- Email: [email protected]
Does drainage of intraabdominal pus reverse multiple organ failure?
Does Drainage of Intraabdominal Pus Reverse Multiple Organ Failure? Lawrence W. Norton, MD, Tucson, Arizona
The most frequent cause of multiple organ failure is intraahdominal sepsis [1,2]. The mechanisms by which an intraperitoneal abscess can induce sequential failure of distant organs are unknown, although many explanations have been offered. Clinical evidence of multiple organ failure might include respiratory insufficiency, jaundice, renal failure, coma, generalized bleeding, or stress ulceration. Organs fail at unpredictable intervals and in variable sequence. The reported m o r t ~ i t y rates of sepsisinduced multiple organ failure have ranged between 45 and 90 percent [1,3]. In this report, our assumption was that drainage of pus would seem to be the best means of reversing multiple organ failure in the patient with an intraabdominal abscess. Persistence or recurrence of an abscess after drainage would explain continued organ failure and eventual death. Although detection of intraabdominal pus was once difficult, diagnosis is now facilitated by the use of ultrasonography and computerized tomographic scanning. A combination of prompt detection of an intraperitone~l abscess with early and, if necessary, repeated drainage should reduce mortality in patients with multiple organ failure. To evaluate this assumption, the outcomes in patients with multiple organ failure who underwent laparotomy for drainage of an intraabdominal abscess have been reviewed retrospectively. Material and Methods
During 6 years (1977 through 1982) in which ultrasonography and computerized tomographic scanning were used routinely to detect intraperitoneal pus, 21 patients with sepsis-induced multiple organ failure underwent one or more laparotomies for drainage of an intraabdominal abscess with the hope of reversing multiple organ failure. The ages of these patients ranged from 3 months to 78 years with a mean age of 47 years. Fourteen of the 21 patients (67 percent) were males. Spontaneous perforation of the gastrointestinal tract occurred in 11 patients or 52 percent (appendicitis four patients, diverticulitis 3 patients, perforated peptic ulcer From the Department of S~gen/. University of Arizona Heelth Sciences Center. Tucson, Arizona. Requests for reprints should be addressedto Lawrence W. Norton, MD. ~ of Surgery, U n ~ " of CokxadoHee~ ~ Cereer, 4200 Ea3t NinthAvenue, Denver, ~ 80262.
Vol~',~ 149, March 11115
3 patients, and spontaneous bowel perforation I patient) Only one had traumatic bowel perforation. Severe pan creatitis, masenteric vascular occlusion, and anastomoti( leak caused an abscess in two patients each. One patienl each had perinephritic abscess and an abscess after cesarian section. The single patient with an undetermined cause was undergoing leukocyte transfusion for leukemia. Fourteen patients (67 percent) entered the hospital with existing intraperitoneal infection. Sepsis developed among " the others as a result of primary disease or a postoperative complication. The most common site of abscess was the left lower quadrant (five patients). The right upper quadrant, the right lower quadrant, and the lesser sac were the next most common locations (three patients each), followed by subphrenic spaces, the pelvis, and the infrahepatic space (two patients each) and the perinephric space (one patient). The principal abscess was found to be in a different location during second laparotomy for drainage in six patients. In these instances, the frequency of left lower quadrant abscess was lower, whereas that of subphrenic abscess was higher. Multiple abscess sites were discoveredduring hfitial exploration for drainage in only one patient but were eventually confirmed in 1I patients (52 percent). Most of these patients had pus in only two locations, but one had an abscess in three sites. The volume of the initial abscess ranged from 10 to 2,000 ml. The mean volume of initial abscesses was 75 ml and that of recurrent abscesses in the same site, 65 ml. Of microorganisms recovered from the principal abscess during the initial drainage procedtrre, Escherichia coli was the most frequently cultured coliform bacteria (71 percent of the specimens). Pseudomonas and Klebsiella were also recovered frequently (52 and 38 percent, respectively). Bacteroides species, notably B. fragilis, were recovered from six of the specimens (29 percent). Beta-hemolytic streptococci were cultured from abscesses in two patients (10 percent) and were the only bacteria recovered in one of them. A variety of antibiotics were employed to treat intraabdominal infection; however, a second generation cephalosporin was most often prescribed. Four patients had only one operation for drab'rage of an intraperitoneal abscess, and 17 (81 percent) underwent at least two laparotomies for drainage. Five of these patients had three explorations, four had four, two had five, and one had seven. The mean number of operations for drainage among these ].7 patients was 3.4 per patient. Including other nondrainage operations during the same hospitalization, the mean number of total operations was 4.2 per patient. The time interval between the first and second
347
Norton
TABLE I
Criteria for Determining Failure of an Organ or System
Organ or System Kidney Lung Liver
Brain Stomach (mucosa)
Coagulation
