Introduction

At the end of the twentieth century, mortality in surgi-cally treated patients with sepsis remains high in spiteof the advanced diagnostic methods and interventionaltherapies now available. Depending on the specific pa-tient group, morbidity ranges from 25% to 90% [1, 2,3, 4, 5]. Endotoxin as a membrane component in gram-negative bacteria may induce activation of cytokine cas-cades and produce both endothelial and cellular damag-

es. This pathway is relevant particularly in multiple or-ganic failure (MOF) [6]. Endotoxin may also induce se-cretion of tumor necrosis factor (TNF), which subse-quently promotes organic failure as well [7, 8, 9]. Severalauthors have reported that high levels of TNF are corre-lated with septic shock and outcome of septic patients[8, 10, 11, 12, 13].

During the past 10 years means of extracorporealdetoxification such as continuous venovenous hemofil-tration (CVVHF) and plasmapheresis (total plasmatic

J. SchmidtS. MannV. D. MohrR. LampertU. FirlaH. Zirngibl

Plasmapheresis combined withcontinuous venovenous hemofiltrationin surgical patients with sepsis

Received: 28 May 1999Final revision received: 29 February 2000Accepted: 1 March 2000

J. Schmidt ()) ´ V.D. Mohr ´ U. Firla ´H. ZirngiblDepartment of Surgery,University Witten-Herdecke,Klinikum Wuppertal, Heusnerstrasse 40,D-42 283 Wuppertal, Germanye-mail: Johannes.Schmidt@klinikum-wuppertal.deTel.: + 49-202-8 962927Fax: + 49-2 02-896 2943

S.MannDepartment of Surgery,University Regensburg,Franz-Josef-Strauss-Allee 11,D-93 042 Regensburg, Germany

R. LampertDepartment of Anesthesiology,University Witten-Herdecke,Klinikum Wuppertal, Heusnerstrasse 40,D-42 283 Wuppertal, Germany

Abstract Objective: To examine theeffect of continuous venovenoushemofiltration (CVVHF) combinedwith plasmapheresis (TPE) in criti-cally ill surgical patients after treat-ment of the septic focus.Design: Observational pilot study.Setting: University teaching hospitalintensive care unit.Interventions: TPE and CVVHFwere administered 24 h after surgi-cal and/or interventional treatmentof septic focus. Arterial blood pres-sure, cardiac output, and systemicvascular resistance values weremonitored. We examined the effectof the combined extracorporealdetoxification on outcome related toage, morbidity, organic failure rate,and initial APACHE II score.Measurements and results: Forty-three patients with sepsis weretreated; 19 received TPE in combi-nation with CVVHF, and 24 did notreceive extracorporeal therapy.Overall mortality was 44.2 %. In thetherapy group mortality was lower(42.1 vs. 45.8%), but the primary

organic failure rate was higher. Therelationship between mortality andage was similar in the two groups.There was also no difference be-tween the groups in the course ofscores on APACHE II, multiple-or-gan failure, and sepsis severity. Onlypatients with an initial APACHE IIscore of 21±25 had a significant re-duction in mortality after combinedextracorporeal detoxification. Mor-tality of 17% in TPE/CVVHF pa-tients with single- (pulmonary) anddouble-organ failure (renal/pulmo-nary) was significantly lower(P < 0.0001) than in untreated pa-tients.Conclusions: Reduction in mortalityin single- and double-organ failurewas as high as 28% in septic patientswith combined extracorporealdetoxification. A prospective ran-domized trial in sepsis and double-organ failure should be projected.

Key words Plasmapheresis ´Hemofiltration ´ Sepsis ´ Mortality ´Surgery

Intensive Care Med (2000) 26: 532±537Ó Springer-Verlag 2000 ORIGINAL

exchange, TPE) have been used in cases of sepsis andseptic shock to remove the pathogenic mediators fromthe patient in distress. Our own team has previouslydemonstrated the positive effect of TPE on patientswith acute necrotizing pancreatitis [14], and othershave supported high-flow CVVHF in patients with sep-tic and nonseptic MOF [15]. Only few case reportshave been published regarding plasmapheresis andhemofiltration as a combined therapy [16, 17, 18]. Nosubstantial data are presently available on indicationsfor this invasive extracorporeal therapy.

