Bacteriologic status of necrotic tissue in necrotizing pancreatitis

A contemporary series of patients undergoing open debridement for necrotizing pancreatitis

BACKGROUND

For patients with acute pancreatitis complicated by infected necrosis, minimally invasive techniques have taken hold without substantial comparison with open surgery. We present a contemporary series of open necrosectomies as a benchmark for newer techniques.

METHODS

Using a prospective database, we retrospectively identified consecutive patients undergoing debridement for necrotizing pancreatitis (2006 to 2009). The primary endpoint was in-hospital mortality.

RESULTS

Sixty-eight patients underwent debridement for pancreatic/peripancreatic necrosis. In-hospital mortality was 8.8% (n = 6). Infection (n = 43, 63%) and failure-to-thrive (n = 13, 19%) comprised the most common indications for necrosectomy. The false negative rate (FNR) for infection of percutaneous aspirate was 20.0%. Older age (P = .02), Acute Physiology and Chronic Health Evaluation II score upon admission (P = .03) or preoperatively (P < .01), preoperative intensive care unit admission (P = .01), and postoperative organ failure (P = .03) were associated with mortality.

CONCLUSIONS

Open debridement for necrotizing pancreatitis results in a low mortality, providing a useful comparator for other interventions. Given the high FNR of percutaneous aspirate, debridement should not be predicated on proven infection.

[Changes in diagnostic and therapeutic standards of acute pancreatitis in clinical practice. Epidemiologic analysis of data from a regional center of internal medicine and surgery]

Epidemiological data analysis of a tertiary (regional) medical and surgical center. Diagnostic and therapeutic standards of patients with acute pancreatitis have changed significantly in the last few decades. Progress in laboratory and imaging diagnostics and achievements in experimental research resulted in a significant modification of the guidelines related to the care of pancreatitic patients. The aim was to analyse and compare the data of patients with acute pancreatitis treated in 1996 (period I) and 2004 (period II) at the Departments of Internal Medicine and Surgery, University of Szeged, to evaluate the concordance with international guidelines during medical and surgical treatment.

RESULTS

The authors analysed the clinical data of 126 and 124 patients, respectively, with acute pancreatitis observed during the two periods. An increase in the incidence of biliary acute pancreatitis, more frequent use of antibiotics, a higher frequency of therapeutic endoscopies (papillotomy and biliary stone extraction), the general application of ultrasonography-guided fine needle aspiration and bacterial culturing in cases of suspected infected necrosis, and higher effectiveness in complex surgical and supportive management of infected necrosis cases were detected in period II.

CONCLUSION

Although most of the achievements suggested in international guidelines on medical/endoscopic and surgical treatment of acute pancreatitis have been implemented during the observation period, no significant changes in the morbidity and mortality data of patients were found.

Severe acute pancreatitis: Pathogenetic aspects and prognostic factors

Approximately 20% of patients with acute pancreatitis develop a severe disease associated with complications and high risk of mortality. The purpose of this study is to review pathogenesis and prognostic factors of severe acute pancreatitis (SAP). An extensive medline search was undertaken with focusing on pathogenesis, complications and prognostic evaluation of SAP. Cytokines and other inflammatory markers play a major role in the pathogenesis and course of SAP and can be used as prognostic markers in its early phase. Other markers such as simple prognostic scores have been found to be as effective as multifactorial scoring systems (MFSS) at 48 h with the advantage of simplicity, efficacy, low cost, accuracy and early prediction of SAP. Recently, several laboratory markers including hematocrit, blood urea nitrogen (BUN), creatinine, matrix metalloproteinase-9 (MMP-9) and serum amyloid A (SAA) have been used as early predictors of severity within the first 24 h. The last few years have witnessed a tremendous progress in understanding the pathogenesis and predicting the outcome of SAP. In this review we classified the prognostic markers into predictors of severity, pancreatic necrosis (PN), infected PN (IPN) and mortality.

