lable at ScienceDirect

Clinical Nutrition ESPEN 15 (2016) 114e121

Contents lists avai

Clinical Nutrition ESPEN

journal homepage: http: / /www.cl inicalnutr i t ionespen.com

Original article

A protocol for sustained reduction of Total Parenteral Nutrition andcost savings by improvement of nutritional care in hospitals

Rian van Schaik a, *, Kurt Van den Abeele a, Glenn Melsens a, Peter Schepens a,Truus Lanssens a, Bernadette Vlaemynck a, Maria Devisch a, Theo A. Niewold b

a General Hospital Sint-Lucas, Sint-Lucaslaan 29, 8310 Brugge, Belgiumb Nutrition and Health Unit, Faculty of Bioscience and Engineering, University of Leuven, Kasteelpark Arenberg 30, BE-3001 Heverlee, Belgium

a r t i c l e i n f o

Article history:Received 15 January 2016Accepted 25 July 2016

Keywords:Parenteral nutritionTPNCost savingHospitalDietitianMalnutrition

Abbreviations: EN, enteral nutrition via tube; TPESPEN, European Society for Parenteral and EnteralSociety for Parenteral and Enteral Nutrition; EB, evid* Corresponding author. Dietetics Department, AZ S

8310 Brugge, Belgium. Fax: þ32 50 37 01 27.E-mail addresses: rian.vanschaik@stlucas.be (R. va

kuleuven.be (T.A. Niewold).

http://dx.doi.org/10.1016/j.clnesp.2016.07.0022405-4577/© 2016 European Society for Clinical Nutr

s u m m a r y

Background and aims: Malnutrition and the use of Total Parenteral Nutrition (TPN) contribute consid-erably to hospital costs. Recently, we reported on the introduction of malnutrition screening andmonitoring of TPN use in our hospital, which resulted in a large (40%) reduction in TPN and improvedquality of nutritional care in two years (2011/12). Here, we aimed to assure continuation of improvedcare by developing a detailed malnutrition screening and TPN use protocol involving instruction tools forhospital staff, while monitoring the results in the following two years (2013/14).Methods: A TPN decision tree for follow up of TPN in patients and a TP-EN instruction card for caregiverswas introduced, showing TPN/EN introduction schedules based on the energy needs of patients ac-cording to EB guidelines, also addressing the risk of refeeding syndrome. TPN patients were monitoredby dietitians and TPN usage and costs were presented to the (medical) staff. Screening and treatment ofmalnourished patients by dietitians is simultaneously ongoing.Results: In 2014 48% of patients, hospitalized for at least 48 h, were screened on malnutrition, 17% ofthemwere diagnosed at risk, 7.9% malnourished and treated by dietitians. TPN usage dropped by 53% andcost savings of 51% were obtained due to 50% decrease of TPN users in 2014 versus 2010. TPN over ENratio dropped from 2.4 in 2010 to 1.2 in 2014.Conclusion: Sustained improvement of nutritional care and reduction of TPN usage and costs is possibleby introduction of procedures embedded in the existing structures.

© 2016 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rightsreserved.

1. Introduction

The general hospital Sint-Lucas in Bruges, Belgium is a hospitalof 415 beds (Table 1). Whereas patient numbers remained fairlyconstant, since 2007 a steady increase in the use of TPNwas noticedand questions arose about the associated costs, and medical andnutritional necessity. The exact reason for the increase in TPN use isunclear, but it coincided with the start of a federal project to in-crease doctors awareness for malnutrition. This may have

N, total parenteral nutrition;Nutrition; ASPEN, Americanence based.int-Lucas, Sint-Lucaslaan 29,

n Schaik), theo.niewold@biw.

ition and Metabolism. Published b

prompted increased prescription in the absence of clear guidelines.In any case, a project was started to analyze the causes and to tryand find a remedy based on the latter analysis. This was done forTPN and EN use and simultaneously malnutrition risk screeningwas introduced. The aim of the project was to improve quality ofnutritional care of patients by screening and treatment of malnu-trition and monitoring patients on TPN and EN during hospital staybased on EB guidelines by dietitians [1,2]. The initial target wasreduction in consumption and costs of TPN for at least 10% in thefirst year (2010) or 20% in the first 2 consecutive years (2010e2011)after the introduction of the new approach. A concomitant increasein the use of EN was expected. The results of the new approachexceeded expectations, and the first results have been publishedsummarily [3]. Briefly, the number of TPN users decreased by 29%during hospitalization, TPN usage and associated costs werereduced by 40% in 2011 compared to 2010. The proceduresencompassed detailed malnutrition screening with the NRS 2002

y Elsevier Ltd. All rights reserved.

