- Email: [email protected]
Outcome and Quality of Life of Patients with Acute Pancreatitis Requiring Intensive Care
Journal of Surgical Research 91, 89 –94 (2000) doi:10.1006/jsre.2000.5925, available online at http://www.idealibrary.com on
Outcome and Quality of Life of Patients with Acute Pancreatitis Requiring Intensive Care Atilla Soran, M.D.,* Lakshmipathi Chelluri, M.D.,† Kenneth K. W. Lee, M.D.,‡ and Samuel A. Tisherman, M.D.† ,‡ ,1 ‡Department of Surgery and †Department of Anesthesiology/Critical Care Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania 15213; and *Department of Surgery, Ankara Numune Hospital, Ankara, Turkey Presented at the Annual Meeting of the Association for Academic Surgery, Philadelphia, Pennsylvania, November 18 –20, 1999
life (SF-36) was similar to that of an age-matched population. Twenty of twenty-one felt their general health was at least as good as it had been 1 year previously. Conclusions. Patients with severe acute pancreatitis need prolonged ICU and hospital stays. APACHE II may be a good predictor of outcome; further, prospective evaluation is needed. Although resource utilization is high, most patients survive and have good longterm quality of life. © 2000 Academic Press Key Words: pancreatitis; outcome; quality of life; intensive care.
Background. Patients with severe acute pancreatitis often require intensive care unit (ICU) admission, have multiple complications, spend weeks to months in the hospital, and consume a large amount of resources. The aim of this study was to evaluate the ICU course, costs, mortality, and quality of life of patients who require ICU admission for acute pancreatitis. Methods. Patients with acute pancreatitis requiring ICU admission were identified retrospectively. Data regarding in-hospital morbidity, mortality, and hospital costs were obtained. Long-term quality of life was assessed using the Short Form-36 Health Survey (SF-36). Results. Fifty-two patients were identified. There were 31 men and 21 women: the mean age was 53 years (range, 22– 89). The most common causes of acute pancreatitis were gallstones (44%) and alcoholism (17%). Pulmonary failure (52% required mechanical ventilation) and renal failure (21% required dialysis) were common. There were 39 (75%) hospital survivors and 13 (25%) nonsurvivors. In the first 24 h, the mean Acute Physiology and Chronic Health Evaluation (APACHE) II scores were 10 ⴞ 6 in survivors and 16 ⴞ 4 in the nonsurvivors (<0.01). Mean length of ICU (15 ⴞ 18 and 28 ⴞ 31 days) and hospital (40 ⴞ 34 and 38 ⴞ 34 days) stays were similar in survivors and nonsurvivors, respectively (NS). The mean hospital cost for survivors was $83,611 ⴞ 88,434 and that for nonsurvivors was $136,730 ⴞ 95,045 (P ⴝ 0.09). The estimated cost to obtain one hospital survivor was $129,188. Of the 39 hospital survivors, 5 died later, 21 completed the SF36, and 13 were lost to follow-up. Long-term quality of
INTRODUCTION
Acute pancreatitis can be a severe illness requiring prolonged hospitalization and, sometimes, multiple operative interventions, which result in high hospital costs. Patients who develop infected pancreatic necrosis have a risk of mortality of almost 100% without surgical debridement [1]. Even with debridement, mortality approaches 40% in some series, with morbidity rates of nearly 100%. Although patients often spend weeks or months in the intensive care unit (ICU) and hospital, and consume a large amount of resources, mortality has improved and the majority of these patients survive to hospital discharge [2]. To determine whether this consumption of health care resources is justified, the long-term results of treatment must be assessed. One study [3] found good long-term quality of life after pancreatitic debridement for necrosis, but did not look at severely ill patients who did not require debridement. Another study [4] found generally good physiologic outcomes after treatment for severe pan-
1 To whom correspondence should be addressed at Division of Critical Care Medicine, 646A Scaife Hall, 200 Lothrop Street, Pittsburgh, PA 15213. Fax: (412) 647-8060. E-mail: tishermansa@ msx.upmc.edu.
89
0022-4804/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
90
JOURNAL OF SURGICAL RESEARCH: VOL. 91, NO. 1, JUNE 1, 2000
creatitis, but did not examine quality of life from the patient perspective. Outcome assessment should consider not only the short-term complications and survival, but also longterm quality of life, to assess therapeutic strategies correctly. The purpose of this study was to analyze the outcome and hospital costs of all patients with acute pancreatitis who required ICU stay, regardless of whether or not they require debridement, and to assess the quality of life after discharge from the hospital. METHODS This study was approved by the University of Pittsburgh Institutional Review Board. Informed consent was obtained from the subjects. The medical records of 52 patients (31 women, 21 men) admitted between January 1992 and December 1996 who required intensive care for acute pancreatitis were reviewed. The severity of illness was assessed by Acute Physiology and Chronic Health Evaluation (APACHE) II score on admission [5, 6]. Because many of the patients were transferred to our institution several days after the onset of acute pancreatitis, use of the Ranson score [7] was usually not appropriate. Data collection included information on patient demographics, cause of pancreatitis, laboratory values, complications, length of stay (ICU and hospital), and results of intravenous contrast-enhanced computed tomography (CT) scans. Details of operative interventions were also recorded. A cost calculation was done for each patient according to hospitalization, operations and procedures, and diagnostic studies. To determine the cost effectiveness, we evaluated the effective cost per survivor by adding the costs incurred by all patients (survivors and nonsurvivors) and dividing the sum by the total number of survivors [8]. The quality of life after discharge from the hospital was assessed by using the Short Form-36 General Health Survey (SF-36). The SF-36 health survey includes questions related to both physical and mental health. Questions regarding physical health include physical functioning (PF), role physical (RP), bodily pain (BP), and general health (GH). Mental health issues include vitality (V), social functioning (SF), role emotional (RE), and mental health (MH). This survey also includes questions for the subject to compare his or her current condition with that of 1 year previously. The results of this survey were compared with published data on a similar-aged control group (ages 45–54) [9]. Statistical analysis was performed by using BMPD software. Comparisons were made between survivors and nonsurvivors using a two-tailed Student t test for continuous variables and Kruskal– Wallis test for nonparametric variables.