Severe acute pancreatitis: case-oriented discussion of interdisciplinary management

The clinical course of an episode of acute pancreatitis varies from a mild, transitory illness to a severe often necrotizing form with distant organ failure and a mortality rate of 20-40%. Patients with severe pancreatitis, representing about 15-20% of all patients with acute pancreatitis, need to be identified as early as possible after onset of symptoms allowing starting intensive care treatment early in the disease process. An episode of severe acute pancreatitis progresses in two phases. The first 10-14 days are characterized by a systemic inflammatory response syndrome maintained by the release of various inflammatory mediators. The second phase, beginning about 10-14 days after the onset of the disease is dominated by sepsis-related morbidity due to infected peripancreatic and pancreatic necrosis. This state is associated with septic multiple organ systemic failure. The importance of infection on the outcome of necrotizing pancreatitis has been clearly delineated and the pre-emptive use of broad-spectrum antibiotics that achieve effective tissue concentrations is considered standard management of patients with severe necrotizing pancreatitis, especially if associated with organ failure or extended necrosis. Patients with infected necrosis should undergo a surgical intervention. The standard open technique consisting of an organ preserving necrosectomy followed by a postoperative concept of lavage and/or drainage to evacuate necrotic debris occurring during the further course has recently been challenged by various minimally invasive approaches.

Pneumoretroperitoneum in Two Patients with Clostridium perfringens Necrotizing Pancreatitis

Pancreatic gas gangrene is an uncommon and often fatal complication of acute pancreatitis, due to the sporulating anaerobe Clostridium perfringens. C. perfringens is a normal constituent of colonic flora, but infects the pancreas by either transmural spread from the colon or via the biliary tree. Only three reported cases in the world literature describe acute pancreatitis with pneumoretroperitoneum and clostridial infection. Two separate cases, at the same institution, of acute pancreatitis complicated by C. perfringens were analyzed. The records of patients were reviewed for admission history, laboratory and radiology results, intensive care support, surgical intervention, and outcome. Retroperitoneal air was visualized early in the clinical course of both patients by computed tomography. Early surgical debridement, drainage, parental antibiotics, and re-exploration resulted in an uncomplicated recovery. Early computed tomography in patients with suspected necrotizing pancreatitis contributes to early intervention and may advantageously enhance survival.

Is late antibiotic prophylaxis effective in the prevention of secondary pancreatic infection?

BACKGROUND

Secondary infection of the inflamed pancreas is the principal cause of death after severe acute pancreatitis (AP). Although patients are not always managed early in the course of AP in clinical practice, prophylactic antibiotics that were used in experimental studies in rats were always initiated early after induction of pancreatitis. The effectiveness of antibiotics initiated later is unknown.

AIM

The aim of this study was to compare the effectiveness of ciprofloxacin and meropenem initiated early versus later in the course of acute necrotizing pancreatitis (ANP) in rats.

METHODS

100 Sprague-Dawley rats were studied. ANP was induced in rats by intraductal injection of 3% taurocholate. Rats were divided randomly into five groups: group I rats received normal saline as a placebo, group II and IV rats received three times daily meropenem 60 mg/kg i.p. at 2 and 24 h, respectively and group III and V rats received twice daily ciprofloxacin 50 mg/kg i.p. at 2 and 24 h, respectively, after induction. At 96 h, all rats were killed for quantitative bacteriologic study. A point-scoring system of histological features was used to evaluate the severity of pancreatitis.

RESULTS

Meropenem and ciprofloxacin initiated 2 h after induction of pancreatitis significantly reduced the prevalence of pancreatic infection (p < 0.001 and p < 0.04, respectively) as compared to controls. Neither of the antibiotics initiated later during the course of AP caused a significant decrease in pancreatic infection in rats (p > 0.05). Although the rats treated early infected less frequently than the rats treated later, the comparison reached statistical significance only in the meropenem group (p < 0.02).

CONCLUSION

Early antibiotic treatment reduces pancreatic infection more efficiently than late antibiotic treatment in ANP in rats.

[Acute pancreatitis]

While interstitial acute pancreatitis usually takes a benign course, necrotizing acute pancreatitis takes a severe course, mainly because of severe local and systemic complications. After a quick diagnosis it is necessary to rapidly assess a degree of severity of the disease and thus the prognosis. The clinical picture and the result of imaging procedures do not always correspond. The management basically includes to treat pain as well as to administer fluid, electrolyte, protein and calories. In addition, systemic treatment of complications such as shock or respiratory and renal insufficiency--if occurring--is necessary. In case of pancreatic necrosis, prophylactic administration of pancreas-penetrable antibiotics is recommended to avoid infection. In the severely ill with infected pancreatic necrosis, surgery is the treatment of choice. In approximately 10% of all patients with alcohol-induced pancreatitis, there is a gradual transition to chronic pancreatitis.

