Clinical trial: oral feeding with a soft diet compared with clear liquid diet as initial meal in mild acute pancreatitis

Transcription

1 Alimentary Pharmacology & Therapeutics Clinical trial: oral feeding with a soft diet compared with clear liquid diet as initial meal in mild acute pancreatitis E.SATHIARAJ*,S.MURTHY*,M.J.MANSARD, G.V.RAO, S. MAHUKARà &D.N.REDDY *Departments of Clinical Nutrition; Surgical Gastroenterology; àbasic Sciences and Gastroenterology, Asian Institute of Gastroenterology, Somajiguda, Hyderabad, India Correspondence to: Ms E. Sathiaraj, Department of Clinical Nutrition, Asian Institute of Gastroenterology, , Somajiguda, Hyderabad , India. esther.sathiaraj@gmail.com Publication data Submitted 29 May 2008 First decision 6 June 2008 Resubmission 13 June 2008 Accepted 1 July 2008 Epub Accepted Article 4 July 2008 SUMMARY Background In mild acute pancreatitis, traditional treatment has been initial fasting and oral refeeding with clear liquids to prevent adverse gastrointestinal events such as pain. The diet is gradually progressed to soft solids and hospital discharge is planned based on patients tolerance to a solid diet. Aim To determine the length of hospitalization and tolerance to oral refeeding when initiated on a soft diet as compared to a clear liquid diet. Methods One hundred and one patients with mild acute pancreatitis were randomized to receive either a clear liquid diet or soft diet when oral feeding was initiated. Frequency of pain, total and postrefeeding length of hospitalization, and dietary intake were monitored. Hospital discharge was decided by the medical team without input from the study coordinators. Results A statistically significant decrease in the length of hospitalization (total and postrefeeding) of a median of 2 days was seen in patients receiving a soft diet (P < 0.001). No significant difference in the need for cessation of diet because of pain was observed between the two groups. Patients initiated on a soft diet consumed significantly more calories and fats on study day 1 (P < 0.001). Conclusion Oral refeeding with a soft diet in patients with mild acute pancreatitis can be considered safe and can result in shorter length of hospitalization. Aliment Pharmacol Ther 28, ª 2008 The Authors 777 doi: /j x

2 778 E. SATHIARAJ et al. INTRODUCTION Acute pancreatitis is an inflammatory disease of the pancreas resulting in hospitalizations. In all, 80% of the patients admitted to the hospital with acute pancreatitis are classified as mild, which resolves in an average of 5 days. 1, 2 The treatment of mild acute pancreatitis is largely supportive including effective pain control, fluid therapy and discontinuation of oral intake until there is near complete resolution of abdominal pain and tenderness. 3 The practice of discontinuing oral intake until pain is resolved is based on the fact that premature feeding may cause an exacerbation. 4 Reestablishment of oral feeding is considered as soon as abdominal pain, nausea and vomiting have subsided and bowel sounds have returned. Feeding is begun using a clear liquid diet (CLD), which is gradually advanced to a soft, low fat diet as tolerated by the patient (e.g. no recurrence of pain or vomiting). Subsequently, if the patient continues to do well without increase in abdominal pain, the hospital discharge is planned. 5, 6 The idea to start oral feeding with liquids is based on a presumption that clear liquids should not exert marked stimulatory effects on pancreatic exocrine function if feeding is provided gradually according to individual tolerance. 7 This sequence, however, may prolong patient hospitalization and incur costs. A previous study hypothesized that patients with mild acute pancreatitis initiated on low fat soft diet (SD) would have a shorter length of hospitalization (LOH), although it did not show significant results 8. We assumed that patients with mild acute pancreatitis would be able to tolerate oral SD as the initial meal and would be discharged from hospital sooner compared with CLD. MATERIALS AND METHODS Consecutive patients with mild acute pancreatitis admitted for reasons of abdominal pain from September 2007 to February 2008 at the Asian Institute of Gastroenterology, which is a single specialty large referral centre, were included in the study based on the following inclusion criteria: (i) Amylase and or lipase greater than three times the upper limit of normal or greater than two times the upper limit and a computerized tomography scan showing unequivocal acute pancreatitis and peripancreatic inflammation. (ii) Mild acute pancreatitis [absence of pancreatic necrosis, abscess and pseudocyst, absence of organ dysfunction, hypotension (systolic <90 mmhg), hypoxaemia (oxygen saturation <90%)]. 