RESULTS
There were 39 survivors and 13 nonsurvivors. The mean age of the patients was 52.5 (range, 22– 89) years, 52 for survivors and 58 for nonsurvivors [not significant (NS)]. Sixty percent of the patients were men (22 of 39 survivors and 9 of 13 nonsurvivors, NS). The most common etiologies of pancreatitis were gallstones (44%) and alcohol (17%) (Fig. 1). The etiology could not be identified in 6 patients (12%). APACHE II scores during the first 24 h after admission to the ICU averaged 11 ⫾ 6. These scores were significantly worse in nonsurvivors than in survivors (16 ⫾ 4 vs 10 ⫾ 6,
FIG. 1.
Causes of pancreatitis in survivors and nonsurvivors.
respectively, P ⬍ 0.05). Laboratory values on admission were similar in survivors and nonsurvivors, except for alkaline phosphate and ␦GTP, which were statistically higher in survivors on admission (Table 1). Bolus intravenous contrast-enhanced CT revealed pancreatic necrosis or phlegmon in 29 (56%) patients (Table 2). The remainder had edema. Twenty-four (46%) patients were treated with supportive medical therapy only; and 28 (54%) patients underwent surgical procedures, with 17 undergoing multiple debridements (Table 3). The indications for debridement were infected necrosis in 12 patients, necrosis (not proven to be infected) and deterioration despite aggressive supportive care in 6, abscess in 3, and bleeding in 1. There were no statistical differences between survivors and nonsurvivors in CT findings or treatment modalities. Pulmonary failure (defined as requiring greater than 24 h of mechanical ventilation) and renal failure (requiring dialysis) were both common complications observed in patients with acute pancreatitis: 27 patients (52%) required mechanical ventilation, and 13 patients (25%) underwent dialysis (Table 4). Maximum serum creatinine level was significantly higher in nonsurvivors (4.4 ⫾ 2.7 mg/dl) than in survivors (2.3 ⫾ 2.5 mg/dl) (P ⬍ 0.05). Hospital mortality was 25% (n ⫽ 13). The cause of death was sepsis in 8 patients (4 retroperitoneal/ pancreatic, 1 abdominal, 3 other sources), multiple organ system dysfunction in 6, and the systemic inflam-
91
SORAN ET AL.: QUALITY OF LIFE AFTER SEVERE ACUTE PANCREATITIS
TABLE 1 Laboratory Values of Patients on Admission to the Intensive Care Unit
Glucose (mg/dl) Total bilirubin (mg/dl) LDH (U/dl) ␥GTP (U/dl) a SGOT (U/dl) SGPT (U/dl) Alkaline phosphatase (U) a Amylase (U) Calcium (mg/dl) Sodium (meq/ml) Potassium (meq/ml) Platelets (⫻ 1000) WBC count (⫻ 1000) Hematocrit (%) Creatinine (mg/dl) Blood urea nitrogen (mg/dl) a
Overall (n ⫽ 52)
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
170.9 ⫾ 84.9 2.4 ⫾ 3.1 454.4 ⫾ 402.4 325.5 ⫾ 408.4 184.9 ⫾ 542.1 107.9 ⫾ 145.8 182.3 ⫾ 166.2 716.6 ⫾ 743.1 7.7 ⫾ 1.5 136.3 ⫾ 7.2 4.2 ⫾ 0.7 311 ⫾ 176.6 17.8 ⫾ 9.3 33.2 ⫾ 7.2 2.1 ⫾ 1.8 34 ⫾ 23.1
160.5 ⫾ 76.9 2.7 ⫾ 3.4 382.2 ⫾ 224.1 376.9 ⫾ 456.1 153.6 ⫾ 483.7 97.7 ⫾ 113.6 207.9 ⫾ 184.5 727.1 ⫾ 726.1 7.7 ⫾ 1.5 135.5 ⫾ 6.3 4.2 ⫾ 0.7 328 ⫾ 185 17.9 ⫾ 9 32.6 ⫾ 7.1 1.8 ⫾ 1.8 31.4 ⫾ 22.2
201.8 ⫾ 102.1 1.5 ⫾ 1.3 647.2 ⫾ 684.1 171.3 ⫾ 125.3 278.8 ⫾ 703.8 138.5 ⫾ 219.5 107.5 ⫾ 43.5 686.2 ⫾ 820.9 7.7 ⫾ 1.6 138.7 ⫾ 9.4 4.2 ⫾ 0.6 260 ⫾ 141 17.2 ⫾ 10.3 34.9 ⫾ 7.5 2.8 ⫾ 1.7 41.9 ⫾ 24.9
P ⬍ 0.05 comparing survivors with nonsurvivors.
