Support and Strategies for Change Among Small Patient-Centered Medical Home Practices

Sampling Approach

NCQA identified a sample of small practices (those having <5 physicians) with the Physician Practice Connections’ Patient-Centered Medical Home recognition.¹⁶ NCQA defines a practice as a single geographic site that shares medical records and key office systems. To be considered for recognition, practices voluntarily complete a Web-based questionnaire and attach documentation to support their responses. Trained surveyors review the documentation, and 5% of applications undergo on-site audit. All applications undergo 3 rounds of internal review. Practices could achieve 3 levels of recognition based on the total score as well as their performance on 10 “must pass” elements. On July 15, 2011, when we selected our study sample, there were 2,418 recognized practices: 30.2% were Level 1, 4.3% were Level 2, and 65.5% were Level 3.

To be able to make comparisons between levels of recognition across a number of states, we stratified the sample by Level 1 and 3 (dropping the small number of Level 2 practices). We included only states having both Level 1 and Level 3 recognized practices, and decreased the sampling weight for 2 states with very large numbers of recognized practices (New York and Pennsylvania), so that they did not have greater representation than the distribution of practices nationally. We surveyed the lead physician at the practice; where that was not documented or the practice indicated there was not a single lead physician, we randomly selected a physician.

Of the 2,418 recognized practices, 915 were excluded because they had 5 or more physicians or did not have any physicians. Another 202 practices were excluded because they did not provide primary care services to adult patients, and 59 of the remaining practices were excluded for having Level 2 recognition. Of the 1,242 eligible practices on July 15, 2011, we selected a random sample of 430 practices (split between Levels 1 and 3) for surveying. Twenty-three practices that we later discovered had closed or merged with other practices were replaced randomly. Nine practices that no longer met the definition of small on the basis of data collected after selection or survey completion were excluded.

Practices received a gift basket for participation. The study was reviewed, approved, and monitored by Chesapeake Research Review Inc, an independent, accredited institutional review board.

Variables

The questionnaire was developed using insights from in-depth case studies conducted in 6 PCMH-recognized small practices. The questionnaire items were revised on the basis of pretesting and cognitive interviews.

To assess practice priorities, we adapted a single item rating on a 10-point scale used in previous studies to assess the priority for QI to address priority for making the practice more of a PCMH.⁹ Higher ratings indicated higher priority.

We adapted items assessing motivations and barriers for QI used in the National Survey of Physician Organizations.⁵ For motivations, we asked respondents to rate the importance of 6 factors in the practice’s decision to seek NCQA recognition as a PCMH on a 5-point scale (1 = not at all to 5 = very important). We also asked practices to report on the extent to which 6 factors were barriers to implementation of the PCMH model also on a 5-point scale (1 = no barrier to 5 = great barrier).

We asked respondents if they had ever been involved in a demonstration project, pilot project, or other program including those sponsored by state, federal, or health plan entities, or in other initiatives. We also inquired about whether practices received any type of payment for being a PCMH (eg, enhanced fee-for-service remuneration, monthly payments, performance incentives).

Respondents reported whether their practice received any type of help when implementing the PCMH and how useful this help was (very, somewhat, or not at all) for their practice.

Finally, to determine strategies and capabilities for practice change, we used the Change Process Capability Questionnaire (CPCQ) of Solberg et al⁹ to assess strategies used to implement the PCMH as well as aspects of organizational culture. Those investigators showed that the CPCQ distinguishes among medical groups and is associated with both change priorities and implemented systems for depression.⁹ The strategies scale includes 18 items. The composite is a sum of items rated “yes, worked well” (1 point), “yes, did not work well” (0.5 point), and “no” (0 points). The CPCQ has 14 items addressing capabilities for change; the score is a sum of items rated 4 or 5 on scale from 1 (strongly disagree) to 5 (strongly agree).

