Recognition as a Patient-Centered Medical Home: Fundamental or Incidental?

Setting

We studied a large nonprofit multispecialty group practice in northern California (hereafter referred to simply as the group; per institutional review board protocol, all names are pseudonyms) that serves 4 counties and more than 700,000 patients who mirror Northern California demographically. The group has approximately 4,500 employees and 1,000 physicians, one-half primary care and one-half specialists. Payment mix is 80% preferred provider organization and 20% health maintenance organization. Of 17 primary care clinics, 11 were recognized by NCQA in 2007–2010—6 at Level 2 and 5 at Level 3 (the highest rating).

As part of a mixed methods case study of PCMH at the group, a multidisciplinary team of social scientists carried out qualitative research led by 2 ethnographers (D.D., M.H.M.) with extensive qualitative experience. In addition, 2 economists involved in the larger study provided comments, and an additional sociologist participated in coding.

Sample

We conducted 38 semistructured in-person interviews at practice sites with front-line staff (physicians, nurses, medical assistants) and administrators (department leads, division heads, and senior executives). Participants were chosen purposively to include personnel involved with PCMH implementation across locations and at different organizational levels.

Data Collection

Interviews took approximately 60 minutes and focused on PCMH implementation. Participants provided written consent but were not compensated. Interviews were audio-recorded and transcribed by a professional service. The interview guide addressed the informant’s concept of PCMH, decision to seek and then not renew NCQA recognition, organizational and cultural meaning of recognition, and implementation barriers or facilitators. The guide is given in Supplemental Appendix 1 (available online http://annfammed.org/content/11/Suppl_1/S14/suppl/DC1). One member of the team (M.H.M.) attended all but 2 interviews, and most interviews included a second team member (D.D., D.Y.H., M.T-S., or D.L.F.).

Data Management and Analysis

All transcripts were entered into ATLAS.ti (ATLAS.ti Scientific Software Development GmbH). Two authors (D.D., M.H.M.) led biweekly meetings to discuss data collection, emergent themes, and sampling strategy. We drafted a codebook to reflect study aims and modified it iteratively based on ongoing analysis. The codebook had 2 tiers: a component code identified particular PCMH components and a substantive code tagged content. We also included codes for cross-cutting themes. Transcripts were coded at the paragraph level. Five interviews were coded by 2 independent coders to establish reliability, and coding differences were discussed and reconciled. When one-half of the interviews had been completed, the team held a day-long retreat to finalize codes, discuss emergent themes, and plan final interviews and analyses that were carried out by the 2 qualitative leads.

PCMH Principles and Organizational Values

Key informants told us that patient centeredness and other key principles of PCMH had always been part of the group’s organizational culture. Executives said founders of the group wanted patients to have a “1-stop shop” for primary and specialty care. Nearly all clinic staff agreed with the sentiment a senior administrator expressed: “the idea of the patient at the center I think has always been at the core of how this group operates.” Informants cited this orientation as responsible for the organization’s willingness to experiment with new initiatives and innovations, as a physician-administrator put it, “We want to improve processes so that we take better care of patients.” Another physician leader cited a number of care redesign projects—“a patient care medical home, primary care redesign or innovation center, MyHealth Online [a Web-based patient portal]”—that he felt illustrated how “there’s always been attempts to improve care, always.” As a nurse summarized: “To me, process improvement and making things better for our patients should be incentive enough.”

The organization did not embrace patient centeredness blindly, however, as leaders recognized that “our model of patients first” had to bend to practical constraints and limits. Fee for service was a common concern, as this administrator noted:

The current primary care medical home concept is really an expansion of talent to support the primary care physician, whether it’s pharmacy, social work, etcetera, to enable the primary care physician to really have a larger panel with more support and be able to be more focused on medical decision making and have the support needed, in a team delivery concept...[However] it is not cost-effective in a fee-for-service environment. In fact, it’s unsustainable in a fee-for-service environment.

