Patient Experience Over Time in Patient-Centered Medical Homes

In this study, the Patient-Centered Medical Home was associated with improvements in patients' experience with access to care but not other domains of care.
Published Online: May 17, 2013
Lisa M. Kern, MD, MPH; Rina V. Dhopeshwarkar, MPH; Alison Edwards, MStat; and Rainu Kaushal, MD, MPH
Objectives: Although the Patient-Centered Medical Home (PCMH) model is being implemented across the country to transform primary care, it is not yet clear whether this model actually improves patients’ experiences with healthcare. Our objective was to measure patients’ experiences over time in practices that transformed into PCMHs.

Study Design: We conducted a prospective study, using 2 serial cross-sectional samples, in a multipayer community.

Methods: We surveyed 715 patients: 346 at baseline, when practices had just completed transformation, and 369 at follow-up, which was a median of 15 months later. These patients received care from 120 primary care providers at 10 ambulatory practices (20 sites) that achieved Level III PCMH, as defined by the National Committee for Quality Assurance. We measured patient experience, as defined by the 7 domains of the Clinician and Group-Consumer Assessment of Healthcare Providers and Systems (CG-CAHPS) Adult Primary Care Questionnaire.

Results: Patients’ self-reported experience with access to care improved significantly over time, with 61% of respondents giving access to care the highest rating at baseline versus 69% at follow-up (P = .02). There were no significant changes over time for the other domains.

Conclusions: The PCMH was associated with improvements in patients’ experience with access to care but not other domains of care. This study, which took place in a multi-payer community, is one of the first to find a positive effect of the PCMH on patient experience.

Am J Manag Care. 2013;19(5):403-410
Although the Patient-Centered Medical Home (PCMH) model is being implemented widely, it is not yet clear whether this model actually improves patient experience.

  • We found that patients’ experience with access to care improved significantly over time, with 61% of respondents giving access to care the highest rating at baseline versus 69% at follow-up (P = .02).

  • There were no significant changes for the other domains.

  • Improvements in access to care may be one of the earliest effects of the PCMH.
Understanding patients’ experience of healthcare is critically important.1 If patients are not satisfied and engaged with their healthcare providers, then healthcare is unlikely to be successful in improving health. Aspects of patients’ experience of healthcare include their perceptions of: the quality of their relationships with providers, the quality of disease management, access to care, and communication with office staff.2

Efforts are under way across the country to improve healthcare, in part through improving patients’ experience.3 One of the most common approaches, with more than 100 demonstration projects under way, is the Patient-Centered Medical Home (PCMH).4 The PCMH is being promoted as a strategy for improving patient-centeredness, improving quality, and decreasing cost.3 This model is also being used to test alternative reimbursement strategies for primary care.5,6

The PCMH is a set of practice standards that emphasize coordination of care and management of chronic disease over time. PCMH standards developed by the National Committee for Quality Assurance (NCQA) include processes of care for: optimizing access to care and patient communication, patient registries, care management, patient self-management support, electronic prescribing, tracking of tests and referrals, and performance reporting.2 Practices receive NCQA recognition when they have implemented a specified number and pattern of processes of care, with different levels of recognition depending on the number and pattern implemented.2

Whether these processes translate into better experiences for patients is not yet clear, in part because evaluations of most demonstration projects are still unfolding.4,7 There have been a few studies that have considered the effects of the PCMH model on patient experience over time, but these studies have been conducted in specific populations or settings that limit generalizability. A recent systematic review found a small positive effect of the PCMH on patient satisfaction; however, this finding was mostly driven by studies in pediatric and geriatric populations. 8 Among those studies in general adult populations, 1 previous study found a positive effect9,10 and another found no effect.11 The first of these studies took place in an integrated delivery system that provided both organization and financing of healthcare,9,10 while the other took place in multiple communities around the country with coordination by a national organization.11

We sought to measure patients’ experiences over time in primary care practices for general adult populations, which transformed into PCMHs in a multi-payer community12 with extensive locally driven quality improvement efforts.



We conducted a prospective study, using 2 serial crosssectional samples of patients receiving primary care in the Hudson Valley region of New York. The Institutional Review Board of Weill Cornell Medical College approved the study. In a separate paper, we describe baseline patient experience compared with national benchmarks.13 In this paper, we measure whether patient experience changes over time, as practices gain experience with the PCMH.


