The medical home functions of primary care practices that care for adults with psychological distress: a cross-sectional study

Adults with poor mental health or psychological distress (PD) have complex primary care needs. Primary care physicians are often the only mental health providers for adults with common mental health conditions, like anxiety and depression [1, 2]. For these patients, primary care physicians are responsible for diagnosing and managing both physical and mental health conditions. Adults with severe mental illnesses (SMI), like schizophrenia and bipolar disorder, are at significantly higher risk of having diabetes, cardiovascular diseases, and chronic respiratory diseases than the general adult population [3‐6]. Adults with PD experience psychosocial and economic barriers that make it difficult to follow treatment recommendations, including unstable housing, unpredictable schedules, lack of transportation, limited social supports, and inability to afford copayments [7‐9]. Adults with PD also have complex care coordination needs; they may receive various services from community mental health clinics and substance abuse treatment programs, which have long been siloed from medical care providers [10]. Primary care providers struggle to sufficiently care for patients with multiple medical and psychiatric comorbidities in setting of socioeconomic challenges and care coordination needs under the constraints of the traditional fifteen minute office visit [11, 12].

Furthermore, adults with PD experience numerous barriers to accessing primary care and receiving patient-centered primary care, even when they have regular primary care physicians. Adults with PD have reported experiencing anxiety and paranoia while waiting for long periods of time in crowded waiting rooms or alone in examination rooms [11, 13] and have been found to be less likely to receive a same-day response when they contact their primary care practice for a medical issue [14]. They have reported difficulty developing trusting relationships with primary care providers and feeling like their providers do not understand their concerns [15]. Nearly 15% of adults with SMI who have primary care physicians report that they still use the emergency department as their main source of routine care [16]. These barriers to primary care utilization may contribute to the high rate of potentially preventable hospitalizations in this population [17].

Recent studies have shown that new models of primary care may be more effective for persons with PD. These new models of care rely on flexible access to primary care services, a primary care team that includes case managers, and technologies that continue the patient-provider relationship outside of the office visit [18, 19]. Adults with depression and poorly-controlled chronic illness have been found to have greater improvement in glycemic control, cholesterol levels, systolic blood pressure, and depression scores when they receive primary care from practices where nurse case managers meet with patients regularly and monitor mood symptoms, medical diseases, self-care activities, and medical adherence [20, 21]. Veterans with posttraumatic stress disorder have been found to have higher rates of primary care utilization and lower rates of specialty care utilization and hospitalization when they receive primary care from practices that reduce barriers to care by offering same day appointments, telephone appointments, and electronic messaging as well as track quality metrics, provide care through multidisciplinary teams, and assign care coordinators to high-risk patients [22]. Adults with chronic medical conditions and comorbid mental health conditions who receive care from primary care practices that are certified as patient-centered medical homes (PCMHs) have been found to lower expected mean counts of emergency department (ED) visits and psychiatric hospitalizations [23].

PCMHs or medical homes are sites of comprehensive primary care that optimize patient-centeredness, access, timeliness, care coordination, treatment planning, patient and caregiver engagement, and population health [24, 25]. Medical homes offer same-day and non-business hours appointments to reduce barriers to accessing primary care. They use care coordinators, registries, and electronic decision support to ensure that all patients in the practice receive appropriate screenings and preventative services. Medical homes also use team-based models of care delivery in order to maximize time that providers are able to spend with patients [18, 19]. The medical home model has been supported by a number of medical professional societies and is being encouraged by the federal government under the alternative payment model provisions of the Patient Protection and Affordable Care Act (PPACA) and the Medicare Access and CHIP Reauthorization Act (MACRA) [26, 27]. This model is specifically encouraged for adults with SMI or common mental illness and comorbid chronic medical conditions under Medicaid Health Home programs [28]. Primary care practices that have many of these medical home functions have generally been found to have higher rates of patient satisfaction, staff satisfaction, and preventative care utilization and lower rates of ED utilization and potentially preventable hospitalizations [29‐32].

