|
Article Excerpt
Patient experience surveys have become central to performance measurement activities nationwide, including pay-for-performance and public reporting initiatives (Cleary 1999; Damberg et al. 2005). Little is known about the extent to which characteristics and activities of medical groups and market factors are related to individual physician performance on patient care experience measures. For example, physicians belonging to medical groups that provide stronger financial incentives and organizational support for improving performance on patient experience measures may be more likely to perform well.
Previous studies indicate that patients of staff/group model health maintenance organizations (HMOs) consistently report worse experiences of care and lower visit continuity compared with patients of network HMOs (Safran, Tarlov, and Rogers 1994; Safran et al. 2000; Safran et al. 2002). Over the past decade, physicians belonging to staff/group model HMOs diversified their health plan contracts in order to gain access to patients who were opting for health plans that offered broader physician networks (Robinson and Casalino 1996; Robinson 2001). In response to competitive pressures, many integrated medical groups also reorganized primary care practices to meet consumer demands for patient-centeredness and choice (Grumbach and Bodenheimer 2004; Stein, Frankel, and Krupat 2005; Grembowski et al. 2008). No study, however, has assessed the association of medical group factors and individual physician performance on patient experience measures. Medical groups have a stronger influence on patients' experiences of care than health plans (Solomon et al. 2002; Safran et al. 2006a), so assessing medical group influences on physician performance on patient experience measures might offer important insight regarding the most effective methods for stimulating improvement.
Performance on patient experience survey measures might also be influenced by factors that cannot easily be modified by physicians, such as the availability of physicians in an area. An assumption when employing physician-level performance incentives is that physicians have some control over the care processes and outcomes being measured and rewarded. Case mix adjustment of performance measures is a common method for ensuring equitable comparisons (Zaslavsky et al. 2001) and most survey initiatives currently employ these statistical adjustments. There is increasing concern, however, that pay-for-performance programs might exacerbate racial and ethnic disparities in care quality (Casalino and Elster 2007; Felt-Lisk, Gimm, and Peterson 2007; Millett et al. 2007; Snyder and Neubauer 2007) because physicians working with disadvantaged populations might face difficulties improving performance. For example, patients who receive care in markets with primary care physician (PCP) shortages may experience difficulty scheduling appointments, longer office wait times, and clinical interactions that focus on resolving immediate medical problems rather than eliciting concerns and negotiating an agenda (Kroenke 1998; Marvel et al. 1999; Peltenburg et al. 2004; Rodriguez et al. 2008). As a result, lower financial resources for practices in resource constrained environments compared with other practices might cause or exacerbate performance differences and racial and ethnic disparities in care quality.
It is not yet known, however, the extent to which individual PCP performance on patient experience measures is related to market factors that are difficult to modify or the characteristics and activities of medical groups. This study assesses the extent to which organizational and market factors are related to individual physician performance on patient experience measures using survey data from the largest pay-for-performance program in the United States--the Integrated Health Association's statewide initiative in California (Damberg et al. 2005).
METHODS
Patient Sampling and Survey Administration
The study draws on commercially insured patients who had visits with 2,286 PCPs of adult patients from 43 medical groups in California during 2005 and 2006. Medical groups included a diverse range of physician organizations, including 27 independent practice associations (IPAs), 12 integrated medical groups, and four "hybrid" groups that are composed of a core integrated group and an associated IPA. During each of the 2 survey years (2006 and 2007), a random sample of approximately 100 patients per physician who had at least one visit with their PCP during the prior year were mailed a survey with items from the Clinician & Group CAHPS survey, a previously validated instrument that measures patients' experiences with a specific, named physician and that physician's practice (Agency for Healthcare Research Quality et al. 2006; Safran et al. 2006a). Mailings included an invitation letter, a printed survey, and a postage-paid return envelope. The survey invitation included a personal online code that gave respondents the option of completing the survey using the web. Previous work demonstrated the absence of web survey mode effects for questions in the Clinician & Group CAHPS survey (Rodriguez et al. 2006). The invitation listed a toll-free number for patients to obtain surveys in Spanish. A second invitation and questionnaire were sent to nonrespondents 2 weeks after the initial mailing. Each annual data collection effort spanned a period of approximately 8 weeks.
Patient Survey Content
For this study, we analyzed four composite measures: physician communication (six items), access to care (five items), care coordination (two items), and office staff interactions (two items) (Table 1). The physician communication, access to care, and office staff interactions measures represent core item and composite content of the Clinician & Group CAHPS Survey (Agency for Healthcare Research Quality et al. 2006), which was endorsed by the National Quality Forum for use in evaluating ambulatory care received from individual physicians and their practices. Survey questions use a six-point response scale ranging from "Never" to "Always" and reference care received from the patient's PCP and the PCP's practice over the prior 12 months. All composite measures analyzed achieve physician-level reliability of 0.70 or higher with samples of 40 established patients per physician (Safran et al. 2006a; Rodriguez et al. 2007).
Patient Sample
Of 332,326 outgoing surveys over the 2 study years, 11,964 (3.6 percent) were undeliverable because of bad address information or patient death. Surveys were received from 120,952 respondents, yielding an unadjusted response rate of 36.4 percent and an adjusted response rate of 37.8 percent. The analytic sample included 112,650 respondents (average per physician = 50.1) who confirmed having seen their PCP during the prior 12 months. Respondents who did not confirm the named physician as their PCP or indicated that they did not visit the physician during the prior 12 months (n = 8,302) were excluded from the analysis.
Patient Survey Composite Scoring
As detailed elsewhere (Safran et al. 2006a), the survey composite scores could range from to 100 points, with higher scores indicating more favorable performance. For example, a response of "Never" would be scored as "0" and a response of "Always" would be scored as "100". Composite scores were computed for each respondent based on the unweighted average of responses to all items comprising the measure. Following the half-scale rule (Nunnelly and Bernstein 1994), respondents had to answer at least 50 percent of questions comprising the composite for a score to be computed. Survey question wording, response scales, and placement in the survey were identical during the 2 study years.
Medical Group Interview
Medical group director interviews were conducted via telephone between April and June 2007. The eligible medical groups consisted of groups that participated in the individual physician-level patient care experience survey in 2007 (n = 43). The interview assessed the medical groups' current financial incentives, including...
|