9 Exclusion criteria were: (i) Patients with organ dysfunction and neoplasms, postsurgical patients, pregnant women, patients with infections such as TB, HIV AIDS, severe acute pancreatitis. (ii) Patients with acute pancreatitis who received enteral support via tube feeding or parenteral nutrition and who received parenteral narcotics for abdominal pain on the day of refeeding. (iii) Patients with acute on chronic pancreatitis who were on enzyme supplementation. Patients who met the inclusion criteria and consented for the study were randomly assigned to receive either CLD or SD. Randomization was done using a computergenerated random number list. Random allocation sequence, enrolment and assigning participants to the two groups were done by separate individuals. All the patients were kept on initial fasting as per the conventional treatment and reestablishment of oral feeding was considered as soon as abdominal pain, nausea and vomiting had subsided and bowel sounds had returned. Patients who were randomized to the CLD group initially received CLD, which progressed to the standardized SD when clear liquids were tolerated (e.g. there is no recurrence of pain or vomiting). The study day 1 CLD and SD were standardized to provide 458 kcal day, 11 g fat day and 1040 kcal day, 20 g fat day respectively (see Table S1). The diet advancement or cessation and hospital discharge were at the discretion of the medical team which was not involved in deciding the type of diet to be received by the patient. Patients were monitored daily until discharge by the clinical nutritionist who recorded dietary and clinical information, including diet tolerance, daily calorie and fat intake and pain scores using a 10-point Likert-type scale. Outcomes The primary outcome of the study was the LOH from the time of refeeding until discharge. Secondary outcomes included the frequency that the subjects discontinued oral feeding because of intolerance such as pain, nausea and vomiting. Statistical methods A sample size of 100 total patients was calculated as required to detect a reduction in the duration of

3 CLINICAL TRIAL: ORAL REFEEDING IN MILD ACUTE PANCREATITIS 779 hospital stay of 1 day with 85% power and a twosided P-value of The postrefeeding LOH and total LOH was compared between the study groups by using a Wilcoxon rank-sum test. The proportion of patients in each group who required cessation of diet postrefeeding was compared using Fisher exact test. In an exploratory analysis, univariate odds ratio with 95% confidence interval were calculated by Fisher Exact test and t-test methods to identify variables potentially related to cessation of feeding owing to pain, nausea or vomiting after study enrolment. Such variables included age, gender, body mass index, cause of pancreatitis and abdominal pain scores at the time of refeeding. Other variables compared between the study groups included the time between hospital admission and consumption of the first meal, calories consumed in the first meal and on study day 1, grams of fat consumed in the first meal and on study day 1 and the total number of meals consumed on study day 1. All statistical analyses were performed using SPSS version 13 (SPSS Inc., Chicago, IL, USA) and two-sided P-values <0.05 were considered significant. RESULTS AND DISCUSSION One hundred and one patients with mild acute pancreatitis were enrolled as per the inclusion criteria and randomized to either CLD (n = 52) or SD (n = 49) at the time of refeeding. All the patients received standardized diets and were not permitted to consume anything apart from the standardized diet on study day 1 and 2. Patient characteristics The baseline characteristics of the study participants are shown in Table 1. The patients age, gender, body mass index, cause of pancreatitis and type of pancreatitis (e.g. acute vs. acute on chronic) were similar in the CLD and SD groups. Length of hospitalization (LOH): primary outcome There was a significant difference in the postrefeeding LOH and total LOH between the two diets (Table 2). The median postrefeeding LOH was 6 days in CLD and 4 days in SD (IQR 4 8 in CLD; 2 5 in SD). The corresponding mean postrefeeding LOH were 6.75 (s.d. 3.37) and 4.18 (s.d. 2.86) days for the CLD and SD groups respectively. The difference in the mean postrefeeding Table 1. Baseline characteristics of the study participants Characteristic CLD (n = 52) SD (n = 49) Mean age (s.d.), years 39 (13) 37 (12) Male (%) 44 (86.3) 39 (79.5) Mean body mass 20.9 (3) 21.3 (4) index (s.d.) Cause of pancreatitis (%) Gallstones 9 (17.3) 7 (14.3) Alcohol 25 (48) 26 (53) Unknown 18 (34.6) 12 (24.5) Others 4 (8.2) Type of pancreatitis (%) Acute 41 (78.8) 40 (81.6) Acute on chronic* 11 (21.2) 9 (18.4) CLD, clear liquid diet; SD, soft diet; s.d., standard deviation. * Acute on chronic pancreatitis was defined as an exacerbation of inflammation superimposed on known chronic pancreatitis. 9 LOH was therefore 2.57 days (t = )413, 95% CI = )3.8 to )1.332, P < 0.001). The total LOH from hospital admission was also significantly different in both groups with SD group having a shorter total LOH (CLD: median 8 days, IQR 6 9 days, mean 8.71, s.d ; SD: median 5 days, IQR 4 7 days, mean 5.92, s.d ; P < 0.001). Despite all optimism and sufficient evidence that nutritional support via the gastrointestinal tract is important in the management of patients with acute pancreatitis, the role of the routine nutrition support, the way of delivery, and the type of diet are questions that Table 2. Outcomes of the clinical trial Outcome CLD (n = 52) SD (n = 49) P Cessation of feeding Because of pain (n) Because of nausea, vomiting (n) Length of hospitalization 6 (4 8) 4 (2 5) <0.001 postrefeeding [median (IQR)] Total length of hospitalization [median (IQR)] 8 (6 9) 5 (4 7) <0.001 CLD, clear liquid diet; SD, soft diet; IQR, interquartile range.