Progress in the management and treatment of infected pancreatic necrosis

Infected pancreatic necrosis and sepsis are the leading causes of mortality in necrotizing pancreatitis. A review has been undertaken of the results of the past two decades relating to different surgical treatments of infected pancreatic necrosis. During the period 1978-85, the surgical treatment of necrotizing pancreatitis and its complications in our department consisted of the 'conventional' therapy (resection of the involved pancreatic tissue, or necrosectomy and drainage) in 61 patients, with a mortality rate of 36% (22 patients died). Since 1986, we have performed necrosectomy and other surgical interventions combined with continuous widespread lavage in 142 patients with infected pancreatic necrosis. The overall mortality decreased significantly to 6.3% (9 patients died). This result was achieved by means of aggressive surgical treatment, continuous, prolonged washing and suction drainage and supportive therapy, including immunonutrition, modifying the cytokine production and adequate antibiotic and antifungal medication. This surgical strategy provides the possibility for recovery in cases of necrotizing pancreatitis with septic complications.

Hemoconcentration as an early risk factor for necrotizing pancreatitis

OBJECTIVE

The aim of our study was to determine whether measurement of serum hematocrit during the first 24 h helps in distinguishing necrotizing from mild pancreatitis.

METHODS

From May 1992 to June 1996, a case-control study was performed with cases of patients with necrotizing pancreatitis. We selected as a control the next patient admitted with mild pancreatitis.

RESULTS

There were 32 patients in each group. Logistic regression identified an admission hematocrit of > or = 47% and a failure of admission hematocrit to decrease at 24 h as the best binary risk factors for necrotizing pancreatitis. At admission, more patients with necrotizing pancreatitis than with mild pancreatitis had a hematocrit > or = 47% (11/32 vs 3/32; p = 0.03). At 24 h, 15 additional patients with necrotizing pancreatitis versus only one with mild pancreatitis showed no decrease in admission hematocrit (p < 0.01). Thus, by 24 h, 26 of 32 patients with necrotizing pancreatitis versus only four of 32 patients with mild pancreatitis met one or the other criterion (p < 0.01). The sensitivity and specificity at admission were 34% and 91%; at 24 h, 81% and 88%.

CONCLUSIONS

Hemoconcentration with an admission hematocrit > or = 47% or failure of admission hematocrit to decrease at approximately 24 h were strong risk factors for the development of pancreatic necrosis.

Value of early blood Th-1 cytokine determination in predicting severity of acute pancreatitis

BACKGROUND

Early evaluation of the severity of acute pancreatitis (AP) requires measurement of many variables within 48 h after admission. Septic complications (SC) are frequent, and preliminary studies have highlighted the value of prophylactic antibiotherapy; however, single and reliable predictive markers of sepsis are not yet available. The aim of this study was to assess the value of determining early blood Th-1 cytokines and their natural antagonists (interleukin-6 (IL-6), IL-1, IL-1ra, and the soluble form of tumor necrosis factor (sTNF) receptors RI and RII) to predict the severity and SC during AP.

METHODS

Thirty-seven patients with AP were prospectively included; 25 of them had severe AP, including 8 with SC. Serum cytokines were measured 48 h and 72 h after the onset of AP with an enzyme-linked immunosorbent assay. The optimal severity or SC diagnostic thresholds was determined using receiver operative curves.

RESULTS

Severe AP in accordance with the Atlanta criteria were better predicted by C-reactive protein and IL-6 serum determination, albeit these levels could not predict absolutely the death of two patients. In severe AP cases (n = 25) the IL-1 to IL-1-ra ratio was lower in cases further complicated by sepsis ((6+/-4) 10(-3) versus (34+/-13) 10(-3), P < 0.05); moreover, sTNF RI (2497+/-270 pg/ml versus 2133+/-611 pg/ml, P < 0.05) and RII (3751+/-400 pg/ml versus 3045+/-509 pg/ml, P < 0.05) were higher in AP characterized by further SC. The IL-1 to IL-1-ra ratio and IL-1 concentration were dramatically decreased within the first 48 h ((0.4+/-0.4) 10(-3) versus (30+/-11) 10(-3), P < 0.05, and 0.3+/-0.3 versus 15+/-3 ng/l, P < 0.05) in patients with further infection of the pancreatic necrosis (n = 3). The SC diagnosis was better anticipated by an IL-1 to IL-1-ra ratio lower than 5 x 10(-3) or by an sTNF RI higher than 1750 pg/ml and sTNF RII higher than 2750 pg/ml, and the infection of the pancreatic necrosis by an IL-1 concentration <2 ng/l or an IL-1 to IL-1-ra ratio <2 x 10(-3).