Table 1Number of recognized hospital beds per unit in 2014.

Department/unit Recognized hospital beds

Geriatrics þ rehabilitation 56 þ 9Orthopedics 43ICU 10Cardiac and Medium care 20General surgery, including

abdominal surgery51

Internal medicine, includinginternal medicine 1 (general),internal medicine 2 (oncology)and internal medicine3 (cardiology and nephrology þ rehabilitation)

65 þ 8

Gynecology 8Maternity, including neonatal ICU 22Neurology þ rehabilitation 8 þ 23Psychiatry, including Day clinic 38Psychiatry children and adolescents 25Pediatrics 24Total number of recognized beds 412

R. van Schaik et al. / Clinical Nutrition ESPEN 15 (2016) 114e121 115

[4,5] and TPN use protocols involving instruction tools for hospitalstaff. After publication, the project received a lot of attention fromother hospitals, and many requests were received for more detailedinformation of the procedures followed.

Therefore, in the present study, a detailed description is given ofthe protocols used in our approach. Furthermore, the results of theproject over the period 2009e2014 are given.

2. Methods

During the TPN project (2010e2011) the project leader/dietitianevaluated EN and TPN usage and attitudes of staff towards EN andTPN at the different nursing units in 2010. Causes and contributorsto incorrect TPN use and increased costs were identified and ob-jectives and actions were set for the nutrition team as describedearlier [3].

In the current period (2012e2014), further actions were taken tosecure sustained improvement of the quality of nutritional care ofpatients during hospital stay:

I) introduction of computerized medical records to improvescreening, detection and monitoring of malnourished patientsadmitted to the hospital. Units excluded from screening onmalnutrition risk were cardiac and medium care because of fasttransfers to other units, psychiatry andmaternity because of lowincidence of malnutrition risk and pediatric unit because allpatients are followed up closely by the pediatricians. Terminallyill patients were excluded from screening too.II) continuous follow up of TPN and EN use.III) development of a TPN decision tree (Fig. 1) for follow up ofpatients on TPN by dietitians based on EB Best practice guide-lines [6e16].IV) development of a TP-EN instruction card for caregivers(physicians, dietitians) (Figs. 2, 3) to start EN and TPN in hos-pitalized patients according EB best practice TPN guidelines[6e9,17,18] and treatment of refeeding syndrome risk [10e16].

Much effort has been put in the structural organization of thewhole process as described before [3] with a central role for theproject leader/dietitian as manager of the dietitians team. The di-etitians moved from the kitchen to the clinic and became part ofmultidisciplinary treatment teams. The manager of the dieteticdepartment was made formally responsible for: educating andsupporting dietitians about clinical nutrition concerning TPN and

EN patients, the development of clinical nutrition protocols andprocedures, and the malnutrition screening process, coordinationof the nutritional management team and taking part in the clinicalnutrition team of the ICU. Furthermore, responsibilities involvedfollow up and presentation of TPN and EN usage and malnutritionscreening results to the staff.

During the years 2012e2014 the manager of the dieteticdepartment started sharing the new approach through training ofdietitians teams in other hospitals in Belgium and by presentingthis approach and results at conferences of dietetic associations inBelgium and other European countries.

2.1. TPN and cost calculations

Since 2011 TPN used in our hospital is Smofkabiven®, FreseniusKabi, consisting of three compartment bags, containing aminoacids, glucose and a new generation fat emulsion, including elec-trolytes (or electrolyte free), delivered in 986 ml, 1477 ml and1970 ml bags. Vitamins, minerals and trace elements are addedbefore starting TPN. TPN administered via central venous routeprovides 1100 kcal total energy, contains 8 g Nitrogen or 50 gprotein (18.2% of protein of total energy), 38 g lipids and 125 gglucose per 986 ml bag. In case of peripherally administered TPN,Smofkabiven® Peri, Fresenius Kabi, is used providing 1300 kcal,9.8 g Nitrogen or 63 g protein (19.4% of total energy) per 1904 mlbag.

According to EB guidelines, TPN protein goals of 1.2e1.5 g per kgbody weight [19] per day are recommended except for patientswith end stage kidney disease (0.8 g per kg per day). Protein goalsof 1.2e1.5 g per kg bodyweight could not be met with TPN onlybecause of the fixed % of nitrogen in the used TPN, but could bereached, if patients tolerate additional protein enriched EN, basedon energy needs of patients according ESPEN guidelines. EN used isNutrison® protein plus, Nutricia Advanced Medical Benelux, deliv-ering 63 g of protein, 49 g fat (20% of total energy), 142 g carbo-hydrates and 1250 kcal per liter bag.