Analysis

We present descriptive analyses of individual items and composites. We explored associations between level of recognition and practice type using contingency tables and Pearson χ² tests for categorical survey variables. For continuous items and composite scores, associations were tested with independent samples t tests. We did not make any adjustment for multiple comparisons. The study was planned with a recruitment sample size of 430 that was expected to attain 300 completed questionnaires at a 70% response rate in order to achieve 80% power (α = .05, 2-sided tests) to detect a difference by level of 25% vs 40% for binary questionnaire items and a difference in means by Level 1 vs 3 of the Cohen d of 0.3 for continuously scored survey items.

Study Group

Of the 421 eligible practices randomized that met eligibility criteria at the end of the study, 249 practices (59.1%) in 23 states responded. The respondents included 91 Level 1 practices, 125 Level 3 practices, and 33 practices that changed from Level 1 to Level 3 between the time of sampling and completion of the survey period. Because this group that changed generally showed greater resemblance to Level 3 practices, we combined these 2 groups in our analysis. We found no significant differences in response based on practice size or recognition level. Table 1 shows characteristics of the practices and the responding physicians. Level 3 practices were more likely to have a fully electronic health record; the respondents for Level 3 practices were more likely to have graduated from medical school more recently than those in Level 1 practices.

Priorities, Motivators, and Barriers

As shown in Table 2, less than one-third of practices rated PCMH implementation as a very high priority. Improving quality and patient experiences were the strongest motivators for PCMH implementation (across all levels). Time and resources were the biggest barriers, and compared with Level 3 practices, Level 1 practices rated these factors and information systems as significantly greater barriers.

Demonstration Projects, Payment, and Help With PCMH Implementation

More than 75% of the practices participated in demonstration or pilot projects or received financial rewards for PCMH, and almost one-half (45.4%) did both (Table 3). The large majority of practices received training for staff (85.5%) or clinicians (84.2%). More than one-half of these practices received consultation specific to their practice (63.9%) or access to a learning collaborative (59.3%). Most practices received training on how to meet NCQA’s recognition requirements (81.3%) or help with preparing their application (81.0%); they also rated such help related to the recognition program as most useful. Level I practices were more likely to receive training about NCQA recognition requirements. Training and engagement in improvement for patient representatives or partners was both much less common and less valued.

Practice Change Strategies and Ability

The most commonly used strategies for practice transformation were developing systems for care, training, and quality measurement/goal setting (Table 4). The least commonly endorsed strategies were involving patients in QI and using formal QI strategies. Only 15.5% of practices said they included patients on QI committees or teams. Many practices reported using some elements of QI methods but seemed not to recognize them as such (eg, 50% said they used pilot testing of new methods, but only 30% said they used formal methods). Compared with Level 1 practices, Level 3 practices scored significantly higher on the CPCQ strategies scale (P = .01) and were more likely to use measurement-based strategies such as pilot testing or providing feedback to clinicians.

High ratings of organizational change ability were seen in the areas of clinicians adhering to practice policies, leaders showing enthusiasm for QI, and clinicians espousing a shared mission (Table 5). Lower ratings were seen in the areas of receiving feedback from patients about benefit from changes, practices prioritizing quality of care over finances, and practice operations relying on organized systems. Level 3 practices showed more capability for practice change in the areas of enthusiastic leadership for care, use of organized systems, and availability of resources for quality improvement.

Key Findings

This is the first study to examine transformation help and strategies among a national sample of small practices that have independent documentation of PCMH implementation. Of these 249 practices in 23 states, the large majority are participating in demonstration projects or receiving financial rewards for using the PCMH model, and most have received training or other kinds of help. Some demonstration projects require obtaining NCQA recognition for participation. We found some key differences between Level 3 and Level 1 practices—primarily in the availability of fully electronic health records, the types of barriers reported, strategies used for QI, and in particular, the use of measurement-based strategies, On other topics, including priority and motivation for PCMH implementation, types of support received, and organizational capabilities for change, practices with different degrees of PCMH implementation were similar. These findings shed light on the strategies that are important to PCMH implementation and suggest the types of support that all small practices need for the “epic whole-practice reimagination and redesign” that PCMH transformation requires.¹³

First, small practices that have achieved recognition did so in the presence of and likely because of local demonstration projects, financial incentives, and various supports for change. After the desire to improve quality and patient experiences, and improve efficiency, respondents reported that the availability of financial incentives was the strongest motivator to change. Time, resources, and information systems were reported to be barriers for all practices but particularly for Level 1 practices—suggesting that addressing these barriers is critical for helping practices to seek higher levels of PCMH implementation. Importantly, most practices did not view PCMH implementation as the highest priority compared with other priorities such as financial goals, QI, or implementing meaningful use requirements for electronic health records.

