INTRODUCTION

Unmet need for mental health treatment remains, despite significant attention to the issue. National surveys indicate that one-third of individuals with serious mental illness did not receive treatment in the past year.1 Common reasons for this unmet need include cost, lack of perceived need for treatment, and stigma.2 However, even among privately insured with unmet need, 67% experience structural barriers to receiving treatment including issues related to cost, lack of available in-network providers, and not knowing where to go for treatment.3,4

Broad consensus supports the use of primary care in conjunction with access to specialty mental health providers to fill this gap.5,6,7 There is a strong evidence base for the collaborative care model which includes care management, supervision and consultation with a mental health specialist, and a patient symptom registry.8 Advantages of mental health treatment through a primary care provider (PCP) include patient convenience, leveraging an established and trusted provider relationship, and a holistic approach that acknowledges that medical conditions improve with treatment of co-morbid mental health conditions. Furthermore, patients that avoid mental health treatment because of stigma may be more willing to accept care from their PCP than a specialist.5,9

Nevertheless, concerns remain about the quality of mental health care delivered by primary care providers. In particular, PCP uptake of evidence-based models of behavioral health integration has been slow.10 For example, less than half of primary care providers collocate with a specialty mental health provider.11 Rates of depression screening remain low, and under one-third of Accountable Care Organizations (ACOs) report using registries to support treatment of mental health in primary care.12,13 Some of these shortfalls are due to barriers PCPs face including incomplete information flow between behavioral and non-behavioral health clinicians, billing difficulties and poor reimbursement for some components of the collaborative care model, inadequate training, and that primary care visits by necessity are of short duration.5,14,15 However, less attention has been paid to the patient perspective in primary care–provided mental health care, especially for those already motivated to seek care.

We aimed to better understand patient satisfaction, preferences, and experiences in primary care–provided mental health care and whether primary care filled the gap when individuals were unable to access specialty care. First, we conducted a national internet survey of privately insured patients reporting serious psychological distress and a need for mental health services (N = 428) to identify whether patients that attempted but were unable to access specialty mental health services obtained care through their primary care provider. If they did obtain care, we determined the types of care received (e.g., medication, counseling) and whether they rated their experiences with this primary care provider positively. Next, we conducted interviews (N = 30) to explore patient-reported barriers and facilitators to engagement and satisfaction with mental health care provided by their PCP.

METHODS

We used a sequential explanatory mixed methods study design to examine access to outpatient mental health services (Appendix Fig. 1).16 We chose this design because our research aim was to (1) quantify the extent to which patients needing services are able to receive them, and in what setting, and (2) characterize the nature of those patient experiences.17 We integrated the strands of data in order to generate a rich description of patient perspectives regarding PCP-provided mental health services. Quantitative survey and qualitative interview data were integrated at several points: through “connecting” in the sampling phase, “expansion” in the interpretation phase, and “contiguous” narrative at the reporting phase.16 Participant consent was obtained prior to both the survey and interviews, and the study was approved by the NYU and Yale Institutional Review Boards.

Survey

Survey methods were previously described in detail.18 In brief, we conducted a national survey in 2018 using KnowledgePanel, a validated online panel of ~ 50,000 households.19 The novel survey was tested through cognitive interviewing to ensure understandability of question wording.20 Survey questions are available in the Appendix Fig. 2. Serious psychological distress was defined by a score of 13 or higher on the Kessler 6 scale.21

Our sample consisted of two separate groups, privately insured English-speaking adults ages 18–64 with serious psychological distress reporting that they (1) used an outpatient mental health provider in the last year or (2) attempted to use a mental health provider but ultimately did not use specialty mental health services (Appendix Fig. 1). Mental health providers were defined as “professionals specifically trained to diagnose and treat emotional or mental health problems, including psychiatrists, therapists, psychologists, mental health nurse practitioners, and social workers.” Participants who did not use specialty mental health services were asked, “In the last 12 months, did you try to make an appointment with a mental health provider? When we say “try,” we mean you called or contacted at least one mental health provider or clinic, contacted your insurer, or looked on your insurer website for a provider.” Those responding yes were considered to have attempted but not accessed specialty services.

Both groups were asked if they received mental health care from a primary care provider. Those that responded affirmatively were then asked what types of services they received from their PCP (i.e., medication, counseling, care management); patients were allowed to choose more than one care type. Respondents then rated the mental health care provided by their PCP on a 1 to 10 scale. For analysis, we dichotomized ratings using the “top box” method (i.e., rated 9–10 versus 1–8).22,23 Demographic information and presence of a usual source of care was previously obtained by KnowledgePanel.

Weights were applied to match respondents to the US population based on demographic characteristics and account for panel recruitment, attrition, oversampling, and survey non-response. Analyses were completed using Stata version 16.

In-Depth Interviews

A purposeful random sampling approach was used to select interview participants from survey participants who agreed to be contacted for telephone interviews.24 A stratified sample was created to ensure equal representation in four groups (used both specialty and primary care for mental health, only specialty care, only primary care, and no receipt of care). KnowledgeNetworks randomly selected participants meeting criteria and arranged times for telephone interviews. Given the nature of our research question, we estimated we would reach saturation with a sample of 30, based on guidance.25 One participant was excluded after audio was disconnected early in the interview and unreachable on callback. A small monetary incentive ($10) was provided through KnowledgeNetworks.