4 780 E. SATHIARAJ et al. still need to be answered. 10 Moreover, little attention has been paid to optimizing the dietary management of mild acute pancreatitis, although this form represents a majority of cases. 11 When to begin oral feeding is a question that occurs in every patient recovering from acute pancreatitis. Many clinicians initiate oral feeding when abdominal pain is absent and abdominal tenderness 8, 12 improves. Traditionally, patients resume feeding with a CLD, which is gradually advanced to a soft, low fat diet as tolerated by the patient, although when to refeed and the choice of diet is not based on good evidence in mild acute pancreatitis. 8 We hypothesized that patients with mild acute pancreatitis could be discharged from the hospital sooner if they resumed oral refeeding with SD as compared with CLD and it was seen that the postrefeeding LOH and total LOH were significantly lower in patients initiated on SD than in patients on CLD (P < 0.001). The CLD group was kept on liquid diet for an average of 2 days before SD was started, whereas the SD group directly began a SD and, having tolerated it well, discharge was planned. This could have contributed to a shorter LOH in the SD group compared with the CLD group. This suggests that initial refeeding with clear liquids may not be essential. Although our primary outcome (LOH) was significantly lower in SD group, the basis of hospital discharge is on subjective discretion of the medical team without any defined objective discharge criteria. We are unaware of validated criteria for either initiation of feeding or hospital discharge after mild acute pancreatitis. Therefore, based on the hospital practice, the timing of discharge was left to the medical team without inputs from the study coordinators. Oral feeding was not delayed, with a mean of 1.5 days between hospital admission and introduction of oral diet, suggesting that we did not wait for an excessive length of time before refeeding. The mean duration between hospital admission and refeeding was similar in both groups (P = 0.462). A previous study showed that pain occurred in 25% of patients during oral refeeding in patients with mild acute pancreatitis was associated with increased hospital stay and increased overall costs on disease treatment. 13 We observed failure to tolerate initial oral feeding because of pain in 11% of our study group. We were careful to exclude patients with organ failure, pancreatic necrosis and parenteral narcotics during refeeding to be assured of feeding tolerance of oral diet. We also hypothesized that a SD would not cause an excess number of patients to suffer pain as compared with a CLD and observed that there was no significant difference in the proportion of patients failing to tolerate SD as the initial meal in mild acute pancreatitis (four of SD, zero of CLD; P = 0.852). Cessation of an oral diet after refeeding because of pain, nausea and vomiting occurred in 11 patients (seven of CLD patients, four of SD patients). Neither diet was associated with significantly increased rate of feeding cessation (Table 2). There were five patients (4.9%) who stopped an oral diet after refeeding because of radiographic and endoscopic procedures as part of the inpatient management that was under the discretion of the medical team. The number of patients who had cessation of their diet after refeeding due to procedures was similar in both groups (three patients in CLD; two patients in SD group). We also did a subgroup analysis between the diabetic and nondiabetic patients of the two groups and found that there was no significant difference in the tolerance of the diets (P = 0.792) and LOH (P = 0.861). No such difference was Dietary factor CLD (n = 52) SD (n = 49) P Time between hospital admission 1.4 (1.3) 1.6 (1.3) 0.46 and first meal Total number of meals consumed 4 (0.82) 3 (0.61) per patient on study day 1 Calories consumed in first meal 137 (40.6) 262 (46.1) <0.001 Fat (grams) consumed in first meal 4 (2) 3 (1.3) 0.34 Calories consumed on study day (107) 921 (65.3) <0.001 Fat (grams) consumed on study day 1 8 (3.3) 15 (3.05) <0.001 Table 3. Dietary data of patients Values are expressed as mean (s.d.). CLD, clear liquid diet; SD, soft diet.

