow up � Patient telephone survey � Patient postal survey � Surgeon survey � Direct observation by health practitioner � Other (Please specify) � .. ………………………………………………………………………………. ………………………………………………………………………………… ………………………………………………………………………………….. …………………………………………………………………………………. Thank you for your co-operation with this survey. Please indicate whether you would be willing to be contacted again in the future. Yes � No �

92

Appendix D Telephone interview 1st contact Telephone interview schedule to survey responders carrying out PDS: 1st contact Date: Contact name: Position: Name of Trust: Introduction & confirm their details correct/that they are responsible for monitoring PDS etc. …Thank you for responding so promptly to our recent survey regarding PD monitoring of SSI. Very interested to learn that you are currently carrying out a system of surveillance at your Trust. Indeed, yours was selected as one of the more interesting ones and we’d like to talk to you about it in more detail. You’ve already indicated that you might be willing to talk to us further. Are you still happy to do this?

Y / N (Recap: purpose of this is audit is to understand more clearly what systems are already in place and whether these might be evaluated/replicated elsewhere - NOT to review performance of your Trust in any way). If yes, would like to arrange a convenient time for a telephone interview (5-30 minutes of your time, depending on how much to talk about). Briefly go through proposed topics & send copy of schedule in advance. email/fax details: For ease of data processing, we’d like to be able to record the interview – would you be happy for us to do this? (for use only as transcription aid, but see also below on dissemination plan) Y / N ….Because our aim is to identify what is happening in different parts of the country, it would be useful to us to have your agreement to the identification of your Trust when writing our results (It is not an anonymous report - Trusts (not individuals) will be identified in a report to the National Co-ordinating Centre for Research Methodology at University of Birmingham – no paper publications, but the NCCRM may put the report on the web and will certainly be sent out for peer review. (NB However, in certain circumstances eg, failure of PDS, might be able to accommodate some anonymity). Are you willing to agree to this? Y / N

93

Appendix E Telephone interview 2nd contact Telephone interview schedule to survey responders carrying out PDS SSI (Post Discharge Surveillance of Surgical Site Infection) 2nd contact – interview Date: Start time: Finish time: Contact name: Position: Name of Trust: From survey response, confirm/ascertain the following, as appropriate: What is the practice setting? (distinguish between setting where surveillance initiated and setting where patients observed) What population is followed up? All or sample? Ascertain details of: On what basis does the Trust select the sample (where applicable): What surgical speciality is included in the PDS system? What surgical procedure is included in the PDS system? What is the risk status of patients followed up? Confirm what method of PDS & expand:

For how long are patients followed up?

How frequently are they seen? Does this vary?

Where are they seen? Does this vary?

94

Who sees the patients to assess their wound? (NB. As part of the surveillance

programme - not relevant if not seen as part of surveillance)

What training, if any, is given to staff on PDS?

Who has overall responsibility for PDS?

IF patients are self assessing – has the Trust compared patient assessment with

professional assessment. IF NOT, does the Trust collect data on false positives, ie, those

patients who think they have a wound infection, but professionals think have not.

Is patient information given as part of the surveillance programme (in relation to

reporting wound infection or the whole programme)?

How is surgical wound infection defined? (eg, CDC guidelines – what aspect is used: classification/diagnostic/risk index etc) (What scoring system, if any, is used? (eg, ASA, ASEPSIS..) What classification system, if any, is used? (eg, clean, contaminated, dirty…does this refer to the wound or the operation?) What is the Trust’s view on the practicality of method(s) currently used? Has the Trust tried any other methods/compared/discarded? Expand. Has there been any evaluation of the PDS system? (what measures eg, increased identification of SSI)

Have you ever calculated the cost of this surveillance system to your Trust? What resource use, eg, staff time/grade etc Is the surveillance system linked with any wider local or national audit? Do you give the data to anyone? (eg, feedback to surgeons) + any other specific issues about survey response….. Would it be OK for me to call you again for clarification on any of these points?