Furthermore, we used TPN based on energy goals instead ofprotein goals. By indicating TPN based on protein goals of 1.2e1.5 gper kg per day, too much energy per kg is supplied according tocurrent opinion, especially in ICU patients [20], increasing the riskof refeeding syndrome [10e16]. Instructions for starting TPN or ENare summarized on the TP-EN instruction card for caregivers(Figs. 2,3). If TPN patients started tolerating EN, TPN was decreased,according to available TPN-EN protocols, based on required energygoals of the patients and protein goals are more likely to be met.

TPN used before 2011 in our hospital was Oliclinomel®, Baxter,Eigenbrakel, Belgium, costing 37 V per pack/day (10% of TPN use)and Kabiven®, Fresenius Kabi, Schelle, Belgium average cost 49 V

per pack/day (90% of TPN use). Before 2011 the average cost of allTPN used was approximately 47.70 V per pack/day. From 2011 on,Smofkabiven®, Fresenius Kabi, was used only. Based on data pro-vided by the hospital pharmacy the average cost of TPN of FreseniusKabi was approximately 49 V per pack/day over the period since2011. All cost calculations are based on the figures above.

3. Results

Over the period 2011e2014 screening on malnutrition risk of allpatients admitted to the hospital for at least 48 h by initial NRS2002 [4] by nurses increased from 18.5% to 48%. Malnutrition riskdetected by nurses remained high, between 54% and 58% ofscreened patients admitted to the hospital between 2010 and 2012and decreased to 34% in 2014. Over the period 2011e2014 thenumber of patients screened by nurses, diagnosed malnourishedand treated by the dietitians increased from 649 or 6.6% of patients

Fig. 1. TPN, total parenteral nutrition, EN, enteral nutrition.

R. van Schaik et al. / Clinical Nutrition ESPEN 15 (2016) 114e121116

admitted to the hospital for at least 48 h in 2011 to 784 or 7.9% in2014 (Table 2). In 2011e2012, screening of patients at risk formalnutritionwas carried out mainly at the units with higher risk ofmalnutrition: internal medicine, ICU, geriatric units, rehabilitation,neurology, abdominal surgery and orthopedics. Since introductionof computerized records in 2013e2014, screening was no longerlimited to high risk departments.

In 2014, 82% of all 190 TPN patients were followed up by thedietitians. 4% (8 patients) of the remaining 18% of patients on TPNused just 1 TPN bag and were not followed up by the dietitian.

Total annual TPN use and costs over the period 2007e2009showed a steady increase. The first 8 month of 2010 procedureshad to be developed and introduced and teaching about properTPN and EN use was carried out. Meanwhile TPN use stillincreased further but in the last four months of 2010 TPN usagedropped. So the overall costs were even higher in 2010 comparedto 2009 due to the rise in TPN use in the first 8 months of 2010.After introduction of the new approach, TPN use in packsdecreased in 2011 by 46% compared to 2010 and decreasedfurther by 53% in 2014. TPN costs decreased by 44% in 2011compared to 2010 and decreased further by 51% in 2014compared to 2010, with a peak in TPN use and cost in 2013,mainly caused by an increase at the surgery department. Theoverall cost reduction was mainly due to a decrease by 50% of TPN

users in 2014 compared to 2010. The average duration of dailyTPN use per patient is, after a drop to 10.7 days in 2012, in 2014similar to 2010 (Table 3). Analysis per department over the period2009e2014 (Table 4) showed the clusters ICU, cardiac care andmedium care, internal medicine and surgery to be responsible for72% of TPN usage and costs in 2010. TPN usage and costsincreased until 2010 and showed since 2011 a substantialdecrease of over 41% in 2014. The oncology unit and other unitsshowed a rise in TPN usage of 32% and of 34% in costs in 2010compared to 2009 but dropped by 71% in usage of TPN packs andin costs in 2014 compared to 2010.

Another pattern, a direct decrease from 2009 on was found forthe neurology department with a decrease in TPN packs usage andcosts of 49% in 2010 compared to 2009 due to the shift to EN insteadof TPN use and costs for CVA patients and the new awareness TPNuse of 1 pack a daywas half the price of 2 one liter packs a day. Since2011 TPN use and costs at this department even dropped to 8% oftheir TPN use in 2009. Also the geriatric, maternity þ neonatal ICU,cardiology and nephrology, the emergency and psychiatry units allshowed a decrease from 2009 on and dropped to 31% in 2014compared to the TPN use in 2009.

