Arch Orthop Trauma Surg (1996) 115:149-152 © Springer-Verlag 1996

J. S c h m i d t • M. H. H a c k e n b r o c h . D. K u m m V. Taravat i

Is instillation drainage for the treatment of infected joints, bones and soft tissues still up to date?

Received: 2 July 1995

Abstrac t For the treatment of infected joints, bones and soft tissues, either an instillation drainage therapy or the use of gentamicin-polymethyl-methacrylate (PMMA) chains (Septopal) in addition to surgical revision and systemically given antibiotics is currently in clinical use. We investi- gated 102 patients treated in our clinic by means of instil- lation drainage and compared the results with those ob- tained with gentamicin-PMMA chains by other authors. The overall long-term success with non-recurrence of the infection is nearly comparable (80% instillation drainage, 84% gentamicin-PMMA chains); however, with instilla- tion drainage the duration of hospitalisation is signifi- cantly longer (mean 42.26 days vs 15.3 - 33 days). Addi- tionally, there was a high rate of germ shifts with instilla- tion drainage (33.3%) and a disappointing result in 20 in- fected endoprostheses (9 recurrences with 11 revisions). Regarding the intensive nursing care requised and the nec- essary isolation from other patients, instillation drainage can only be recommended for the therapy of infected joints, bones and soft tissues if the results are better in compari- son with gentamicin-PMMA chains. An improvement may be achieved with the closed instillation drainage system.

Introduction

For the therapy of infected joints, bones and soft tissues, surgical revision plays the main role combined with sys- temically given antibiotics [14, 16]. For additional local therapy instillation drainage in different forms [2, 10] or the use of gentamicin-polymethylmethacrylate (PMMA) chains (Septopal) [8-10, 12] is recommended.

Instillation drainage is normally seen as a mechanical cleaning complementary to the surgical procedure [6, 11], and therefore no antibiotics or antiseptics are added to the NaC1 fluid I15]. The problem with instillation drainage is

J. Schmidt (N~) . M. H. Hackenbroch - D. Kumm • V. Taravati Klinik und PoIiklinik fiir Orthopfidie der Universitfit zu K61n, Joseph-Stelzmann-Strasse 24, D-50931 K61n, Germany

the risk of secondary contamination of the wound along the drains, the necessary intensive care and the long hos- pitalisation [1, 2, 7]. Especially the last makes this therapy very expensive: therefore instillation drainage can nowa- days only be recommended, if the results are significantly better than those of alternative methods.

Gentamicin-PMMA chains are implanted in the wound either completely, with primary closure of the wound, or with parts of the chains penetrating the skin. The chains are made of PMMA balls containing gentamicin. Gentam- icin penetrates continuously into the wound and kills sen- sitive germs [3].

In the literature the results of Septopal therapy are mentioned several times. In a review of 16 articles, Grie- ben [4] found a rate of 85% (range 63%-100%) for the short-term success within 2 months (Table 1 a) and a rate

Table 1 a Early postoperative results with Septopal in bone infec- tions (modified according to Grieben [4])

Author n Control of infection (%)

Ackermann 31 20 (65%) De Groote et al. 24 15 (63%) Eberle 34 31 (90%) Guo Fen 10 10 (100%) Hfirle 35 33 (94%) Jenny et al. 86 58 (67%) Klemm et al. 147 136 (93%) Lambiris et al. 88 8l (92%) Lidgren 6 5 (83%) a M/filler et al. 171 159 (93%) Probst et al. 205 183 (89%) Schulte et al. 40 27 (68%) Shipley et al. 18 14 (78%) VEscei 80 60 (75%) Weise and Weller 50 44 (88%) Winkelmann et al. 20 14 (70%)

Total 1045 890 (63%-100%, mean 85%)

One patient still under observation

150

Table l b Follow-up results with Septopal in bone infections Table 3 Localization and aetiology of diseases from patients (modified according to Grieben [4]) treated by instillation drainage, n = 102 (60 male, 42 female)

Author n No re- Observation Localization currence time (months)

