K Radack1, S Park. 1. Department of Internal Medicine, University of Cincinnati College of Medicine, Ohio 45267.
Abstract
OBJECTIVE: To assess the reported association between skin tags and colonic polyps and to evaluate the methodologic rigor of the studies. DESIGN AND DATA IDENTIFICATION: English-language literature search using MEDLINE, Index Medicus, and bibliographic reviews of texts and all pertinent articles to perform a quantitative and methodologic analysis of all studies from 1983 (the original publication) assessing the association under study. Eligible studies were independently assessed using explicit methodologic guidelines for validity and generalizability of observational research. Two appraisers independently performed tests for heterogeneity and used meta-analytic methods in an attempt to provide summary estimates of the overall strength of association. MEASUREMENTS AND MAIN RESULTS: Significant statistical heterogeneity across studies indicated sharp differences in the direction and magnitude of the odds ratios for the association between skin tags and colonic polyps (chi-square test of homogeneity = 37.42, 9 degrees of freedom; p < 0.005). This marked disparity prevented meaningful pooling of the individual data. The degree of statistical heterogeneity was not significantly reduced after an analysis of specific subgroups of studies. Limitations potentially responsible for the varying outcomes included lack of blinded ascertainment of clinical information, noncomparability of subjects, differing diagnostic evaluations of the colon, and uncontrolled confounding. In addition, all but one study were performed in a tertiary care setting, seriously limiting the results to the "average" subject seen in primary care settings. CONCLUSION: Methodologic limitations and inconsistencies in study outcomes preclude the aggregation of data necessary to compute a valid and meaningful summary estimate of association. Sufficient variability prevents any consensus regarding the association between skin tags and colonic polyps. In addition, the applicability of the results is limited primarily to subjects seen in tertiary care centers, limiting the overall clinical usefulness of skin tags as "biomarkers" of colonic polyps. Recommendations for further research are provided.
OBJECTIVE: To assess the reported association between skin tags and colonic polyps and to evaluate the methodologic rigor of the studies. DESIGN AND DATA IDENTIFICATION: English-language literature search using MEDLINE, Index Medicus, and bibliographic reviews of texts and all pertinent articles to perform a quantitative and methodologic analysis of all studies from 1983 (the original publication) assessing the association under study. Eligible studies were independently assessed using explicit methodologic guidelines for validity and generalizability of observational research. Two appraisers independently performed tests for heterogeneity and used meta-analytic methods in an attempt to provide summary estimates of the overall strength of association. MEASUREMENTS AND MAIN RESULTS: Significant statistical heterogeneity across studies indicated sharp differences in the direction and magnitude of the odds ratios for the association between skin tags and colonic polyps (chi-square test of homogeneity = 37.42, 9 degrees of freedom; p < 0.005). This marked disparity prevented meaningful pooling of the individual data. The degree of statistical heterogeneity was not significantly reduced after an analysis of specific subgroups of studies. Limitations potentially responsible for the varying outcomes included lack of blinded ascertainment of clinical information, noncomparability of subjects, differing diagnostic evaluations of the colon, and uncontrolled confounding. In addition, all but one study were performed in a tertiary care setting, seriously limiting the results to the "average" subject seen in primary care settings. CONCLUSION: Methodologic limitations and inconsistencies in study outcomes preclude the aggregation of data necessary to compute a valid and meaningful summary estimate of association. Sufficient variability prevents any consensus regarding the association between skin tags and colonic polyps. In addition, the applicability of the results is limited primarily to subjects seen in tertiary care centers, limiting the overall clinical usefulness of skin tags as "biomarkers" of colonic polyps. Recommendations for further research are provided.