The reduction of TPN use in the departments responsible forthe largest proportion (72%) of TPN use continued since 2011 afterintroduction of the new approach. Among those are the ICU and

Fig. 2. TPN, total parenteral nutrition; EN, enteral nutrition; p, protein; ICU, intensive care unit; EF, Electrolytes free; N, Nitrogen. CRP, C-reactive protein; Na, Sodium; K, Potassium; Cl, Chlorine; Ca, Calcium; Mg, Magnesium; P,Phosphorus; AP, Alkaline phosphatase; ALAT, Alanine transaminase; TAG, Triacyl glycerol; Pt/INR, Prothrombin time; HCO3-, Bicarbonate; Cu, Copper; Fe, Iron.

R.vanSchaik

etal./

ClinicalNutrition

ESPEN15

(2016)114

e121

117

EN instruction card

1. Supply 30 minutes before (re)start nutrition (EN, TPN, glucose): 100 - 300 mg thiamine / day first 3 days and day 4 - 10 continue 100 mg 2. Supply 1 dose of vitamins (preferably 200% DRI), minerals and trace elements (100% DRI) /day during 10 days extra, in case of dialysis: contact nephrologist3. Start feeding by 10 kcal/kg/day, increasing by 5-10 kcal/kg/day in 4-10 days until full needs covered 4. Avoid extra glucose (via oral, enteral or intravenous route) during phase of feeding increase (or compensate this by decreasing energy supply)

Day 3

Day 1-10: minimum 4 days (or continue in case of abnormal lab results): K, P, Mg, Ca in case of severe hypocalcemia and/or other electrolyte disorders, Albumin 6. Fluid balance and weight curve follow up necessary

Assessm

ent stomach residue according procedure

Standard tube feeding 63 ml / h over 24 h= 1500 ml = 1500 kcal and 60 g protein

Standard tube feeding 84 ml / h over 24 h = 2000 ml = 2000 kcal and 80 g protein

Tube feeding can be adjusted further depending on nutritional and fluid needs to schedule day 4 or changed

to another type of tube feeding and kept on the same dose of day 3

Day 4

Day 1

Day 2

Standard tube feeding 21 ml /h over 24 h = 500 ml = 500 kcal and 20 g protein

5. Before start EN (Glucose): Na, K, Ca, Mg, P, Creatinine, Albumin, Glucose

In case of malnutrition or cachexia: beware of risk of refeeding syndrome and follow instructions

below:10,12

Start tube feeding according assessment of nutritional status and needs by dietitian

Protein needs: 1.2 - 1.5 g / kg ( BMI > 30 : follow BMI 27.5 max)

Energy needs: 25 - 30 kcal / kg (BMI > 30 : follow BMI 27.5 max)

Standard tube feeding 42 ml /h over 24 h= 1000 ml = 1000 kcal and 40 g protein

Fig. 3. EN, enteral nutrition; TPN, total parenteral nutrition; DRI, daily recommended intake; Na, Sodium; K, Potassium; Ca, Calcium; Mg, Magnesium; P, Phosphorus.

Table 2Overview of the different steps of introduction of screening on malnutrition of patients hospitalized > 48 h and results.

Year Totalhospitaladmissions

Hospitaladmission > 48 h

Percentage screenedof hospitaladmissions > 48 h

Percentage at riskof malnutritionof screened patients

Percentage diagnosedmalnourished of patientsat risk of malnutrition

Percentage diagnosedmalnourished of hospitaladmissions > 48 h

Departments

2010 14,035 58 34 internal medicine 12011 13,752 9780 18.5 54 36 6.6 þ internal medicine 2, ICU2012 13,848 9904 34.0 54 32 10.9 þ geriatric units, rehabilitation,

neurology, internal medicine 3,abdominal surgery, orthopedics

2013 14,242 9753 33.8 39 29 9.6 all above units þ general surgery,some units introduced electronicmedical records

2014 14,708 9913 47.9 34 17 7.9 all above units and fullintroduction of electronicmedical records

Excluded from screening on malnutrition: terminally ill patients, and those admitted to short stay surgery, maternity, psychiatry and pediatrics.

Table 3Annual use and costs of TPN during hospital admission between 2007 and 2014.

