Amid growing concern about the quality of health care in the

Transcription

1 Cottage To Kaiser Physician Organization And Care Management In California: From Cottage To Kaiser How can the quality chasm be closed without retooling small-office physician practice in the United States? by Diane R. Rittenhouse, Kevin Grumbach, Edward H. O Neil, Catherine Dower, and Andrew Bindman ABSTRACT: Data from a survey of practicing physicians in California s thirteen largest urban counties were used to ascertain differences in care management processes, financial incentives for quality, and practice pressures by type of practice setting. Physicians in the Permanente medical groups have adopted and value quality-oriented, system-level care management tools to a much greater degree than physicians in independent practice association (IPA) networks or traditional cottage-industry practices. Our findings raise disturbing questions about how the health system will close the quality chasm in medical care without transforming the underlying organization of physician practices. Amid growing concern about the quality of health care in the United States, increased attention has turned to the role of organizations. 1 Large physician organizations and integrated systems of health care are developing and implementing disease management programs, distributing physician practice pattern information, and offering physicians quality-based financial incentives to stimulate quality improvement. 2 Yet physicians in these organizations, facing pressures to contain costs and to see more patients in less time, may neither participate in these quality improvement efforts nor find them useful. How physicians experience quality improvement efforts might vary greatly according to the type of physician organization they participate in. Physician organizations encompass a broad spectrum of practice sizes and settings. At one end of the spectrum is the traditional cottage industry of solo or small-group practices, Diane Rittenhouse (DianeR@itsa.ucsf.edu) is an assistant professor in the Department of Family and Community Medicine, University of California, San Francisco (UCSF). Kevin Grumbach is a professor and chair of that department and chief of Family and Community Medicine at San Francisco General Hospital. Ed O Neil directs the Center for the Health Professions at UCSF. Catherine Dower is associate director, health law and policy, of the that center. Andrew Bindman is a professor of medicine, epidemiology, and biostatistics at UCSF; chief of the division of general internal medicine at San Francisco General Hospital; and vice chief for population health in that hospital s Department of Medicine. HEALTH AFFAIRS ~ Volume 23, Number 6 51 DOI /hlthaff Project HOPE The People-to-People Health Foundation, Inc.

2 Physician Practice representing a collection of autonomous professionals providing largely selfdefined expert care within organizational, payment, and regulatory environments involving conflicting incentives, goals, and objectives. 3 Attheotherendofthe spectrum are the Permanente medical groups: large, integrated, prepaid multispecialty medical groups that provide or arrange to provide a coordinated continuum of services to a defined population and are willing to be held clinically and fiscally accountable for the outcomes and health status of the population served. 4 Between these two polarities exist a variety of other practice models. As capitated payment gained popularity in the 1980s and 1990s, independent medical groups of varying scale proliferated to accept not only financial risk from health plans but also responsibility for managing care. Alongside the integrated multispecialty medical groups grew virtually integrated independent practice associations (IPAs), which promised some of the benefits of large medical groups while allowing physicians to remain in their small-practice settings. 5 Despite the potential for physician organizations to improve physician practice by implementing care management processes, aligning financial incentives, and fostering practice environments that emphasize quality over cost containment, there have been no studies of individual physicians perceptions of these organizations efforts. 6 This study used data from a survey of physicians in California to compare physicians experiences with care management processes, financial incentives, and practice environment across practice settings. We focused on care management tools that have been shown to positively affect physician practice: disease management programs (DMPs); the receipt of useful, quality-oriented practice pattern information; and the receipt of financial bonuses for quality. We also assessed factors that could negatively affect quality of care such as pressure to see more patients daily or to limit referrals to specialists. Ultimately, we were interested in physicians perceptions of the value of large physician organizations. Do physicians at Kaiser Permanente perceive that this traditional prepaid group practice model confers advantages for managing patient care more effectively? If so, do independent medical groups have similar advantages from the point of view of their physicians? What about IPAs? Might they represent the best of both worlds by providing organizational benefits of group practice while preserving the character and tradition of the small-practice setting? Study Data And Methods Data. Study data were derived from the University of California, San Francisco (UCSF), California Physician Survey. Self-administered questionnaires were mailed in the winter of to physicians practicing in California s thirteen largest urban counties, accounting for 78 percent of the state s physicians. Physicians were identified from the American Medical Association (AMA) Physician Masterfile. To be eligible for the survey, physicians had to be listed as providing direct patient care, not in training, and not employed by the federal government. Sampled primary care 52 November/December 2004