Encouraging results arising from our own researchon TPE noting higher efficacy of combined extracorpo-real detoxification in patients with persistent renal in-sufficiency led us to carry out a pilot study in cases ofpersistent severe sepsis after sufficient surgical treat-ment of the septic source [14].

Materials and methodsWe investigated the effect of combined extracorporeal detoxifica-tion therapy in 43 patients with sepsis treated in a surgical intensivecare unit. We compared 19 patients receiving CVVHF plus TPEwith 24 patients who had been treated earlier and received no addi-tive extracorporeal therapy. Randomization was thus not perform-ed. The mean age was 60.3 years in the therapy group and 64.2 yearsin the untreated group; there was no difference in sex distribution.All patients with sepsis had either surgical (n = 21), interventional(n = 12), or both types (n = 10) of treatment of the septic focus, sothat at the start of the study the septic source had been already elim-inated or at least substantially diminished. The placement of drain-ages in interventional treatment was guided by computed tomogra-phy and/or sonography. Patients in both groups had standard inten-sive care with monitoring of arterial blood pressure, cardiac output,and determination of systemic vascular resistance values.

TPE was performed by administration of two fresh-frozen plas-ma units per 10 kg body weight over a period of 2±3 h. During thisprocedure a polypropylene filter (Gambro) was used. Hemofiltra-tion was carried out in high flow mode (2 l substitute per hour) us-ing the 100 Multiflow filter from Hospal. The substitute presentedthe following electrolyte concentrations: 140 mmol/l sodium,2 mmol/l potassium, 2.13 mmol/l calcium, 0.75 mmol/l magnesium,112 mmol/l chloride, 35.75 mmol/l lactate.

Organic failure was defined according to internationally estab-lished agreements in intensive care medicine [19, 20, 21, 22, 23,24]. Pulmonary failure was present when the Horowitz ratio(PaO2/FiO2) was lower than 300 mmHg, or when ventilation thera-py lasted longer than 12 h [20]. Renal failure was diagnosed whencreatinine level was higher than 2.5 mg/dl, and urea was higherthan 150 mg/dl or urinary excretion was lower than 500 ml/day[21, 22]. Hepatic failure was diagnosed when the bilirubin levelwas higher than 3.0 mg/dl or the elevation in transaminase activitymore than twice the normal level ( > 36 U/l) [23]. Cardiac failurewas diagnosed when the dopamine dose was higher than 20 mg/hfor at least 12 h or it was lower than 20 mg/h with additional admin-istration of inotropic drugs (such as dobutamine, adrenaline, nora-drenaline) for at least 12 h [24].

During TPE the CVVHF was either recirculated or completelystopped. Combined extracorporeal therapy had to be started with-in 24 h after sufficient treatment of the septic focus. TPE was in-stalled first and repeated once every following day. CVVHF was

started immediately after the plasmapheresis course was complet-ed. TPE was withdrawn when hyperdynamic circulation dimin-ished at least 1.0 l/min and remained constant for more than 12 h.CVVHF was stopped when the septic situation had stabilized,with a drop in C-reactive protein values and leukocyte counts to-gether with restoration of renal function (creatinine < 2.0 mg/dl,production of urine > 100 ml/h, urea < 100 mg/dl). At the end ofeach day (12 p.m.) we calculated the scores for the revised AcutePhysiology and Chronic Health Evaluation (APACHE II), MOF,and sepsis severity [25]. The total doses of dopamine, dobutamine,adrenaline, and noradrenaline were determined every day, and thecorrelation was examined between these values and the course ofcardiocirculatory data. In the tables they are presented as meanvalues for the entire group at the start, 48 h after therapy onset,and when therapy ended.