matory response syndrome in 1. One patient had necrosis of the cecum. Other contributing factors included respiratory failure, renal failure, coagulopathy, hypovolemia, and cardiogenic shock. Some patients had multiple causes. The median hospital stay was 27 days with a mean of 39 ⫾ 33 days (40 ⫾ 34 in survivors and 38 ⫾ 34 in nonsurvivors, NS). Median intensive care unit stay was 9 days, with a mean of 18 ⫾ 23 days (15 ⫾ 18 in survivors and 28 ⫾ 31 in nonsurvivors, NS). The overall mean hospital cost was $96,891 ⫾ 92,152 ($136,730 ⫾ 95,045 for nonsurvivors and $83,611 ⫾ 88,434 for survivors, NS). Overall, the ICU and nonICU rooms accounted for 41% of all hospital costs (Fig. 2). Operation and anesthesia (exclusive of professional fees) was only 6%. The estimated cost to obtain one hospital survivor was $129,188. Twenty-one of thirty-nine survivors were contacted by telephone survey. Of the 18 who were not contacted, 5 had died in long-term follow-up, and 13 were lost to follow-up. The SF-36 surveys were conducted a median of 42 (range, 17– 69) months after discharge. Data from
the SF-36 survey show that the quality of life of acute pancreatitis survivors was generally good. When compared with an age-matched (45–54 years) control population, there were no significant differences in either the physical or mental health domains (Fig. 3); i.e., quality of life was as expected for their ages. Twenty of twenty-one patients (95%) stated that their general health was about the same (10) or better (8 much better, 2 somewhat better) than it had been 1 year previously (1 much worse).
TABLE 2
Operative and Medical Management of Patients with Severe Acute Pancreatitis
DISCUSSION
In this study of patients who required ICU admission for severe, acute pancreatitis, we found that these patients often required operative intervention and often developed infectious complications. They frequently developed organ system failure, particularly pulmonary and renal, necessitating mechanical ventilation and hemodialysis, respectively. Hospital costs, TABLE 3
Computed Tomography Findings during Severe Acute Pancreatitis
Phlegmon Necrosis Edema/inflammation
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
9 (23%) 13 (33%) 17 (44%)
2 (15%) 5 (39%) 6 (46%)
Treatment
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
Overall (n ⫽ 52)
Nonoperative Debridement Drainage Resection Laparotomy
19 (49%) 17 (44%) 2 (5%) 1 (2%) —
5 (38%) 5 (38%) 1 (8%) 1 (8%) 1 (8%)
24 (46%) 22 (42%) 3 (6%) 2 (4%) 1 (2%)
92
JOURNAL OF SURGICAL RESEARCH: VOL. 91, NO. 1, JUNE 1, 2000
TABLE 4 Short- and Long-Term Complications of Patients with Acute Pancreatitis Complications
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
Pulmonary failure a Renal failure b Sepsis/septic shock Infected pancreatic necrosis Pancreatic abscess Retroperitoneal abscess Intraabdominal abscess Pancreatic pseudocyst Pancreatic fistula Encephalopathy Others c
16 6 4 10 2 2 3 12 8 5 4
11 7 8 3 1 2 0 2 0 1 18
Required ⬎24 h of mechanical ventilation. Required hemodialysis. c Others (one patient each): upper gastrointestinal bleeding, bile duct obstruction, colonic obstruction, gastric outlet obstruction, deep venous thrombosis, pulmonary embolism, acute cholecystitis, cholangitis, ascites, postoperative hernia, ischemic necrosis of cecum, unstable angina, coagulopathy, persistent hypercalcemia. a b
particularly room costs, were high. This reflects prolonged ICU and hospital stays. Mortality was 25%, which is similar to the results of others [10 –13]. Despite prolonged hospitalization and multiple complications, survivors had long-term functional outcome that
FIG. 2. Distribution of hospital costs for survivors and nonsurvivors of severe acute pancreatitis.
FIG. 3. Results of the SF-36 survey for survivors of severe acute pancreatitis. Comparison is made to a historical normal control group, ages 45–54 [9].
was quite acceptable. The data from the SF-36 survey suggest that compared with their peers, these patients have no significant limitations secondary to physical or emotional problems. Only one respondent noted that his general health was worse than it had been 1 year previously. Since most physicians caring for patients with severe pancreatitis only see the patients when they are quite ill and debilitated, it is encouraging to know that, despite multiple operations and complications, they can recover and return to their preillness lifestyles. This study is limited by the fact that it was retrospective and did not include large numbers of patients. With greater numbers of patients, perhaps, more subtle quality of life issues would have been delineated from the SF-36 data. In addition, follow-up in this study was not complete. It is possible that those patients who were lost to follow-up did not have as positive an outcome as those who responded to our inquiry. At worst, if all of these patients had poor outcomes, this would mean that 33% of long-term survivors did not have acceptable outcomes. The fact that five patients who initially survived their hospitalization died later is worth noting. Patients with severe pancreatitis frequently develop multiple complications and organ system dysfunction. Even if they leave the hospital, they often require continuing nursing care and rehabilitation. They remain at risk for complications and even death. Longterm survival (at least 6 months), not just survival to discharge or 30 days, needs to be considered. There is a sense that survival of these patients has improved in recent years. Most likely this improvement is the result of a combination of factors, including aggressive surgical management of the necrotic pancreas and better supportive care of the other vital or-
SORAN ET AL.: QUALITY OF LIFE AFTER SEVERE ACUTE PANCREATITIS
