What is the Optimal Timing of Refeeding in AP?
Optimal time of refeeding in AP has been investigated in several trials over the years. The traditional approach is to start oral refeeding after relief of abdominal pain and once serum levels of pancreatic enzymes return to normal. Several studies have questioned this approach. In a prospective and randomized trial our group compared two different protocols of refeeding and two different schedules for refeeding [11]. Four groups of patients with AP were defined according to the refeeding protocol (Table 11.1). The primary outcome of the study was length of hospital stay (LOHS); in addition, gastrointestinal symptoms after refeeding were evaluated. We did not find any significant difference in tolerance or gastrointestinal symptoms when comparing initiation of refeeding early, as soon as bowel sounds were present, or at standard time; when bowel sounds were present, there was no abdominal pain, no fever, decreasing serum lipase levels, and blood leukocyte counts had decreased to below 15 × 109/l. However, we observed a significant reduction in LOHS by two days in the early refeeding group. It seems therefore that refeeding after bowel sounds are present is a safe approach and well tolerated for patients with mild AP. Similar findings were reported from a Chinese randomized clinical trial, where refeeding started once patients felt hungry resulted in shorter LOHS compared to refeeding started at routine times, without any significant difference in adverse gastrointestinal events [12]. A German multicenter trial was unable to demonstrate any difference in LOHS when comparing initiation of refeeding in mild AP at the time self‐selected by patients or when serum lipase levels were below twice the upper limit of normal [3]. Eckerwall et al. [2] compared two protocols of oral refeeding in mild AP: immediate oral feeding ad libitum and traditional management by initial fasting followed by stepwise reintroduction of oral intake. LOHS was significantly shorter in the early refeeding group (4 vs. 6 days; P <0.05). However, this study does not allow differentiation of the individual importance of early reintroduction of refeeding and rapid step‐up protocol. All studies together clearly demonstrate that normalization of pancreatic enzyme levels is not a prerequisite to restart feeding. Early refeeding may shorten LOHS, but this was not consistently observed in all studies. The different definitions used for early and standard time for refeeding may explain this discrepancy to some extent.
Table 11.1 Oral refeeding protocols after acute pancreatitis.
| Protocol | Time of refeeding | Refeeding schedule |
|---|---|---|
| Standard time and stepwise schedule | Standarda | Stepwise increase from a liquid 1200 kcal/day diet, to a soft 1500 kcal/day diet and a solid 1800 kcal/day diet over at least three days |
| Early time and stepwise schedule | Earlyb | |
| Standard time and direct schedule | Standarda | Initial solid diet containing 1800 kcal/day |
| Early time and direct schedule | Earlyb |
a Once the following criteria are fulfilled: bowel sounds are present, no abdominal pain, no fever, no leukocytosis, and decreasing serum pancreatic enzyme levels.
b Once bowel sounds are present and pain in controlled with non‐opioid analgesics.
In our center, oral refeeding after AP is started as soon as bowel sounds are present and abdominal pain is controlled with non‐opioid analgesics.
How Should Oral Refeeding be Scheduled?
The initial meal given to patients with AP is considered to be important in determining whether reintroduction of oral intake is tolerated. Patients following the conventional stepwise refeeding protocol are traditionally started on a hypocaloric clear liquid diet and, if this first meal is well tolerated, soft diet (modified in texture, and in caloric and fat content) and solid low‐fat diet are introduced in a stepwise manner until the patient can tolerate a normal oral diet [13].
In the study performed by our group [11], a protocol for refeeding with a solid low‐fat diet (about 1800 kcal, 19 g of fat) versus standard stepwise increasing caloric diet over three days was evaluated (see Table 11.1). We found that a solid low‐fat diet from the start was similarly tolerated compared to stepwise increasing caloric intake, and it was associated with a shorter LOHS if associated with early refeeding.
Different protocols for refeeding in subjects with mild AP have been investigated in five previous randomized clinical trials [2,4–7]. Jacobson et al. [5] compared a clear liquid diet (588 kcal, 2 g of fat per day) to a low‐caloric, low‐fat diet (1200 kcal, 35 g of fat per day) in patients with mild AP and showed no difference in tolerance or LOHS. Moraes et al. [6] performed a study with three treatment arms comparing a hypocaloric clear liquid diet, an intermediate hypocaloric soft diet (both around 250 kcal, 2 and 4 g of fat, respectively) and a full solid diet (around 1200 kcal, 30 g fat per day) in patients with mild AP. No differences in pain relapse rates or LOHS between the three treatment arms were found. Sathiaraj et al. [7] compared refeeding with a clear liquid diet (458 kcal, 11 g of fat) to a soft diet (1040 kcal, 20 g fat per day) in patients with mild AP. LOHS was significantly reduced in the soft diet group. Finally, Rajkumar et al. [4] investigated clear liquid diet compared to soft diet. Total and post‐refeeding LOHS was shorter in the soft diet group. None of the previously published randomized clinical trials observed any increased risk of refeeding intolerance or other adverse events related to the more aggressive refeeding protocols [2,4,6,7].
What are the Predictors of Oral Feeding Intolerance in AP Patients?
There is significant concern about the relapse of gastrointestinal symptoms and pancreatitis following oral refeeding after AP since, the burden of oral feeding intolerance can be high. Some studies have shown that patients with oral feeding intolerance have significantly longer length of hospitalization [14–16], while others have demonstrated a reduced quality of life during hospitalization [17]. There is also evidence suggesting that these patients are at increased risk of early readmission if they are discharged with ongoing gastrointestinal symptoms, or are unable to tolerate a full diet at discharge [18].
A recent systematic review analyzed the current body of evidence and the incidence and predictors of oral feeding intolerance [19]. By evaluating 2024 patients in 22 studies these authors showed a global incidence of oral feeding intolerance of 16% (Table 11.2) [2–6,11,12,14–17,20–29]. The study found no relationship between the risk of developing oral feeding intolerance and age, sex, duration of symptoms before hospital admission, or etiology of AP. However, patients with blood lipase levels prior to refeeding of more than 2.5 times the upper limit of normal and those with (peri)pancreatic collections and pleural effusions were at increased risk of developing oral feeding intolerance [19]. However, daily monitoring of serum lipase levels is not currently recommended in clinical practice and is associated with additional time and financial costs. Furthermore, the impact of monitoring serum lipase levels on risk of developing oral feeding intolerance has not been shown in other studies [11]. On the other hand, the practical significance of (peri)pancreatic collections as potential predictors of oral feeding intolerance