Published surgical success rates in pediatric urology--fact or fiction?

Parents counseled for surgery are quoted operative risks based on published results in the literature. However, outcomes from single surgeon or single institution retrospective studies are not generalizable. We assessed whether published outcomes were perceived to be representative of personal practice by pediatric urologists. We also correlated patterns of perceptions with surgical volumes and demographic variables.

A survey of 26 questions on 8 topics was e-mailed to 269 members of the American Academy of Pediatrics Section on Urology. Topics studied included distal and single stage proximal hypospadias, pyeloplasty, ureteral reimplantation with or without tapering, bladder neck reconstruction, and single and 2-stage bladder exstrophy repair. Participants were asked whether their rates of results were lower, the same or higher than specific published outcomes on a 5-point Likert scale. Study participants were also requested to provide demographic data and to estimate their annual case volume for each condition.

Of the 269 section members who were contacted 110 (40.9%) completed the survey. More than 80% of respondents practice pediatric urology exclusively with 50.9% reporting operative volumes of more than 400 cases per year. A greater proportion of participants reported worse outcomes than published reports for hypospadias fistula rates (distal p = 0.001, proximal p = 0.023), bladder neck repair (p = 0.018) and exstrophy repair continence rates (single and 2-stage p <0.001). Improved outcomes compared to published data were reported for ureteral reimplantation (p = 0.013) and pyeloplasty (p = 0.003). However, these findings did not correlate consistently with case volume or other demographic characteristics.

A significant proportion of pediatric urologists perceive their personal outcomes to be different than those in the published literature, regardless of practice setting, operative volume or time in practice. In an era of pay for performance and quality improvement, publication bias can have implications for patient care, reimbursement and malpractice.