gans that are frequently dysfunctional as a result of the systemic inflammatory response initiated by the pancreatitis. Regarding debridement of the necrotic pancreas, our general policy has been to debride patients with infected necrosis and those who deteriorate despite aggressive supportive care, even if infection has not been proven. Previous studies of patients with severe pancreatitis have focused on surgical complications. Nonsurgical complications have rarely been examined. In this study, most patients who required ICU admission for severe acute pancreatitis had failure of at least one organ system and suffered at least one complication not directly related to the pancreas per se. These complications were the major reason for mortality since deaths were almost always after withdrawal of support when it was apparent that continued therapy was futile in terms of the potential for restoring an acceptable quality of life. One recent study by Broome et al. [3] examined the quality of life (using the SF-36) after pancreatic debridement for pancreatic necrosis, compared with that after medical management of chronic pancreatitis and after operative management of pancreatic ductal abnormalities. They found that care for patients with pancreatic necrosis was expensive, but that the quality of life was good, thus justifying the expense. The comparison groups could be questioned. Patients undergoing elective operation for ductal abnormalities would not be expected to have the physiologic derangements seen in patients with acute pancreatitis, nor would they be expected to require postoperative rehabilitation. Similarly, patients treated medically for chronic pancreatitis would also present a significantly different patient population, particularly because of the psychologic overlay that complicates treatment of chronic diseases causing pain. These authors also compared their patients with patients with chronic medical diseases (congestive heart failure and chronic obstructive pulmonary disease). Issues with these chronic diseases would be expected to be different from those associated with acute pancreatitis. Doepel et al. [4] found good long-term functional outcomes in survivors of severe acute pancreatitis. This study, however, focused on physiologic outcome, not quality of life from the patient’s perspective. Neither of these studies [3, 4] analyzed resource utilization in the nonsurvivors. In the present study, nonsurvivors of acute pancreatitis, like survivors, had prolonged ICU and hospital lengths of stay and accumulated hospital costs greater than $100,000 on average. Since most patients survived to discharge, the average cost to obtain a hospital survivor was not much greater than this. Looking at the issue solely from the perspective of survivors, Fenton-Lee and Imrie [14] also found that the costs
93
were reasonable considering the good long-term outcomes. They further estimated that care for these patients resulted in a benefit of 8.55 quality-adjusted life years on average per patient. Although this study examined long-term quality of life, it did not look at the related question of long-term costs of care after discharge from the hospital. Future studies of long-term outcome from severe pancreatitis need to evaluate these costs, e.g., skilled nursing care and rehabilitation. Because of the frequency of complications and high costs of care noted above, predicting outcome from severe pancreatitis has been of interest for more than 20 years. Ranson and Pasternack [7], developed a system based on 11 parameters at the time of admission and over the first 48 h of care for predicting risk of complications and death from acute pancreatitis. The more generic APACHE II system has also been successfully applied to this patient population [6]. Because patients often spent a significant period at a community hospital prior to transfer to our institution, in this study, data for the Ranson criteria were usually not available. Consequently, we used APACHE II. We found that survivors had significantly better APACHE scores than nonsurvivors. More data on the predictive value of APACHE in these patients are needed, particularly with the current availability of APACHE III [15]. In conclusion, patients with severe acute pancreatitis often require admission to an ICU. These patients often spend weeks or months in the ICU and hospital, resulting in hospital costs that can be enormous, frequently greater than $100,000. They also need longterm care after discharge from the hospital and are still at risk for complications and death. The overall expense is justified given that the great majority of patients who survive have a good long-term quality of life. REFERENCES 1.
Banks, P. A. Infected necrosis: Morbidity and therapeutic consequences. Hepatogastroenterology 38: 116, 1991.
2.
Bradley, E. L. III. A fifteen year experience with open drainage for infected pancreatic necrosis. Surg. Gynecol. Obstet. 177: 215, 1993.
3.
Broome, A. H., Eisen, G. M., Harland, R. C., et al. Quality of life after treatment for pancreatitis. Ann. Surg. 223: 665, 1996.
4.
Doepel, M., Eriksson, J., Halme, L., et al. Good long-term results in patients surviving severe acute pancreatitis. Br. J. Surg. 80: 1583, 1993.
5.
Knaus, W. A., Draper, E. A., Wagner, D. P., and Zimmerman, J. E. APACHE II: a severity of disease classification system. Crit. Care Med. 13: 818, 1985.
6.
Larvin, M., and McMahon, M. J. APACHE-II score for assessment and monitoring of acute pancreatitis. Lancet July: 201, 1989.
7.
Ranson, J. H., and Pasternack, B. S. Statistical methods for
94
8.
9. 10. 11.
JOURNAL OF SURGICAL RESEARCH: VOL. 91, NO. 1, JUNE 1, 2000 quantifying the severity of clinical acute pancreatitis. J. Surg. Res. 22: 79, 1977. Chalfin, D. B., Cohen, I. L., and Lambrinos, J. The economics and cost-effectiveness of critical care medicine. Intensive Care Med. 21: 952, 1995. Ware, J. E., Jr. (Ed.). SF-36 Health Survey Manual and Interpretation Guide. Boston: Nimrod Press, 1993. Warshaw, A. L., and Gongliang, J. Improved survival in 45 patients with pancreatic abscess. Ann. Surg. 202: 408, 1985. Fugger, R., Schultz, F., Rogy, et al. Open approach in pancreatic and infected pancreatic necrosis: Laparostomies and preplanned revisions. World J. Surg. 15: 516, 1991.
12.
Miller, B. J., Henderson, A., Strong, R. W., et al. Necrotizing pancreatitis: Operating for life. World J. Surg. 18: 906, 1994.
13.
D’Egidio, A., and Schein, M. Surgical strategies in the treatment of pancreatic necrosis and infection. Br. J. Surg. 78: 133, 1991.
14.
Fenton-Lee, D., and Imrie, C. W. Pancreatic necrosis: Assessment of outcome related to quality of life and cost of management. Br. J. Surg. 80: 1579, 1993.
15.
Knaus, W. A., Wagner, D. P., Draper, E. A., et al. The APACHE III prognostic system: Risk prediction of hospital mortality for critically ill hospitalized adults. Chest 100: 1619, 1991.