95

APPENDIX F Individual components of wound definition systems not previously described in Bruce et al (2001). First author (year) Purulent discharge without

culture Culture mandatory Other mandatory criteria, used alone or in combination

Andreasen (2002) Yes - Superficial: Purulent discharge without culture OR surgical wound revision with positive culture from evacuated material or bacteria isolated from subcutaneous collection in primarily closed wounds. Deep: Purulent discharge without culture OR wound rupture or wound revision with findings of infection involving fascia and muscle (culture must be positive) and/or finding of subfascial abscess on puncture, surgical revision or reoperation.

Higgins (1999) Superficial SSI:Yes above the fascia Deep SSI:Yes deep to the fascia or near the gallbladder fascia.

- Superficial: Purulent discharge without culture OR erythema.

Jamali (2001) - Yes NS Mahathardol (2001) Superficial SSI: Yes above the

fascial layer Deep SSI: Yes at or beneath the fascial layer

_ Superficial: Purulent discharge without culture OR erythema

Melling (2001) Yes ? Purulent discharge without culture OR a painful erythema that lasted for 5 days and was treated with antibiotics within 6 weeks of surgery.

The Australian Council on Healthcare Standards (1995), published in Reid (2002)

Yes - Purulent discharge without culture OR evidence of inflammation around the wound, OR if they were deemed to be infected by surgical staff.

Rishi (2001) ? ? Each wound was given a score from 0 to 7 with zero representing the optimal, physiologic postoperative wound appearance. Wound infection was defined as a score of 4 or more that meant postoperative wound conditions with pus combined with other signs and symptoms such as erythema, oedema, or increased pain.

Vrijland (2000) ? ? A mild wound infection was defined as redness surrounding the laparotomy wound. A moderate wound produced pus and a severe infection was defined as a non-healing defect of the wound extending into the subcutaneous tissue.

?-Unclear SSI-Surgical site infection

96

Appendix G Excluded Studies

Duplicated reporting of post-discharge surveillance (In some of these studies

further follow-up had taken place in more recent studies)

Creedy, D. K. and D. L. Noy (2001). Postdischarge surveillance after cesarean section.

Birth 28(4): 264-9.

Golliot, F., P. Astagneau, et al. (1999). [Surveillance of surgical-site infections: results of

the INCISO 1998 Network]. Annales de Chirurgie 53(9): 890-7.

Leaper, D. J. and A. G. Melling (2001). Antibiotic prophylaxis in clean surgery: clean

non-implant wounds. Journal of Chemotherapy(1): 96-101.

Reilly, J. S. (1999). The effect of surveillance on surgical wound infection rates. Journal of

Tissue Viability 9(2): 57-60.

Reilly, J. S., D. Baird, et al. (2001). The importance of definitions and methods in surgical

wound infection audit. Journal of Hospital Infection 47(1): 64-66.

Reilly, J., S. Twaddle, et al. (2001). An economic analysis of surgical wound infection.

Journal of Hospital Infection 49(4): 245-9.

Reilly, J., J. McIntosh, et al. (2002). Changing surgical practice through feedback of

performance data. Journal of Advanced Nursing 38(6): 607-14.

Taylor, E. W., K. Duffy, et al. (2003). Telephone call contact for post-discharge

surveillance of surgical site infections. A pilot, methodological study. Journal of

Hospital Infection 55(1): 8-13.

Vilar-Compte, D., et al., [Surveillance of surgical wound infections. 18-month experience in the

Instituto Nacional de Cancerologia]. Salud Publica de Mexico, 1999. 41(1): p. S44-50.

Editorial

Hall, J. C. (1999). Monitoring wound infection after surgery: the quest for useful

information at a reasonable cost.[comment]. Australian & New Zealand Journal

97

of Surgery 69(2): 84.

Mitchell, D. H. (2001). Post-discharge surgical wound surveillance. ANZ Journal of

Surgery 71(10).

Spearing, N. and D. Olesen (2001). Editorial comment. Surveillance: the next step in the

process. Australian Infection Control 6(2): 40-1.

Study already included in Bruce et al (2001)

Sands, K., G. Vineyard, et al. (1999). Efficient identification of postdischarge surgical site

infections: use of automated pharmacy dispensing information, administrative

data, and medical record information. Journal of Infectious Diseases 179(2): 434-

41.

Weiss, I. C., C. L. Statz, et al. (1999). Six years of surgical wound infection surveillance at

a tertiary care center: Review of the microbiologic and epidemiological aspects of

20 007 wounds. Archives of Surgery 134(10): 1041-1048.