Year Hospital admissionsper year

TPN costin euro

Patientson TPN

Average durationof TPN use (d)

Total TPNdays

Discharged patientsusing TPN at home

Discharged patientsusing EN at home

Total TPN packs used per yearaccording to hospital pharmacy

2007 13,349 229,266 nda nd nd nd nd nd2008 13,652 231,331 nd nd nd nd nd nd2009 13,822 241,000 377 12.8 4826 9 1 50512010 14,035 255,352 381 13.4 5105 7 6 53592011 13,752 141,478 271 10.7 2900 4 19 29012012 13,848 137,905 249 11.3 2814 3 12 28242013 14,242 147,131 229 13.0 2977 1 8 29822014 14,721 125,702 190 13.4 2546 1 16 2545

a nd: no reliable data available.

R. van Schaik et al. / Clinical Nutrition ESPEN 15 (2016) 114e121118

Table 4Annual use of total parenteral nutrition and costs per department/nursing unit during hospital admission in 2009e2014.

Department/nursing unit 2009 2010 2011 2012 2013 2014

Packs Costs in V Packs Costs in V Packs Costs in V Packs Costs in V Packs Costs in V Packs Costs in V

Recovery 25 996 32 1347 14 695 0 0 0 0 0 0Psychiatry adults 31 1539 11 512 6 303 0 0 0 0 12 584Emergency unit 41 2011 14 695 2 100 3 155 0 0 1 50Othera 55 2208 98 4637 12 607 19 977 2 98 5 248Maternity þ neonatal ICU 94 4798 46 2369 26 1310 13 644 0 0 27 1314Internal medicine 1 (general) 549 26,944 888 43,733 294 14,433 391 19,210 439 21,629 321 15,834Internal medb 2 (oncology) 223 11,150 311 15,472 282 14,069 32 1581 66 3328 113 5565Internal medb

3 (cardiology þ nephrology þ rehabilitation)113 5562 53 2616 34 1651 95 4689 79 3880 16 813

Geriatrics þ rehabilitation 363 17,102 481 23,478 300 14,908 245 12,089 225 11,144 138 6841Surgery (general þ abdominal) 593 28,944 710 34,672 409 20,356 475 23,519 766 37,990 499 24,726Neurology þ rehabilitation 870 41,958 444 21,584 95 4607 106 5209 38 1896 65 3209ICU þ cardiac- þ medium care 2094 97,772 2271 104,237 1427 68,450 1445 69,832 1367 67,167 1348 66,518Total 5051 240,984 5359 255,352 2901 141,486 2824 137,905 2982 147,132 2545 125,702

a Gynecology, orthopedics, dialysis, patients resident elsewhere.b Internal med; internal medicine.

R. van Schaik et al. / Clinical Nutrition ESPEN 15 (2016) 114e121 119

cardiac- and medium care units which showed a decrease of TPNusage of 40% in packs and 36% in costs in 2014 compared to 2010which had a big impact on overall TPN spending, because thiscluster is the largest single user of TPN with 53% of total TPN costsin 2014. A similar pattern, a large decrease in use in 2011compared to 2010 followed by further decrease in 2014 by 30% inpacks and costs was noted for surgery and a 64% decrease in packsand costs for internal medicine compared to 2010. A peak in TPNuse in packs and costs in 2013 was seen, mainly at the surgerydepartment, but the total number of TPN users over all de-partments declined further (Tables 4 and 6). Other departmentsshowed also a further decrease of use in 2014 compared to 2010(recovery, geriatrics, and psychiatry) (Table 4). The percentage ofTPN users of hospital admissions in 2014 dropped by 55% to 1.3%in 2014 compared to 2.7% in 2010.

Concerning EN use, no sizeable changes in users was seen(Table 5). EN used in packs per department showed neurology,followed by ICU and surgery to be the main users through the years2009e2014. The department of geriatrics showed a steady decreasein EN use over the years from 605 packs in 2009 to 30 packs used in2014. The percentage of EN users of hospital admissions over theyears 2010 through 2014 remained at a similar level with 1.1%. The

Table 5Annual use and costs of enteral nutrition during hospital admission in 2009e2014.

Year Number of packs Costs in V/year

2009 3595 10,8412010 3793 11,0002011 3460 10,9912012 3499 11,4612013 3425 11,3442014 3488 9198

Table 6Percentage of TPN and EN users of hospital admissions and ratio TPN/EN users per year.

year Hospital admissions TPN users % TPN users of hospital admissio

2009 13,822 377 2.732010 14,035 381 2.712011 13,752 271 1.972012 13,848 249 1.802013 14,242 229 1.612014 14,721 190 1.29

TPN: total parenteral nutrition.EN: enteral nutrition.

ratio of TPN over EN users of hospital admissions per year droppedfrom 2.4 in 2010 to 1.2 in 2014 (Table 6).

