LAUSR

retrospective observational studies did not demonstrate causality. Preoperative anemia may be a mediator or confounder of worse surgical outcomes rather than the incipient cause. In addition, none of these studies demonstrated that treating preoperative anemia improved outcomes. Finally, previous high-quality studies have cautioned that treatment of perioperative anemia with red blood cell transfusions may harm cancer surgery patients—a substantial portion of the cohort in our study—and should be avoided. Notably, preoperative optimization of anemia with intravenous iron may be a more attractive option with early evidence supporting its use for select outcomes such as length of stay reduction. Although this early work has not shown a reduction in complications, trials are currently ongoing that may expand on the benefits of preoperative intravenous iron therapy (Preoperative Intravenous Iron to Treat Anaemia in Major Surgery, PREVENTT, NCT01692418). An important caveat for these promising interventions is that current payment practices and preoperative access to patients in the American health care environment may preclude further use of such an approach in the population we studied. Preoperative anemia can be addressed acutely via red blood cell transfusions or long-term pro-erythopoetic agents such as intravenous iron. Given the controversy with the former and potential limitations in use of the latter secondary to health care system factors (reimbursement, which clinician to administer) in the perioperative care setting for the latter, we felt that the inclusion of anemia would confound results and tend to dilute the effects of other more clearly defined comorbidities and their opportunities for preoperative optimization. In the future, we look forward to prospective studies that demonstrate correcting preoperative anemia improves postoperative outcomes. This question raised by the authors is important and along with our study highlights the need to devote future research to understanding in more detail the following preoperative optimization considerations: (1) which comorbidities need to be corrected; (2) to what extent should comorbidities be corrected; and (3) valuation of the incremental benefit per optimization effort. Research in this area must also consider how to best implement and disseminate our findings to improve the slow adoption and knowledge transfer process of patientfocused interventions that improve outcomes. We would advocate for such an approach for any further incremental comorbidity additions to the original model described in our manuscript.