3 Cottage To Kaiser physicians (PCPs) listed their primary specialty as family practice, general practice, general internal medicine, general pediatrics, or obstetrics/gynecology (OB/GYN). Specialists were eligible who listed their primary specialty as cardiology, endocrinology, gastroenterology, general surgery, neurology, ophthalmology, or orthopedics. Physicians were drawn using a probability sample stratified by county and by physician s race/ethnicity with an oversampling of nonwhite physicians. Completed questionnaires were obtained from 1,365 of the 2,240 eligible physicians (61 percent). In addition to these cross-sectional data from 2001, we included data from a longitudinal cohort of physicians surveyed in 2001 using the same questionnaire. This cohort comprised physicians who had responded to prior surveys of California primary care and specialist physicians in 1996 and 1998, respectively. These earlier studies used the same sampling methodology. Physicians responding to the 1996 and 1998 surveys were resurveyed in 2001 using the same questionnaire as used for the 2001 cross-sectional survey. Of 413 eligible primary care physicians in the longitudinal sample, 334 responded in 2001 (81 percent response rate), and 408 of 610 eligible specialists responded (67 percent). We pooled results from the 2001 cross-sectional and longitudinal samples after determining that the results by practice setting were similar in both groups; pooling the results increased the overall sample size and the stability of the results for each practice setting group. The survey included a wide range of questions investigating physicians experiences and perceptions, including type and size of main practice setting; experience with DMPs and practice pattern information; bonus payments and the factors used to determine them; practice pressures; and perception of the value of physician organizations. Analyses. We created a practice setting variable with four mutually exclusive categories. The first category comprised physicians in the Northern and Southern California Permanente medical groups two large, prepaid multispecialty medical groups. The second category included physicians in medical groups of eleven or more physicians, other than Permanente. This category is the most heterogeneous and may include physicians from academic medical centers and some of California s large, long-standing group practices, in addition to physicians from smaller medical groups. 7 The third category consisted of physicians in small offices (one to ten physicians) who participated in at least one IPA. The final category, cottage industry, consisted of physicians in small offices who did not participate in any IPAs. The small number of physicians in other settings, such as community health centers, were excluded from the analyses. We analyzed measures of experience with care management processes, financial incentives, and practice environment for physicians in each of these four practice-setting categories. Because the experiences of primary care and specialist physicians in managed care are known to differ, we separated these two groups of physicians in our initial analyses. Where results were found to differ meaningfully, the results for both groups are displayed. Otherwise, in the interest of space we HEALTH AFFAIRS ~ Volume 23, Number 6 53