Criteria of Failure Renal dialysis on at least one occasion for elevated serum creatlnino level Hypoxia requiring resp'ratorassisted ventilation Serum billrubln level greater t ~ n 3 mg/dl; serum glutamtc oxaiacetic transaminase level twice normal value Coma Upper gastrointestirml bleeding due to acute gastric ulcerations requiring transfusion of blood Laboratory evidence of dissemln.~ted intravascular coagulation
operations for drainage ranged from 1 to 42 days with a mean interval of 10 days. Intervals between the second and third operations in 12 patients ranged from 3 to 35 days with a mean interval of 11 days. In every patient, drainage of intraabdominal abscess was accomplished by midline laparotomy, aspiration of pus, local debridement of the abscess site, irrigation with saline solution, placement of Penrose or sump drains or both, and closure of the fascia but not the skin. All 21 patients had failure of two or more organ systems associated with intraabdominal sepsis. Criteria for determining organ system failure are shown in Table L Two organs failed in 14 patients (67 percent), and three or more organs failed in the remaining seven patients. Renal, pulmonary, and hepatic failure were most common, occurring in about two thirds of the patients (Table II). Coma, stress ulcer bleeding, and coagulation defects occurred less often. The time interval between the event that catmed sepsis and failure of the first organ ranged between 2 m~d 30 days with a mean of 8 days, the interval between failure of the first and second organs ranged from I to 15 days with a mean of 3.8 days, and the interval between failure of the second and third organs was an average 5 days. The first organ to fail was the kidney in 43 percent of the patients, the liver in 29 percent, the lung in 23 percent, and the gastric mucosa in 5 percent. Results
Sixteen of the 21 patients (76 percent) died with multiple organ failure despite one or m o r e operative p r o c e d u r e s for drainage of an i n t r a a b d o m i n a l abscess. D e a t h occurred an average of 8 days a f t e r the last procedure. Drainage of pus i m p r o v e d organ function t e m p o r a r i l y in only one of the patients who died. In this p a t i e n t who h a d a p e r f o r a t e d d u o d e n a l ulcer, dialysis for renal failure was discontinued after drainage of a right u p p e r q u a d r a n t abscess. W h e n the abscess recurred a t the s a m e site I m o n t h later, di-
348
TABLE II
Organ or System Failure Associated With Intraabdomlnal Abscess
Organ or System Failure
n
%
Kidney Lung Liver Brain Stomach Coagulatlon
t4 14 12 4 2 1
67 67 57 19 10 5
Patients
......
_
alysis was again required a n d was m a i n t a i n e d until death, which occurred despite additional drainage attempts. T h e m e a n age of the surviving p a t i e n t s (40 years) was less t h a n t h a t of those who died (49 years). Survivors u n d e r w e n t an average of 3.2 drainage operations c o m p a r e d with 3.5 for nonsurvivors. T h e n u m b e r , location, a n d bacteriologic characteristics of abscesses a m o n g survivors did not a p p e a r to be more favorable t h a n a m o n g the nonsurvivors. Small abscess size correlated negatively with survival. Abscess volume averaged nearly 25 ml more in survivors t h ~ a in nonsurvivors. Survivors a p p e a r e d to have no a d v a n t a g e in t e r m s of nutrition, incidence of h y p o tension due to sepsis, or history of chronic alcoholism. In only 3 of the 16 p a t i e n t s who died did a u t o p s y reveal a persistent or r e c u r r e n t i n t r a a b d o m i n a l abscess t h a t h a d b e e n o v e r l o o k e d clinically. One h a d i n f r a h e p a t i c and lesser sac collections of pus after undergoing three l a p a r o t o m i e s for drainage of pelvic pus, a n o t h e r had a p e r s i s t e n t abscess in the right u p p e r q u a d r a n t of the a b d o m e n despite three efforts to drain it, and a t h i r d died with i n f r a h e p a t i c a n d i n t r a h e p a t i c abscesses l w e e k a f t e r drainage of the s a m e sites. Survivors h a d failure o f a n average of 2.1 organs or systems c o m p a r e d with 2.4 for nonsurvivors. P u l m o n a r y insufficiency occurred in each of the five survivors. Dependence on ventilatory s u p p o r t ended an average of 7 days after the last operation to drain pus. Only two of the five h a d c o m c o m i t a n t p u l m o nary a n d renal failure. T h i s combination occurred in 15 (71 percent) of the nonsurvivors. T h e m e a n t i m e interval between detection o f the first i n t r a a b d o m i n a l abscess a n d failure of the first organ was significantly longer a m o n g the survivors (13 days) t h a n a m o n g t h e n o n s u r v i v o r s (5.6 days). Survivors, however, h a d a s h o r t e r interval between the first a n d second organ failure (1.4 days) t h a n did nonsurvivors (4.5 days). T h e m e a n d u r a t i o n of multiple organ failure was m u c h longer a m o n g the nonsurvivors (2:3 days) t h a n a m o n g the survivors (8 days). T h e longest d u r a t i o n o f multiple organ failure
The An'~"l~n Journal o~ ,~r~r,/
Intraabdomlnal Drainage for Multiple Organ Fa|l~e
was only 12 days in a survivor compared with 60 days in a nonsurvivor. Intervals between detection of an abscess by ultrasonography or computerized tomographic scanning and laparotomy for drainage were similar for survivors (1.2 day) and nonsurvivors (1.5 days). Between In-stand second explorationsfor pus, survivors averaged 15 days and nonsurvivors averaged 9.