Letter

Kennon, J., P. Russo, et al. (2001). A comparison of two methods for identifying surgical

site infections following orthopaedic surgery (multiple letters) [6]. Journal of

Hospital Infection 49(4): 302-304.

Narrative review

Anon (2000). SSI core measure: look harder and look worse? JCAHO may add post-

discharge tracking. Hospital Infection Control 27(4): 51-3.

Anon (2000). Tracking infections after outpatient surgery: form streamlines process,

improves communication. Hospital Infection Control 27(8): HEALTHCARE

INFECTION PREVENTION: insert 3-4.

Anon (2001). Survey program monitors surgical site infections. Same-Day Surgery

98

25(11): 131-2.

Anon (2001). The top five. Capitalize on the advantages of semiautomated data

collection methods to keep on top of surgical site infections. Joint Commission

Benchmark 3(9): 11.

Embry, F. C. and L. F. Chinnes (2001). Draft definitions of surveillance of infection in

home healthcare. Home Healthcare Nurse 19(7): 439-44.

Gaynes, R. P. (2000). Surveillance of surgical-site infections: the world coming together?

Infection Control & Hospital Epidemiology 21(5): 309-10.

Gaynes, R. P. (2000). Surgical-site infections and the NNIS SSI Risk Index: room for

improvement. Infection Control & Hospital Epidemiology 21(3): 184-5.

Heeg, P. (2003). Surveillance of Postoperative Wound Infections. [German]. Aktuelle

Traumatologie 33(6): 292-294.

Howard, J. (2000). Communicable disease and infection control: the surveillance

contribution. British Journal of Community Nursing 5(12): 619-20.

Krajacic, A. (1999). Health care surveillance: what should be included? Today's Surgical

Nurse 21(1): 29-32.

Martone, W. J. and R. L. Nichols (2001). Recognition, prevention, surveillance, and

management of surgical site infections: introduction to the problem and

symposium overview. Clinical Infectious Diseases 1(33): S67-8.

Noy, D., D. Creedy, et al. (2001). Post discharge surveillance methods: a critique.

Australian Infection Control 6(3): 81-8.

Platt, R., D. S. Yokoe, et al. (2001). Automated methods for surveillance of surgical site

infections. Emerging Infectious Diseases 7(2): 212-6.

99

Post discharge surveillance not undertaken as part of study or unclear that

undertaken.

Anon(2001). Surveillance of nosocomial infections and registration of bacterial isolates

with important antimicrobial resistancies. [German]. Hygiene + Medizin 26(3):

89-92.

Aschl, G., A. Kirchgatterer, et al. (2003). [Indications and complications of percutaneous

endoscopic gastrostomy]. Wiener Klinische Wochenschrift 115(3-4): 115-20.

Borer, A., J. Gilad, et al. (2001). Impact of active monitoring of infection control

practices on deep sternal infection after open-heart surgery. Annals of Thoracic

Surgery 72(2): 515-20.

Bradshaw, C., C. Pritchett, et al. (1999). Information needs of general day surgery

patients. Ambulatory Surgery 7(1): 39-44.

Codina, C., A. Trilla, et al. (1999). Perioperative antibiotic prophylaxis in Spanish

hospitals: results of a questionnaire survey. Hospital Pharmacy Antimicrobial

Prophylaxis Study Group. Infection Control & Hospital Epidemiology 20(6):

436-9.

Delgado-Rodriguez, M., A. Gomez-Ortega, et al. (1999). The effect of frequency of chart

review on the sensitivity of nosocomial infection surveillance in general surgery.

Infection Control & Hospital Epidemiology 20(3): 208-12.

Delgado-Rodriguez, M., A. Gomez-Ortega, et al. (2001). Efficacy of surveillance in

nosocomial infection control in a surgical service. American Journal of Infection

Control 29(5): 289-94.

Dhaliwal, J. K., A. M. El-Shafei, et al. (2000). Hospital morbidity due to post-operative

infections in obstetrics and gynecology. Saudi Medical Journal 21(3): 270-273.

Douglas, P., M. Asimus, et al. (2001). Prevention of orthopaedic wound infections: a

quality improvement project. Journal of Quality in Clinical Practice 21(4): 149-53.

