Which Type of Medical Group Provides Higher-Quality Care?
Which Type of Medical Group Provides Higher-Quality Care?
复制标题
哪种类型的医疗集团提供更高质量的护理？
DOI：
--
复制
发表时间：
2006
影响因子：
39.2
通讯作者：
L. Casalino
中科院分区：
文献类型：
--
作者：
L. Casalino
The quality of medical care depends on the individual physician and on the organization in which he or she works (13). In this issue, Mehrotra and colleagues (4) compare 3 types of physician organizationsmedical groups, independent practice associations (IPAs), and hybridson 6 measures of quality. Medical groups are usually defined as organizations in which 3 or more physicians share facilities, staff, and income. Independent practice associations are organizations that contract with health maintenance organizations (HMOs) on behalf of large numbers of medical groups and 1- and 2-physician practices (5, 6). Hybrids, a term coined by the investigators, are organizations that consist of a medical group and an IPA. Mehrotra and coworkers (4) interviewed the leaders of 119 California physician organizations97% of the medical groups and IPAs that contracted with PacifiCare (a large HMO) in 1999querying them about their organization's structure and use of quality improvement processes and electronic medical records (EMRs). They linked their results to PacifiCare data that scored each organization on 6 measures of quality. The mean number of patients per measure per organization ranged from 25 to 3329, depending on the measure. The authors conclude that medical groups scored substantially higher than did IPAs on 4 of the 6 measures: A higher percentage of patients treated by medical groups had mammography, Papanicolaou smears, chlamydia screening, and diabetic retinal examinations. There was no difference between medical groups and IPAs on measures of the use of asthma-control medications and -blockers after acute myocardial infarction. Hybrid organizations scored better than IPAs but worse than medical groups. Medical groups used more quality improvement processes and were more likely to use EMRs; however, use of these processes and EMRs was not associated with higher quality scores. This study should not be considered evidence that, in general, medical groups provide higher-quality care than do IPAs. There were only 19 groups, and they averaged 242 physicians each, making them extremely large compared with the small groups that predominate in the United States (7, 8). It would be more precise to conclude that very large medical groups scored higher than IPAs. The study's data suggest another interesting conclusion, not explicitly mentioned in the article: If large medical groups had higher quality scores than did IPAs, they presumably also had higher quality scores than did the smaller medical practices (most of which had between 1 and 10 physicians) (9) that comprised the IPAs. Quality scores were calculated at the level of the IPA, but in fact represent the aggregated scores, such as the percentage of women who had mammography, of the practices in the IPA. These 2 conclusions, that large groups provide better-quality care than do IPAs and better quality care than smaller practices, are important but should be interpreted with caution because they are based on only 19 large groups and 6 measures of quality. Generalizability is also limited by the California setting, in which large groups and IPAs were paid mainly through capitation, which is not typical of most of the United States (10). The study was observational and cross-sectional, which means that we cannot infer that the size and structure of an integrated group cause the differences in quality. For example, the results could be due to unobserved differences between patients who seek care in large medical groups and those who seek care in smaller practices. The authors' other main findingthat use of quality improvement processes and EMRs was not associated with quality scoresshould also, as they suggest, be interpreted carefully. In addition to the limitations mentioned, this measure does not capture the extent of EMR use in an organization, and the quality improvement measures do not distinguish intensive use from minimal use of quality improvement processes. Some of the measures of quality improvement used (for example, collect data on time patients spend in waiting room) may not be related to quality, and some (for example, use of a diabetes disease management program) may be overly general. Despite these limitations, the study by Mehrotra and colleagues is important for the information it provides and because it sought the Holy Grail of studies of physicians and quality: to analyze the associations between physician organization structure, quality improvement processes, and outcomes. Only 1 previous, much smaller study presented data on structure, processes, and outcomes; it provided little information on structure (IPAs seem not to have been included) and found a minimal association between structure or quality improvement processes and quality (11). Three previous studies linked physician organization structures to quality improvement processes. One found that large medical groups and IPAs did not differ, nationally or in California, in their use of quality improvement processes for 4 chronic diseases (12), but that large medical groups were more likely than IPAs to provide health promotion (13) and smoking cessation programs (14) and patient reminders for preventive services (15). A study of California organizations that serve large numbers of Medicaid patients found that community clinics and hospital-based clinics use more quality improvement processes than do IPAs (16). A third study found that medical groups with 10 or more physicians were more likely to use such processes than were smaller practices (17). Three previous studies linked structure to outcomes. One study found that patients treated by groups with 3 or more physicians were more likely to receive preventive services, but did not find that this likelihood increased as groups grew larger (18). The other 2 studies did not compare medical groups and IPAs, did not compare groups by size, and did not find consistent relationships between structural characteristics and quality (19, 20). Many analysts believe that large medical groups provide better care than do small groups and IPAs (21, 22), but the evidence for or against this hypothesis has been sparse. The study by Mehrotra and coworkers and the other studies cited previously are the best we have to date; they provide limited and somewhat inconsistent support for the hypothesis. Large groups potentially have economies of scale and scope useful for improving quality: scale to employ physician and nonphysician staff to implement quality improvement processes and to use information technology to assist with these processes, and scope because physicians from multiple specialties can collaborate in the design of quality improvement processes and can consult in the care of individual patients. However, large size can also produce diseconomies of scale (23); small groups might find it easier to coordinate activities within the group, and the small group setting might facilitate closer relationships of physicians and staff with patients (23, 24). The most important contribution of the study by Mehrotra and colleagues may be that it highlights critical, but largely unexplored, questions about the organization of physician practice. Which types of physician organization (for example, large, medium, or small; single specialty or multispecialty; or physician-owned or hospital-owned) provide better-quality care? To what extent does quality depend on the use of quality improvement processes and on other group characteristics, such as the culture of the group or the quality of its individual physicians and staff? We need to answer these fundamental questions about how the organization of physician practices affects quality of care.