Physicians voiced concerns about PCMH components perceived to be burdensome. One physician-administrator recalled a plan to expand e-mail so “every single patient could just e-mail the doctor for free.” This was among the “worst fears” of physicians who worried that it would “open the floodgates” and to overwhelming demand that would amount to “shooting ourselves in the foot.” Fiscal constraints impeded adoption of other patient-centered innovations such as care coordination. Coordination was seen as a benefit for patients, but, as a senior administrator articulated, “how do you build patient-centered medical home when the adjunct people are almost as costly as primary care doctors?” Informants also noted the difficulty of implementing the vague principles of PCMH. In the words of a physician administrator:

How do you incentivize right care? Or how do you incentivize avoiding care, unnecessary care, or investing in nonreimbursed stuff in a fee-for-service environment, even when that nonreimbursed stuff provides value? How do you do that? That’s the most important question.

Implementing PCMH components that allowed physicians to see more patients was strategically important. In the compelling image offered by one executive, the group has:

…a foot in each of 2 canoes, going down the rapids, and the canoes are diverging…The 2 canoes are fee for service and some form of capitation…Ultimately, we will go risk sharing, so that’s why we’re developing the panel size…it’s preparation for the future.

In sum, PCMH recognition was consistent with the group’s fundamental values: being patient oriented, innovating to keep up with local competitors, and increasing efficiency in preparation for a future that requires lower total costs.

PCMH Recognition and Incidental Circumstances

Although the group valued PCMH principles, it pursued NCQA recognition only after being offered a financial incentive. A physician-administrator explained the reasoning of a group of local employers who put up to $150,000 on the table: “‘Let’s do another type of pay-for-performance program with our local medical groups, and let’s make it around IT [information technology] infrastructure because we know that that’s associated with better care.’” Another physician leader noted that the group responded to the employer offer in part to improve care but also because it offered the chance “to talk to a coalition of large employers, in a committee without contracting any insurance company intermediaries.” The employers accepted NCQA recognition as evidence of IT infrastructure and tied the bonus to PCMH level. Even though the employer group bonus “wasn’t a huge factor” (midlevel administrator) and “barely covered our costs because this was such a costly thing in terms of people’s time” (physician-administrator), the employer challenge still prompted the group to seek recognition.

Recognition per se, was not an organizational priority however. An administrator stated, “If you said we’re going to start [a] project, and we’re going to improve the way we treat patients, well, I’m all there. That’s interesting to me. But I don’t really care if there’s a certification [sic].” Some administrators were even unaware that the group had received NCQA recognition as a PCMH. For example, after being told by the interviewer that recognition had lapsed, one executive said, “I didn’t know that, but okay.”

Group leaders involved in applying for recognition recalled “this was a painful process” where it was “not clear exactly what we have that would really fit [NCQA requirements]…we would have roundtables where we would say, ‘Okay this is what they’re looking for.’” No centralized effort was made to seek recognition; each division relied on its own staff and resources. Some divisions made do with fewer resources as this informant recalled proudly: “It doesn’t matter if we have less staff than everybody else, it doesn’t matter if we have less implementation resources than everybody else. We just figure out how to do it and we do it.” Nevertheless, without an organization-wide commitment to apply for recognition for all clinics, recognitions were sought for only one-half of the primary care clinics, clinics received different levels of recognition, and higher levels of recognition seemed to go to clinics that were better staffed and resourced.

In 2010, as the NCQA recognition obtained in 2007 was set to expire, the group explored whether payers remained interested in PCMH recognition. A physician-administrator recalled asking contracting staff, “when you’re sitting down with health plans or PPOs [preferred provider organizations] and you’re around the bargaining table, is anybody asking about certification (sic)…as a patient-centered medical home? The answer was no.” This information was taken into account as group leaders weighed the costs and benefits of recognition against competing organizational priorities, such as strategic planning for redesigning chronic care and an organization-wide integration of disparate scheduling and electronic health record systems into one common system. The absence of enthusiasm from payers and pressing internal competing demands resulted in the decision to let the recognition lapse.

As noted above, contextual factors played a key role in the findings we observed. These factors are summarized in Supplemental Appendix 2 (available online http://annfammed.org/content/11/Suppl_1/S14/suppl/DC1).