Outcome and Quality of Life of Patients with Acute Pancreatitis Requiring Intensive Care Atilla Soran, M.D.,* Lakshmipathi Chelluri, M.D.,† Kenneth K. W. Lee, M.D.,‡ and Samuel A. Tisherman, M.D.† ,‡ ,1 ‡Department of Surgery and †Department of Anesthesiology/Critical Care Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania 15213; and *Department of Surgery, Ankara Numune Hospital, Ankara, Turkey Presented at the Annual Meeting of the Association for Academic Surgery, Philadelphia, Pennsylvania, November 18 –20, 1999
life (SF-36) was similar to that of an age-matched population. Twenty of twenty-one felt their general health was at least as good as it had been 1 year previously. Conclusions. Patients with severe acute pancreatitis need prolonged ICU and hospital stays. APACHE II may be a good predictor of outcome; further, prospective evaluation is needed. Although resource utilization is high, most patients survive and have good longterm quality of life. © 2000 Academic Press Key Words: pancreatitis; outcome; quality of life; intensive care.
Background. Patients with severe acute pancreatitis often require intensive care unit (ICU) admission, have multiple complications, spend weeks to months in the hospital, and consume a large amount of resources. The aim of this study was to evaluate the ICU course, costs, mortality, and quality of life of patients who require ICU admission for acute pancreatitis. Methods. Patients with acute pancreatitis requiring ICU admission were identified retrospectively. Data regarding in-hospital morbidity, mortality, and hospital costs were obtained. Long-term quality of life was assessed using the Short Form-36 Health Survey (SF-36). Results. Fifty-two patients were identified. There were 31 men and 21 women: the mean age was 53 years (range, 22– 89). The most common causes of acute pancreatitis were gallstones (44%) and alcoholism (17%). Pulmonary failure (52% required mechanical ventilation) and renal failure (21% required dialysis) were common. There were 39 (75%) hospital survivors and 13 (25%) nonsurvivors. In the first 24 h, the mean Acute Physiology and Chronic Health Evaluation (APACHE) II scores were 10 ⴞ 6 in survivors and 16 ⴞ 4 in the nonsurvivors (<0.01). Mean length of ICU (15 ⴞ 18 and 28 ⴞ 31 days) and hospital (40 ⴞ 34 and 38 ⴞ 34 days) stays were similar in survivors and nonsurvivors, respectively (NS). The mean hospital cost for survivors was $83,611 ⴞ 88,434 and that for nonsurvivors was $136,730 ⴞ 95,045 (P ⴝ 0.09). The estimated cost to obtain one hospital survivor was $129,188. Of the 39 hospital survivors, 5 died later, 21 completed the SF36, and 13 were lost to follow-up. Long-term quality of
INTRODUCTION
Acute pancreatitis can be a severe illness requiring prolonged hospitalization and, sometimes, multiple operative interventions, which result in high hospital costs. Patients who develop infected pancreatic necrosis have a risk of mortality of almost 100% without surgical debridement [1]. Even with debridement, mortality approaches 40% in some series, with morbidity rates of nearly 100%. Although patients often spend weeks or months in the intensive care unit (ICU) and hospital, and consume a large amount of resources, mortality has improved and the majority of these patients survive to hospital discharge [2]. To determine whether this consumption of health care resources is justified, the long-term results of treatment must be assessed. One study [3] found good long-term quality of life after pancreatitic debridement for necrosis, but did not look at severely ill patients who did not require debridement. Another study [4] found generally good physiologic outcomes after treatment for severe pan-
1 To whom correspondence should be addressed at Division of Critical Care Medicine, 646A Scaife Hall, 200 Lothrop Street, Pittsburgh, PA 15213. Fax: (412) 647-8060. E-mail: tishermansa@ msx.upmc.edu.
89
0022-4804/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
90
JOURNAL OF SURGICAL RESEARCH: VOL. 91, NO. 1, JUNE 1, 2000
creatitis, but did not examine quality of life from the patient perspective. Outcome assessment should consider not only the short-term complications and survival, but also longterm quality of life, to assess therapeutic strategies correctly. The purpose of this study was to analyze the outcome and hospital costs of all patients with acute pancreatitis who required ICU stay, regardless of whether or not they require debridement, and to assess the quality of life after discharge from the hospital. METHODS This study was approved by the University of Pittsburgh Institutional Review Board. Informed consent was obtained from the subjects. The medical records of 52 patients (31 women, 21 men) admitted between January 1992 and December 1996 who required intensive care for acute pancreatitis were reviewed. The severity of illness was assessed by Acute Physiology and Chronic Health Evaluation (APACHE) II score on admission [5, 6]. Because many of the patients were transferred to our institution several days after the onset of acute pancreatitis, use of the Ranson score [7] was usually not appropriate. Data collection included information on patient demographics, cause of pancreatitis, laboratory values, complications, length of stay (ICU and hospital), and results of intravenous contrast-enhanced computed tomography (CT) scans. Details of operative interventions were also recorded. A cost calculation was done for each patient according to hospitalization, operations and procedures, and diagnostic studies. To determine the cost effectiveness, we evaluated the effective cost per survivor by adding the costs incurred by all patients (survivors and nonsurvivors) and dividing the sum by the total number of survivors [8]. The quality of life after discharge from the hospital was assessed by using the Short Form-36 General Health Survey (SF-36). The SF-36 health survey includes questions related to both physical and mental health. Questions regarding physical health include physical functioning (PF), role physical (RP), bodily pain (BP), and general health (GH). Mental health issues include vitality (V), social functioning (SF), role emotional (RE), and mental health (MH). This survey also includes questions for the subject to compare his or her current condition with that of 1 year previously. The results of this survey were compared with published data on a similar-aged control group (ages 45–54) [9]. Statistical analysis was performed by using BMPD software. Comparisons were made between survivors and nonsurvivors using a two-tailed Student t test for continuous variables and Kruskal– Wallis test for nonparametric variables.