4 Physician Practice present the results for primary care and specialist physicians combined. Chisquare tests for bivariate comparisons of categorical data were performed. Analyses used weighted data to account for the disproportionate sampling method. We also used logistic regression models to assess the effect of age, sex, and race/ethnicity as potential confounders of the effect of practice setting on care management, financial incentives, and physicians experiences. Study Results In , 22 percent of PCPs in California practiced with Permanente; 16 percent practiced with other medical groups of eleven or more physicians; and 44 percent were in smaller settings (one to ten physicians) and participated in at least one IPA. Eighteen percent practiced in small settings and did not participate in any IPAs, representing the cottage-industry sector. For specialists, these proportions were 15 percent, 12 percent, 41 percent, and 32 percent, respectively. Care management processes. The proportion of PCPs with patients enrolled in DMPs varied greatly, from a high of 67 percent among Permanente physicians to a low of 17 percent among cottage-industry physicians (Exhibit 1). DMP participation among PCPs in non-permanente medical groups and IPAs fell between these extremes. Overall, specialists were much less likely than PCPs to participate in DMPs (Exhibit 2), although the trends across practice setting among specialists were similar to those for PCPs. Availability of practice pattern information also varied by practice setting (Exhibit 1). Sixty percent of Permanente PCPs and 67 percent of PCPs in other medical groups reported receiving practice pattern information from their main medical group or practice over the prior year, compared with less than 25 percent of their colleagues in small-practice settings. However, 69 percent of physicians practicing in small settings and participating in at least one IPA received practice pattern information from their main IPA. Almost all Permanente PCPs who received practice pattern information described the information as very or somewhat useful (Exhibit 1). This proportion was higher than for PCPs in other medical groups (79 percent). By contrast, fewer than half of IPA PCPs who received practice pattern information from their main IPA described the information as very or somewhat useful. When asked, Why do you think these organizations provided you with practice pattern information? 40 percent of PCPs at Permanente and other medical groups responded mostly to improve quality, compared with 16 percent of IPA physicians. Five percent of PCPs at Permanente responded mostly to lower cost, compared with 29 percent in non- Permanente medical groups and 44 percent in IPAs. Physicians assessments of organizations motives for distributing this information were consistent with the types of information received by physicians in different practice settings. Physicians at Permanente and other medical groups were less likely than IPA physicians to receive information on costs of care and referrals to specialists. 54 November/December 2004

5 Cottage To Kaiser EXHIBIT 1 Participation In Disease Management Programs (DMPs) And Receipt Of Practice Pattern Information By Practice Setting, California Primary Care Physicians, Permanente groups (n = 173) Groups with 11+ physicians (n = 113) with IPA (n = 346) no IPA (n = 113) Patients are enrolled in DMPs 67.3% 33.9% 39.1% 17.2% Practice pattern information received From medical groups From IPA Usefulness of practice pattern information From medical groups Very useful Somewhat useful Not useful From IPA b Very useful Somewhat useful Not useful Perceived motive behind practice pattern information From medical groups Improve quality Cost and quality Lower cost From IPA c Improve quality Cost and quality Lower cost Received practice pattern information, by type Prescribing Cost of care Referrals to specialists Preventive care Disease performance Patient satisfaction SOURCE: University of California, San Francisco (UCSF), California Physician Survey sponsored by the California HealthCare Foundation and the Bureau of Health Professions, NOTES: P <.001 for all comparisons across practice settings except where indicated. IPA is independent practice association. a Not applicable. b p =.015 c p =.418. In contrast, much higher proportions of Permanente PCPs reported receiving information on patient satisfaction, preventive care, and disease-specific performance compared with their colleagues in other settings. Overall, there was no difference in the receipt of practice pattern information between PCPs and specialist physicians at Permanente; in all other practice settings, specialists were about half as likely as PCPs to receive practice pattern information and were less likely to describe the information as useful (Exhibit 2). Specialists in all practice settings received more practice pattern information on patient satisfaction, preventive care, and disease-specific performance than on HEALTH AFFAIRS ~ Volume 23, Number 6 55

