Advertisement

Do we really have a reproducible classification system for pelvic organ prolapse?

        To the Editors: We welcome the reproducibility study by Hall et al. (Hall AF, Theofrastous JP, Cundiff GW, Harris RL, Hamilton LF, Swift SE, et al. Interobserver and intraobserver reliability of the proposed International Continence Society, Society of Gynecologic Surgeons, and American Urogynecologic Society pelvic organ prolapse classification system. Am J Obstet Gynecol 1996;175:1467-71) because reproducibility studies may avoid the potential research, clinical, and medicolegal hazards of using unreproducible methods for clinical measurements and classification.
        • Bland JM
        • Altman DG.
        Statistical methods for assessing agreement between two methods of clinical measurement.
        • Grant JM.
        The fetal heart rate is normal, isn't it? Observer agreement of categorical assessments.
        Hall et al. conclude that the proposed International ContinenceSociety, Society of Gynecologic Surgeons, and American Urogynecologic Society pelvic organ prolapse classification system is highly reproducible. This conclusion should be confronted with two classic papers by Bland and Altman
        • Bland JM
        • Altman DG.
        Statistical methods for assessing agreement between two methods of clinical measurement.
        and Grant
        • Grant JM.
        The fetal heart rate is normal, isn't it? Observer agreement of categorical assessments.
        on the assessment of observer agreement of continuous and categoric variables. These authors showed that observer agreement of continuous and categoric variables should be, respectively, assessed by limits of agreement and proportions of agreement, with 95% confidence intervals, rather than by correlation, which measures association but not agreement. If, for example, Table I in the article of Hall et al. is analyzed with the proportions of agreement with 95% confidence intervals, it will be evident that, although the 95% confidence interval for each stage is too large for a final conclusion, the study suggests that prolapse staging may be reproducible for stages II, III and IV, but not for stages 0 and I (Table I) or even for a merging of stages 0 and I. By use of the same data it can also be concluded that, by merging stages 0, I, and II in a single stage, a reproducible stage will result (proportion of agreement 0.89, 95% confidence interval 0.75 to 0.97).
        Table IProportions of agreement with 95% confidence intervals regarding genital prolapse staging according to data provided by Hall et al. in Table I.
        StageNo. of agreements/No. of trialsProportion of agreement95% Confidence interval
        01/20.500.1–0.99
        I5/150.330.12–0.62
        II18/310.580.39–0.75
        III9/130.690.39–0.91
        IV1/11.000–1
        Proportions of agreement were calculated by dividing number of agreements by total number of agreement trials.
        *According to Grant,2 significant if lower limit of 95% confidence interval >0.50.
        It is important that Hall et al. provide more data about the reproducibility of the pelvic organ prolapse classification system analyzed. This may help to decide whether a five-stage classification system should be recommended in clinical practice or, instead, a simplified three-stage version (e.g., merging stages 0, I, and II into a single stage) should be adopted.
        João Bernardes, MD, PhD, Department of Gynecology and Obstetrics, Faculdade de Medicina, Hospital de S. João, 4200 Porto, Portugal, Altamiro Costa Pereira, MD, PhD, Serviço de Bioestatística e Informática Médica, Faculdade de Medicina, 4200 Porto,
        6/8/82729
        NO LABEL6/8/82729

        References

          • Bland JM
          • Altman DG.
          Statistical methods for assessing agreement between two methods of clinical measurement.
          Lancet. 1986; 1: 307-310
          • Grant JM.
          The fetal heart rate is normal, isn't it? Observer agreement of categorical assessments.
          Lancet. 1991; 337: 215-218