Ackermann 31 20 (65%) 2-28 (mean 18.5) De Groote et al. 24 15 (63%) 6-21 Jenny et al. 58 47 (81%) 7-24 Klemm et al. 128 120 (94%) 7-24 Lambiris et al. 81 70 (86%) 5-20 Lidgren 6 5 (83%) 4- 6 Miiller et al. 146 137 (94%) mean 21 Shipley et al. 14 12 (86%) 4-15 V6scei 80 69 (86%) 3-54 Weise and Weller 47 26 (55%) mean 15 Winkelmann et al. 20 14 (70%) 1-10

Total 635 525 (55-94, mean 84%)

Table 2 Duration of hospitalisation of patients treated with Sep- topal

Source: Duration of hospitalisation

University of Arkansas (3) University of Louisville (3) University of West Virginia (3) University of Pennsylvania (3) H66k, M. and Lindberg, L. (11) Asche, G. (2) Ltitje, H.C. and Penschuck, C. and

Aydin, V. (19)

7-37 (mean 15.3 days) 3-55 (mean 30.4 days) mean 16 days 3 4 6 (mean 12.5 days) 1-92 (mean 33 days) mean 32.4 days mean 20.4 days

of 84% (range 55%-94%) for the long-term results (Table 1 b). The duration of hospitalisation is shown in Table 2.

The aim of this study is to compare the results of instilla- tion drainage with those of Septopal implantation as a local adjuvant therapy of infected joints, bones and soft tissues.

Materials and methods

From 1980 to 1991 at the Clinic and Polyclinic of Orthopaedics of the University of Cologne we used a continuous installation drainage in all infected joints, bones and soft tissues. One or two rinsing drains were combined with preferably two suction drains. Care was taken that the fluid reached the entire infected area and not rinsing only a small canal. The bottles had to be changed when filled, and therefore it was not a closed drainage system [11].

Of 120 consecutive patients with instillation drainage 102 could be followed up by 36 months later through personal investigation or questionnaire, The results of these patients are compared with the re- sults of Septopal-treated patients mentioned in the literature.

Results

Our 102 patients consisted of 60 men and 42 women with an average age of 40.75 years (range 1-86 years). The lo- calization and the aetiology are given in Table 3. The dif- ferent germs found when establishing the diagnosis are given in Table 4.

Osteomyelitis 38 Infected joints 34 Infected prostheses 20 Infected soft tissues 10

Aetiology n

Postoperatively 56 After trauma 8 After i.a. injections 4 Haematogenous 8 Unknown reason 26

The rinsing suction devices were used for an average period of 13.15 days (range 1-35 days). In 84 patients we only used 0.9% NaC1, 12 patients received additionally polividoniod as an antiseptic fluid (Betaisodona), and in 6 of our first cases we added neomycin and bacitracin (Neba- cetin).

In 20 of the 102 patients with primary application of an instillation drainage, 27 operations for revision were nec- essary: 15 patients required one revision, 3 patients re- quired two, and 2 patients required three revisions (Table 5). The number of revisions in comparison with the dif- ferent indications in our patients is also shown in Table 4. The cumulative probability for the non-recurrence of in- fection after instillation drainage was calculated by the ac- tuarial method (Life Table Analysis [5, 13]) with the worst and best cases in Table 6; the curve is shown in Fig. 1. The rate for short-term success (within 3 months) is ac- tuarial 93.93% (worst 86.27%, best 94.12%), and for long-term success (up to 36 months) is actuarial 79.77% (worst 71.56%, best 81.37%).

Different problems arose from the instillation drainage itself, as shown in Table 7, especially the complication of 34 germ shifts. Of the 82 patients with no necessity for re- vision, 27 suffered a germ shift; of the 20 revised patients, 7 had a germ shift.

The average duration of hospitalisation o f our patients was 42.26 days (range 9 -199 days) as shown in Table 8.

Discussion

As mentioned also in other papers, the problem of the comparison of different therapies in infected joints, bones and soft tissues is the inhomogenei ty of the patient collec- tives. Our material with 102 patients comprising different infections and a significant variety of age, sex distribu- tion, aetiology, germ specimens and localisation permits a critical discussion and comparison with the results pub- lished elsewhere.