4. Discussion

In a previous study, we showed that follow up of patients on TPNby dietitians resulted in a decline of patients on TPN by 29% andrelated costs by 40% in the period 2010e2011 [3]. Here, we aimed toassure continuation of the new approach by providing caregiversnewly developed TPN instruction tools and by follow up of TPNpatients by dietitians during hospitalization and bymonitoring TPNusage. Malnutrition and risk of refeeding syndrome needed to bedetected and treated, so we introduced guidelines to address thisproblem in instruction tools for caregivers [10e16]. According tothe literature, 20e45% of patients admitted to the hospital is at riskfor malnutrition [21]. It is not clear what the prevalence is ofrefeeding risk in hospitals because it is often not detected. How-ever, one study showed 9% of acute care patients admitted to thehospital to be at risk for refeeding syndrome, but the ICU wasexcluded [22], and another study demonstrated 34% of patients atICU at risk for refeeding syndrome [23].

Number of users/year Average number of packs/user/year

166 21.7159 23.9174 19.9181 19.3152 22.5157 22.2

ns EN users % EN users of hospital admission Ratio TPN/EN users

166 1.20 2.3159 1.13 2.4174 1.27 1.6181 1.31 1.4152 1.07 1.5157 1.07 1.2

R. van Schaik et al. / Clinical Nutrition ESPEN 15 (2016) 114e121120

Concerning TPN, the TP-EN instruction card for caregivers showsTPN and EN introduction schedules based on energy needs perpatient per day according to EB guidelines, and ESPEN/ASPENguidelines for energy needs in 4 different situations: 1) refeedingrisk in malnourished patients [10e16], 2) basic needs [7], 3)increased energy needs [8] and 4) acute stress [6,9]. In case ofBMI > 30, energy and protein goals are based onweight of BMI 27.5,because overfeeding has to be avoided [9,20,24]. It also provides amonitoring schedule of laboratory tests in case of refeeding risk inTPN or EN patients and provides information about types of TPNand EN, nutritional value and specified applications. The TPN de-cision tree aids caregivers in the process of starting and monitoringpatients on TPN, also in cases at risk of refeeding syndrome.

The introduction of TPN instruction tools and the follow up ofTPN patients by dietitians resulted in a further decline of total TPNusage and in the number of users of 53% and 51% respectively in2014 compared to 2010. TPN/EN ratio of hospital admissions peryear dropped from 2.4 in 2010 to 1.2 in 2014. This reduction inoverall TPN usage far exceeded the aims set at the beginning of theproject (10% in 1 year or 20% in 2 years). It is concluded that ourfindings corroborate earlier reports that nutrition support teamsrecommendations can reduce hospital costs [25e27]. Our findingsonly concern savings in TPN costs, but one could also expect adecrease in TPN-related complications and associated costs. Un-fortunately, we were not able to establish that because althoughcentral and peripheral catheter complications are registered in ourhospital, the system does not allow for differentiation between TPNand other causes. Reduction of TPN was expected to increase ENusage but this was not the case. This can possibly be at least partlyexplained by improved overall nutritional care and better aware-ness of the caregivers. There was an overall decrease over theperiod 2011e14 for TPN usage and costs, with the notable exceptionof 2013 which was mainly caused by the surgery department,probably caused by the severe illness of patients there. A furtherreduction of TPN use could be reached by planned introduction in2016 of post gastric tube placement in ICU patients for provision ofEN in case of delayed gastric emptying. This is expected to increaseEN usage [9].

The decreased usage of TPN and number of users in the geriatricpopulation as well as in end stage oncology patients is probably duein part to recently introduced new ethical guidelines regarding TPNuse. Gastro-enterologists, intensivists and neurologists offered pa-tients EN instead of TPN according to EB guidelines, in which pro-cess the weekly follow up of ICU patients by the dietitian was likelyan important factor. Delayed start of TPN in patients at ICU couldalso have played a role in the decline of TPN usage, because since2011 the ICU in our hospital delayed TPN provision to critical illpatients until 72 h after ICU admittance according to EB guidelines[9] and recent studies [20,24] also to reduce the risk of refeedingsyndrome [10e16].