One researcher (KK), a primary care physician and health services researcher, conducted interviews using a semi-structured interview guide exploring patient experiences in obtaining mental health treatment through their PCP, the role of the PCP in mental health treatment, and coordination of care with specialists (Appendix Fig. 3). Interviews lasted approximately 10–20 minutes, and were recorded, professionally transcribed, and reviewed for accuracy. Data were analyzed using the constant comparison method.26 Three coders independently reviewed and coded all transcripts using Dedoose software version 8.3.35. Coders met regularly to review the code structure and resolve discrepancies. Once initial coding was complete, the final code structure was reapplied to all transcripts by one member of the coding team (RS or KN) then reviewed by the third member (KK), and consensus was reached for remaining discrepancies. The code structure was reviewed regularly with senior methodologists (LC and DK). We then used coded data to identify overarching recurrent themes.

RESULTS

Survey Analyses

Survey completion rate was 66% (Appendix Fig. 1) based on the American Association for Public Opinion Research standard definition for probability-based internet panels.27 Of the 2131 qualifying survey respondents, 428 met the current study inclusion criteria. Participants were predominately young (52% ages 18–34), female (63%), and non-Hispanic White (66%) (Table 1).

Table 1 Respondent Characteristics
Full size table

Use of Specialty and PCP Mental Health Treatment

Although 78% of respondents accessed specialty mental health treatment, 22% reported that they tried to access specialty mental health care but did not access services (Table 1). Those attempting but not accessing specialty care were significantly more likely to be younger (65% vs. 49%, p = 0.05) and be without a usual source of care (12% vs. 5%, p = 0.05). Among these, 53% reported receiving mental health care from a PCP, indicating that 11% of those trying to access services did not receive either specialty or primary care–based mental health services.

Ratings and Mental Health Treatment Received from PCP

Of participants receiving mental health care from their PCP (N = 190), 41% highly rated their care (Table 2). Those only receiving care through their PCP were less likely to rate their care highly as compared to those also receiving specialty treatment (21% versus 48%, p = 0.01). Most participants in primary care–only treatment received just medication (73%), whereas among those receiving treatment from both a primary care and specialty provider, most received both medication and counseling (73%). Access to care management services was low overall (9%) and did not significantly differ between those receiving specialty care and those only in PCP-provided treatment.

Table 2 Among Those Receiving Mental Health Care from a Primary Care Provider, Rating of Care and Treatment Received
Full size table

Interview Analyses

Among participants in the study sample, 63% (N = 282/428) of survey respondents agreed to be contacted for a telephone interview. Those agreeing to be contacted were older and more likely to be college educated but did not differ by whether they accessed specialty care or used their PCP for mental health treatment (Appendix Table 1). Interviewees were mostly female, non-Hispanic White, and college educated (Appendix Table 2).

Participants described three primary factors that shaped their engagement and satisfaction with their PCP-provided mental health care: PCP engagement in mental health care, patient relationship and trust with the PCP, and views on PCP responsibilities and role in mental health care (Table 3).

Table 3 Categories of Themes and Subthemes Identified in Interviews: Barriers and Facilitators to Patient Engagement and Satisfaction with Mental Health Care from Primary Care Provider
Full size table

PCP Engagement

Patients appreciated when their PCP was proactive in screening and initiating mental health conversations and believed these practices supported a holistic view of their health. Conversely, patients expressed dissatisfaction when they felt the burden was on them to initiate mental health discussions and when there was not a clear follow-up plan on subsequent visits for their mental health issues. Furthermore, some patients felt their PCPs minimized their mental health complaints as not valid or concerning (Table 3, a).

Relationship and Trust with PCP

Continuity and trust with an established PCP facilitated engagement in PCP-directed mental health care. Some patients found it difficult to address mental health with a new PCP and “retell their story.” Others reported holding back from engaging in mental health discussions over worry that their PCP would judge them (Table 3, b).

Responsibilities and Role of PCP

Patient views varied on what role their PCP should have in their mental health treatment. Many saw their PCP as a trusted source for specialist recommendations and wanted their PCP to remain in communication with their specialist and act as a care coordinator. Others preferred their care to be siloed and viewed their PCP as only addressing medical problems. The largest contrast in views was around psychiatric medication prescribing. Some patients viewed their PCP as an easy source of prescriptions, which they often directed—they might specifically request a medication previously prescribed by a specialist and direct the PCP when they felt the dose needed adjustment. Others preferred only counseling but felt that their PCP did not have the time or training to provide it and could do little for their mental health other than prescribe. Some described PCPs as not having adequate expertise to treat their problems; they viewed mental health treatment as a specialized skill set; some described negative experiences with PCP-directed medication management (Table 3, c).