5 CLINICAL TRIAL: ORAL REFEEDING IN MILD ACUTE PANCREATITIS 781 observed between alcoholic and non-alcoholic patients too (LOH: P = 0.58; intolerance: P = 0.068). The total number of meals consumed on study day 1 was similar in both groups (Table 3). Dietary data on calories and grams of fat consumed on the first meal and study day 1 were assessed in all patients in both groups. Patients randomized to SD consumed significantly more calories in both their first meal and study day 1 and significantly more grams of fat on study day 1 compared with patients randomized to CLD (P < 0.001) (Table 3). In summary, resumption of oral feeding after mild acute pancreatitis with a SD significantly shortened the LOH when compared with a CLD. We also found that SD did not appear to cause significant feeding intolerance, provided significantly more calories and nutrients when compared with CLD and therefore may be considered as an option to initiate patients with mild acute pancreatitis. ACKNOWLEDGEMENTS The authors express their gratitude to all the participants in the study. The help rendered by Ms Saba Firdous, Ms C. S. Madhulika and Ms Gowher Fatima in data collection and analysis is gratefully acknowledged. Declaration of personal and funding interests: None. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Table S1. Nutritional values of (a) clear liquid diet and (b) soft diet 14 Please note: Blackwell Publishing are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article. REFERENCES 1 Uomo G. Pancreatic head mass: How can we treat it? Acute pancreatitis: conservative treatment. J Pancreas 2000; 1: O Keefe SJD, Lee RB, Anderson FP, et al. The physiologic effects of enteral and parenteral feeding on pancreatic enzyme secretion in humans. Am J Physiol 2003; 284: Marshall JB. Acute pancreatitis. Review with an emphasis on new developments. Arch Intern Med 1993; 153: Petrov MS, van Santvoort HC, Besselink MGH, et al. Oral refeeding after onset of acute pancreatitis: a review of literature. Am J Gastroenterol 2007; 102: Abou-Assi S, O Keefe SJD. Nutrition support during acute pancreatitis. Nutrition 2002; 18: Scalapio JS, Chowla NM, Ukleja A. Nutrition supplementation in patients with acute and chronic pancreatitis. Gastroenterol Clin North Am 1999; 28: Pupelis G, Snippe K, Plaudis H, Rudakovska M. Early oral feeding in acute pancreatitis: an alternative approach to tube feeding. Preliminary report. Acta Chir Belg 2006; 106: Jacobson BC, Vliet MBV, Hughes MD, et al. A prospective randomized trial of clear liquids versus low fat solid diet as the initial meal in mild acute pancreatitis. Clin Gastroenterol Hepatol 2007; 5: Mark F, Lawrence FS, Lawrence BJ. Sleisenger and Fordtran s Gastrointestinal and Liver Diseases, 8th edn. Philadelphia, PA: WB Saunders Company, 2006: 1242, Dervenis CH. Enteral nutrition in severe acute pancreatitis: future development. J Pancreas 2004; 5: Eckerwall GE, Tingstedt BB, Bergenzaun PE, Andersson RG. Immediate oral feeding in patients with mild acute pancreatitis may accelerate recovery a randomized clinical study. Clin Nutr 2007; 26: Whitcomb D. Acute pancreatitis. N Engl J Med 2006; 354: Chebli JMF, Gaburri PD, DeSouza AFM, et al. Oral refeeding in patients with mild acute pancreatitis: prevalence and risk factors of relapsing abdominal pain. J Gastroenterol Hepatol 2005; 20: Gopalan CS, Rama Sastri BV, Balasubramanian SC. Nutritive Value of Indian foods, 2nd edn. Printed by National Institute of Nutrition (NIN), Indian counsel of Medicinal Research (ICMR), Hyderabad, 1971: Revised and updated by Narsinga Rao BS, Dosthale YG, Pant KC, reprint, 2004.