Objective
To determine whether blinded and unblinded Pelvic Organ Prolapse Quantification (POP-Q) examinations differ in a randomized trial.
Study Design
Blinded POP-Q examinations performed at 3 months and 1 year were compared with unblinded examinations performed by the surgeon in a randomized trial of vaginal mesh for pelvic organ prolapse.
Results
Sixty-five patients were included in the study. Correlations between the blinded and unblinded POP-Q points and stages varied from low to moderate (rho = 0.29-0.78). At 3 months, the blinded overall prolapse recurrence rate was 45.3% compared with 39.1% based on unblinded staging ( P = .34). At 1 year, the blinded overall recurrence rate was significantly higher than the unblinded recurrence rate: 68.3% vs 53.3% ( P = .004). The 1-year blinded anterior wall recurrence rate was also higher than the recurrence based on unblinded staging: 56.7% vs 43.3% ( P = .021).
Conclusion
Use of unblinded POP-Q staging resulted in underestimation of 1-year overall recurrence after prolapse repair.
Knowledge of the intervention type at the time of follow-up examination can potentially lead to bias in randomized clinical trials. Blinding, or masking, is a procedure in which 1 or more of the parties in a clinical trial are kept unaware of the treatment assignments. Blinding of the examiner in postoperative examinations eliminates the risk that the evaluations will be biased by the examiners’ treatment preferences. In surgical studies, blinding necessitates the use of examiners other than the surgeons who performed the study procedures, because surgeons recognize their own patients and will remember which procedures they performed. This eliminates the risk that surgeons will inflate the success of their procedures.
Many surgical studies do not use blinded evaluations, and the impact of their failure to do so is often unknown. We are unaware of any published studies in the urogynecology literature that compare blinded and unblinded examination results in a clinical trial. This study was designed to determine whether blinding reduces bias when the Pelvic Organ Prolapse Quantification (POP-Q) examination is performed in a clinical trial. The POP-Q examination is an objective, standardized, validated measure of prolapse recognized by the International Continence Society. Previous studies have shown good interrater and intrarater reliability for the POP-Q examination. This examination is often used as an objective outcome in urogynecologic surgical studies.
Our primary aim was to determine whether there was a difference between blinded examiners and surgeons who evaluated their own patients with respect to POP-Q examination points, total POP-Q stage, and overall recurrence in a randomized clinical trial of vaginal prolapse repair with and without mesh.
Materials and Methods
Institutional review board exemption was obtained for this secondary data analysis of a multicenter, double-blinded randomized clinical trial evaluating traditional vaginal prolapse repair to vaginal mesh repair of prolapse (VAMP study). The details and results of this study have been described elsewhere. Both patients and outcome assessors were blinded to the surgical intervention that was performed. Patients had POP-Q examinations performed 3 months and 1 year, postoperatively, by both blinded examiners and the surgeons who performed their procedures. All POP-Q examinations were performed in a standardized fashion with patients in 45-degree upright position during straining. The blinded examiners included faculty urogynecologists, fellows, nurse practitioners, and resident physicians. Residents, fellows, and nurse practitioners were instructed and supervised on how to correctly perform a POP-Q examination before blinded assessments. The unblinded examinations were performed by the patient’s surgeon at postoperative visits.
Previously collected deidentified data from this study were obtained for the 3-month and 1-year blinded and unblinded POP-Q examinations, including the POP-Q point measurements, POP-Q stage by anterior, posterior, and apical compartments, and overall POP-Q stage. Overall, prolapse recurrence was defined as a total POP-Q stage of 2 or greater. Anterior, posterior, and apical prolapse recurrence were defined as a corresponding POP-Q stage of 2 or greater.
The number of patients in this study was determined by the original randomized controlled trial and therefore a sample size calculation was not performed for this ancillary study. SPSS for Windows (version 16; SPSS, Inc, Chicago, IL) was used for data management and statistical analysis. Because the data had statistically nonnormal distributions, nonparametric statistical methods were used for analysis. Scatterplots, Spearman correlations, and Friedman tests were used to compare the blinded and unblinded POP-Q stages and point measurements. The McNemar test was performed to compare the blinded and unblinded percentages of patients with recurrence. A .05 significance level was used for all statistical tests. No 1-sided tests were performed. Means are presented as mean ± standard deviation (SD).
Results
Of 65 patients in the original study, 64 patients with 3-month data and 60 patients with 1-year data were included in the current study. Six surgeons performed unblinded POP-Q examinations on their own patients, and 34 blinded examiners performed examinations. The blinded examiners were recorded in the data for 62 of the 3-month cases and 58 of the 1-year cases. At 3 months, 51.6% (32) of the blinded examinations were performed by faculty urogynecologists, 29.0% (18) by fellows, 12.9% (8) by resident physicians, and 6.5% (4) by nurse practitioners. At 1 year, 41.4% (24) of the blinded examinations were performed by faculty urogynecologists, 31.0% (18) by fellows, 17.2% (10) by resident physicians, 8.6% (5) by nurse practitioners, and 1.7% (1) by a faculty physician who was not an urogynecologist.
For both the 3-month and 1-year examinations, Spearman correlations between the blinded and unblinded POP-Q points and stages were low to moderate (rho = 0.29-0.78), indicating a wide variation in agreement between blinded and unblinded examiners ( Table 1 ). Correlations between blinded and unblinded POP-Q points were highest for perineal body measurement Pb at 3 months and anterior wall measurements points Aa and Ba at 3 months and 1 year. Correlations in POP-Q stage were lowest for the apical compartment (POP-Q C stage). Results of blinded and unblinded examinations by POP-Q points, stages, and recurrence at 3 months and 1 year after surgery are shown in Table 2 . At 3 months, the vaginal cuff (C) measurement was worse for the blinded examination (median −6.5 vs −7.0, P = .047). At 1 year, the perineal body (Pb) measurement was significantly smaller for the blinded examination (mean 4.0 ± 0.8 vs 4.2 ± 0.8, P = .011), and the POP-Q anterior wall (Ba) stage was significantly higher for the blinded examination (mean 1.5 ± 0.7 vs 1.3 ± 0.8, P = .039).