Concerning malnutrition, in the interpretation of the data thefollowing should be kept in mind: the number of patients screenedfor malnutrition risk, hospitalized for at least 48 h increased overthe years to 48% in 2014. However, detection of malnutrition risk bynurses decreased from 58% in 2010 to 34% in 2014 (Table 2). Thisdecline might be explained by the fact that in 2010e2011 only highrisk units were screened, whereas in the following years units withlower risk of malnutrition were included. Furthermore, afterintroduction of electronic screening an estimated 10e15% doublescreenings of patients due to transfers to other units during hos-pitalization became visible and was eliminated, whereas the totalnumber of screenings by nurses still increased. Dietitians diagnosed17% of patients detected at risk of malnutrition in 2014, a 50%decline compared to 2010. The apparent lower incidence ofmalnutrition might be due to incomplete screening by nurses after

introduction of electronic screening in 2013 at several nursingunits. In 2015, incomplete screening will be addressed and we aimto increase malnutrition risk screening percentage to 60% of allpatients admitted to the hospital and 80% of patients admitted tounits taking part in screening at risk for malnutrition. Furthermore,pre-admission malnutrition screening of surgery patients is plan-ned from 2016 on to detect patients at risk in an earlier stage.

Essential for the success of this approach in our hospital is thecontinuing involvement and interest of the hospital management,which ensures cooperation of essentially all levels needed e.g. di-etitians, doctors, nurses, hotel service and kitchen personnel,pharmacists, ICT-, medical registration- and hospital administra-tion personnel. The procedures are embedded in the care plans andhospital structures. The method and procedures are part of theinternal quality indicators for dietitians and will be reviewed everyother year according instructions of the external quality indicator.This approach is the first described in Belgium, resulting inimproved quality of care and a considerable decrease in TPN con-sumption and associated costs over 5 years.

Conclusion: The current study shows that sustained improve-ment of quality of nutritional care and cost savings is assured in ourhospital. Reason for sustained improvement of quality of nutri-tional care is not only the introduction of protocols, but also of aTPN decision tree and TP-EN instruction card for caregivers, fol-lowed up by the multidisciplinary team including the medical staff,thus embedded in the existing structures. Close monitoring ofprocedures and results remains necessary to ensure ongoingimproved nutritional care of patients on TPN. We hope by sharingour results with dietitians teams in other hospitals, nutritionalstatus of patients on TPN by this approach becomes more wide-spread. In the mean time, we also started to train dietitian teams inother hospitals upon request. Similar cost savings can be expectedin similar circumstances, however TPN usage and costs will alsodepend on the severity of illness of the local hospital population,which might differ from our hospital. Indeed, from the limited datawe could obtain from other Flemish hospitals it appeared that thereis a large variety in TPN use, ranging from as high as our hospital atthe start of the project to as low as we are currently.

Statement of the authorship

RS collected data, organized and carried out the project,analyzed the data and wrote the draft article, PS, TL, BV and NCcarried out the project, PS, KA and GM worked out protocols andprocedures and TN assisted in the analysis, critically revised andfinally approved the manuscript.

Conflict of interest

None.

Sources for funding

None.

Acknowledgements

We would like to thank Dr. D. De Coninck (retired by 2014) andDr D. Bernard (current), medical directors, for supervision andsupport during the study. A. Surmont, pharmacist; for providing ENand TPN data and members of the nutritional management team,for their feedback on the current protocols, procedures and TPNapproach.

R. van Schaik et al. / Clinical Nutrition ESPEN 15 (2016) 114e121 121

References

[1] Cove ME, Pinsky MR. Early or late parenteral nutrition: ASPEN vs. ESPEN. CritCare 2011;15:317.

[2] Soguel L, Revelly JP, Schaller MD, Longchamp C, Berger MM. Energy deficit andlength of hospital stay can be reduced by a two-step quality improvement ofnutrition therapy: the intensive care unit dietitian can make the difference.Crit Care Med 2012;40:412e9.

[3] van Schaik R, Niewold T. Quality improvement and cost savings by dietitiansthrough follow-up of patients with total parenteral nutrition during hospitaladmission. e-SPEN J 2014;9:e59e62.

[4] Kondrup J, Rasmussen HH, Hamberg O, Stanga Z. Ad Hoc ESPEN WorkingGroup. Nutritional risk screening (NRS 2002): a new method based on ananalysis of controlled clinical trials. Clin Nutr 2003;22:321e36.

[5] Kondrup J, Allison SP, Elia M, Vellas B, Plauth M. ESPEN guidelines for nutritionscreening. Clin Nutr 2002;2003(22):415e21.

[6] Singer P, Berger MM, Van den Berghe G, Biolo G, Calder P, Forbes A, et al.ESPEN guidelines on parenteral nutrition: intensive care. ESPEN. Clin Nutr2009;28:387e400.

