Physician-researchers at a women's health clinic that caters to an underserved population are reviewing the results of a study they conducted regarding the effect of support systems during and after pregnancy on the development of postpartum depression. The researchers found that women with a poor support system had a statistically significant greater chance to develop postpartum depression than women with a good support system. If poor support system is the exposure of interest, which of the following represents the most likely measure (ie, odds ratio [OR] or relative risk [RR]) and magnitude of the effect and corresponding 95% confidence interval (CI)?

Question

Physician-researchers at a women's health clinic that caters to an underserved population are reviewing the results of a study they conducted regarding the effect of support systems during and after pregnancy on the development of postpartum depression. The researchers found that women with a poor support system had a statistically significant greater chance to develop postpartum depression than women with a good support system.  If poor support system is the exposure of interest, which of the following represents the most likely measure (ie, odds ratio [OR] or relative risk [RR]) and magnitude of the effect and corresponding 95% confidence interval (CI)?

Quizplus · Accepted Answer

Explanation: In cohort studies, 2 groups of subjects (ie, cohorts) are selected based on their exposure status (ie, type of support system during and after pregnancy). The cohorts are then followed over time, and the incidence of the disease (ie, postpartum depression [PPD]) is compared between groups. The main measure of association between exposure and outcome in observational or experimental follow-up studies is relative risk (RR). RR = Incidence of disease in exposed group / Incidence of disease in unexposed group RR <1.0 indicates decreased risk in the exposed group; RR = 1.0 indicates no difference in risk between the groups; and RR >1.0 indicates increased risk in the exposed group. In this question, the exposure is defined as having a poor support system. Therefore, RR is calculated as: RR = Incidence among women with poor support system / Incidence among women with good support system RR = 0.36 / 0.06 = 6.0 Because RR >1.0, there is an increased risk of PPD among women with a poor support system. Given that the researchers found a statistically significant association between poor support system and PPD, the 95% confidence interval (CI) must exclude the null value (1.0 in the case of RR) (Choice D). Therefore, RR = 6.0 (>1.0) with a 95% CI: 4.75-7.25 (excluding the null value of 1.0) is a likely expression of the RR and CI for this exposure in this study. (Choices A and B) Odds ratio (OR) is the main measure of association between exposure and outcome in case-control and cross-sectional studies; a statistically significant OR will be accompanied by a CI that excludes the null value (OR = 1.0). However, the abstract describes a cohort study, and in a cohort study, RR is the most appropriate measure of association. (Choice C) RR ≈ 0.17 results from defining the exposure as having a good (rather than a poor) support system: RR = 0.06 / 0.36 ≈ 0.17. Educational objective: When comparing the relative risk (RR) of the same outcome in 2 cohorts (eg, exposed, nonexposed), RR is calculated as: RR = Rate of outcome in cohort 1 / Rate of outcome of cohort 2

Physician-Researchers at a Women's Health Clinic That Caters to an Underserved