Comments The results of this study show that drainage of an abdominal abscess reverses multiple organ failure in a disappointingly small proportion of patients. No dependable predictors of survival are identifiable. The advantage appears to lie with the younger patient in whom multiple organ failure develops relatively late after sepsis and who needs ventilatory assistance for less than ! week. Adequate drainage of all pus does not ensure that the function of failed organs willimprove. Location, size,and bacteriologic characteristicsof an abscess do not appear to influence outcome. These findings substantiate the serious threat of intraperitoneal sepsis. A m o n g the determinants of death in patients with intraabdominal abscess, Fry et al [4] identified organ failureas the most lethal. Single organ failurewas associated with a fatal outcome in 23 percent of patient~, failureof two organ systems was associated with a 53 percent mortality, and failure of three organs, with a 79 percent mortality.Another determinant of death in Fry's study was recurrent or persistent abscess. The increased mortality rate under these circumstances was attributed to the additive effect of repeated major operative procedures, insufficient immunologic competence, or inadequate initialdrainage. In contrast to this conclusion, survivors in our series had nearly as m a n y explorations for persistentor recurrent abscess as nonsurvivors. Furthermore, of four patients drained adequately on only one occasion,all died from irreversiblemultiple organ failure.Fry [4] reported other factors associated with death in patients with intrabdominal abscess to be lesser sac abscess, m~tiple abscesses, and subhepatic abscess. A m o n g our patients, the only two with a principal abscess in the subhepatic space survived, and one of the three with a lessersac abscess recovered. Since the earliest descriptions of the multiple organ failuresyndrome and its frequent association with intraabdominal sepsis,the view has been perpetuated that rapid and adequate drainage of pus will increase survival.There is only scant evidence that this istrue. P o ~ and Shields [5] attributedthe death of three of six patientswith refractorymultiple organ failure and intraperitoneal abscess to incomplete drainage and continuing infection.In referringto this experience, Fry noted that the time-honored principlesof drainage and debridern,nt m a y be the only
Volume14g,MarchlgSS
hope for patients with multiple organ failure to reverse a progressive decline in organ system function. What is adequate drainage of an intraabdominal abscess? In this series, drainage was accomplished by the traditional method of laparotomy, aspiration, irrigation, and placement of drains. This approach would be challenged by those advocating radical surgical debridement [6], postoperative peritoneal lavage [7], or gravity (open) drainage [8]. The latter is reported to have salvaged seven of nine patients with severe multiple organ failure and intraabdominal sepsis [9]. An argument can be made that midline transabdominal drainage of an abscess should be avoided in patients with multiple organ failure because of the attendant risks of anesthetic complications, blood loss, and bowel injury. An alternative to open drainage is percutaneous aspiration. This technique is based on the high accuracy of computerized tomographic scanning in localizing intraabdominal abscesses [I0]. Experience with closed drainage suggests that 75 to 80 percent of abscesses are drained successfully and that mortality is low [11,12]. Use of percutaneous aspiration to drain intraabdominal abscesses in patients with multiple organ failure has not been reported. The role of antibiotics in reducing mortality from multiple organ failure was not evaluated in this study. Most patients received antibiotics throughout the course of sepsis. Polk and Shields [5] have raised the point that prolonged antibiotic therapy may predispose patients to infection by allowing development of resistant organisms or, more importantly, may obscure diagnosis. Terminating antibiotic treatment promptly might unmask an abscess that was not suspected clinically. If early and repeated drainage will not reverse multiple organ failure in a majority of patients with sepsis, what alternative therapy is available? None exists at present, but approaches might be suggested by a review of various explanations of the multiple organ failure syndrome. Endotoxin has been shown to have an inhibitory effect on cellular function in organs such as the liver but has not been implicated as a cause of multiple organ failure [13]. Recent experimental evidence suggests that redllced oxygen consumption in the liver in the presence of peritoneal infection is due to altered nutrient blood flow [14]. Eiseman et al [15] postulate that multiple organ failure is produced by formation of immune complexes in response to infection with deposition of the complexes on the liver, lungs, and kidneys. The immune complexes then trap macrophages, which can directly damage the endothelium. A popular theory of the pathogenesis of multiple organ failure is that of Saba et al [16], which holds that opsonic fibronectin deficiency in septic patients
349
Norton
allows local action on the microcirculation by bacteria, immune complexes, endotoxin, ~ibrin microaggregates, and other debris. Reversal of opsonic deficiency by administration of cryoprecipitate in septic patients has resulted in improvement of organ function in some [17]. A new hypothesis explains the development of multiple organ failure during sepsis on the basis of accelerated proteolysis of muscle with depletion of protein, deterioration of cellular structure, insufficient production of acute phase reactants, and reduced synthesis of other necessary proteins [18]. Another recent study has postulated a hepatic energy deficit associated with a decreasing blood to ketone body ratio as the metabolic basis of postoperative failure [I9]. None of the theories of the cause of multiple organ failure provides a firm foundation for therapy. Until better understanding occurs, effortsto reverse multiple organ failure in patients with intraabdominal abscess will remain largelysurgical,that is,drainage of pus, whether by laparotomy or percutaneous aspiration. This study emphasizes the futilityof such drainage in most patients.
