Although health care stakeholders acknowledge that health care treatments can affect different segments of the patient population in different ways, early discussions focused on how comparative effectiveness research (CER) could be applied to traditional demographic segments (e.g., age, gender). Only recently has the discussion shifted to understanding how the findings of CER should be interpreted and applied when significant variations in treatment effects are evident across a population. This study, The Impact of Comparative Effectiveness Research on Health and Health Care Spending, illustrates some of the issues that must be addressed in applying the results of CER studies, particularly to reimbursement policies based on such studies.
This research was funded by the National Pharmaceutical Council and conducted by Drs. Anirban Basu, Anupam B. Jena, and Tomas J. Philipson.It was initially published as a working paper (#15633) by the National Bureau of Economic Research.
- Although CER has been positioned as a way to improve health and possibly reduce costs, the impact of CER on patient health depends on how the research findings are applied by health care decision makers, including those who may use the information in coverage and reimbursement decisions.
- CER may not decrease spending as much as expected and it may adversely impact health if coverage and reimbursement policies ignore the differences in the effects of treatments across a population.
- With rapid scientific advancement in the field of personalized medicine, the study findings support the need for new research that can be used to suggest appropriate polices to ensure that CER can be conducted and applied in a manner that recognizes the importance of variation in treatment effects and helps to ensure timely and efficient access to therapies for all patients who may benefit.
The research provides a quantitative framework for considering both the economic and health implications of using CER results in making decisions about what therapies to cover and reimburse. The authors use this framework, and the results of the well-known Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study comparing the effectiveness of antipsychotic medications, to illustrate how policies that ignores varying, or heterogeneous, treatment effects across individual patients, may reduce pharmaceutical costs but have counter-productive effects on overall medical costs and patient well-being. The economic value of these negative effects on patients’ well-being may, in part, mitigate the expected reduction in overall costs (drug plus medical) as medical costs may increase.
Using the framework, the authors compare two very restrictive, hypothetical coverage policies for Medicaid schizophrenia patients to one in which patients have a choice of four antipsychotic medicines and can switch among the medications to achieve the best therapeutic result. Under the first scenario, only one product, judged to be most cost-effective in previous analyses of the CATIE trial data, is covered and reimbursed. The second scenario covers one additional product. According to the authors, it is important to note that neither of these policies was actually implemented in any Medicaid program although often suggested by policy analysts. They are illustrative of coverage policies that ignore the heterogeneous effects of different medications for individual patients, effects that were clear from the CATIE trial results. Such restrictive coverage policies would result in large numbers of patients going without needed medications because they were deemed less cost-effective than the winner of the CATIE trial, at the risk of adverse medical events and additional costs. Compared to the policy with the greatest choice, both of the hypothetical policies would result in significant loss of quality and patient well-being.
Due to copyright issues, the article must be downloaded from the Journal of Health Economics website.