For statistical calculation of significance the c2 test and Fisher'sexact test were used; significance was set at P < 0.05. Computersoftware was SPSS for Windows. Statistical tests were carried outby a specialized consultant in charge of evaluating ongoing clinicaland experimental studies in the Department of Surgery. The studywas approved by the Institutional Review Board. All relatives ofthe patients concerned were contacted during the ongoing therapy,and written consent was obtained. Data were evaluated accordingto the guidelines for good clinical practice.

Results

Morbidity regarding heart, renal, pulmonary, or hepaticdysfunction was similar in the two groups and did notreach statistical difference. Initial APACHE II scorewas also similar (26.6 vs. 27.3 in the control group;n. s.). In the therapy group 12 of 19 patients had postop-erative peritonitis, whereas in the control group therewere only 9 of 24 patients. The underlying disease wasmalignant in 8 patients in the therapy group and in 11patients in the control group (Table 1). The two groupsalso did not seem to differ in terms of diagnosis (abdom-

533

Table 1 Demographic data at the start of the study

Treatmentgroup

Non-treatmentgroup

n 19 24

Age(yrs) 60.3 64.2

Female/Male 7/12 10/14

Underlying diseasemalignant 8 11non-malignant 11 13

ScoresAPACHE II 27.32 26.59 (n. s.)SSS 11.61 10.77 (n. s.)MOF 6.61 5.58 (p < 0.001)

Prevalence organ failurecardiac 9/19 10/24pulmonary 6/19 3/24renal 3/19 2/24hepatic 2/19 3/24others 6/19 3/24

inal, chest, trauma, vascular, benign and malignant), al-though statistical significance could not be calculatedbecause of the small sample sizes. The two groups alsoshowed no differences in the rate or in the type of organ-ic failure; hepatic failure was only 12.5 % in the controlgroup (vs. 10.5 %; n. s.).

Regarding intravenous catecholamines, the quantityof adrenaline and noradrenaline administered in thetwo groups was equivalent. On the other hand, dopam-ine administration was 40.2 mg/h in the control groupand 44.8 mg/h in the therapy group after 48 h of treat-ment (n. s.). Dobutamine presented a more pronounceddifference with 16.5 mg/h in controls against 29.7 mg/hafter 48 h of therapy onset (P < 0.001). This variancewas present only within the first 48 h of therapy. Laterthe values were not statistically different between the

groups (Fig.1). Systemic vascular resistance (SVR) was580 dyne s±1 cm±5 in the control group and 560 dy-ne s±1 cm±5 in the therapy group 48 h after start of thetherapy presenting (n. s.; Fig. 2). Mean cardiac index(CI) was 4.3 l min±1 m±2 in controls and 5.9 l min±1 m±2

in the therapy group during the first 48 h (Fig.3). Thedifference was statistically relevant (P < 0.01) at thestart and after 48 h of treatment. Nonsurvivors in bothgroups had a mean CI between 3.7 and 3.9 l min±1 m±2

during the first 48 h. In all survivors with extracorporealdetoxification, hypercirculation was reduced by at least2.0 l/min during the first 96 h.

Duration of hepatic failure was identical in the twogroups (6.2 days). Duration of pulmonary failure was15.8 days in controls and 17.7 days in the therapy group(n. s.). Renal malfunction was reduced by a mean of

534

Fig.1 Inotropic drug doses atthe start, 48 h after the begin-ning, and at the end of therapyin both treatment groups (meanvalues). The difference in dosesfor dobutamine was statisticallysignificant between the groupsat the start and after 48 h oftreatment. The differencesfound for dopamine were notsignificant

Fig.2 Systemic vascular resis-tance values during the courseof therapy. There were no sta-tistically significant changes inSVR in the treated group

6.6 days by TPE/CVVHF (P < 0.05). The duration of re-nal failure was 10.6 days in the control group vs. 4.0 daysin the therapy group. The duration of organ failure re-flected the rate at the start of the study. Additional or-gan failures appearing with the ongoing therapy werenot considered for calculation. The mean duration ofCVVHF was 10.2 days and 6.4 days for TPE.