100

Farinas-Alvarez, C., M. C. Farinas, et al. (2000). Analysis of risk factors for nosocomial

sepsis n surgical patients. British Journal of Surgery 87(8): 1076-81.

Geel, C. W. and A. S. Flemister, Jr. (2001). Standardized treatment of intra-articular

calcaneal fractures using an oblique lateral incision and no bone graft. Journal of

Trauma-Injury Infection & Critical Care 50(6): 1083-9.

Gerberding, J., R. Gaynes, et al. (1999). National Nosocomial Infections Surveillance

(NNIS) system report, data summary from January 1990-May 1999, issued June

1999. AJIC: American Journal of Infection Control 27(6): 520-532.

Hebden, J. (2000). Use of ICD-9-CM coding as a case-finding method for sternal wound

infections after CABG procedures. American Journal of Infection Control 28(2):

202-3.

Maddalozzo, J., T. K. Venkatesan, et al. (2001). Complications associated with the

Sistrunk procedure. Laryngoscope 111(1): 119-23.

Murphy, C. L. and M. L. McLaws (1999). Methodologies used in surveillance of surgical

wound infections and bacteremia in Australian hospitals.[erratum appears in Am

J Infect Control 2000 Apr;28(2):203]. American Journal of Infection Control

27(6): 474-81.

Murphy, C. L. and M. L. McLaws (2000). Erratum: Methodologies used in surveillance of

surgical wound infections and bacteremia in Australian hospitals (American

Journal of Infection Control (1999) 27 (474-481)). AJIC: American Journal of

Infection Control 28(2).

Murphy, G. J., R. Pararajasingam, et al. (2001). Methicillin-resistant Staphylococcus

aureus infection in vascular surgical patients. Annals of the Royal College of

Surgeons of England 83(3): 158-63.

Przybylski, G. J. and A. D. Sharan (2001). Single-stage autogenous bone grafting and

internal fixation in the surgical management of pyogenic discitis and vertebral

101

osteomyelitis. Journal of Neurosurgery 94(1 Suppl): 1-7.

Ratliff, C. R. and A. M. Donovan (2001). Frequency of peristomal complications.

Ostomy Wound Management 47(8): 26-9.

Rioux, C., T. Blanchon, et al. (2002). Audit of preoperative antibiotic prophylaxis in a

surgical site infections surveillance network. [French]. Annales Francaises d

Anesthesie et de Reanimation 21(8): 627-633.

Saleh, K. J., D. C. Dykes, et al. (2002). Functional outcome after total knee arthroplasty

revision: A meta-analysis. Journal of Arthroplasty 17(8): 967-977.

Sawyer, R. G., D. P. Raymond, et al. (2001). Implications of 2,457 consecutive surgical

infections entering year 2000. Annals of Surgery 233(6): 867-74.

Tegnell, A., B. Isaksson, et al. (2002). Changes in the appearance and treatment of deep

sternal infections. Journal of Hospital Infection 50(4): 298-303.

Wasowicz, D. K., R. F. Schmitz, et al. (2000). [Assessment of day surgery in a district

training hospital: safety, efficacy and patient's satisfaction]. Nederlands Tijdschrift

voor Geneeskunde 144(40): 1919-23.

Study did not separate results for infections that occurred and were detected in the post

discharge period

McGreevy, J. M., P. P. Goodney, et al. (2003). A prospective study comparing the

complication rates between laparoscopic and open ventral hernia repairs. Surgical

Endoscopy 17(11): 1778-1780.

Systematic review not limited to post discharge surgical site infection surveillance

(checked for references no new studies located)

Wallace, W. C., M. E. Cinat, et al. (2000). New epidemiology for postoperative

nosocomial infections. American Surgeon 66(9): 874-8.

102

Study unable to be translated (Sufficient information was available to deduce that

this was not a study comparing methods of post-discharge surveillance

van Dalen, T., et al., [Frequency of postoperative wound infections: an unsuita ble parameter for

comparison of hospitals]. Nederlands Tijdschrift voor Geneeskunde, 2000. 144(10): p.

476-9.