The 7 counties of the Hudson Valley are located immediately north of New York City. Physicians in this region provide healthcare with fee-for-service reimbursement from multiple payers. The average practice size is 4 physicians.14

This study took place in the context of an initiative led by THINC,15 a non-profit organization that convened 6 health plans and 1 large employer to provide financial incentives for physicians to implement the PCMH model. Financial incentives ranged from $2.00 to $10.00 per member per month for achieving PCMH Level III, as defined by the National Committee on Quality Assurance’s (NCQA’s) 2008 criteria.16

PCMH transformation took place at 12 adult primary care practices and 1 pediatric primary care practice. The physicians in these practices are members of the Taconic Independent Practice Association (IPA).17 Practices were assisted in their transformation by the Taconic IPA, as well as by 2 external consulting groups. The lead physicians from each practice met at least monthly as a Medical Council to coordinate their efforts and share best practices. Practice transformation consisted of systematically reviewing the NCQA tool, documenting PCMH processes that were already in place, and targeting and implementing those processes that were not initially in place that were also of interest to the practice. Practices were permitted to vary in which aspects of the PCMH they implemented.

Practice-based needs assessments began in January 2009, and actual transformation began in March 2009. All practices submitted their applications to NCQA and were awarded Level III recognition (the highest of 3 levels). The median submission date was December 2009 (range August 2009-January 2010).

Sampling and Recruitment

We excluded the 1 pediatric practice that had undergone transformation, because the patient experience tool (described below) was not applicable to a pediatric population. We excluded 2 adult solo practices that delayed their medical home implementation, leaving 10 practices in total (20 sites).

Recruitment took place in the waiting rooms of the practices, because this was viewed by the Medical Council as the most  patient-centered strategy. (The alternative, providing the research team with patients’ contact information without patients’ explicit approval to release such information, was not viewed as patient-centered.) We prepared 1-page information sheets, which had English and Spanish versions on each side and invited patients to participate. We provided each practice with the same fixed number of information sheets, which practice staff then distributed to consecutive patients. Patients who opted to participate provided their own contact information and the name of their primary care doctor to confirm receipt of primary care at one of the participating practices. The information sheets described the study broadly as a patient experience survey and did not name the PCMH per se. Patients were offered a $5 incentive for participation.

Baseline patient experience data were collected from November 2009 to February 2010. Follow-up data were collected (in a separate sample of patients) from February 2011 to August 2011. The median duration between rounds of the survey was 15 months. We sampled patients at 2 time points, because we sought to include patients with recent visits to their primary care physicians, in order to avoid recall bias.

Measurement of Patient Experience

We measured patient experience, which is a measure of patient-centeredness that is broader than patient satisfaction and includes reports from patients on what they did or did not experience in their interactions with the healthcare system.18 We based our survey tool on the 2007 Clinician & Group - Consumer Assessment of Healthcare Providersand Systems (CG-CAHPS) Adult Primary Care Questionnaire.19 We included the 35 questions from CG-CAHPS and added 14 additional questions, in order to address concepts included in the PCMH model that were not explicitly covered in the CG-CAHPS. These 14 questions were derived from questions in the Ambulatory Care Experiences Survey (ACES),20 the American College of Physicians Center for Practice Innovation Clinician and Staff Survey (unpublished), the Commonwealth Fund Quality of Health Care Patient Survey,21 and the Commonwealth Fund International Health Policy Survey.22 The final survey contained 49 questions.

Survey Administration

Telephone surveys were administered by the Cornell Survey Research Institute (SRI). Cornell SRI attempted to contact each patient up to 5 times. If they did not reach a patient after 5 attempts, the patient was removed from the potential respondent pool. Outreach to patients was stratified by practice; data collection stopped for a given practice if the target number of completed surveys (N = 40) was reached.


We used descriptive statistics to characterize the patientsin the sample. We compared the baseline and follow-up samples, using χ2 or Fisher’s exact tests.

We applied analytical guidelines published by the national CG-CAHPS team23 to aggregate survey responses into 7  nonmutually-exclusive domains: access to care, communication and relationships, disease management, doctor  communication, follow-up of test results, office staff, and overall rating of the doctor. We calculated for each question the proportion of patients that gave the most favorable response, using the questions with 6-point scales. We then averaged this result across questions within each domain to yield the average proportion of patients who gave the most favorable response for that domain. We compared baseline and follow-up by domain using 2-sample tests of proportions. We also analyzed the data at the question level, similarly using 2-sample tests of proportions to measure change over time.

All data analyses were conducted with SAS version 9.3 (SAS Institute Inc, Cary, North Carolina), except for the 2-sample tests of proportions, which were conducted using Stata/IC 12.0 (Stata Corp LP, College Station, Texas). We considered P values <.05 to be significant.

PDF is available on the last page.
Adult ADHD Compendium
COPD Compendium
Dermatology Compendium
Diabetes Compendium
GI Compendium
Hematology Compendium
Immuno-oncology Compendium
Lipids Compendium
MACRA Compendium
Oncology Compendium
Pain Compendium
Reimbursement Compendium
Rheumatoid Arthritis Compendium
Know Your News
HF Compendium
Managed Care PODCAST