Summary Intraabdominal abscess induced multiple organ failure in 21 patients. Sepsis was most often due to perforation of the colon and was located with almost equal frequency in the upper and lower abdomen. Four patients died after single laparotomy for drainage. Seventeen were drained operatively more than once (average 3.4 operations) at mean intervals of 10 days. Sixteen of the 21 patients (76 percent) died with multiple organ failure despite drainage. Organ function improved temporarily in only one patient. Autopsy showed that pus had persisted or recurred in three patients. No significant predictors of survival were identified, although the advantage appears to lie with the younger patient in whom multiple organ failure develops relatively late after sepsis (mean 13 days) and who needs ventilatory assistance for less than i week. The location, size, and bacteriologic characteristics of abscesses do not appear to influence outcome. This study shows that early and repeated drainage of intraabdominal pus will not reverse multiple organ failure in the majority of patients.
Acknowledgment: I acknowledge the capable assistance of Elizabeth Criss, RN in reviewing patient records.
References 1. EtsernanB, Beart R, Norton L. Multiple organ failure. Surg Gynecol Obstet 1977;14:323-6. 2. Fry DE, Pearlstein L, Fulton RL, Polk HC. Multiple system organ failure. Arch Surg 1980;115:136-40. 3. StephenM, LoewenthalJ. GeneralizedInfective peritonitis,Surg Gynecol Obstet 1978;147:231-4. 4. Fry DE, Garrison RN, Heltsch RC, Calhoun K, Polk HC. Deter-mlnants of death in patients with Intraabdomlnal abscess. Surgery 1980;88:517-23. 5. PolkHC, Shields CL. Remoteorgan failure: a valid sign of occult Intraabdomlnal infection. Surgery 1977;81:310-3. 6. Hudspeth AS. Radical surgical debrldement in the treatment of advanced generalized peritonitis. Arch Surg 1975;110: 233-6. 7. Stephen M, Lowenthal J. Continuing peritoneal lavage in high-risk peritonitis. Surgery 1979;85:603-6. 8. Duff J, Moffat J. Abdominal sepsis managed by leaving the abdomen open, Surgery 1981;90:775-7. 9. Vasquez JM, Speare OR. Gravity drainage In the treatment of patients with near fatal rectrrent Intraabdomlnal sepsis, Miltt Mad 1983;148:597-9. 10. Satnl S, Kellum JM, O'Leary MP, et al. Improved localization and survival lq patients with Intraabdomtnal abscesses. Am J Surg 1983;145:136-42. 11. Percutaneouscatheter ~alnage of abdominal abscesses guided by ultrasound and computed tomogrephy. Am J Radtol 1979;133:1-8. 12. MandelSR, Boyd D, Jaques PF, Mandell V, Staab EV. Drainage of epatlc, Intraabdomtnaland medlastinal abscesses guided by computerized axial tomography. Am J Surg 1983;145: 120-5. 13. Harken AH, LIIIoRS, HufnagelHV. Direct influence of endotoxln on cellular respiration. Surg Gynecol Obstet 1975;140: 858-60. 14. C-~lson RN, Ratcllffe DJ, Fry DE. Hepatocellular function and nutrient blood flow in experimental peritonitis. Surgery 1982;92:713-9. 15. Elseman B, Sloan R, HansbroughJ, Molntosh R. Multiple organ failure: clinical and experimental Am Surg 1980;46:149. 16. Saba TM, McC,afferty MH, Lansey ME. Depressed retlculoendothelial function in the surgical patient. Infect Surg 1983;2:124-33. 17. Scov|ll WA, Saba TM, Blumenstock FA, Bernard H, Powers SR Jr. Opsonic a surface binding glycoprotetn therapy during sepsis. Ann $urg 1976;188:521-9. 18. Clowes Jr GH, George BC, VIIlee CA, Saravis CA. Muscle proteoly,~ls induced by a circulatlnFj p,aptldein patients with sepsis or trauma. N Engl J Mad 1383;308:545-52. 19. Ozawa K, Aoyama H, Yasuda K, et al. Metabolic abnormalities associated with postoperative organ failure. Arch Surg 1983;118:1245-51.
