LAUSR

TO THE EDITOR: Vogrin and colleagues (1) evaluated the comparative effectiveness of routine invasive coronary angiography (ICA) for managing unstable angina by analyzing hospital discharge data. Because the study was observational, ICA was clinically judged without randomization and 2 groups compared in the study may not be regarded as selected from same population. If factors related to outcome determined the intervention, these would be confounding factors. To adjust for confounders, the authors developed a propensity score for coronary angiography and used it to adjust for effect in addition to age, sex, and Charlson comorbidity score. Forty-four covariates were mentioned in the Appendix of the article. Could the authors shared whether they believe that confounding was adequately controlled for or unmeasured covariates were potentially important?