CONCLUSION

Besides severity markers, IL-1, IL-1-ra, and sTNF RI and RII should be considered in base-line AP assays and, if confirmed by larger studies, could help to screen patients at risk for SC and candidates for prophylactic antibiotherapy with a good negative predictive value.

Role of ultrasonographically guided fine-needle aspiration cytology in the diagnosis of infected pancreatic necrosis

BACKGROUND

Early detection of infected pancreatic necrosis has a major impact on further management and outcome in acute pancreatitis. The aim of this study was to evaluate the clinical value of ultrasonographically guided fine-needle aspiration cytology (FNAC) in patients with necrotizing pancreatitis over an 8-year period.

METHODS

From January 1988 to September 1996 193 (mean 2.0 (range 1-9) aspirations per patient) prospectively assessed FNACs guided ultrasonographically were performed in 98 patients with necrotizing pancreatitis proven by contrast-enhanced computed tomography. Aspirates were considered infected if either Gram stain and/or culture revealed micro-organisms.

RESULTS

Ultrasonographically guided FNAC correctly diagnosed infection in 29 of 33 patients with infected necrosis a median of 13 days after onset of symptoms. Of 61 patients with sterile necrosis 55 were identified correctly as sterile by FNAC. There were six false-positive and four false-negative aspirates of which nine occurred during the first week of the disease. In four patients who did not undergo operation FNAC revealed Gram-negative organisms; however, in the absence of repeated aspirations, the positive results remained unconfirmed. An overall sensitivity of 88 per cent and a specificity of 90 per cent was obtained. No difference was found in biochemical and clinical parameters indicating systemic inflammatory response syndrome before each FNAC between patients with proven sterile or infected necrosis. All patients tolerated the procedure well and no major complications were observed.

CONCLUSION

Ultrasonographically guided FNAC is a fast and reliable technique for the diagnosis of infected necrosis. As complication rates are very low, the procedure can be repeated at short intervals to improve the diagnostic accuracy. Ultrasonographically guided FNAC is recommended for all patients with necrotizing pancreatitis in whom systemic inflammatory response syndrome persists beyond the first week after onset of symptoms.

Surgical strategy and management of infected pancreatic necrosis

Infected pancreatic necrosis and sepsis are the leading causes of death in patients with necrotizing pancreatitis. Between 1986 and 1993, 123 patients with infected pancreatic necrosis were treated; in all cases the infected necrosis extended to the retroperitoneal area. Surgical treatment was performed a mean of 18.5 days after the onset of acute pancreatitis. Operative management consisted of wide-ranging necrosectomy through all the affected area, combined with continuous widespread lavage and suction drainage applied for a mean of 39.5 days, with a median of 6.5 litres of normal saline per day. In 56 cases (46 per cent), another surgical intervention (distal pancreatic resection, splenectomy, cholecystectomy, sphincteroplasty or colonic resection) was also performed. Bacteriological findings revealed mainly enteric bacteria, but Candida infection was detected in 21 per cent of patients. The overall hospital mortality rate was 7 per cent (nine patients died). Infected pancreatic necrosis responds well to aggressive surgical treatment, continuous, long-standing lavage and suction drainage, together with supportive therapy combined with adequate antibiotic and antifungal medication.

Pathogenesis and prevention of early pancreatic infection in experimental acute necrotizing pancreatitis

OBJECTIVE

The authors test antibiotic strategies aimed at either mitigating bacterial translocation from the gut or delivering antibiotics specifically concentrated by the pancreas for prevention of early secondary infection after acute necrotizing pancreatitis.

BACKGROUND

Infection currently is the principal cause of death after severe pancreatitis. The authors have shown that the risk of bacterial infection correlates directly with the degree of tissue injury in a rodent model of pancreatitis. Bacteria most likely arrive by translocation from the colon.