6 Physician Practice EXHIBIT 2 Participation In Disease Management Programs (DMPs) And Receipt Of Practice Pattern Information By Practice Setting, California Specialists, Permanente groups (n = 202) Groups with 11+ physicians (n = 132) with IPA (n = 535) no IPA (n = 366) Patients are enrolled in DMPs 36.8% 11.3% 12.0% 3.0% Practice pattern information received From medical groups From IPA Usefulness of practice pattern information From medical groups Very useful Somewhat useful Not useful From IPA b Very useful Somewhat useful Not useful Perceived motive behind practice pattern information From medical groups Improve quality Cost and quality Lower cost From IPA c Improve quality Cost and quality Lower cost Received practice pattern information, by type Prescribing Cost of care Referrals to specialists Preventive care Disease performance Patient satisfaction SOURCE: University of California, San Francisco (UCSF), California Physician Survey sponsored by the California HealthCare Foundation and the Bureau of Health Professions, NOTES: P <.001 for all comparisons across practice settings except where indicated. IPA is independent practice association. a Not applicable. b p =.004 c p =.094. cost of care or referrals. The gradient of this quality-related information across practice settings for specialists was similar to the gradient for PCPs. Financial incentives. Greater proportions of physicians in medical groups and IPAs received bonus payments based on productivity, use of referrals, and use of prescription drugs (although these proportions were less than 25 percent across all practice settings; see Exhibit 3). More than half of Permanente physicians reported that quality of care was a factor in their bonus payment, and 61 percent reported that patient satisfaction was a factor. These proportions were much higher than for all other practice settings. 56 November/December 2004

7 Cottage To Kaiser EXHIBIT 3 Financial Incentives By Practice Setting, California Physicians, Permanente groups (n = 375) Groups with 11+ physicians (n = 245) with IPA (n = 881) no IPA (n = 479) Receive bonus payments 68.9% 40.8% 29.6% 14.0% Factors in calculation of year 2000 bonus Productivity Referrals to specialists Pharmacy prescribing a Profit sharing Quality of care Patient satisfaction SOURCE: University of California, San Francisco (UCSF), California Physician Survey sponsored by the California HealthCare Foundation and the Bureau of Health Professions, NOTES: P <.001 for all comparisons across practice settings except where indicated. IPA is independent practice association. a p =.006 Practice pressures. More than 80 percent of Permanente and other medical group physicians experienced pressure to see more patients per day; these proportions were lower among IPA physicians (Exhibit 4). However, fewer than one-third of Permanente physicians felt pressure to limit the number of tests ordered, and one-quarter felt pressure to limit the number of referrals to specialists; these proportions were much higher for physicians in non-permanente medical groups and IPAs. Cottage-industry physicians were the least likely to experience practice pressures in all three categories. Perceived value of large medical groups or IPAs. The perceived value of EXHIBIT 4 Experience Of Practice Pressures By Practice Setting, California Physicians, Practice pressures Permanente groups (n = 375) Groups with 11+ physicians (n = 245) with IPA (n = 881) no IPA (n = 479) See more patients No Yes, care not compromised Yes, care compromised Limit the number of referrals to specialists No Yes, care not compromised Yes, care compromised Limit the number of tests ordered No Yes, care not compromised Yes, care compromised 17.0% % % % SOURCE: University of California, San Francisco (UCSF), California Physician Survey sponsored by the California HealthCare Foundation and the Bureau of Health Professions, NOTES: P <.001 for all comparisons across practice settings. IPA is independent practice association. HEALTH AFFAIRS ~ Volume 23, Number 6 57

8 Physician Practice EXHIBIT 5 Value Of Participation In Large Medical Group Or IPA By Practice Setting, California Physicians, Perceived advantage of participating in large medical group or IPA Permanente groups (n = 375) Groups with 11+ physicians (n = 245) with IPA (n = 881) no IPA (n = 479) Overall practice viability Advantage Neither Disadvantage Attracting patients Advantage Neither Disadvantage Shared practice responsibilities Advantage Neither Disadvantage Quality of care Advantage Neither Disadvantage 77.0% % % % SOURCE: University of California, San Francisco (UCSF), California Physician Survey sponsored by the California HealthCare Foundation and the Bureau of Health Professions, NOTES: P <.001 for all comparisons across practice settings. IPA is independent practice association. large medical groups and IPAs varied by practice setting (Exhibit 5). Seventy-seven percent of Permanente physicians perceived an advantage for overall practice viability, compared with smaller proportions of their colleagues in other medical groups (56 percent), IPAs (40 percent), and the cottage industry (18 percent). Four-fifths of physicians at Permanente and two-thirds of physicians in other medical groups thought that large medical groups or IPAs were an advantage for shared practice responsibilities, compared with roughly one-quarter of physicians in smaller settings. Large medical groups and IPAs were perceived as an advantage for quality of care by 57 percent of Permanente physicians, 32 percent of other medical group physicians, and 7 percent of cottage-industry physicians. Physicians experiences might differ across practice settings because of differences in physicians characteristics across these settings; that is, younger physicians who gained experience with managed care in their medical training might choose to work in particular practice settings. However, in regression models that examined the effect of physicians age, sex, and race/ethnicity as potential confounders, only practice setting emerged as a significant predictor of care management experiences and practice pressures (data not shown). Summary We found compelling differences in physicians experiences with care management processes, financial incentives, and practice environments according to 58 November/December 2004