The objective of this study was to examine whether adults with PD are as likely to receive primary care from practices that have implemented the medical home functions as adults without PD. Previous work has shown variation in adoption of these functions across primary care practices [33]. Patients who are uninsured and live in the South have been found to be less likely to receive care from practices with medical home functions [34]. We were interested in determining if persons with psychological distress are less likely to receive primary care from practices with the medical home functions than persons without psychological distress, because previous literature has found that adults with SMI are less likely to receive high quality medical care [35]. Persons with psychological distress are also more likely to be low-income and publicly insured than persons without psychological distress. Low income patients have previously been found to receive care from physicians with lower quality ratings and to report greater ease accessing care at the emergency room than from their primary care physicians [36, 37].

Table 1 presents the population characteristics of our linked data set. We found that while 6.22% (n = 267) of participants in our study had a K6 score of 13 or greater indicating PD, only 4.6% of participants in the MEPS household survey met this criterion.

As compared to household survey participants without PD, a significantly greater percentage of participants with PD had medical conditions or risk factors that require close primary care management, including hypertension (62% vs. 46%), diabetes (29% vs. 15%), coronary heart disease (13% vs. 8%), stroke (16% vs. 5%), myocardial infarction (13% vs. 5%), asthma (26% vs. 12%), and fair or poor perceived health status (87% vs. 49%). In contrast, a significantly lower percentage of participants with psychological distress had enabling factors for primary care engagement and compliance with follow-up recommendations, including marriage (71% vs. 78%), a high school diploma (88% vs. 95%), and income above 400% of the federal poverty line (10% vs. 42%).

Table 2 presents the different characteristics between primary care practices that care for adults with and without PD. A significantly greater percentage of participants with PD received their primary care from nonprofit or hospital-owned practices (46% vs. 34%), while a significantly smaller percentage received their care from physician-owned primary care practices (45% vs. 59%). A significantly greater percentage of participants with PD received care from practices that utilize nurse practitioners and physician assistants (79% vs. 72%). There were no statistical differences between the percentage of participants with and without PD that receive their primary care from practices that have multiple locations.

Although we found that participants with PD receive care from different types of practices than participants without PD, we did not find much difference between the two groups as related to the medical home functionality of their site of primary care. We found no significant between-group differences in the implementation of electronic health records (92% vs. 91%), secure messaging with patients (75% vs. 81%), same-day appointment availability (94% vs. 94%), follow-up appointments within forty-eight hours of hospitalization (75% vs. 77%), preventative care reminders (92% vs. 89%), or case managers (50% vs. 52%). We also found no differences in multi-specialty practice (45% vs. 40%) and PCMH certification (47% vs. 48%). However, we did find that a significantly greater percentage of participants with PD receive their primary care from practices where physicians routinely receive quality reports (94% vs. 89%, p = 0.003) and that utilize clinical decision support technologies through their electronic health record systems (93% vs 89%, p = 0.021).

Table 3 shows the results of our multivariable linear regression using the composite measure of eight medical home functions. Our regression model found that there is no significant difference between participants with PD and participants without PD in their likelihood to be seen by a practice that has implemented medical home functions (Coefficient = − 0.206, p = 0.141). This model did show that hospital-owned or non-profit practices (Coefficient = − 0.486, p < 0.001), practices with 2–10 or 11 prescribers (Coefficients = 0.721, 1.298, p < 0.001 for both outcomes), and practices with multiple locations (Coefficient = 0.272, p = 0.001) are more likely to have implemented the medical home functions. We also found that practices in southern states were significantly less likely to implement medical home functions than practices in northeastern states (Coefficient = − 0.260, p = 0.018).

Table 4 displays the final results of the multivariate logistic regressions for each of the ten medical home functions that we included in our analysis. After controlling for patient-level predisposing, enabling, and need factors, as well as provider-level characteristics, we found the participants with PD had no greater or less likelihood of receiving primary care from practices that have implemented any of the medical home functions as compared with participants without PD.