RESULTS
There were 39 survivors and 13 nonsurvivors. The mean age of the patients was 52.5 (range, 22– 89) years, 52 for survivors and 58 for nonsurvivors [not significant (NS)]. Sixty percent of the patients were men (22 of 39 survivors and 9 of 13 nonsurvivors, NS). The most common etiologies of pancreatitis were gallstones (44%) and alcohol (17%) (Fig. 1). The etiology could not be identified in 6 patients (12%). APACHE II scores during the first 24 h after admission to the ICU averaged 11 ⫾ 6. These scores were significantly worse in nonsurvivors than in survivors (16 ⫾ 4 vs 10 ⫾ 6,
FIG. 1.
Causes of pancreatitis in survivors and nonsurvivors.
respectively, P ⬍ 0.05). Laboratory values on admission were similar in survivors and nonsurvivors, except for alkaline phosphate and ␦GTP, which were statistically higher in survivors on admission (Table 1). Bolus intravenous contrast-enhanced CT revealed pancreatic necrosis or phlegmon in 29 (56%) patients (Table 2). The remainder had edema. Twenty-four (46%) patients were treated with supportive medical therapy only; and 28 (54%) patients underwent surgical procedures, with 17 undergoing multiple debridements (Table 3). The indications for debridement were infected necrosis in 12 patients, necrosis (not proven to be infected) and deterioration despite aggressive supportive care in 6, abscess in 3, and bleeding in 1. There were no statistical differences between survivors and nonsurvivors in CT findings or treatment modalities. Pulmonary failure (defined as requiring greater than 24 h of mechanical ventilation) and renal failure (requiring dialysis) were both common complications observed in patients with acute pancreatitis: 27 patients (52%) required mechanical ventilation, and 13 patients (25%) underwent dialysis (Table 4). Maximum serum creatinine level was significantly higher in nonsurvivors (4.4 ⫾ 2.7 mg/dl) than in survivors (2.3 ⫾ 2.5 mg/dl) (P ⬍ 0.05). Hospital mortality was 25% (n ⫽ 13). The cause of death was sepsis in 8 patients (4 retroperitoneal/ pancreatic, 1 abdominal, 3 other sources), multiple organ system dysfunction in 6, and the systemic inflam-
91
SORAN ET AL.: QUALITY OF LIFE AFTER SEVERE ACUTE PANCREATITIS
TABLE 1 Laboratory Values of Patients on Admission to the Intensive Care Unit
Glucose (mg/dl) Total bilirubin (mg/dl) LDH (U/dl) ␥GTP (U/dl) a SGOT (U/dl) SGPT (U/dl) Alkaline phosphatase (U) a Amylase (U) Calcium (mg/dl) Sodium (meq/ml) Potassium (meq/ml) Platelets (⫻ 1000) WBC count (⫻ 1000) Hematocrit (%) Creatinine (mg/dl) Blood urea nitrogen (mg/dl) a
Overall (n ⫽ 52)
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
170.9 ⫾ 84.9 2.4 ⫾ 3.1 454.4 ⫾ 402.4 325.5 ⫾ 408.4 184.9 ⫾ 542.1 107.9 ⫾ 145.8 182.3 ⫾ 166.2 716.6 ⫾ 743.1 7.7 ⫾ 1.5 136.3 ⫾ 7.2 4.2 ⫾ 0.7 311 ⫾ 176.6 17.8 ⫾ 9.3 33.2 ⫾ 7.2 2.1 ⫾ 1.8 34 ⫾ 23.1
160.5 ⫾ 76.9 2.7 ⫾ 3.4 382.2 ⫾ 224.1 376.9 ⫾ 456.1 153.6 ⫾ 483.7 97.7 ⫾ 113.6 207.9 ⫾ 184.5 727.1 ⫾ 726.1 7.7 ⫾ 1.5 135.5 ⫾ 6.3 4.2 ⫾ 0.7 328 ⫾ 185 17.9 ⫾ 9 32.6 ⫾ 7.1 1.8 ⫾ 1.8 31.4 ⫾ 22.2
201.8 ⫾ 102.1 1.5 ⫾ 1.3 647.2 ⫾ 684.1 171.3 ⫾ 125.3 278.8 ⫾ 703.8 138.5 ⫾ 219.5 107.5 ⫾ 43.5 686.2 ⫾ 820.9 7.7 ⫾ 1.6 138.7 ⫾ 9.4 4.2 ⫾ 0.6 260 ⫾ 141 17.2 ⫾ 10.3 34.9 ⫾ 7.5 2.8 ⫾ 1.7 41.9 ⫾ 24.9
P ⬍ 0.05 comparing survivors with nonsurvivors.