Studies unable to be located within time frame of review

Esposito, S., F. Iannielio, et al. (2003). Multicentre survey of post-surgical infections in

Campania (Italy). [Italian]. Infezioni in Medicina 11(3): 146-152.

Cortes-Ramas, A. M., J. M. Abad-Diez, et al. (2001). Underresgistration of postoperative

infections in the Minimum Basic Data Set (MBDS). [Spanish]. Revista de Calidad

Asistencial 16(6): 378-383.

103

References Amiel, G. E., I. Sukhotnik, et al. (1999). Use of N-butyl-2-cyanoacrylate in elective

surgical incisions--longterm outcomes. Journal of the American College of

Surgeons 189(1): 21-5.

Andreasen, J. J., B. Korsager, et al. (2002). Postoperative wound infection: indicator of

clinical quality? Danish Medical Bulletin 49(3): 242-4.

Arbman, G., H. Krook, et al. (2000). Closed vs. open hemorrhoidectomy--is there any

difference?[see comment]. Diseases of the Colon & Rectum 43(1): 31-4.

Astagneau, P., C. Rioux, et al. (2001). Morbidity and mortality associated with surgical

site infections: Results from the 1997-1999 INCISO surveillance. Journal of

Hospital Infection 48(4): 267-274.

Avato, J. L. and K. K. Lai (2002). Impact of postdischarge surveillance on surgical-site

infection rates for coronary artery bypass procedures. Infection Control &

Hospital Epidemiology 23(7): 364-367.

Beaujean, D., S. Veltkamp, et al. (2002). Comparison of two surveillance methods for

detecting nosocomial infections in surgical patients. European Journal of Clinical

Microbiology & Infectious Diseases 21(6): 444-448.

Bhatia, J. Y., K. Pandey, et al. (2003). Postoperative wound infection in patients

undergoing coronary artery bypass graft surgery: A prospective study with

evaluation of risk factors. Indian Journal of Medical Microbiology 21(4): 246-251.

Bitzer, E. M., H. Dorning, et al. (2000). [The success of hernia surgery in routine care

from the patient's perspective]. Chirurg 71(7): 829-34.

Bland, M. and D. Altman (1986). Statistical methods for assessing agreement between

two methods of clinical measurement. Lancet: 307-310.

Bruce, J., E. M. Russell, et al. (2001). The measurement and monitoring of surgical

adverse events. Health Technology Assessment 5(22).

Cadwallader, H. L., M. Toohey, et al. (2001). A comparison of two methods for

identifying surgical site infections following orthopaedic surgery. Journal of

Hospital Infection 48(4): 261-6.

Callaghan, J. J., R. P. Katz, et al. (1999). One-stage revision surgery of the infected hip. A

minimum 10-year followup study. Clinical Orthopaedics & Related Research 369:

139-43.

Canonico, S., F. Campitiello, et al. (2001). Sutureless skin closure in varicose vein surgery:

Preliminary results. Dermatologic Surgery 27(3): 306-308.

104

Cruse, P. J. and R. Foord (1973). A five year prospective study of 23, 649 wounds.

Archives of Surgery 107: 206-10.

Dagan, O., P. N. Cox, et al. (1999). Nosocomial infection following cardiovascular

surgery: comparison of two periods, 1987 vs. 1992. Critical Care Medicine 27(1):

104-8.

de Boer, A. S., E. L. Geubbels, et al. (2001). Risk assessment for surgical site infections

following total hip and total knee prostheses. Journal of Chemotherapy(1): 42-7.

Delgado-Rodriguez, M., A. Gomez-Ortega, et al. (2001). Epidemiology of surgical-site

infections diagnosed after hospital discharge: a prospective cohort study.

Infection Control & Hospital Epidemiology 22(1): 24-30.

Derzie, A. J., F. Silvestri, et al. (2000). Wound closure technique and acute wound

complications in gastric surgery for morbid obesity: a prospective randomized

trial. Journal of the American College of Surgeons 191(3): 238-43.

Eriksen, H. M., S. Chugulu, et al. (2003). Surgical-site infections at Kilimanjaro Christian

Medical Center. Journal of Hospital Infection 55(1): 14-20.

Ferraz, A. A., B. G. da Nobrega Junior, et al. (2001). Late results on the surgical

treatment of Chagasic megaesophagus with the Thal-Hatafuku procedure.