We are encouraged by the high levels of adoption of medial home functions by primary care practices participating in the 2015 MEPS MOS component. Nearly 90% of practices reported offering same day appointments, electronic health records with clinical decision support, and individualized physician quality reports. Nearly half of practices reported using care coordinators and having PCMH certification. We did also not find any disparities as they relate to the type of patients that receive care from practices with the medical home functions; we did not find any differences by age, gender, race, insurance type, socioeconomic status, and English language proficiency.

However, our results showed regional variation in the adoption of medical home functions. We found the South was significantly less likely to adopt medical home functions than the North, and there was no difference between the West, the Midwest, and the Northeast. This finding aligns with state-level legislation promoting medical home functionality in primary care practices. Patient-Centered Primary Care Collaborative initiatives indicated that all states in the Northeast have a law promoting medical home functionality, while very few states in the South have state-level law [46]. The results of this failure to adopt medical home functions in the South may be contributing to delays in care as persons with PD in the South have been found to be more likely to delay care than persons with PD living in other regions. This is especially notable because the highest proportion of white and black persons with PD live in the South [47]. Results of our study suggest the importance of regulation to promote the adoption of medical home functions. Future research should further investigate the effectiveness of these laws in terms of promoting health care access and quality.

We are also encouraged by the finding that adults with PD receive their primary care at practices that have implemented medical homes functions to the same degree as adults without PD. Due to their complex medical and psychosocial comorbidities, adults with mental health issues have significant potential to benefit from primary care that is delivered through a medical home. Furthermore, we found that a greater proportion of adults with PD are cared for in practices that have structures that align with the medical home model, including the use of advanced practitioners, which fosters a team-based approach, and participation in risk-bearing payment contracts, which increases provider accountability.

We also found that adults with PD are somewhat more likely to receive care in practices with two of the medical home functions: use of EHR clinical reminders and individual quality reports for physicians. However, these differences disappeared when controlling for patient-level and practice-level characteristics. These differences in medical home adoption are likely due to differences in the characteristics of medical practices that provide care to adults with PD; adults with PD are more likely to receive care from larger practices and non-profit or hospital-based practices and these practices are also more likely to have implemented the medical home functions. One could posit that practices that are owned by physicians would be less likely to implement clinical practice reminders in the EHR and personalized physician quality reports as these functions could be interpreted as threats to physician autonomy [48]. Smaller practices may also lack the infrastructure and health information technologies to provide personalized quality reports [49, 50].

Persons with mental illness have historically received lower quality medical care. We are very encouraged that although persons with PD receive care in practices with different organizational characteristics than persons without PD, they have not been left behind in this movement towards medical homes. The MOS survey did not assesses whether the practices surveyed were Federally Qualified Health Centers (FQHCs) or contracted as Medicaid Health Homes, but we wonder if policies promoting the adoption of medical home functions in these settings has ensured equal dissemination of functions across practice types [28, 51]. As of 2015, the year of the MOS, 20 states had established Medicaid Health Home programs [52]. Additional research is needed to look at state-level variation in order to determine if there are differences in states with and without Medicaid Health Home programs.

Our study is among the first to explore adoption of medical home practices among persons with and without PD. While it is encouraging that there is no difference in adoption of the medical home practices, future research should investigate if adoption of these practices is associated with improvements in access, satisfaction, utilization, and health outcomes for persons with PD.

The way that primary care delivered is undergoing major changes and we are glad to see that adults with mental health issues are not being left behind. Medical home functions provide a structure for reducing cognitive overload and exhaustion, strengthening the patient-clinician relationship, promoting patient-provider communication about health, and facilitating patient involvement in their own care [63]. Ongoing policy initiatives, such as the Accountable Care Organization, the Comprehensive Primary Care model, and value-based financing model, are also needed and should be implemented to truly transform their practice into a medical home. Due to the complex needs of adults with mental health issues, the delivery of primary care through medical home models has significant potential to improve the health outcomes of this population. We expect that adoption of these medical home functions as well as behavioral health integration will continue to grow among primary care practices that care for this patient population.