Pancreatic surgery and length of hospital stay: meta-analysis of randomized clinical trials.

Abstract

Background: Length of hospital stay (LOS) is a commonly reported postoperative outcome measure in pancreatic surgery. However, LOS may be influenced by various factors, including healthcare systems, cultural traditions, and readmission rate. This systematic review and meta-analysis investigated LOS in pancreatic surgery.

Methods: A meta-analysis was conducted to identify all randomized clinical trials (RCTs) of pancreatic surgery reporting LOS. Analyses were conducted using a random-effects model with mean differences as the effect estimator for the continuous outcome LOS.

Results: In all, 186 RCTs published between 1994 and 2025, comprising 28 381 patients, were included. The mean LOS differed significantly by type of resection (P = 0.003) and by country (P < 0.001). Country remained a statistically significant factor after adjusting for readmission rate, year of study, and type of resection. The shortest LOS was observed in the USA, with a pooled value of 11 days (95% confidence interval (c.i.) 9.93 to 12.32) and the longest was seen in Japan, with a pooled value of 28 days (95% c.i. 25.60 to 31.67). Subgroup analysis of studies comparing minimally invasive and open surgery showed a mean LOS difference of 2 days (95% c.i. -2.79 to -0.85; prediction interval -7.70 to 4.06). However, this effect varied across countries, with mean LOS differences ranging from 1 to 6 days.

Conclusion: LOS after pancreatic surgery exhibited substantial statistical heterogeneity based on country, the type of operation, and access. This highlights the importance of considering contextual factors when evaluating and comparing LOS as an outcome measure in pancreatic surgery.