Journal of the American College of Surgeons 193(5): 493-8.

Fields, C. L. (1999). Outcomes of a postdischarge surveillance system for surgical site

infections at a Midwestern regional referral center hospital. American Journal of

Infection Control 27(2): 158-64.

Fowler Jr, V. G., K. S. Kaye, et al. (2003). Staphylococcus aureus bacteremia after median

sternotomy: Clinical utility of blood culture results in the identification of

postoperative mediastinitis. Circulation 108(1): 73-78.

Friedman, C., L. K. Sturm, et al. (2001). Electronic chart review as an aid to

postdischarge surgical site surveillance: increased case finding. American Journal

of Infection Control 29(5): 329-32.

Gerbitz, J., N. Rose, et al. (2001). [Quality control after hernia operation]. Swiss Surgery

7(3): 105-9.

Goldsborough, M. A., M. H. Miller, et al. (1999). Prevalence of leg wound complications

after coronary artery bypass grafting: determination of risk factors. American

Journal of Critical Care 8(3): 149-53.

Grimshaw, J. M., R. E. Thomas, et al. (2004). Effectiveness and efficiency of guideline

dissemination and implementation strategied. Health Technology Assessment

105

8(6).

Habib, F. A. (2002). Incidence of post cesarean section wound infection in a tertiary

hospital, Riyadh, Saudi Arabia. Saudi Medical Journal 23(9): 1059-1063.

Heah, S. M., J. E. Hartley, et al. (2000). Laparoscopic suture rectopexy without resection

is effective treatment for full-thickness rectal prolapse. Diseases of the Colon &

Rectum 43(5): 638-43.

Henderson, A. and W. Zernike (2001). A study of the impact of discharge information

for surgical patients. Journal of Advanced Nursing 35(3): 435-41.

Henriksen, T. F., L. R. Holmich, et al. (2003). Incidence and Severity of Short-Term

Complications after Breast Augmentation: Results from a Nationwide Breast

Implant Registry. Annals of Plastic Surgery 51(6): 531-539.

Higgins, A., J. London, et al. (1999). Prophylactic antibiotics for elective laparoscopic

cholecystectomy: are they necessary? Archives of Surgery 134(6): 611-3.

Holtz, T. H. and R. P. Wenzel (1992). Post discharge surveillance from nosocomial

wound infection: a brief review. American Journal of Infection Control 20: 206-

213.

Horan, T. C., R. P. Gaynes, et al. (1992). CDC definitions of nosocomial surgical site

infections, 1992: a modification of CDC definitions of surgical wound infections.

American Journal of Infection Control 20(5): 271-4.

Hui, T. T., D. I. Giurgiu, et al. (1999). Iatrogenic gallbladder perforation during

laparoscopic cholecystectomy: etiology and sequelae. American Surgeon 65(10):

944-8.

Jamali, A. R., G. Mehboob, et al. (2001). Postoperative wound infections in orthopaedic

surgery. Journal of the College of Physicians & Surgeons Pakistan 11(12): 746-

749.

Jonkers, D., T. Elenbaas, et al. (2003). Prevalence of 90-days postoperative wound

infections after cardiac surgery. European Journal of Cardio Thoracic Surgery

23(1): 97-102.

Kent, P., M. McDonald, et al. (2001). Post-discharge surgical wound infection

surveillance in a provincial hospital: Follow-up rates, validity of data and review

of the literature. ANZ Journal of Surgery 71(10): 583-589.

Killian, C. A., E. M. Graffunder, et al. (2001). Risk factors for surgical-site infections

following cesarean section. Infection Control & Hospital Epidemiology 22(10):

613-7.

106

Krupova, Y., J. C. Gould, et al. (2001). Postoperative neurosurgical site infection

surveillance. Journal of Chemotherapy 13(2): 210-2.

Kumar, A., D. R. Bickerstaff, et al. (2001). Day surgery anterior cruciate ligament

reconstruction: Sheffield experiences. Knee 8(1): 25-7.

Letrilliart, L., M. Guiguet, et al. (2001). Postdischarge nosocomial infections in primary

care. Infection Control & Hospital Epidemiology 22(8): 493-8.