matory response syndrome in 1. One patient had necrosis of the cecum. Other contributing factors included respiratory failure, renal failure, coagulopathy, hypovolemia, and cardiogenic shock. Some patients had multiple causes. The median hospital stay was 27 days with a mean of 39 ⫾ 33 days (40 ⫾ 34 in survivors and 38 ⫾ 34 in nonsurvivors, NS). Median intensive care unit stay was 9 days, with a mean of 18 ⫾ 23 days (15 ⫾ 18 in survivors and 28 ⫾ 31 in nonsurvivors, NS). The overall mean hospital cost was $96,891 ⫾ 92,152 ($136,730 ⫾ 95,045 for nonsurvivors and $83,611 ⫾ 88,434 for survivors, NS). Overall, the ICU and nonICU rooms accounted for 41% of all hospital costs (Fig. 2). Operation and anesthesia (exclusive of professional fees) was only 6%. The estimated cost to obtain one hospital survivor was $129,188. Twenty-one of thirty-nine survivors were contacted by telephone survey. Of the 18 who were not contacted, 5 had died in long-term follow-up, and 13 were lost to follow-up. The SF-36 surveys were conducted a median of 42 (range, 17– 69) months after discharge. Data from
the SF-36 survey show that the quality of life of acute pancreatitis survivors was generally good. When compared with an age-matched (45–54 years) control population, there were no significant differences in either the physical or mental health domains (Fig. 3); i.e., quality of life was as expected for their ages. Twenty of twenty-one patients (95%) stated that their general health was about the same (10) or better (8 much better, 2 somewhat better) than it had been 1 year previously (1 much worse).
TABLE 2
Operative and Medical Management of Patients with Severe Acute Pancreatitis
DISCUSSION
In this study of patients who required ICU admission for severe, acute pancreatitis, we found that these patients often required operative intervention and often developed infectious complications. They frequently developed organ system failure, particularly pulmonary and renal, necessitating mechanical ventilation and hemodialysis, respectively. Hospital costs, TABLE 3
Computed Tomography Findings during Severe Acute Pancreatitis
Phlegmon Necrosis Edema/inflammation
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
9 (23%) 13 (33%) 17 (44%)
2 (15%) 5 (39%) 6 (46%)
Treatment
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
Overall (n ⫽ 52)
Nonoperative Debridement Drainage Resection Laparotomy
19 (49%) 17 (44%) 2 (5%) 1 (2%) —
5 (38%) 5 (38%) 1 (8%) 1 (8%) 1 (8%)
24 (46%) 22 (42%) 3 (6%) 2 (4%) 1 (2%)
92
JOURNAL OF SURGICAL RESEARCH: VOL. 91, NO. 1, JUNE 1, 2000
TABLE 4 Short- and Long-Term Complications of Patients with Acute Pancreatitis Complications
Survivors (n ⫽ 39)
Nonsurvivors (n ⫽ 13)
Pulmonary failure a Renal failure b Sepsis/septic shock Infected pancreatic necrosis Pancreatic abscess Retroperitoneal abscess Intraabdominal abscess Pancreatic pseudocyst Pancreatic fistula Encephalopathy Others c
16 6 4 10 2 2 3 12 8 5 4
11 7 8 3 1 2 0 2 0 1 18
Required ⬎24 h of mechanical ventilation. Required hemodialysis. c Others (one patient each): upper gastrointestinal bleeding, bile duct obstruction, colonic obstruction, gastric outlet obstruction, deep venous thrombosis, pulmonary embolism, acute cholecystitis, cholangitis, ascites, postoperative hernia, ischemic necrosis of cecum, unstable angina, coagulopathy, persistent hypercalcemia. a b
particularly room costs, were high. This reflects prolonged ICU and hospital stays. Mortality was 25%, which is similar to the results of others [10 –13]. Despite prolonged hospitalization and multiple complications, survivors had long-term functional outcome that
FIG. 2. Distribution of hospital costs for survivors and nonsurvivors of severe acute pancreatitis.
FIG. 3. Results of the SF-36 survey for survivors of severe acute pancreatitis. Comparison is made to a historical normal control group, ages 45–54 [9].
was quite acceptable. The data from the SF-36 survey suggest that compared with their peers, these patients have no significant limitations secondary to physical or emotional problems. Only one respondent noted that his general health was worse than it had been 1 year previously. Since most physicians caring for patients with severe pancreatitis only see the patients when they are quite ill and debilitated, it is encouraging to know that, despite multiple operations and complications, they can recover and return to their preillness lifestyles. This study is limited by the fact that it was retrospective and did not include large numbers of patients. With greater numbers of patients, perhaps, more subtle quality of life issues would have been delineated from the SF-36 data. In addition, follow-up in this study was not complete. It is possible that those patients who were lost to follow-up did not have as positive an outcome as those who responded to our inquiry. At worst, if all of these patients had poor outcomes, this would mean that 33% of long-term survivors did not have acceptable outcomes. The fact that five patients who initially survived their hospitalization died later is worth noting. Patients with severe pancreatitis frequently develop multiple complications and organ system dysfunction. Even if they leave the hospital, they often require continuing nursing care and rehabilitation. They remain at risk for complications and even death. Longterm survival (at least 6 months), not just survival to discharge or 30 days, needs to be considered. There is a sense that survival of these patients has improved in recent years. Most likely this improvement is the result of a combination of factors, including aggressive surgical management of the necrotic pancreas and better supportive care of the other vital or-
SORAN ET AL.: QUALITY OF LIFE AFTER SEVERE ACUTE PANCREATITIS