METHODS

Severe acute necrotizing pancreatitis was induced in rats by a combination of low-dose controlled intraductal infusion of glycodeoxycholic acid superimposed on intravenous cerulein hyperstimulation. At 6 hours, animals were randomly allocated to five treatment groups: controls, selective gut decontamination (oral antibiotics and cefotaxime), oral antibiotics alone, cefotaxime alone, or imipenem. At 96 hours, surviving animals were killed for quantitative bacterial study of the cecum, pancreas, and kidney.

RESULTS

The 96-hour mortality (35%) was unaffected by any treatment regimen. Cecal gram-negative bacteria were significantly reduced only by the oral antibiotics. Pancreatic infection was significantly reduced by full-gut decontamination and by imipenem, but not by oral antibiotics or by cefotaxime alone. Renal infection was reduced by both intravenous antibiotics.

CONCLUSIONS

Early pancreatic infection after acute necrotizing pancreatitis can be reduced with a full-gut decontamination regimen or with an antibiotic concentrated by the pancreas (imipenem) but not by unconcentrated antibiotics of similar spectrum (cefotaxime) or by oral antibiotics alone. These findings suggest that 1) both direct bacterial translocation from the gut and hematogenous seeding interplay in pancreatic infection while hematogenous seeding is dominant at extrapancreatic sites and 2) imipenem may be useful in clinical pancreatitis.

Time course of bacterial infection of the pancreas and its relation to disease severity in a rodent model of acute necrotizing pancreatitis

BACKGROUND

Bacterial infection of pancreatic necrosis is thought to be a major determinant of outcome in acute necrotizing pancreatitis. The determinants and possibilities for prophylaxis are unknown and difficult to study in humans.

OBJECTIVE

The time course of bacterial infection of the pancreas in a rodent model of acute necrotizing pancreatitis was characterized. The authors ascertained if there is a correlation with the degree of necrosis.

METHODS

Acute pancreatitis (AP) of graded severity was induced under sterile conditions by an intravenous infusion of cerulein (5 micrograms/kg/hr) for 6 hours (mild AP), or a combination of intravenous cerulein with an intraductal infusion of 10-mM glycodeoxycholic acid (0.2 mL for 2 min for moderate AP, 0.5 mL for 10 min for severe AP). Sham-operated animals (intravenous and intraductal NaCl 0.9%) served as controls. Ninety-six hours after induction, animals were killed for quantitative bacterial examination and histologic scoring of necrosis. In addition, groups of animals with severe AP were investigated at 12, 24, 48, 96, and 144 hours.

RESULTS

No significant pancreatic necrosis was found in control animals (0.3 +/- 0.1) or animals with mild AP (0.6 +/- 0.1) killed at 96 hours. Necrosis scores were 1.1 +/- 0.2 for animals with moderate AP and 1.9 +/- 0.2 for animals with severe AP. Control animals did not develop significant bacterial infection of the pancreas (> or = 10(3) CFU/g). At 96 hours, the prevalence of infection was 37.5% in animals with mild AP and 50% in animals with moderate AP. In animals with severe AP, infection of the pancreas increased from 33% in the first 24 hours to 75% between 48 and 96 hours (p < 0.05). The bacterial counts and the number of different species increased with time and was maximal (> 10(11) CFU/g) at 96 hours.

CONCLUSION

Bacterial infection of the pancreas in rodent AP increases during the first several days, and its likelihood correlates with the severity of the disease. This model, which closely mimics the features of human acute pancreatitis, provides a unique opportunity to study the pathogenesis of infected necrosis and test therapeutic strategies.

Pancreatic infection complicating acute pancreatitis

Pancreatic infection is the leading cause of death from acute pancreatitis. Patients with severe necrotizing pancreatitis are most at risk. Early computed tomography and percutaneous fine-needle aspiration microbiology of areas of pancreatic necrosis enable early diagnosis. Pancreatic infection should be treated surgically, although sterile necrosis may be managed conservatively. The role of antimicrobial drugs is uncertain.

