Evaluating the literature on preoperative androgen stimulation for hypospadias repair using the fragility index – can we trust observational studies?

BACKGROUND: Preoperative androgen stimulation (PAS) is typically used in hypospadias repair for patients with a proximal meatus or small glans size. Hypospadias PAS literature suffer from small sample sizes and lack of power to claim robust conclusions. Small changes in the number of events may completely change the statistical significance, making the conclusions drawn unreliable. Fragility index (FI) is the number of additional events needed to occur in either the control or experimental group to turn a statistically significant result to a non-significant result. The objective of the report was to assess the quality of available literature revolving around PAS use in hypospadias repair and its effects on post-operative complication rates using FI. METHODS: A comprehensive search of MEDLINE, EMBASE, and grey literature (ESPU and SPU abstracts) was conducted to identify RCTs and observational studies investigating the effect of PAS on complications post-hypospadias repair between 1990 and 2020. The FI was calculated for each study. Postoperative complications were defined as: fistula, stricture/stenosis, diverticula, and dehiscence. The odds ratio (OR), 95% confidence intervals (CI), corresponding p-values was calculated for each study. A random effects mixed model was implemented to combine the ORs for each study design. RESULTS: Fourteen studies qualified for inclusion, of which nine were observational studies and five were RCTs (Figure 1). The median sample size was 110 patients (IQR 69-171). The summary ORs for observational studies was 1.74 (95% CI: 1.10 to 2.74; p = 0.020) and for RCTs was 0.71 (9% CI: 0.34 to 1.47; p = 0.350). The median FI was 0 (IQR 0-2) of the included studies. DISCUSSION: PAS use does not appear to significantly affect complication rates shown in RCTs, however, observational studies cumulatively suggested significantly greater odds of complications after PAS. The FI is best used for RCTs with 1-to-1 randomization and binary data. Observational studies are rarely balanced for demographics and comorbidities with unequal sample size between comparable groups. The study was limited by substantial variability in how PAS was delivered to patients, leading to restricted comparability. CONCLUSION: Strong conclusions regarding the influence of PAS on hypospadias repair outcomes cannot be properly drawn based on the current literature due to deficits from either a statistical or methodological standpoint. The current PAS literature has shown inconclusive results, calling for well-designed RCTs, involving standardized surgical techniques and PAS protocols, to evaluate the true effect of PAS on complications post-hypospadias repair.