[7] Bozzetti F, Arends J, Lundholm K, Micklewright A, Zürcher G, Muscaritoli M.ESPEN guidelines on parenteral nutrition: non-surgical oncology. Clin Nutr2009;28:445e54.

[8] Braga M, Ljungqvist O, Soeters P, Fearon K, Weimann A, Bozzetti F.ESPEN guidelines on parenteral nutrition: surgery. Clin Nutr 2009;28:378e86.

[9] McClave SA, Martindale RG, Vanek VW, McCarthy M, Roberts P, Taylor B, et al.Guidelines for the provision and assessment of nutrition support therapy inthe adult critically ill patient: Society of Critical Care Medicine (S.C.C.M.) andAmerican Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). J ParenterEnter Nutr 2009;33:277e316.

[10] Stanga Z, Brunner A, Leuenberger M, Grimble RF, Shenkin A, Allison SP, et al.Nutrition in clinical practice-the refeeding syndrome: illustrative cases andguidelines for prevention and treatment. Eur J Clin Nutr 2008;62:687e94.

[11] Mehanna HM, Moledina J, Travis J. Refeeding syndrome: what it is, and how toprevent and treat it. BMJ 2008;336:1495e8.

[12] Ten Dam S, Jonkers C, Visser S, Noordhoff H, Hoekstra M, Vedder K, et al.NVOnderwerp refeedingsyndroom. Nederlands Voedings Team Overleg. 2012.Accessed November 29, 2015, at, http://nederlandsvoedingsteamoverleg.nl/content/media/NVOnderwerp%20Refeedingsyndroom%20-%20Eindversie%20jan%202013.pdf.

[13] Boateng AA, Sriram K, Meguid MM, Crook M. Refeeding syndrome: treatmentconsiderations based on collective analysis of literature case reports. Nutrition2010;26:156e67.

[14] Khan LUR, Ahmed J, Kahn S, Macfie J. Refeeding syndrome: a literature review.Gastroenterol Res Pract 2011;2011. Article ID 410971.

[15] Kraft M, Btaiche IF, Sacks GS. Review of the refeeding syndrome. Nutr ClinPract 2005;20:625e33.

[16] Mehanna HM, Nankivell PC, Moledina J, Travis J. Refeeding syndrome e

awareness prevention and management. Head Neck Oncol 2009;1:4.[17] Arends J, Bodoky G, Bozzetti F, Fearon K, Muscaritoli M, Selga G, et al. ESPEN

guidelines on enteral nutrition: non-surgical oncology. Clin Nutr 2006;25:245e59.

[18] Kreymann KG, Berger MM, Deutz NE, Hiesmayr M, Jolliet P, Kazandjiev G, et al.ESPEN guidelines on enteral nutrition: intensive care. Clin Nutr 2006;25:210e23.

[19] Singer P, Hiesmayr M, Biolo G, Felbinger TW, Berger MM, Goeters C, et al.Pragmatic approach to nutrition in the ICU: expert opinion regarding whichcalorie protein target. Clin Nutr 2014;33:246e51.

[20] Preiser J, van Zanten ARH, Berger MM, Biolo G, Casaer MP, Doig GS, et al.Metabolic and nutritional support of critically ill patients: consensus andcontroversies. Crit Care 2015;19:35.

[21] Stratton RJ, Green CJ, Elia M. Disease-related malnutrition. An evidence-basedapproach to treatment. Oxford: CAB International; 2003. p. 56e8.

[22] Owers EL, Reeves AI, Ko SY, Ellis AK, Huxtable SL, Noble SA, et al. Rates of adultacute inpatients documented as at risk of refeeding syndrome by dietitians.Clin Nutr 2015;34:134e9.

[23] Marik PE, Bedigan MK. Refeeding hypophosphataemia in an intensive careunit: a prospective study. Arch Surg 1996;131:1043e7.

[24] Casaer MP, Mesotten D, Hermans G, Wouters PJ, Schetz M, Meyfroidt G, et al.Early versus late parenteral nutrition in critically ill adults. N Engl J Med2011;365:506e17.

[25] Goldstein M, Braitman LE, Levine GM. The medical and financial costs asso-ciated with termination of a nutrition support nurse. J Parenter Enter Nutr2000;24:323e7.

[26] Roberts MF, Levine GM. Nutrition support team recommendations can reducehospital costs. Nutr Clin Pract 1992;7:227e30.

[27] Jones JS, Tidwell B, Travis J, Spencer T, Phillips P, Burford B. Nutrition supportof the hospitalized patient: a team approach. J Miss State Med Assoc 1995;36:91e9.