DISCUSSION

We found that among our sample of privately insured adults reporting serious psychological distress who attempted to access specialty mental care, 1 in 5 did not ultimately see a specialty mental health provider. However, of these, just over half did obtain mental health treatment from their PCP. Our findings support the use of primary care as a venue for mental health services when patients face structural barriers to specialty care such as cost, limited availability of specialists, or not knowing where to go for treatment.3 For those patients desiring medication therapy, many expressed satisfaction with their PCP, especially as a convenient inexpensive way to continue medication previously recommended by a specialist. In this way, increased PCP involvement can reduce the burden on psychiatrists and other non-physician specialists, improving access to mental health treatment.

However, rating of primary care–provided mental health care was lower among those using their PCP alone versus those also in care with a mental health specialty provider. This suggests patients who attempted but could not obtain specialty care were less satisfied with the quality of their PCP-provided mental health care. Qualitative interviews provide insight into why some did not rate primary care mental health care highly.

First, patients were dissatisfied when they felt the burden was on them to initiate mental health discussions, when a PCP minimized their mental health concerns or even discouraged discussion of mental health care, or if there was not regular follow-up on their mental health plan. The collaborative care model has been shown to be an effective and cost-effective model for treatment of both depression and anxiety in the primary care setting.28,29 Several components of this model—screening, care management with regular consultation with a mental health specialist, and patient registries—facilitate regular symptom assessment and a treat to target approach that can provide systems-based solutions to poor PCP engagement and follow-up.

Second, some patients were reluctant to discuss mental health because of stigma and lack of a trusting established relationship with their PCP. This underscores the need for screening and proactive discussions around mental health, especially for patients in a new PCP relationship. Ongoing education of PCPs to reinforce the importance of mental health care in the scope of primary care practice as well as clinical training can increase confidence and reduce clinical inertia.

Third, patients’ perception of their PCP’s role in mental health care influenced their willingness to engage in PCP-provided mental health care. Patients cited lack of availability of counseling through their PCP as a barrier to care. This perception was supported by survey data; of those only engaged in PCP care, 73% received medication only. For many patients, medication may not be desired nor clinically indicated. Yet PCPs are not routinely trained in brief therapy techniques (behavioral activation; motivational interviewing; problem-solving treatment) nor are they able to provide within the time constraints of a complicated visit when there are competing medical demands.15 In the absence of an integrated behavioral health team member, outside referral is needed if patients are to receive counseling. The significant expansion of telehealth during the COVID-19 pandemic may present a convenient and effective way to improve access to counseling services, especially for those in rural areas with provider shortages.30

Others felt that their PCP did not have adequate training to manage psychiatric medication or preferred their PCP to only focus on their medical care. While this may be patient perception, some PCPs may indeed lack sufficient clinical training.5,9 Increased PCP engagement and education can address some of these concerns. In other cases, pairing primary care providers with specialty consults may be critical to improving medication prescribing and patient experiences. For some patients, especially those with serious mental illness, a referral and open communication with the specialist may be the most critical role of the PCP.

This study has several limitations. First, survey results are subject to non-response bias. Those that chose to respond may be different than non-respondents, even after weighting for demographic characteristics. Second, self-report and recall bias are also inherent in survey data, though we used cognitive interviewing to pretest the survey instrument and a short recall period (12 months) to minimize this limitation. Third, for interview data, the majority of participants were female, non-Hispanic White, and highly educated. Findings from this study may not transfer to non-English-speaking or non-privately insured populations, where it may be even more difficult to access to specialty mental health care. Fourth, those that agreed to be contacted for interviews differed on demographic characteristics, and those who completed interviews may have had exceptional experiences in their mental health care.

It is notable that our population specifically looked at those seeking specialty care, and doesn’t represent the population of patients who did not attempt to access specialty care. Instead, our results are instructive in better understanding how primary care can best serve those unable to access specialty mental health services, either due to workforce shortages, cost, or other issues. Note that some respondents using specialty care may have accessed this care through their PCP. The study population included patients reporting serious psychological distress, while many treated in primary care have mild to moderate symptoms. Results related to use and satisfaction of PCP-provided mental health care may differ for those with less severe symptoms.

Our research also has a number of strengths. Participants were from a national sample and their experiences represent real-world care delivered in primary care practices rather than controlled well-resourced programs found in efficacy studies of integrated care models. We address calls to focus on patients’ needs and preferences to improve mental health care in primary care settings.7 Our study examines those most in need of care—those reporting serious psychological distress. We used rigorous qualitative methods including use of a multidisciplinary team, audio-taping and independent transcription, and standardized coding and analysis.24,25.

In this study of privately insured patients reporting serious psychological distress that actively sought specialty mental health care, many of the known barriers to mental health care—insurance coverage and lack of perceived need for treatment—were addressed, yet almost one-quarter still did not engage in specialty care. About half of those not accessing specialty care received some mental health treatment from their primary care provider, suggesting primary care is successfully filling some of this gap. While ratings of mental health care provided by PCPs were lower among those not engaged in specialty care, interviewees expressed support for their PCP as a referral source, care coordinator, and prescriber. Patient-reported barriers to engagement and satisfaction in PCP-provided mental health care can be addressed by supporting implementation of proven components of the collaborative care model through technical assistance and flexible payment structures that cover upfront costs as well as adequate reimbursement for care.