See page 420 for o related editorial comment.
350
The AmeriCanJoumDI of Surgory
The most frequent cause of multiple organ failure is intraahdominal sepsis [1,2]. The mechanisms by which an intraperitoneal abscess can induce sequential failure of distant organs are unknown, although many explanations have been offered. Clinical evidence of multiple organ failure might include respiratory insufficiency, jaundice, renal failure, coma, generalized bleeding, or stress ulceration. Organs fail at unpredictable intervals and in variable sequence. The reported m o r t ~ i t y rates of sepsisinduced multiple organ failure have ranged between 45 and 90 percent [1,3]. In this report, our assumption was that drainage of pus would seem to be the best means of reversing multiple organ failure in the patient with an intraabdominal abscess. Persistence or recurrence of an abscess after drainage would explain continued organ failure and eventual death. Although detection of intraabdominal pus was once difficult, diagnosis is now facilitated by the use of ultrasonography and computerized tomographic scanning. A combination of prompt detection of an intraperitone~l abscess with early and, if necessary, repeated drainage should reduce mortality in patients with multiple organ failure. To evaluate this assumption, the outcomes in patients with multiple organ failure who underwent laparotomy for drainage of an intraabdominal abscess have been reviewed retrospectively. Material and Methods
During 6 years (1977 through 1982) in which ultrasonography and computerized tomographic scanning were used routinely to detect intraperitoneal pus, 21 patients with sepsis-induced multiple organ failure underwent one or more laparotomies for drainage of an intraabdominal abscess with the hope of reversing multiple organ failure. The ages of these patients ranged from 3 months to 78 years with a mean age of 47 years. Fourteen of the 21 patients (67 percent) were males. Spontaneous perforation of the gastrointestinal tract occurred in 11 patients or 52 percent (appendicitis four patients, diverticulitis 3 patients, perforated peptic ulcer From the Department of S~gen/. University of Arizona Heelth Sciences Center. Tucson, Arizona. Requests for reprints should be addressedto Lawrence W. Norton, MD. ~ of Surgery, U n ~ " of CokxadoHee~ ~ Cereer, 4200 Ea3t NinthAvenue, Denver, ~ 80262.
Vol~',~ 149, March 11115
3 patients, and spontaneous bowel perforation I patient) Only one had traumatic bowel perforation. Severe pan creatitis, masenteric vascular occlusion, and anastomoti( leak caused an abscess in two patients each. One patienl each had perinephritic abscess and an abscess after cesarian section. The single patient with an undetermined cause was undergoing leukocyte transfusion for leukemia. Fourteen patients (67 percent) entered the hospital with existing intraperitoneal infection. Sepsis developed among " the others as a result of primary disease or a postoperative complication. The most common site of abscess was the left lower quadrant (five patients). The right upper quadrant, the right lower quadrant, and the lesser sac were the next most common locations (three patients each), followed by subphrenic spaces, the pelvis, and the infrahepatic space (two patients each) and the perinephric space (one patient). The principal abscess was found to be in a different location during second laparotomy for drainage in six patients. In these instances, the frequency of left lower quadrant abscess was lower, whereas that of subphrenic abscess was higher. Multiple abscess sites were discoveredduring hfitial exploration for drainage in only one patient but were eventually confirmed in 1I patients (52 percent). Most of these patients had pus in only two locations, but one had an abscess in three sites. The volume of the initial abscess ranged from 10 to 2,000 ml. The mean volume of initial abscesses was 75 ml and that of recurrent abscesses in the same site, 65 ml. Of microorganisms recovered from the principal abscess during the initial drainage procedtrre, Escherichia coli was the most frequently cultured coliform bacteria (71 percent of the specimens). Pseudomonas and Klebsiella were also recovered frequently (52 and 38 percent, respectively). Bacteroides species, notably B. fragilis, were recovered from six of the specimens (29 percent). Beta-hemolytic streptococci were cultured from abscesses in two patients (10 percent) and were the only bacteria recovered in one of them. A variety of antibiotics were employed to treat intraabdominal infection; however, a second generation cephalosporin was most often prescribed. Four patients had only one operation for drab'rage of an intraperitoneal abscess, and 17 (81 percent) underwent at least two laparotomies for drainage. Five of these patients had three explorations, four had four, two had five, and one had seven. The mean number of operations for drainage among these ].7 patients was 3.4 per patient. Including other nondrainage operations during the same hospitalization, the mean number of total operations was 4.2 per patient. The time interval between the first and second
347
Norton
TABLE I
Criteria for Determining Failure of an Organ or System
Organ or System Kidney Lung Liver
Brain Stomach (mucosa)
Coagulation