9 Cottage To Kaiser practice setting. Physicians in the Northern and Southern California Permanente medical groups two large, prepaid, multispecialty medical groups reported more use of care management processes than physicians in other organizational settings. Permanente physicians perceived their organization as placing more emphasis on quality improvement than on cost containment. This perception was supported by the fact that greater proportions of Permanente physicians received quality-related practice pattern information and quality-based financial incentives and that smaller proportions felt pressure to limit their number of tests and referrals, compared with their colleagues in other practice settings. Among specialists, Permanente physicians were particularly more likely to use care management processes than their counterparts in other practice settings. We were interested in the degree to which the experiences of Permanente physicians might be similar to those of physicians in other medical groups. This comparison must be interpreted with caution because of the heterogeneity within the other medical group category. Although the receipt of useful practice pattern information from their main medical group was similar, physicians experiences in these two settings also differed in many ways. Permanente physicians were much more likely than physicians in non-permanente medical groups to participate in DMPs. The proportion of Permanente physicians reporting that their organization s main motive for providing practice pattern information was to lower costs was one-sixth that for physicians in other medical groups. Permanente physicians received more practice pattern information on disease-specific care performance, preventive care, and patient satisfaction and less on referrals to specialists. Bonus payments for physicians in Permanente medical groups were much more likely to be based on quality of care and patient satisfaction. Although similar proportions of physicians in Permanente and other medical groups reported feeling pressure to see more patients per day, much lower proportions of physicians in Permanente medical groups reported feeling pressure to limit both the number of tests ordered and the number of referrals to specialists. We were also interested in comparing the experiences of physicians in non- Permanente medical groups with those of physicians in smaller practice settings linked by IPAs. We found similarities in the level of participation in DMPs, receipt of practice pattern information on disease-specific performance and preventive care, and bonus payments. The responses of physicians in non-permanente medical groups and those in IPAs also differed in important ways. Compared with IPA physicians, physicians in non-permanente medical groups were more likely to describe the practice pattern information they received as useful, were less likely to describe the organizational motive behind providing this information as mostly to lower costs, and were much less likely to receive practice pattern information focusing on cost of care and referrals to specialists. The perceived value of participation in a large medical group or IPA was generally more positive for physicians in non-permanente medical groups than for those in IPAs. HEALTH AFFAIRS ~ Volume 23, Number 6 59