Overall mortality was 44.2 %. Patients with TPE/CVVHF presented a lower mortality (8/19 vs. 11/24;n. s.). The relationship between mortality and age was

identical in the two groups. Detoxification had a statisti-cally significant effect on the rate of organic failure andmortality in patients with single- (25 % vs. 0%) or dou-ble- (42 % vs. 17 %) organ failure (P < 0.0001) com-pared to the untreated group. Comparing the subgroupsof patients with single- and double-organ failure to thecontrol group showed a trend to a positive effect ofTPE/CVVHF on outcome (Table 2).

APACHE II, MOF, and sepsis severity scores did notdiffer between the two groups (n. s.) either at the begin-ning of therapy or at the end. When patients' APACHEII scores were subdivided into smaller groups (in stepsof five score points), the correlation with mortalityamong controls did not differ from the average valuesreported in the literature for such subjects (Table 3).On the other hand, patients receiving TPE/CVVHFand having an APACHE II score of 21±25 showed sta-tistically lower mortality (17 % vs. 40%) than compara-ble cases reported in the literature (P < 0.05).

In conclusion, TPE in combination with CVVHF im-proved the clinical course of surgical patients with sepsisin single- and double-organ failure and resulted in anoverall reduction in mortality of 28%. In septic patientswith double-organic failure (renal, pulmonary) the de-crease was as high as 25 %.

Discussion

There are only few case reports dealing with plasma-pheresis and hemofiltration as a combined therapy insepsis or septic shock in surgical patients [16, 17, 18].No substantial data are presently available on indica-tions for this invasive extracorporeal therapy.

This prospective pilot study examined patients withsepsis in a surgical intensive care unit. Results werecompared to historical data on patients receiving treat-

535

Fig.3 Cardiac index (CI) dur-ing the course of therapy. Therewas a statistically significantdifference between treatmentand control groups to 48 h afteronset of TPE/CVVHF(P < 0.01)

Table 2 The table presents the correlation between organic failurerate and mortality in both groups. Single and double organ failureshow a significantly lower mortality in the therapy group(p < 0.0001), whereas patients with triple and higher organic fail-ure rates presented no statistical differences

Mortalityuntreated

MortalityTPE/CVVHF

Single organ failure 1/4 (25%) 0/4 (0 %) p < 0.0001Double organ failure 5/12 (42%) 1/6 (17%) p < 0.00013 organ failure 3/6 (50%) 6/8 (75%) n. s.4 organ failure 2/2 (100%) 1/1 (100 %) n. s.

Total 11/24 (45.8%) 8/19 (42.1%) n. s.

Table 3 The table shows the comparison of the mortality rates insubgroups of APACHE II scores (in steps of 5 points) in treatedand untreated patients. Statistical relevancy is pointed out. Datataken from literature are additionally listed

APACHE II Mortalityliterature

Mortalityuntreated

MortalityTPE/CVVHF

11±20 26% 25% (1/4) 0% (0/0)21±25 40% 33% (2/6)* 17% (1/6)* p < 0.05*26±30 68% 43% (3/7)** 33% (2/6)** n. s.**31±35 79% 67% (4/6)*** 67% (4/6)*** n. s.***> 35 100% 100% (1/1) 100% (1/1)

ment earlier at the same institution. After performingan extensive literature search, the study was started inthe lack of sufficient retrievable information regardingcombined extracorporeal detoxification under septicconditions.