Our former technique of instillation drainage was re- cently improved by means of a closed system [11]. In our

151

Table 4 Microbiological findings at the beginning of therapy (occurrence of one or more isolated germs with regard to the number of all patients, n = 102)

1980 1981 1982 1983 1984 1985 1 9 8 6 1 9 8 7 1 9 8 8 1 9 8 9 1990 1991 n

Staphy lococcus aureus 2 1 2 2 4 2 2 2 3 3 8 7 38 Staphy lococcus epiderrnidis 1 l 1 1 1 1 6

Pseud o m o n a s aeruginosa 1 1 1 3 S treptococcus 1 1 2 3 1 8

Klebsie l la p n e u m o n i a e 1 1 2 Enterobac ter cloacae 1 1 Pep tococcaceae 1 1 1 1 4

Serrat ia marcescens 1 1 Escher ich ia coli 1 1 Enterococcus 1 1 2

Pas teure l la mul toc ida 1 1

Ac ine tobac ter calcoacet icus 1 1

Ci trobacter f r eund i i 1 1

Patients with no germs identified 2 2 1 3 4 6 3 6 7 6 3 7 50 Total number of patients 5 3 4 6 10 9 4 13 12 9 12 15 102

Table 5 Revisions in relation to different locations of infec- tion

Location of infection n Patients with Total Revisions recurrence recurrence

1× 2× 3×

Hip 14 2 2 2 Knee 16 1 1 1 Hip endoprosthesis 14 6 6 6 Knee endoprosthesis 6 3 5 2 Tibia shaft 21 4 6 2 Upper leg soft tissue 4 2 5 Shoulder 3 1 1 1 Femur shaft 13 1 1 1 Upper ankle joint 1 0 0 0 Humerus 2 0 0 0 Radius 2 0 0 0 Other soft tissue 6 0 0 0

Total 102 20 27 15

Table 6 Cumulative probability of non-recurrence of infection af- ter instillation drainage (actuarial method)

t (month) Actuarial Worst Best

0 1 1 1 1 0.9703 0.9510 0.9706 3 0.9393 0.8627 0.9412 6 0.8961 0.8039 0.9019

12 0.8305 0.7450 0.8431 24 0.7977 0.7156 0.8137 36 0.7977 0.7156 0.8137

opinion an antiseptic fluid should not be added; in our 12 patients with an antiseptic additive, no positive effect could be seen, because 4 needed revision surgery and 2 addi- tional systems did not work.

Our r insing and suction t ime (average 13.15 days) was relatively long. Other authers r ecommend a period of 3 -5 days [ 11 ]. Obviously, the high number of germ shifts we

0.8 °.il ~0 .4

~0.

| l - - A lk

" Actuarial • Best • Worst

; ; 1'01? 1'41'6 6 fo 2'2 2'4 2'6 2'8 a'o 3? a'4 3'6 Time (month)

Fig, 1 Cumulative probability of non-recurrence of infection after instillation drainage (actuarial method)

noted was due to the long drainage time. On the other hand, we found no correlation between the germ shift and the necessity for revision.

Overall, our revision rate is comparable to the results of Septopal-treated patients, and our long-term success rate of nearly 80% is satisfactory. The failure rate of in-

152

Table 7 Complications after instillation drainage application Drainage fluid n

0.9% NaC1 84 0.9% NaC1 with 12

Betaisodona 0.9% NaC1 with 6

Refobacin or Nebacitin

Total 102 5

Drainage Germ Germ Loss of Recur- Germ obstruc- persis- coloniza- blood rence shift tion tence tion of the through

drainage drainage

1 3 1 1 13 28 2 4 3

2 3 3

3 1 1 20 34

Table 8 Duration of hospitalisation from patients treated with in- stillation drainage

Minimum Maximum Mean Total (days) (days) (days) mean

Osteomyelitis 9 129 32.11 Joint infection 15 199 50.72 Infected 17 123 50.40

endoprostheses Soft-tissue 10 71 35.82

infection 42.26 days

stal lat ion dra inage in infected implants is d isappoint ing. In 10 patients with deep infections, the prosthesis was re- v ised but not r emoved when insert ing the inst i l la t ion drainage. In these 10 cases we had to per form 6 further re- visions. In the remain ing 10 pat ients the prostheses were p r imar i ly removed , and there were only 3 further revi- sions. Obviously , this is not significant , because the deci- sion to r emove the prosthes is depended on the cl inical as- pect, but we have to recognize that in the severer cases, when it seemed to be necessary to r emove the prostheses, we have had better results.