10 Physician Practice In 2001, one in five California PCPs and one in three specialists practiced in small settings of one to ten physicians, without participating in an IPA. These physicians had a generally negative view of the role of large medical groups and IPAs. Two-thirds of these cottage-industry physicians considered participation to be a disadvantage for quality of care. These physicians reported much less use of care management processes than physicians in other organizational settings. Very low proportions had any patients enrolled in DMPs. Only 17 percent received practice pattern information from their main medical group or practice. Few received bonuses in 2000 that were based on quality of care or patient satisfaction. On the other hand, cottage-industry physicians experienced less practice pressure than physicians in other organizational settings. Study limitations. This study has four major limitations. First, it relied on physicians self-reports, which did not allow us to draw conclusions about the efforts of physician organizations to promote care management activities. Arguably, where there is a difference between administrative efforts and the day-to-day practice of medicine, physicians perceptions are the more important measure of care processes. Second, physicians are not randomly distributed to organizations. This could introduce confounding of our study results for example, if physicians who were more interested in and adept at managing care chose their practice setting based on these preferences and skills. Although the results of our regression models suggest that physicians age, sex, or race/ethnicity did not confound the association of practice setting with care management and practice pressures, there may have been other unmeasured variables that influence these associations. The third major limitation is the heterogeneity of the non-permanente medical group category. Because relatively few physicians in California practice in these types of group practices with eleven or more physicians, to achieve an adequate sample size for this category we pooled physicians whose groups spanned a wide range of sizes and organizational types. The few large, well-established medical groups in California may be performing similarly to Permanente groups; by pooling physicians from these groups with physicians from smaller group practices that have more modest infrastructures, we may have obscured variations within this category. However, this study was not designed to capture this potential degree of variation within non-permanente medical groups. The fourth limitation is that this study focuses primarily on the role of physician organizations; thus, physicians experiences with care management processes implemented by other organizations such as hospitals and health plans were not addressed. In certain cases (DMPs and bonuses), we were unable to attribute the source of a program to a particular physician group, health plan, or other entity. However, the finding that physicians working in the four types of physician organizations had differential exposure to these programs, regardless of source, remains important. Much of the responsibility for implementing health plan initiatives in California is delegated to physician organizations, which further com- 60 November/December 2004

11 Cottage To Kaiser plicates the assessment of physicians experiences. More information is needed about the interfaces between the practicing physician and the various organizations that attempt to influence health care delivery. Policy implications. These results have important policy implications. The first is the success of the Permanente medical groups in creating an environment in which physicians perceive quality to be an organizational priority. Although a small number of other large medical groups in California are similar in structure to Permanente groups and may therefore boast similar performance, taken together, non-permanente medical groups in California have not achieved the same result as the Permanente groups. The Permanente groups exclusive relationship with the Kaiser Foundation Health Plan may allow for efficiencies that improve the organization s ability to implement quality-related processes of care. In addition, information technology capabilities have been shown elsewhere to be associated with the adoption of care management processes; Kaiser Permanente s leadership in this area may be contributing to its success. 8 The second major policy implication is the difference between IPAs and medical groups in the ability to implement care management processes and achieve buy-in from their physicians. Physicians in IPAs perceive their IPA to be more focused on cost than quality, and they report that financial incentives from their IPA are not aligned with a quality improvement agenda. In theory, the organizational structure of the IPA allows physicians in solo or small-group practice to come together to benefit from administrative efficiencies, contracting strength, and a shared clinical culture. 9 Although some IPAs have been successful at developing a shared clinical culture and have placed emphasis on quality improvement, our data suggest that this did not describe the experience of most IPA physicians in California in Our findings raise disturbing questions about how the United States will close the quality chasm in medical care. 11 Physicians in the large, integrated Permanente medical groups have adopted and value system-level care management tools much more so than physicians in IPA settings or traditional cottageindustry practices. And yet most physicians continue to practice in solo and small-group settings, and IPA participation is declining. In California one-third of PCPs and nearly one-half of specialists are in solo practice a proportion that has not changed in recent years. 12 More evidence is needed to firmly establish whether care management tools, deployed in routine practice, truly lead to better patient outcomes. If this evidence is forthcoming, physicians in small-office settings will face increasing pressure from purchasers and the public to demonstrate that they are willing and able to deploy care management tools of proven effectiveness. How small practices will accomplish this without the benefit of a larger organizational infrastructure remains unclear. Health plans and professional organizations may have a role in supporting cottage-industry physicians. Alternatively, the United States might learn from other countries that are exploring decentralized, coopera- HEALTH AFFAIRS ~ Volume 23, Number 6 61