The molecular weight of most proinflammatory cyto-kines in sepsis, especially that of high molecular pro-tease/inhibitor complexes, is higher than 20 kDa, andplasmapheresis should therefore be effective in plas-matic clearance of these mediators. Stegmayr [26] hasdemonstrated that endotoxin and interleukin-6 areeliminated by TPE. Gardlund et al. [11] took samplesof interleukin-1b and TNF-a as well and only found asignificant decrease for TNF after TPE. McClellandet al. [27] showed that with plasmapheresis endotoxinas well as TNF and immunosuppressive prostaglandinsare cleared by TPE whereas only a decrease in endotox-in can be measured. Haupt et al. [28] demonstrated thatpouring separated plasma on human endothelial celllines elicits the same response in the form of secretionof interleukin-6 and prostaglandin E2 as with plasma ofthe same patient. This observation led to the conclusionthat plasmapheresis may eliminate mediators of sepsis.

However, not only promising results have been re-ported in the literature. Natanson et al. [3] observed mor-tality of 80±100 % in dogs with septic shock and plasma-pheresis. Pilz and Werdan [29] studied ten patients withTPE and found total mortality to be raised to 80%.

Our own data from patients with acute necrotizingpancreatitis demonstrate that by performing TPE therelationship between pro- and anti-inflammatory medi-ators may be regulated down to a more normal plasmat-ic level, so that outcome is significantly improved [14].The positive effect of high-flow hemofiltration on pa-tients with septic MOF has been demonstrated by sever-al authors [2, 9, 15]. Therefore it certainly seems rationalto combine TPE and CVVHF in cases of surgical sepsis.

One-third of our patients showed no prior organiccomorbidity, one-third had several other conditions,and 11 % only one. Prognosis worsened with the preva-lence of organic failure. This observation has also beenreported by Volk et al. [5]. The effect of organic morbid-ity prior to the septic shock may completely conceal apossibly positive effect on outcome. Unfortunately,Volk et al. did not analyze the relationship between pri-or dysfunction and mortality.

Mok and Butt [30] reported that TPE improved thecardiocirculatory situation in six of seven children withmeningococcal sepsis. The mean dose of catechol-amines, particularly that of dobutamine, was significant-ly higher and SVR significantly lower in our therapygroup than in the control group. This is in accord withthe results of Natanson et al. [3], who observed experi-mental cardiac depression after TPE in septic dogs.

It was not surprising to us that mortality was higherin patients with renal failure at the beginning of extra-

corporeal detoxification. This may be because 50% ofthese patients initially had triple-organ failure, as op-posed to 35% in controls. This result is partially consis-tent with the observations of Sakellariou [12], whofound the highest mortality in patients with MOF in sep-tic shock and TPE. On the other hand, Stegmayr [26] re-ported that seven patients with MOF had their renalfunction restored after TPE and survived septic shockfrom meningococcal sepsis.

In our study there was no effect of the combineddetoxification seen in cases of hepatic and pulmonaryfailure. Brunner et al. [31] observed that TPE repre-sents an ideal therapeutic concept in isolated hepaticfailure. On the other hand, they also demonstrated thatthe more often TPE is used, the more quickly patientsacquire acute respiratory distress syndrome, with fataloutcome. Although this was not observed in the presentstudy, our subgroup was too small for differences toreach statistical significance.

In our study TPE/CVVHF lowered mortality signifi-cantly in cases of single- and double-organ failure,whereas no effect was seen when more than two organsfailed. This finding points out that a late start of therapyfaces a progressive septic situation together with an in-creasing immunological paralysis and may thereforenot be of any use to the patient. These results are consis-tent with the observations of Stegmayr [26], who report-ed 80 % mortality in patients with triple-organ distress.Also Reeves and Butt [17] have reported 78% mortalityin triple-organ failure and 100% when more organsmalfunctioned [17].