The durat ion of hospi ta l i sa t ion is even longer than in the Septopal group. Therefore, this therapy is more ex- pens ive and harms the pat ient more, even consider ing the need for a second operat ion to remove the chains. The pa- tient also requires much more nursing care. Final ly, the patients have to be isola ted because of the r isk of spread- ing the germs.

In our opinion we can no longer r e c o m m e n d the classi- cal inst i l la t ion dra inage since it p roduces no bet ter results than the use of g e n t a m i c i n - P M M A chains in a far less cost ly and more t ime-saving procedure. Only when instil- la t ion dra inage offers much better results, as descr ibed for the c losed procedure , wil l it still have a p lace in the ther- apy of infected joints , bones and soft t issues.

References

1.Asche G (1978) Spiilsaugdrainage oder Gentamycin-PMMA- Kugeln in der Therapie infizierter Osteosynthesen. Unfall- heilkunde 81 : 463-468

2. Blaha JD, Nelson CL, Frevert LF, Henry SL, Seligson D, Es- terchai JL, Heppenstal RB, Calhoun J, Cobos J, Mader J (1990) The use of Septopal (polymethylmethacrylate beads with gen- tamicin) in the treatment of chronic osteomyelitis. Instr Course Lect 39 : 509-514

3. Dingeldein E (1983) Bakteriologische Untersuchungen bei Pa- tienten unter der Behandlung mit Gentamicin-PMMA-Ketten. Polim Med 13:9-10

4.Grieben A (1983) Klinische Ergebnisse mit Septopal-Kette in Knochen- und Weichteil-Indikation. Polim Med 13:29-30

5. Harms V (1988) Biomathematik, Statistik und Dokumentation. Harms, Kiel

6. Hendrich V, Knner EH (1986) Die Saug-Sptil-Behandlung bei der Behandlung der chronischen Osteomyelitis. Unfallchirur- gie 12:101-103

7. H66k M, Lindberg L (1987) Treatment of chronic osteo- myelitis with gentamicin-PMMA beads. Trop Doct 17:157- 163

8.Jenny G (1988) Local antibiotic therapy using gentamicin- PMMA chains in post-traumatic bone infections, short and long-term-results. Reconstr Surg Traumatol 20:36-46

9. Klemm K (1979) Gentamycin-PMMA-Kugeln in der Behand- lung abszedierender Knochen- und Weichteilinfektionen. Zen- tralbl Chir 104:934-942

10.Ltitje HC, Penschuck C, Aydin V (1988) Local antibiotic treat- ment of soft-tissue infections with gentamicin-PMMA chains. Reconstr Surg Traumatol 20:112-119

11.Pfister A, Ochsner PE (1993) Erfahrungen mit geschlossenen Sptil-Saug-Drainagen und gleichzeitiger Anwendung eines An- tiseptikums. Unfallchirurg 96:332--340

12.Quell M, V6csei V (1986) Mittelfristige Ergebnisse tiber die Behandlung der posttraumatischen Osteitis und Osteomyelitis mit Gentamicin-PMMA-Miniketten. Aktuel Traumatol 16: 158- 160

13. Sachs L (1984) Angewandte Statistik. Springer, Berlin Heidel- berg New York

14. Schuckmann P, Schuckmann W (1989) Unsere Erfahrungen mit dem Einsatz von Septopalketten bei der Behandlung bak- trieller Arthritiden und Osteomyelitiden. Beitr Orthop Trauma- tol 36 : 428-434

15.Schumpelick V, Bleese NM, Mommsen U (1989) Chirurgie. Enke, Stuttgart

16. Weise K, Mtiller HP (1988) Die lokale Antibiotikatherapie bei postoperativer und posttraumatischer Osteitis mit der Sep- topal@-Kette. Unfallchirurg 91:416-421