Building Healthcare Quality: If the Future Were Easy, It Would Be Here by Now

Abstract

Dr. Eugene Nelson's article has provided us with a seductive vision of a healthcare future that works. The image is compelling-small primary care microunits working to apply standardized protocols in a data-driven environment as part of iterative cycles of continuous quality improvement (CQI). We are presented with anecdotal reports of two practices in Florida and New Hampshire that are just now beginning to operate along these lines. But when one realizes that some of these concepts are already a generation old, it is natural to ask why they have not been more widely adopted. Cohen et al. (1986) and Goldberg et al. (1987) first described the introduction of small primary care units (a.k.a. firms) at an academic medical center in 1986. The reorganization served to enhance clinic productivity by 20 percent, while decreasing patient flow (the time spent in completing visits) by 15 percent. Unscheduled walk-ins declined by 39 percent because each unit's nurse practitioner was available to provide just-in-time cross-coverage. Because each firm staffed its own ward and clinic, tightly integrated care could be provided to a defined population of patients. Shifting of care from the inpatient to the outpatient setting decreased total charges by 27 percent. Neither is the notion of standardizing care a new one. In 1973, Sox reported that by following algorithms for 11 acute medical complaints, physician assistants would have accurately evaluated 45 percent of 3,024 patients seen without direct physician involvement (Sox, Sox, and Tompkins 1973). Two years later, Grimm et al. (1975) documented that the application of a symptom-oriented pharyngitis protocol had decreased the ordering of white-cell counts by 37 percent and the prescribing of antibiotics by 63 percent. The answer to why healthcare practices evolve relatively slowly is that curbing longstanding clinical habits is simply hard to do. This is especially so if implementing change initially consumes, rather than saves, resources. During the last decade, much effort was expended on the development and dissemination of clinical practice guidelines (Audet and Greenfield 1990; VanAmringe and Shannon 1992). It was at first believed that physicians and health systems would be easily persuaded to move practice into compliance with the best available scientific evidence. Rigorous reports, however, were later surprisingly consistent in demonstrating that guidelines in isolation had failed to promote voluntary change in practice patterns (Kosecoff et al. 1987; Hill, Levine, and Welton 1988; Lomas et al. 1989; Lomas 1991). Analogous to the experience with guideline implementation, the early literature supportive of clinical CQI teams has remained largely anecdotal (Kaluzny 1996). At least one rigorous trial is now at hand, however, and it suggests that a future where CQI principles are a routine part of mainstream practice is not quite ready for prime time. We recently published results from the first multicenter randomized controlled trial of the use of CQI teams in chronic disease care (Goldberg et al. 1998). Eight teams at two academic and two HMO clinics in Seattle worked for a year to improve guideline compliance and clinical outcomes regarding the primary care of hypertension and depression. A trained CQI facilitator instructed teams in the use of the Shewhart cycle of activities popularly referred to as plan, do, check, act (PDCA). Teams invested a considerable amount of time, with each holding 13 meetings and spending a total of 212 hours on average. Yet across all sites, no positive effects were demonstrated for either disease condition. At only one clinic was an outcome actually improved, a 17-percentage-point increase in the number of hypertensives adequately controlled. The reasons for these negative results were multiple. Some teams experienced difficulty in collecting timely data. The number and type of process changes selected varied greatly. …