Mahatharadol, V. (2001). A reevaluation of antibiotic prophylaxis in laparoscopic

cholecystectomy: a randomized controlled trial. Journal of the Medical

Association of Thailand 84(1): 105-8.

Mahomed, N. N., J. A. Barrett, et al. (2003). Rates and outcomes of primary and revision

total hip replacement in the United States Medicare population. Journal of Bone

& Joint Surgery - American Volume 85A(1): 27-32.

Mangram, A. J., T. C. Horan, et al. (1999). Guideline for prevention of surgical site

infection. Emerging Infectious Diseases 27(2): 97-132.

Marroni, M., M. Fiorio, et al. (2003). [Nosocomial infections in vascular surgery: 1-year

surveillance]. Recenti Progressi in Medicina 94(10): 430-3.

Martini, F., C. Tieben, et al. (2000). [Inpatient and follow-up nosocomial wound

infection in orthopedics]. Zeitschrift fur Orthopadie und Ihre Grenzgebiete

138(1): 74-8.

McCowan, M., A. Balfour, et al. (1999). Wound care. Discharged but not forgotten.

Nursing Times 95(34): 71.

Melling, A. G., B. Ali, et al. (2001). Effects of preoperative warming on the incidence of

wound infection after clean surgery. The Lancet 358: 876-880.

Mitchell, D. H., G. Swift, et al. (1999). Surgical wound infection surveillance: the

importance of wounds that develop after hospital discharge. ANZ Journal of

Surgery 69: 117-120.

Nelzen, O. (2000). Prospective study of safety, patient satisfaction and leg ulcer healing

following saphenous and subfascial endoscopic perforator surgery. British

Journal of Surgery 87(1): 86-91.

Noy, D. and D. Creedy (2002). Postdischarge surveillance of surgical site infections: a

multi-method approach to data collection. American Journal of Infection Control

30(7): 417-24.

Oliveira, A. C., M. A. Martins, et al. (2002). [Comparative study of surgical wound

infection diagnosed in-hospital and post discharge]. Revista de Saude Publica

107

36(6): 717-22.

Paajanen, H. (2003). Groin hernia repair under local anaesthesia: Effect of surgeon's

training level on long-term results. Ambulatory Surgery 10(3): 143-146.

Perencevich, E. N., K. E. Sands, et al. (2003). Health and economic impact of surgical

site infections diagnosed after hospital discharge. Emerging Infectious Diseases

9(2): 196-203.

Pernice, L. M., B. Bartalucci, et al. (2001). Early and late (ten years) experience with

circular stapler hemorrhoidectomy. Diseases of the Colon & Rectum 44(6): 836-

41.

Platt, R., K. Kleinman, et al. (2002). Using automated health plan data to assess infection

risk from coronary artery bypass surgery. Emerging Infectious Diseases 8(12):

1433-41.

Porras-Hernandez, J. D., D. Vilar-Compte, et al. (2003). A prospective study of surgical

site infections in a pediatric hospital in Mexico City. American Journal of

Infection Control 31(5): 302-8.

Powell, S. F., R. A. Brown, et al. (2002). Is there a need for postoperative surveillance

after day case groin surgery in children? South African Journal of Surgery 40(3):

91-4.

Rammelt, S., S. Barthel, et al. (2003). Calcaneus fractures. Open reduction and internal

fixation. [German]. Zentralblatt fur Chirurgie 128(6): 517-528.

Reid, R., J. W. Simcock, et al. (2002). Postdischarge clean wound infections: incidence

underestimated and risk factors overemphasized. ANZ Journal of Surgery 72(5):

339-43.

Reilly, J., S. Twaddle, et al. (2001). An economic analysis of surgical wound infection.

Journal of Hospital Infection 49(4): 245-9.

Reilly, J. (2002). Evidence-based surgical wound care on surgical wound infection. British

Journal of Nursing 11(16 Suppl).

Ridderstolpe, L., H. Gill, et al. (2001). Superficial and deep sternal wound complications:

Incidence, risk factors and mortality. European Journal of Cardio Thoracic

Surgery 20(6): 1168-1175.

Rishi, M. B., M. A. Aboshkiwa, et al. (2001). Post-operative wound infections. Etiology

and follow-up problems [1]. Saudi Medical Journal 22(10): 928-929.