The additional therapy did not significantly alterAPACHE II scores in either group. This finding is con-sistent with those of Volk et al. [5], who found similarscore values in the two therapy groups in their study.Pilz and Werdan [29] noted a decrease in APACHE IIscore from 28 to 22 after TPE. In our trial only patientswith scores between 21 and 25 had significantly lowermortality after TPE/CVVHF. This is consistent withthe finding that patients with APACHE II scores above25 normally suffer an imminent progression of their sep-tic situation, so that any additive therapy would be toolate.

It is extremely important to define at what timedetoxification therapy should be installed, and for howlong it should be performed. In this study all patients inthe therapy group were treated within 24 h after onsetof the septic disease. Treatment lasted a mean of6.4 days for TPE and 10.2 days for hemofiltration. Dur-ing a follow-up of 9±12 months we encountered no in-stances of acute therapy-related complications such asallergy and long-term side effects of massive plasmaticsubstitution (such as hepatitis, cytomegalovirus infec-tion, and acquired immunodeficiency syndrome).

536

Conclusion

Our findings demonstrate that TPE/CVVHF has a posi-tive effect on outcome in surgically treated septic pa-tients with double-organ failure. This effect is especiallypronounced in patients with an initial APACHE II scorebetween 21 and 25. As this study was not a prospective

randomized trial, and subgroups were rather small, thevalidity of the results may thus be criticized as being lim-ited by the scope of an observational study against his-torical controls. Therefore we urgently recommendthat a prospective randomized multicenter trial shouldbe carried out which can produce results that achievestatistical significance.

537

References

1. Atonelli M, Vivino G, Corradini S,Cannata F, Russo A, Stefanuti C (1992)Are extracorporeal techniques a valu-able tool for removal of mediators inseptic patients? Artif Organs 16: 636

2. Groeneveld ABJ (1990) Septic shockand multiple organ failure treatmentwith haemofiltration? Intensive CareMed 16: 489±490

3. Natanson C, Hoffmann WD, Koev LA,Dolan DP, Banks SM, Bacher J, DannerRL, Klein HG, Parillo JE (1993) Plas-ma exchange does not improve survivalin a canine model of human septicshock. Transfusion 33: 243±248

4. Storck M, Hartl WH, Inthorn D (1991)Septic shock and multiple organ failuretreatment with haemofiltration? Inten-sive Care Med 17: 372±373

5. Volk HD, Reinke P, Krausch D, Zuc-kermann H, Asadullah K, Müller JM,Döcke WD, Kox WJ (1996) Monocytedeactivation ± rationale for a new thera-peutic strategy in sepsis. Intensive CareMed 22: 474±481

6. Nieter B, Sisova S, Klug C, Newie T,Manger T, Thieme M, Döcke W, VolkHD, Precht K, Zuckermann H, Kühn F(1993) Plasmaaustausch bei septisch-toxischen Krankheitsbildern. NierenHochdruckkrankheiten 20: 155±159

7. Beutler B, Cerami A (1989) Tumor ne-crosis factor, cachexia, shock and in-flammation: a common mediator.Annu Rev Biochem 57: 505±518

8. Lynn WA, Cohen J (1995) Adjunctivetherapy for septic shock: a review of ex-perimental approaches. Clin Infect Dis20: 143±158

9. Sold M (1992) Hämofiltration beimMultiorganversagen. Anasthesiol In-tensivmed Notfallmed Schmerzther 33:349±353

10. Friedland JS, Porter JC, Daryanani S,Bland JM, Screaton NJ, Vesely MJJ,Griffin GE, Bennet ED, Remick DG(1996) Plasma proinflammatory cyto-kine concentrations, APACHE IIIscores and survival in patients in an in-tensive care unit. Crit Care Med 24:1775±1780

11. Gardlund B, Sjölin J, Nilsson A, RollM, Wickerts CJ, Wretlind B (1995) Plas-ma levels of cytokines in primary septicshock in humans: correlation with dis-ease severity. J Infect Dis 172: 296±301

12. Sakellariou G (1994) Plasmapheresis asa therapy in specific forms of acute re-nal failure. Nephrol Dial Transplant 1:210±218

