What Works in Managed Care—and What Doesn't

P. Nudelman,Ed Wagner
DOI: https://doi.org/10.1097/01974520-199301000-00005
1993-04-01
Frontiers of Health Services Management
Abstract:Managed care used to be simple in the good old days when HMOs were HMOs and the only managed-care organizations (MCOs) were staff-or group-model HMOs. The current alphabet soup of MCOs includes a continuum of practice arrangements with the staff/group-model HMO at one end, and fee-for-service (FFS) indemnity care at the other. While organizations and providers operating at the continuum extremes have been compared repeatedly over the past two decades with consistent findings, the organizations and providers working in between are much newer and have been subjected to less scrutiny. Miller and Luft provide a timely, useful synthesis of the literature on MCO performance, as well as some interesting speculation about the future. They seem convinced that network-and mixed-model MCOs will continue to grow, despite the scarcity of evidence that they can reduce utilization or costs. Conversely, more recent comparisons of staff/group-model HMOs with FFS care continue to confirm older findings that more traditional HMOs reduce costs without jeopardizing quality of care. The more recent data suggest to Miller and Luft that the more an MCO resembles a staff/group-model HMO, the better its performance. This lesson might have been assimilated by the MCO industry as the authors predict that successful MCOs will emulate staff/group-model HMOs by paying greater attention to physician selection, by ensuring a large number of enrollees per provider, and by changing physician behavior through collegiality and education, rather than case-by-case review. If staff/group-model HMOs outperform FFS care and these newer models, why are network-and mixed-model plans accounting for most of the growth in MCOs? And why are staff/group plans having trouble controlling premium inflation? The authors do not directly address these two nagging questions. The answer to the first question would appear to be related to the second. Since HMOs often appear to offer no major price advantage to employers, it is not surprising that plans that provide greater provider choice and less disruption of existing doctor-patient relationships (i.e., network and point-of-service plans) would be most attractive. Miller and Luft document that HMO price inflation has paralleled, and in many instances, equalled or exceeded inflation rates in the FFS sector. Curiously, the authors do not comment on the apparent paradox uncovered by their review of performance and recent trends. Why can't organizations that consistently hold down utilization and costs compete on price in many markets? The authors present intriguing survey data suggesting that HMOs, which actually are less expensive, appear to be more expensive than competitors to employees and employers alike. Since MCOs play a major role in many health care reform proposals, the review of the performance data and the authors' speculation about the future is timely and important. Miller and Luft acknowledge the evident trend for employers to become more sophisticated and powerful influences over local and national health care markets. We hope the authors are correct in predicting that employers will encourage some of the more attractive features of managed competition such as common benefits and flat payments across plans. Under this scenario, the price advantages of the most cost-effective plans are less likely to be obfuscated by subtle differences in coverage or out-of-pocket costs unappreciated by employers or employees, and these MCOs should prosper, grow, and eliminate their less cost-effective competitors. The authors speculate that the winners will be larger, experienced national or regional organizations that will "increasingly look alike" in offering a full range of HMO, FFS, and mixed options. Without disputing the reality of the trend toward diversification of insurance offerings, we wonder whether organizations with their roots at opposite ends of the HMO-FFS continuum will truly look alike in the end. …
What problem does this paper attempt to address?