Unveiling Observation Bias in Pilonidal Sinus Disease Studies: A Comparative Analysis of Study Designs and Follow-Up Protocols.

Abstract

<b>Introduction:</b> The extent of observation bias in long-term surgical outcomes due to inadequate or unplanned follow-up in pilonidal sinus disease (PSD) studies is unclear. We made a hypothesis on the high risk of underreported recurrence in studies relying on the patients returning in case of recurrent disease (return on recurrence; ROR).<b>Aim:</b> To investigate and quantify the amount of bias associated with regular <i>vs.</i> return-on-recurrence follow-up of patients.<b>Materials and methods:</b> A total of 5.485 retrievable PSD publications were screened for eligibility, yielding 1.222 PSD studies with 135.349 patients treated, published between 1833 and 2023, included for analysis. Of these, 139 were Randomized Controlled Trials (RCT), 54 were ROR trials, and 1,029 were non-RCT non-ROR (N) trials. Recurrence rates were compared between groups<b>Results:</b> The five-year recurrence rates across all treatments in PSD were 18.9% for RCTs, 10.4% for N trials, and 12.4% forROR trials. Recurrence rates in the N and the ROR trials were statistically indistinguishable at ten-year follow-up (p = 0.1),whereas RCTs show significantly higher recurrence rates than N or ROR at 10-year follow-ups (p<0.001, p<0.001). Comparingonly primary open treatment, the five-year recurrence rate was 18.6% for RCT trials, 11.5% for N trials, and 4.1% for ROR trials.The recurrence rates at 10-year follow-up in the ROR and N trials were 20.1% and 19.7%, respectively. Notably, there was nodata available for 10-year follow-up in RCTs.<b>Conclusions:</b> Observation bias seems to significantly impact the results of ROR studies. Implementing a well-structured, all-encompassing patient follow-up at specific time intervals can effectively mitigate this bias, which might otherwise compromise the validity of our findings.<b>Significance of work:</b> The extent of observation bias in ROR studies compared to RCTs and non-RCTs remains unknown. Ourscrutiny of excision and primary open therapy, the largest therapeutic cohort globally, reveals ROR recurrence rates potentiallydeflated by three to fourfold relative to RCTs or non-RCT findings.