Second, practices placed the greatest value on tangible supports. Although most practices received training for clinicians and staff, and many also received consultation or access to collaborative support, the type of help they valued the most was training and support related to the NCQA application process. This finding suggests that practices appreciate training directed toward specific goals and help with meeting the requirements for external review more than more general types of assistance.

Third, the use of QI strategies among PCMH-recognized practices highlights the differences between these practices and other small practices. Compared with small practices in the survey by Rittenhouse et al,⁵ a larger proportion of PCMH-recognized practices in our survey used rapid cycle QI (51.4% vs 10.2%) and gave performance feedback to physicians (60.3% vs 18.9%). These strategies also distinguish practices with Level 3 vs Level 1 PCMH recognition. Although the use of a fully electronic health record is not specifically required for PCMH recognition, the greater use of these records may support these QI strategies and other activities of the PCMH such as care coordination and population management.

Fourth, most PCMH-recognized practices have yet to engage patients as partners in transformation efforts. Although 30% of practices said they had trained clinicians and staff on involving patients or consumer advocates, only 15% of practices included patients on QI teams. Along with the lower value ascribed to patient training, it appears that PCMH practices may need more help in navigating the practical steps in successful patient engagement; training on how to work with patients and consumer representatives in QI may need to happen before actual training for the patients and advocates. Moreover, practices may need more convincing evidence or experience about the link between patient engagement in QI efforts and achieving their key motivations of improving quality and patient experiences.¹⁷

Finally, the large proportion of Level 1 practices that switched to Level 3 recognition suggest that many PCMH practices are on a continuing journey of transformation. More than one-quarter of Level 1 practices switched during our time frame, and many of the Level 3 practices in this study had previously upgraded from Level 1 before the study. This finding demonstrates that practices are able to successfully navigate the roadmap to greater PCMH implementation.

The Supplemental Appendix (available online at http://annfammed.org/content/11/Suppl_1/S6/suppl/DC1) details some of the contextual factors that my have influenced our findings and that others may wish to consider when undertaking PCMH initiatives in small practices.

Study Limitations

Several limitations affect the generalizability of our findings. Only 59% of practices responded to this survey. This response rate may affect the representativeness of the results, but we did not detect any differences between those that did or did not respond on the basis of characteristics that we could test (number of physicians, number of employees, and recognition level). It would be preferable to gain perspectives from multiple clinicians or staff on some topics; however, our resources allowed for only 1 respondent per practice. Only practices that have NCQA recognition were included. Several other organizations and states qualify practices as medical homes; however, NCQA’s program is the largest program, representing nearly 24,000 clinicians. Our study may not be representative of all NCQA-recognized practices because we sought to ensure the ability to compare Level 1 and Level 3 practices. The size of these practices may vary because we used the number of physicians as a determinant of practice size rather than the number of clinicians. In addition, we were unable to include practices led by nurse practitioners or physician assistants because we did not have consistent information about those practices in the NCQA data. Despite these limitations, this survey is the first to look across nearly one-half of US states at transformation experiences of small practices where there is independent confirmation of their PCMH implementation.

Implications

Small practices that have achieved recognition as PCMHs generally do so in the context of financial incentives and other supports. To further spread the adoption of the PCMH model, these efforts are likely to continue to be important. Practices value training and help that answers immediate needs; finding ways to make training more useful and oriented to daily work of practices is important. These practices infrequently included patients in their QI efforts. Policy makers who design demonstration projects and organizations that qualify practices as PCMHs should consider how to balance incentives for practices to organized systems as well as to promote patient engagement. Research is needed to demonstrate and identify the most successful ways to engage patients in PCMH implementation.