Pathogenesis of pancreatic sepsis

Although pancreatic sepsis is the most common cause of major morbidity and mortality associated with acute pancreatitis, the pathogenesis of such infections is unknown. Since intraperitoneal foci of inflammation are known to promote bacterial translocation, we hypothesized that acute pancreatitis promotes bacterial translocation that leads to infection of the inflamed pancreas and peripancreatic tissues. Non-lethal acute pancreatitis was induced in rats, and the translocation of live bacteria to the pancreas, mesenteric lymph nodes, liver, and spleen was determined. The presence of orally fed fluorescent beads, sensitive inert markers of translocation, was also determined in the pancreas and mesenteric lymph nodes. Live bacteria were recovered from 33% of the pancreata of rats with acute pancreatitis but from none of the control rats. Beads were visualized in 91% of the pancreata of rats with acute pancreatitis but in none of the pancreata from control rats. Beads were not visualized in the mesenteric lymph nodes of rats with acute pancreatitis, suggesting a transperitoneal route of migration. We conclude that acute pancreatitis promotes bacterial translocation leading to transperitoneal infection of the pancreas. These results support the use of selective decontamination of the gut and peritoneal lavage for the prevention of pancreatic infections in acute pancreatitis.

Prognostic factors in sterile pancreatic necrosis

Although the overall mortality in sterile pancreatic necrosis is low, patients who experience systemic complications may have a higher mortality. To study the impact of systemic complications and other factors on survival, possible prognostic factors were evaluated among 26 patients who experienced at least one systemic complication. Mortality was 38%. Factors that correlated with a fatal outcome were high Ranson's scores during the first 48 hours (P = 0.01), high APACHE-II scores at admission (P = 0.04) and at 48 hours (P = 0.03), shock (P < 0.001), renal insufficiency (P < 0.05), multiple systemic complications (P < 0.001), and high body mass index (P = 0.01). Most systemic complications occurred during the first 2 weeks of illness. Logistic regression analysis showed that shock was the best predictor of a fatal outcome. Patients with favorable prognostic factors survived whether treated medically or surgically, whereas those with unfavorable factors had a fatal outcome whether treated medically or surgically. It is concluded that patients with severe sterile necrosis have a high mortality rate and that shock and other prognostic factors identify which patients are most likely to have a fatal outcome.

Early surgical débridement of symptomatic pancreatic necrosis is beneficial irrespective of infection

In order to assess the recent trend of nonoperative management of pancreatic necrosis, we reviewed 82 variables in 73 consecutive patients with symptomatic necrotizing pancreatitis. The mortality rate for the series was 25% (18 of 73). The only preintervention variables that correlated with mortality were APACHE II score greater than 15 (p = 0.01), preintervention blood transfusion (p less than 0.001), respiratory failure (p less than 0.001), and shock (p less than 0.01). Patients who developed recurrent sepsis following the initial intervention had a significantly higher mortality rate (17 of 34) than those who did not (1 of 39) (p less than 0.001). The rate of recurrent sepsis varied widely among individual surgeons and correlated with APACHE II score. The presence of infected versus noninfected necrosis did not correlate significantly with outcome. When percutaneous radiologically guided drainage was the initial therapeutic modality (n = 6), recurrent sepsis requiring surgical drainage inevitably occurred. Patients treated with percutaneous drainage (often in combination with surgical drainage) had a longer hospital stay (82 versus 42 days, p less than 0.001), spent more days in the intensive care unit (31 versus 6 days, p less than 0.001), and required more days of total parenteral nutrition (57 versus 27 days, p less than 0.001) than those treated solely by surgical means. We conclude that aggressive initial surgical débridement should be the first step in managing symptomatic pancreatic necrosis and that the presence of infection should not be the sole determinant of intervention.

Acute necrotizing pancreatitis: management by planned, staged pancreatic necrosectomy/debridement and delayed primary wound closure over drains

We reviewed our recent experience with management of 23 consecutive patients with acute necrotizing pancreatitis. All patients had documented necrotizing pancreatitis with parenchymal or peripancreatic necrosis. Our method of treatment has evolved from our previous approach of controlled open lesser sac drainage (marsupialization) to staged necrosectomy/debridement with delayed primary closure over drains. With this latter approach, hospital mortality was 4 of 23 patients (17 per cent), but significant morbidity still occurred in 12 of 23 patients (52 per cent). However, recurrent intra-abdominal abscess before discharge occurred in only one patient. We believe that this operative approach toward the severely ill patient with acute necrotizing pancreatitis who requires operative intervention will minimize the occurrence of intra-abdominal sepsis.