Case reports and case series—worth being recognized?

When analyzing a global database, it is advised to exclude all case reports and case series of pilonidal sinus disease (PSD) due to the potential bias posed by small series in consecutive recurrence rates (RR). Emerging data indicate that lower annual caseloads in PSD surgery are associated with poorer outcomes, specifically higher RRs, which may be particularly pronounced in small patient series. Consequently, we investigated the RRs of all case reports and case series involving fewer than ten patients, comparing them in terms of therapy method, time since surgery, and study design against established large studies published to date.

A comprehensive meta-analysis comparing RRs was performed, identifying 1,338 studies encompassing 153,098 patients. Recurrence rates at 5‑year and 10-year follow-ups were compared between cohort studies with ten or more patients, case reports, and case series with fewer than ten patients. Bias estimation was performed using the Newcastle–Ottawa scale for cohort studies.

In total, 1338 studies with 153 randomized controlled trials (RCTs) and 1124 non-RCT and non-retrospective observational (non-ROR) studies and 61 ROR studies were analyzed, including 72 case reports and 133 case series groups, encompassing 978 patients. In the non-RCT analysis, the 5‑year RR across all treatments was 11.4% (confidence interval [CI] 11.1–11.6) for cohort studies with ten or more patients and 20.6% (CI 15.5–25.4) for case reports and case series with fewer than ten patients. Procedure-specific 5‑year RRs were 7.4% and 5.6% in all flap treatments, 7.2% and 20.5% in primary open closure, and 13.8% and 33.5% in primary midline closure for cohort studies and case reports/case series, respectively.

The RRs of case reports and case series differ significantly from those of cohort studies, as evidenced by all four comparisons. Therefore, the exclusion of such studies is justified in order to achieve higher reliability when estimating RRs in PSD surgery meta-analyses.