The use and misuse of matching in case-control studies: the example of polycystic ovary syndrome

Abstract

Matching control selection strategies are often employed in PCOS case-control studies; however, they are infrequently used in an appropriate fashion. When properly applied, matching may offer improved study precision, but this is highly contingent on the causal pathway under consideration, strength of the associations between the matching variable and both the risk factor of interest and PCOS, and use of an appropriate stratified data analysis.Variations in design, including strategies to consider suspected confounding variables, may be in part responsible for discrepancies among study results reported in the polycystic ovary syndrome (PCOS) literature (1). The case-control study design is frequently used to consider hypothesized risk factors for PCOS, with investigators often complementing this framework with a 'matching' strategy in which controls are selected for cases according to the distribution of suspected confounding variables among the latter (2). A recent PubMed search (i.e., on 6/14/06), using the search terms 'polycystic ovary syndrome' and 'case-control studies', limited to the English language and 'published in the last 1 year', generated 40 citations, 23 of which fell under the PCOS case-control rubric. Almost half of these 23 studies, 11 (48%), matched controls to cases by body mass index (BMI), age, or both, but only one (9%) analyzed the data appropriately for its matched nature. In a prior publication, we discussed the merits of the case-control design for studies of PCOS causal risk factors (3); however, the proper

Occupational exposure to arsenic and risk of nonmelanoma skin cancer in a multinational European study

“…ORs were adjusted for the matching variables sex, age and county of residence to avoid selection bias. 25 The results were also adjusted for skin propensity to sunburns, family history of cancer and exposure to arsenic in drinking water that were identified as confounders and continued to be significant predictors for NMSC in multivariable regression models. A number of other risk factors possibly associated with NMSC and occupational exposure to arsenic were also considered (i.e., education level, skin pigmentation, tobacco smoking, self-reported sunlight exposure during the weekends and occupational sunlight exposure).…”

“…Odds ratios (OR) and 95% confidence intervals (95% CI) for NMSC were calculated to estimate the associations with occupational exposure to arsenic using unconditional logistic regression. ORs were adjusted for the matching variables sex, age and county of residence to avoid selection bias . The results were also adjusted for skin propensity to sunburns, family history of cancer and exposure to arsenic in drinking water that were identified as confounders and continued to be significant predictors for NMSC in multivariable regression models.…”

Transoral Robotic Surgery Using the Thulium:YAG Laser<subtitle>A Prospective Study</subtitle>

“…We chose matching criteria a priori to create similar distributions of confounding variables. 10 Bloom et al 10 described the confounding phenomena as highly contingent on the proposed causal pathway between the risk factor (cutting device) and the outcome of interest (feasibility, safety, and functional outcome). We, therefore, chose to match based on tumor size, tumor location, age, and sex, as these factors were considered by the authors to most strongly bias the outcomes of interest in this study.…”

Short and long term mortality rates associated with first pregnancy outcome: Population register based study for Denmark 1980–2004

“…A good deal of insight might also be gained by case-control studies which are carefully designed [18]. In addition to selecting control groups relative to year of the woman’s birth, age at first pregnancy, marital status, and socioeconomic status, cases should be matched to controls with similar psychiatric histories prior to the year of the first pregnancy.…”