age, injury severity, and other characteristics. In fact, one subset of HMO patients, those transferred from the trauma center to the HMO hospital, had the longest LOS. The authors speculate that the possible reasons for the increased LOS among transferred patients are the disruption in continuity of care, problems in discharge planning, and medical complications that occur after transfer. On the other hand, a study of 3,141 admissions from 1990 to 1992 to a Seattle trauma center found LOS to be similar among motor vehicle crash patients with and without commercial insurance (Rhee et al., 1997). (Those without commercial insurance were either Medicaid or self-pay.) Although this study did not specifically compare managed care and fee-for-service patients, it found no effect of payer status on utilization. Part of the difficulty of monitoring LOS among managed care and fee-for-service patients is that LOS has declined in general in response to nationwide trends in cost containment.
Outcomes under managed care have been studied with respect to mortality and morbidity, but studies are sparse and not necessarily comparable. In the study by Campbell and coworkers cited above, HMO patients experienced lower mortality rates (4.1 percent) relative to non-HMO patients (9.9 percent), after adjustment for severity of injury and other factors, but HMO patients were younger, more often male, and more likely to have blunt injuries. The study by Rhee and colleagues (1997) found no differences in mortality between patients with commercial insurance and those without, but as pointed out above, the study was not of managed care per se. On the other hand, Young and Lowe (1997) found gatekeeping by managed care to be associated with adverse outcomes. After reviewing solicited case reports of 29 patients denied authorization by MCOs for ED payment, they determined that 14 percent had adverse outcomes (mortality and morbidity), 14 percent were at increased risk, and 72 percent were near misses (i.e. cases in which adverse outcomes were averted by ED care despite denial of authorization). Their findings confirmed earlier published reports about adults and children being denied authorization for ED payment (Derlet and Young, 1997). These findings are disturbing and warrant examination in a larger, random sample in which there is no selection bias.
The quest for cost containment also is significantly felt in EMS. There are several important and interrelated trends transforming EMS: (1) the consolidation of small, independent public and private EMS providers into large, publicly traded corporations that can realize economies of scale; (2) increased contracting between MCOs and EMS agencies in capitated or risk-sharing agreements; and (3) the development of new triage guidelines for EMS dispatchers designed to steer non-emergency calls away from the EMS system to more appropriate community resources (Neely, 1997; Neely et al., 1997; NHTSA, 1997b). It is important to monitor these trends with respect to their impact on EMS access, utilization, cost, and patient outcomes. The development of new triage guidelines is a special concern because of the potential for undertriage, which occurs when patients are not administered the emergency services they need. Public health professionals are concerned that the development of new triage guidelines is