This systematic review followed the Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines (11). Our protocol was registered with the International Prospective Register of Systematic Reviews (CRD42022310131).
Search strategy
We performed the electronic database search on Pubmed, EMBASE, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, and Ichushi Web (Japanese). We included all relevant studies published from 1968 up to May 3, 2022. The lower range of the publication year was set as 1968 since KP became well-recognized globally after a seminal publication was released during this year (12). MeSH terms including "biliary atresia", "portoenterostomy”, “hepatic", "Kasai procedure", "surgery" and other essential keywords were include in the search strategy. A medical information services/literature search expert developed the search strategy by combining conditions for disease name, intervention name, study design, and search period. The full search strategy used is shown in Supplementary Table S1.
Inclusion & exclusion criteria
Studies which examined BA patients with regards to an intervention or exposure definition related to the timing of KP at age 30, 45, 60, 75, 90, 120, and/or 150 days were included. The main outcome measures of interest were NLS rates at 5, 10, 15, 20, and 30 years post-KP and the hazard ratio (HR) or risk ratio (RR) for NLS in relation to the age of KP. The included studies had to be either a randomized controlled trial or an observational study. We included studies published in English and Japanese. Exclusion criteria included articles that were not published in a peer reviewed journal. Animal studies and case reports were excluded. Unpublished documents and grey literature including conference papers, dissertations, and patents were also excluded.
Study selection and data extraction
The relevant papers from the literature were managed in Rayyan (13) to facilitate the screening and selection process. Screening based on literature abstracts was performed blinded by two independent reviewers (E.H., Y.M.), and inconsistencies between the reviewers that occurred during the process were resolved through discussion or consulting the third reviewer (K.S.). The included studies were assessed for their quality of reporting evidence, and their data was extracted and summarized. Two reviewers (E.H., Y.M.) independently extracted data from the included studies and disagreements were resolved by discussion or consulting the third reviewer (N.S.).
Quality assessment
The quality of the included studies were assessed based on the Cochrane tool known as the ROBINS-I tool (14). Both E.H. and K.S. independently assessed the included papers against these checklists and any discrepancies were resolved by discussion. The response options for an overall assessment were, (1) low risk of bias: the study is comparable to a well-performed randomized trial; (2) moderate risk of bias: the study provides sound evidence for a non-randomized study but cannot be considered comparable to a well-performed randomized trial; (3) serious risk of bias: the study has some important problems; (4) critical risk of bias: the study is too problematic to provide any useful evidence and should not be included in any synthesis; and (5) no information on which to base a judgement about risk of bias.
Data synthesis
If multiple articles were identified from the same country, the national data (e.g., nationwide registry) was given priority for data synthesis to avoid double counting of data. We included a general summary table that quantifies the main characteristics of the included studies, including study design, timing of KP, study year, country, number of patients, gender, mean age of patients and outcome measures. Where appropriate, a meta-analysis was conducted using the DerSimonian and Laird random-effect model for HR analyses and the restricted maximum likelihood random effect model for RR analyses using STATA (17, StataCorp LLC, College Station, TX) (15). Sensitivity analyses to examine the relationship between KP < 30 with KP > 31 days, > 61 days and > 91 days were performed. To assess the degree of heterogeneity between studies, we visually inspected the forest plots and estimated the I2 statistics; ≥75% of I2 statistics was deemed considerable (16). When the heterogeneity was considerable, we attempted to identify the source of heterogeneity.