Criteria of Failure Renal dialysis on at least one occasion for elevated serum creatlnino level Hypoxia requiring resp'ratorassisted ventilation Serum billrubln level greater t ~ n 3 mg/dl; serum glutamtc oxaiacetic transaminase level twice normal value Coma Upper gastrointestirml bleeding due to acute gastric ulcerations requiring transfusion of blood Laboratory evidence of dissemln.~ted intravascular coagulation
operations for drainage ranged from 1 to 42 days with a mean interval of 10 days. Intervals between the second and third operations in 12 patients ranged from 3 to 35 days with a mean interval of 11 days. In every patient, drainage of intraabdominal abscess was accomplished by midline laparotomy, aspiration of pus, local debridement of the abscess site, irrigation with saline solution, placement of Penrose or sump drains or both, and closure of the fascia but not the skin. All 21 patients had failure of two or more organ systems associated with intraabdominal sepsis. Criteria for determining organ system failure are shown in Table L Two organs failed in 14 patients (67 percent), and three or more organs failed in the remaining seven patients. Renal, pulmonary, and hepatic failure were most common, occurring in about two thirds of the patients (Table II). Coma, stress ulcer bleeding, and coagulation defects occurred less often. The time interval between the event that catmed sepsis and failure of the first organ ranged between 2 m~d 30 days with a mean of 8 days, the interval between failure of the first and second organs ranged from I to 15 days with a mean of 3.8 days, and the interval between failure of the second and third organs was an average 5 days. The first organ to fail was the kidney in 43 percent of the patients, the liver in 29 percent, the lung in 23 percent, and the gastric mucosa in 5 percent. Results
Sixteen of the 21 patients (76 percent) died with multiple organ failure despite one or m o r e operative p r o c e d u r e s for drainage of an i n t r a a b d o m i n a l abscess. D e a t h occurred an average of 8 days a f t e r the last procedure. Drainage of pus i m p r o v e d organ function t e m p o r a r i l y in only one of the patients who died. In this p a t i e n t who h a d a p e r f o r a t e d d u o d e n a l ulcer, dialysis for renal failure was discontinued after drainage of a right u p p e r q u a d r a n t abscess. W h e n the abscess recurred a t the s a m e site I m o n t h later, di-
348
TABLE II
Organ or System Failure Associated With Intraabdomlnal Abscess
Organ or System Failure
n
%
Kidney Lung Liver Brain Stomach Coagulatlon
t4 14 12 4 2 1
67 67 57 19 10 5
Patients
......
_
alysis was again required a n d was m a i n t a i n e d until death, which occurred despite additional drainage attempts. T h e m e a n age of the surviving p a t i e n t s (40 years) was less t h a n t h a t of those who died (49 years). Survivors u n d e r w e n t an average of 3.2 drainage operations c o m p a r e d with 3.5 for nonsurvivors. T h e n u m b e r , location, a n d bacteriologic characteristics of abscesses a m o n g survivors did not a p p e a r to be more favorable t h a n a m o n g the nonsurvivors. Small abscess size correlated negatively with survival. Abscess volume averaged nearly 25 ml more in survivors t h ~ a in nonsurvivors. Survivors a p p e a r e d to have no a d v a n t a g e in t e r m s of nutrition, incidence of h y p o tension due to sepsis, or history of chronic alcoholism. In only 3 of the 16 p a t i e n t s who died did a u t o p s y reveal a persistent or r e c u r r e n t i n t r a a b d o m i n a l abscess t h a t h a d b e e n o v e r l o o k e d clinically. One h a d i n f r a h e p a t i c and lesser sac collections of pus after undergoing three l a p a r o t o m i e s for drainage of pelvic pus, a n o t h e r had a p e r s i s t e n t abscess in the right u p p e r q u a d r a n t of the a b d o m e n despite three efforts to drain it, and a t h i r d died with i n f r a h e p a t i c a n d i n t r a h e p a t i c abscesses l w e e k a f t e r drainage of the s a m e sites. Survivors h a d failure o f a n average of 2.1 organs or systems c o m p a r e d with 2.4 for nonsurvivors. P u l m o n a r y insufficiency occurred in each of the five survivors. Dependence on ventilatory s u p p o r t ended an average of 7 days after the last operation to drain pus. Only two of the five h a d c o m c o m i t a n t p u l m o nary a n d renal failure. T h i s combination occurred in 15 (71 percent) of the nonsurvivors. T h e m e a n t i m e interval between detection o f the first i n t r a a b d o m i n a l abscess a n d failure of the first organ was significantly longer a m o n g the survivors (13 days) t h a n a m o n g t h e n o n s u r v i v o r s (5.6 days). Survivors, however, h a d a s h o r t e r interval between the first a n d second organ failure (1.4 days) t h a n did nonsurvivors (4.5 days). T h e m e a n d u r a t i o n of multiple organ failure was m u c h longer a m o n g the nonsurvivors (2:3 days) t h a n a m o n g the survivors (8 days). T h e longest d u r a t i o n o f multiple organ failure
The An'~"l~n Journal o~ ,~r~r,/
Intraabdomlnal Drainage for Multiple Organ Fa|l~e
was only 12 days in a survivor compared with 60 days in a nonsurvivor. Intervals between detection of an abscess by ultrasonography or computerized tomographic scanning and laparotomy for drainage were similar for survivors (1.2 day) and nonsurvivors (1.5 days). Between In-stand second explorationsfor pus, survivors averaged 15 days and nonsurvivors averaged 9.