12 Physician Practice tive models of physician organization. 13 Whatever form of physician organization prevails in the United States, it is difficult to envision the closure of the quality chasm without a major retooling of small-office practice. This study was funded by the California HealthCare Foundation and the Bureau of Health Professions, Health Resources and Services Administration. The authors acknowledge Jean Yoon and Arpita Chattopadhyay for their assistance with programming and Queenie Bin for her assistance with manuscript preparation. NOTES 1. Institute of Medicine, Crossing the Quality Chasm: A New Health System for the Twenty-first Century (Washington: National Academies Press, 2001). 2. D.R. Rittenhouse et al., The Role of Provider Organizations in Medi-Cal Managed Care (Sacramento: California HealthCare Foundation, February 2004); and L. Casalino et al., External Incentives, Information Technology, and Organized Processes to Improve Health Care Quality for Patients with Chronic Diseases, Journal of the American Medical Association 289, no. 4 (2003): S.M. Shortell and J. Schmittdiel, Prepaid Groups and Organized Delivery Systems: Promise, Performance, and Potential, in Toward a Twenty-first Century Health System: The Contributions and Promise of Prepaid Group Practice, ed. A.C. Enthoven and L.A. Tollen (San Francisco, Jossey-Bass, 2004). 4. Ibid. 5. K. Grumbach et al., Independent Practice Association Physician Groups in California, Health Affairs 17, no. 3 (1998): ; J.C. Robinson and L.P. Casalino, Vertical Integration and Organizational Networks in Health Care, Health Affairs 15, no. 1 (1996): ; and B.N. Shenkin, The Independent Practice Association in Theory and Practice: Lessons from Experience, Journal of the American Medical Association 274, no. 22 (1995): Prior studies of physician organizations have interviewed medical directors and chief executive officers, not individual physicians. See L.P. Casalino et al., Benefits of and Barriers to Large Medical Group Practice in the United States, Archives of Internal Medicine 163, no. 16 (2003): ; Casalino et al., External Incentives ; R.R. Gillies et al., How Different Is California? A Comparison of U.S. Physician Organizations, Health Affairs, 15 October 2003, content.healthaffairs.org/cgi/content/abstract/hlthaff.w3.492 (10 August 2004); and M.B. Rosenthal et al., Transmission of Financial Incentives to Physicians by Intermediary Organizations in California, Health Affairs 21, no. 4 (2002): One study of physicians use of care management tools did not consider practice setting. See M. Reed, K. Devers, and B. Landon, Physicians and Care Management: More Acceptance than You Think, Issue Brief no. 60 (Washington: Center for Studying Health System Change, January 2003). Another study analyzed individual physician data according to practice setting to draw conclusions about physician satisfaction. See R. Sturm, Effect Of Managed Care and Financing on Practice Constraints and Career Satisfaction in Primary Care, Journal of the American Board of Family Practice 15, no. 5 (2002): Sample size precluded us from breaking down these categories further. The unweighted N for each of the categories was as follows: Permanente physicians: PCPs, 173; specialists, 202. Other medical group physicians: PCPs, 113; specialists, 132. IPA physicians: PCPs, 346; specialists, 535. Cottage-industry physicians: PCPs, 113; specialists, Casalino et al., External Incentives. 9. Robinson and Casalino, Vertical Integration and Organizational Networks ; and Shenkin, The Independent Practice Association. 10. K. Terry, How IPAs Are Changing, Medical Economics 80, no. 12 (2003): IOM, Crossing the Quality Chasm. 12. K. Grumbach et al., California Physicians 2002: Practice and Perceptions (San Francisco: California Workforce Initiative at the UCSF Center for the Health Professions, December 2002). 13. Primary care trust (PCT) structures in England are an example of this type of decentralized cooperative model operating in a very different health system context. See, for example, S. Campbell, M. Roland, and D. Wilkin, Primary Care Groups: Improving the Quality of Care through Clinical Governance, British Medical Journal 322, no (2001): ; and A.B. Bindman, J.P. Weiner, and A. Majeed, Primary Care Groups in the United Kingdom: Quality and Accountability, Health Affairs 20, no. 3 (2001): November/December 2004