13. Waage A, Halstensen A, Espevik T(1987) Association between tumor ne-crosis factor in serum and fatal outcomein patients with meningococcal disease.Lancet: 355±357

14. Zirngibl H, Schmidt J, Mann S, Agha A,Mann U, Messmann H, Grüne S, Hol-stege A, Schölmerich J (1997) Plasma-pheresis in acute necrotizing pancreatit-is ± a new therapeutical approach. Gas-troenterology 112: 498

15. Gotloib L, Shostak A, Lev A, Fudin R,Jaichenko J (1995) Treatment of surgi-cal and non-surgical septic multiorganfailure with bicarbonate hemodialysisand sequential hemofiltration. Inten-sive Care Med 21: 104±111

16. Barzilay E, Kessler D, Berlot G, GulloA, Geber D, Zeev IB (1989) Use of ex-tracorporeal techniques as additionaltreatment for septic induced multipleorgan failure patients. Crit Care Med17: 634±637

17. Reeves JH, Butt WW (1995) Blood fil-tration in children with severe sepsis:safe adjunctive therapy. Intensive CareMed 21: 500±504

18. Pollack M (1992) Editorial response:blood exchange and plasmapheresis insepsis and septic shock. Clin Infect Dis15: 431±433

19. ACCP/SCCM Members of the Ameri-can College of Chest Physicians/Societyof Critical Care Medicine ConsensusConference Committee (1992) Defini-tions of sepsis and organ failure andguidelines for the use of innovativetherapies in sepsis. Crit Care Med 20/6:864±874

20. Schramm JC, Neveling D, Muhr G(1990) Prognostische Parameter desakuten Lungenversagens (ARDS) beipolytraumatischen Patienten. Unfall-chirurg 93: 573±577

21. List WF, Kulier A, Kiesling A, Semu J(1990) Hämofiltration bei akutem Nie-renversagen. Erfahrungen einer operat-iven Intensivstation. Anasthesist 39:540±546

22. Vianello A, Collodel L, Calconi G, Teo-dori T, da Porto A (1994) Analysis ofthe assumption of a linear decline of re-ciprocal serum creatinine over time dur-ing chronic kidney graft failure. J Ne-phrol 7: 222±228

23. Zauner C, Grimm G, Kranz A, Madl C,Schneeweiss B, Lenz K (1990) Das Mul-tiorganversagen auf einer internisti-schen Intensivstation. Ursachen, In-zidenz ± Prognose. Intensivmed Semin:29±32

24. List WF, Kulier A, Kiesling A, Semu J(1990) Hämofiltration bei akutem Nie-renversagen. Erfahrungen einer operat-iven Intensivstation. Anasthesist 39:540±546

25. Wisner DH (1992) History and currentstatus of scoring systems for criticalcare. Arch Surg 127: 352±356

26. Stegmayr BG (1996) Plasmapheresis insevere sepsis or septic shock. Blood Pu-rif 14: 94±101

27. McClelland P, Williams PS, Yaqoob M,Mostafa SM, Bone M (1990) Multipleorgan failure ± a role for plasma ex-change? Intensive Care Med 16:100±103

28. Haupt W, Fritzsche H, Hohenberger W,Zirngibl H (1996) Selective cytokine re-lease induced by serum and separatedplasma from septic patients. Eur J Surg162: 769±776

29. Pilz G, Werdan K (1990) Cardiovascu-lar parameters and scoring systems inthe evaluation of response to therapyin sepsis and septic shock. Infection 18:253±262

30. Mok Q, Butt W (1996) The outcome ofchildren admitted to intensive carewith meningococcal septicaemia. Inten-sive Care Med 22: 259±263

31. Brunner G, Lösgen H, Schmidt FW(1980) Plasmapheresis treatment forsupport of the failing liver and otherforms of liver disease. Plasmaexchange:329