Nearly 2 years ago I speculated in this column that health planners or health economists would attempt to manipulate the patterns of patient care to influence the cost and/or quality of clinical care. At that time I suggested that, in that event, the intervention should be managed as we have with drug or device trials to ensure the authenticity and accuracy and most of all assuring the safety of the patient. Furthermore, the design should be incorporated in the intervention, that equipoise be present in the arms of the trial and that a safety monitoring board be in place to alert investigators when and if patient safety is threatened. Patient consent should also be obtained.
Little did I know that an example was in play at the time of publication. A study presented at the Heart Failure Society of America meeting indicates that the Centers for Medicare & Medicaid Services, as part of the Affordable Care Act, was carrying out such an experiment in the attempt to lower cost and improve the quality of the care of heart failure patients by decreasing the occurrence of readmissions. On the surface, that appears to be a laudable goal and one that we can all support. In an attempt to decrease the readmissions, CMS had incentivized the process by financially rewarding hospitals if they decreased repeat admissions after discharge. Much to the surprise of the health planners, the intervention reported that, although 30-day readmission decreased as the result of the financial incentives, 30-day mortality increased. This was particularly surprising since in numerous drug trials, notably MERIT-HF (Lancet. 1999 Jun 12;353:2001-7), readmission usually tracked closely with mortality.
Beginning in 2012, CMS, using claims data from 2008 to 2012, penalized hospitals if they did not achieve acceptable readmission rates. At the same time, the agency established the Hospital Admission Reduction Program to monitor 30-day mortality and standardize readmission data. The recent data indicate that the incentives did achieve some decrease in rehospitalization but this was associated with a 16.5% relative increase in 30-day mortality. It was of particular concern that in the previous decade there had been a progressive decrease in 30-day mortality (Circulation 2014;130:966-75). The increase in 30-day mortality observed in the 4-year observational period appears to have interrupted the progressive decrease in 30-day mortality, which would have decreased to 30% if not impacted by the plan.
My previous concerns with this type of social experimentation and manipulation of health care was carried out, and as far as I can tell, continues without any oversight and little insight into the possible risks of this process. A better designed study would have provided better understanding of these results and might have mitigated the adverse effects and mortality events. It is suggested that some hospitals actually gamed the system to their economic advantage. In addition, no oversight board was or is in place as we have with drug trials to allow monitors to become aware of adverse events before there any further loss of life occurs.
I would agree that a randomized trial in this environment would be difficult to achieve. Obtaining consent from thousands of patients would also be difficult. Nevertheless,.
Dr. Goldstein, medical editor of Cardiology News, is professor of medicine at Wayne State University and division head emeritus of cardiovascular medicine at Henry Ford Hospital, both in Detroit. He is on data safety monitoring committees for the National Institutes of Health and several pharmaceutical companies.