Risnes, I., M. Abdelnoor, et al. (2001). Sternal wound infections in patients undergoing

open heart surgery: randomized study comparing intracutaneous and

108

transcutaneous suture techniques. Annals of Thoracic Surgery 72(5): 1587-91.

Rovera, F., A. Imperatori, et al. (2003). Infections in 346 consecutive video-assisted

thoracoscopic procedures. Surgical Infections 4(1): 45-51.

Saeed, I., A. C. Anyanwu, et al. (2001). Subjective patient outcomes following coronary

artery bypass using the radial artery: results of a cross-sectional survey of harvest

site complications and quality of life. European Journal of Cardio Thoracic

Surgery 20(6): 1142-6.

Sands, K. E., D. S. Yokoe, et al. (2003). Detection of postoperative surgical-site

infections: comparison of health plan-based surveillance with hospital-based

programs. Infection Control & Hospital Epidemiology 24(10): 741-3.

Schneeberger, P. M., M. H. Smits, et al. (2002). Surveillance as a starting point to reduce

surgical-site infection rates in elective orthopaedic surgery. Journal of Hospital

Infection 51(3): 179-84.

Scottish Cente for Infection and Environmental Health (2003). Surveillance of surgical

site infection, For procedures carried out from: 4/04/02-30/06/03.

Seaman, M. and R. Lammers (1991). Inability of patients to self-diagnose wound

infection. The journal of emergency medicine 9: 215-219.

Sewonou, A., C. Rioux, et al. (2002). Incidence of surgical-site infections in ambulatory

surgery: Results of the INCISO surveillance network in 1999-2000. [French].

Annales de Chirurgie 127(4): 262-267.

Sorensen, L. T. and T. Jorgensen (2003). Short-term pre-operative smoking cessation

intervention does not affect postoperative complications in colorectal surgery: A

randomized clinical trial. Colorectal Disease 5(4): 347-352.

Taylor, E. W., K. Duffy, et al. (2004). Surgical site infection after groin hernia repair.

British Journal of Surgery 91(1): 105-111.

Tegnell, A., C. Aren, et al. (2002). Wound infections after cardiac surgery--a wound

scoring system may improve early detection. Scandinavian Cardiovascular Journal

36(1): 60-4.

Thibon, P., J. J. Parienti, et al. (2002). Use of censored data to monitor surgical-site

infections. Infection Control & Hospital Epidemiology 23(7): 368-71.

Vilar-Compte, D., R. Roldan, et al. (2001). Surgical site infections in ambulatory surgery:

a 5-year experience. American Journal of Infection Control 29(2): 99-103.

Vrijland, W. W., J. Jeekel, et al. (2000). Intraperitoneal polypropylene mesh repair of

incisional hernia is not associated with enterocutaneous fistula. British Journal of

109

Surgery 87(3): 348-352.

Whitby, M., M. L. McLaws, et al. (2002). Post-discharge surveillance: can patients reliably

diagnose surgical wound infections? Journal of Hospital Infection 52(3): 155-60.

Whiting, P., A. W. S. Rutjes, et al. (2003). The development of QUADAS: a tool for the

quality assessment of studies of diagnostic accuracy included in systematic

reviews. BMC Medical Research Methodology 3: 25.

Whitney, J. D., S. Heiner, et al. (2001). Tissue and wound healing effects of short

duration postoperative oxygen therapy. Biological Research for Nursing 2(3):

206-15.

Yerdel, M. A., E. B. Akin, et al. (2001). Effect of single-dose prophylactic ampicillin and

sulbactam on wound infection after tension-free inguinal hernia repair with

polypropylene mesh. The randomized, double-blind, prospective trial. Annals of

Surgery 233(1): 26-33.

Yokoe, D. S., C. L. Christiansen, et al. (2001). Epidemiology of and surveillance for

postpartum infections. Emerging Infectious Diseases 7(5): 837-41.

Zhu, X. L., W. K. Wong, et al. (2001). A randomized, double-blind comparison of

ampicillin/sulbactan and ceftriaxone in the prevention of surgical-site infections

after neurosurgery. Clinical Therapeutics 23(8): 1281-91.