Comments The results of this study show that drainage of an abdominal abscess reverses multiple organ failure in a disappointingly small proportion of patients. No dependable predictors of survival are identifiable. The advantage appears to lie with the younger patient in whom multiple organ failure develops relatively late after sepsis and who needs ventilatory assistance for less than ! week. Adequate drainage of all pus does not ensure that the function of failed organs willimprove. Location, size,and bacteriologic characteristicsof an abscess do not appear to influence outcome. These findings substantiate the serious threat of intraperitoneal sepsis. A m o n g the determinants of death in patients with intraabdominal abscess, Fry et al [4] identified organ failureas the most lethal. Single organ failurewas associated with a fatal outcome in 23 percent of patient~, failureof two organ systems was associated with a 53 percent mortality, and failure of three organs, with a 79 percent mortality.Another determinant of death in Fry's study was recurrent or persistent abscess. The increased mortality rate under these circumstances was attributed to the additive effect of repeated major operative procedures, insufficient immunologic competence, or inadequate initialdrainage. In contrast to this conclusion, survivors in our series had nearly as m a n y explorations for persistentor recurrent abscess as nonsurvivors. Furthermore, of four patients drained adequately on only one occasion,all died from irreversiblemultiple organ failure.Fry [4] reported other factors associated with death in patients with intrabdominal abscess to be lesser sac abscess, m~tiple abscesses, and subhepatic abscess. A m o n g our patients, the only two with a principal abscess in the subhepatic space survived, and one of the three with a lessersac abscess recovered. Since the earliest descriptions of the multiple organ failuresyndrome and its frequent association with intraabdominal sepsis,the view has been perpetuated that rapid and adequate drainage of pus will increase survival.There is only scant evidence that this istrue. P o ~ and Shields [5] attributedthe death of three of six patientswith refractorymultiple organ failure and intraperitoneal abscess to incomplete drainage and continuing infection.In referringto this experience, Fry noted that the time-honored principlesof drainage and debridern,nt m a y be the only
Volume14g,MarchlgSS
hope for patients with multiple organ failure to reverse a progressive decline in organ system function. What is adequate drainage of an intraabdominal abscess? In this series, drainage was accomplished by the traditional method of laparotomy, aspiration, irrigation, and placement of drains. This approach would be challenged by those advocating radical surgical debridement [6], postoperative peritoneal lavage [7], or gravity (open) drainage [8]. The latter is reported to have salvaged seven of nine patients with severe multiple organ failure and intraabdominal sepsis [9]. An argument can be made that midline transabdominal drainage of an abscess should be avoided in patients with multiple organ failure because of the attendant risks of anesthetic complications, blood loss, and bowel injury. An alternative to open drainage is percutaneous aspiration. This technique is based on the high accuracy of computerized tomographic scanning in localizing intraabdominal abscesses [I0]. Experience with closed drainage suggests that 75 to 80 percent of abscesses are drained successfully and that mortality is low [11,12]. Use of percutaneous aspiration to drain intraabdominal abscesses in patients with multiple organ failure has not been reported. The role of antibiotics in reducing mortality from multiple organ failure was not evaluated in this study. Most patients received antibiotics throughout the course of sepsis. Polk and Shields [5] have raised the point that prolonged antibiotic therapy may predispose patients to infection by allowing development of resistant organisms or, more importantly, may obscure diagnosis. Terminating antibiotic treatment promptly might unmask an abscess that was not suspected clinically. If early and repeated drainage will not reverse multiple organ failure in a majority of patients with sepsis, what alternative therapy is available? None exists at present, but approaches might be suggested by a review of various explanations of the multiple organ failure syndrome. Endotoxin has been shown to have an inhibitory effect on cellular function in organs such as the liver but has not been implicated as a cause of multiple organ failure [13]. Recent experimental evidence suggests that redllced oxygen consumption in the liver in the presence of peritoneal infection is due to altered nutrient blood flow [14]. Eiseman et al [15] postulate that multiple organ failure is produced by formation of immune complexes in response to infection with deposition of the complexes on the liver, lungs, and kidneys. The immune complexes then trap macrophages, which can directly damage the endothelium. A popular theory of the pathogenesis of multiple organ failure is that of Saba et al [16], which holds that opsonic fibronectin deficiency in septic patients
349
Norton
allows local action on the microcirculation by bacteria, immune complexes, endotoxin, ~ibrin microaggregates, and other debris. Reversal of opsonic deficiency by administration of cryoprecipitate in septic patients has resulted in improvement of organ function in some [17]. A new hypothesis explains the development of multiple organ failure during sepsis on the basis of accelerated proteolysis of muscle with depletion of protein, deterioration of cellular structure, insufficient production of acute phase reactants, and reduced synthesis of other necessary proteins [18]. Another recent study has postulated a hepatic energy deficit associated with a decreasing blood to ketone body ratio as the metabolic basis of postoperative failure [I9]. None of the theories of the cause of multiple organ failure provides a firm foundation for therapy. Until better understanding occurs, effortsto reverse multiple organ failure in patients with intraabdominal abscess will remain largelysurgical,that is,drainage of pus, whether by laparotomy or percutaneous aspiration. This study emphasizes the futilityof such drainage in most patients.