gans that are frequently dysfunctional as a result of the systemic inflammatory response initiated by the pancreatitis. Regarding debridement of the necrotic pancreas, our general policy has been to debride patients with infected necrosis and those who deteriorate despite aggressive supportive care, even if infection has not been proven. Previous studies of patients with severe pancreatitis have focused on surgical complications. Nonsurgical complications have rarely been examined. In this study, most patients who required ICU admission for severe acute pancreatitis had failure of at least one organ system and suffered at least one complication not directly related to the pancreas per se. These complications were the major reason for mortality since deaths were almost always after withdrawal of support when it was apparent that continued therapy was futile in terms of the potential for restoring an acceptable quality of life. One recent study by Broome et al. [3] examined the quality of life (using the SF-36) after pancreatic debridement for pancreatic necrosis, compared with that after medical management of chronic pancreatitis and after operative management of pancreatic ductal abnormalities. They found that care for patients with pancreatic necrosis was expensive, but that the quality of life was good, thus justifying the expense. The comparison groups could be questioned. Patients undergoing elective operation for ductal abnormalities would not be expected to have the physiologic derangements seen in patients with acute pancreatitis, nor would they be expected to require postoperative rehabilitation. Similarly, patients treated medically for chronic pancreatitis would also present a significantly different patient population, particularly because of the psychologic overlay that complicates treatment of chronic diseases causing pain. These authors also compared their patients with patients with chronic medical diseases (congestive heart failure and chronic obstructive pulmonary disease). Issues with these chronic diseases would be expected to be different from those associated with acute pancreatitis. Doepel et al. [4] found good long-term functional outcomes in survivors of severe acute pancreatitis. This study, however, focused on physiologic outcome, not quality of life from the patient’s perspective. Neither of these studies [3, 4] analyzed resource utilization in the nonsurvivors. In the present study, nonsurvivors of acute pancreatitis, like survivors, had prolonged ICU and hospital lengths of stay and accumulated hospital costs greater than $100,000 on average. Since most patients survived to discharge, the average cost to obtain a hospital survivor was not much greater than this. Looking at the issue solely from the perspective of survivors, Fenton-Lee and Imrie [14] also found that the costs
93
were reasonable considering the good long-term outcomes. They further estimated that care for these patients resulted in a benefit of 8.55 quality-adjusted life years on average per patient. Although this study examined long-term quality of life, it did not look at the related question of long-term costs of care after discharge from the hospital. Future studies of long-term outcome from severe pancreatitis need to evaluate these costs, e.g., skilled nursing care and rehabilitation. Because of the frequency of complications and high costs of care noted above, predicting outcome from severe pancreatitis has been of interest for more than 20 years. Ranson and Pasternack [7], developed a system based on 11 parameters at the time of admission and over the first 48 h of care for predicting risk of complications and death from acute pancreatitis. The more generic APACHE II system has also been successfully applied to this patient population [6]. Because patients often spent a significant period at a community hospital prior to transfer to our institution, in this study, data for the Ranson criteria were usually not available. Consequently, we used APACHE II. We found that survivors had significantly better APACHE scores than nonsurvivors. More data on the predictive value of APACHE in these patients are needed, particularly with the current availability of APACHE III [15]. In conclusion, patients with severe acute pancreatitis often require admission to an ICU. These patients often spend weeks or months in the ICU and hospital, resulting in hospital costs that can be enormous, frequently greater than $100,000. They also need longterm care after discharge from the hospital and are still at risk for complications and death. The overall expense is justified given that the great majority of patients who survive have a good long-term quality of life. REFERENCES 1.
Banks, P. A. Infected necrosis: Morbidity and therapeutic consequences. Hepatogastroenterology 38: 116, 1991.
2.
Bradley, E. L. III. A fifteen year experience with open drainage for infected pancreatic necrosis. Surg. Gynecol. Obstet. 177: 215, 1993.
3.
Broome, A. H., Eisen, G. M., Harland, R. C., et al. Quality of life after treatment for pancreatitis. Ann. Surg. 223: 665, 1996.
4.
Doepel, M., Eriksson, J., Halme, L., et al. Good long-term results in patients surviving severe acute pancreatitis. Br. J. Surg. 80: 1583, 1993.
5.
Knaus, W. A., Draper, E. A., Wagner, D. P., and Zimmerman, J. E. APACHE II: a severity of disease classification system. Crit. Care Med. 13: 818, 1985.
6.
Larvin, M., and McMahon, M. J. APACHE-II score for assessment and monitoring of acute pancreatitis. Lancet July: 201, 1989.
7.
Ranson, J. H., and Pasternack, B. S. Statistical methods for
94
8.
9. 10. 11.
JOURNAL OF SURGICAL RESEARCH: VOL. 91, NO. 1, JUNE 1, 2000 quantifying the severity of clinical acute pancreatitis. J. Surg. Res. 22: 79, 1977. Chalfin, D. B., Cohen, I. L., and Lambrinos, J. The economics and cost-effectiveness of critical care medicine. Intensive Care Med. 21: 952, 1995. Ware, J. E., Jr. (Ed.). SF-36 Health Survey Manual and Interpretation Guide. Boston: Nimrod Press, 1993. Warshaw, A. L., and Gongliang, J. Improved survival in 45 patients with pancreatic abscess. Ann. Surg. 202: 408, 1985. Fugger, R., Schultz, F., Rogy, et al. Open approach in pancreatic and infected pancreatic necrosis: Laparostomies and preplanned revisions. World J. Surg. 15: 516, 1991.
12.
Miller, B. J., Henderson, A., Strong, R. W., et al. Necrotizing pancreatitis: Operating for life. World J. Surg. 18: 906, 1994.
13.
D’Egidio, A., and Schein, M. Surgical strategies in the treatment of pancreatic necrosis and infection. Br. J. Surg. 78: 133, 1991.
14.
Fenton-Lee, D., and Imrie, C. W. Pancreatic necrosis: Assessment of outcome related to quality of life and cost of management. Br. J. Surg. 80: 1579, 1993.
15.
Knaus, W. A., Wagner, D. P., Draper, E. A., et al. The APACHE III prognostic system: Risk prediction of hospital mortality for critically ill hospitalized adults. Chest 100: 1619, 1991.