Summary Intraabdominal abscess induced multiple organ failure in 21 patients. Sepsis was most often due to perforation of the colon and was located with almost equal frequency in the upper and lower abdomen. Four patients died after single laparotomy for drainage. Seventeen were drained operatively more than once (average 3.4 operations) at mean intervals of 10 days. Sixteen of the 21 patients (76 percent) died with multiple organ failure despite drainage. Organ function improved temporarily in only one patient. Autopsy showed that pus had persisted or recurred in three patients. No significant predictors of survival were identified, although the advantage appears to lie with the younger patient in whom multiple organ failure develops relatively late after sepsis (mean 13 days) and who needs ventilatory assistance for less than i week. The location, size, and bacteriologic characteristics of abscesses do not appear to influence outcome. This study shows that early and repeated drainage of intraabdominal pus will not reverse multiple organ failure in the majority of patients.
Acknowledgment: I acknowledge the capable assistance of Elizabeth Criss, RN in reviewing patient records.
References 1. EtsernanB, Beart R, Norton L. Multiple organ failure. Surg Gynecol Obstet 1977;14:323-6. 2. Fry DE, Pearlstein L, Fulton RL, Polk HC. Multiple system organ failure. Arch Surg 1980;115:136-40. 3. StephenM, LoewenthalJ. GeneralizedInfective peritonitis,Surg Gynecol Obstet 1978;147:231-4. 4. Fry DE, Garrison RN, Heltsch RC, Calhoun K, Polk HC. Deter-mlnants of death in patients with Intraabdomlnal abscess. Surgery 1980;88:517-23. 5. PolkHC, Shields CL. Remoteorgan failure: a valid sign of occult Intraabdomlnal infection. Surgery 1977;81:310-3. 6. Hudspeth AS. Radical surgical debrldement in the treatment of advanced generalized peritonitis. Arch Surg 1975;110: 233-6. 7. Stephen M, Lowenthal J. Continuing peritoneal lavage in high-risk peritonitis. Surgery 1979;85:603-6. 8. Duff J, Moffat J. Abdominal sepsis managed by leaving the abdomen open, Surgery 1981;90:775-7. 9. Vasquez JM, Speare OR. Gravity drainage In the treatment of patients with near fatal rectrrent Intraabdomlnal sepsis, Miltt Mad 1983;148:597-9. 10. Satnl S, Kellum JM, O'Leary MP, et al. Improved localization and survival lq patients with Intraabdomtnal abscesses. Am J Surg 1983;145:136-42. 11. Percutaneouscatheter ~alnage of abdominal abscesses guided by ultrasound and computed tomogrephy. Am J Radtol 1979;133:1-8. 12. MandelSR, Boyd D, Jaques PF, Mandell V, Staab EV. Drainage of epatlc, Intraabdomtnaland medlastinal abscesses guided by computerized axial tomography. Am J Surg 1983;145: 120-5. 13. Harken AH, LIIIoRS, HufnagelHV. Direct influence of endotoxln on cellular respiration. Surg Gynecol Obstet 1975;140: 858-60. 14. C-~lson RN, Ratcllffe DJ, Fry DE. Hepatocellular function and nutrient blood flow in experimental peritonitis. Surgery 1982;92:713-9. 15. Elseman B, Sloan R, HansbroughJ, Molntosh R. Multiple organ failure: clinical and experimental Am Surg 1980;46:149. 16. Saba TM, McC,afferty MH, Lansey ME. Depressed retlculoendothelial function in the surgical patient. Infect Surg 1983;2:124-33. 17. Scov|ll WA, Saba TM, Blumenstock FA, Bernard H, Powers SR Jr. Opsonic a surface binding glycoprotetn therapy during sepsis. Ann $urg 1976;188:521-9. 18. Clowes Jr GH, George BC, VIIlee CA, Saravis CA. Muscle proteoly,~ls induced by a circulatlnFj p,aptldein patients with sepsis or trauma. N Engl J Mad 1383;308:545-52. 19. Ozawa K, Aoyama H, Yasuda K, et al. Metabolic abnormalities associated with postoperative organ failure. Arch Surg 1983;118:1245-51.
See page 420 for o related editorial comment.
350
The AmeriCanJoumDI of Surgory