Receipt of Depression Treatment From General Medical Providers and Specialty Mental Health Providers

The important role that primary care providers play in the treatment of mental disorders has long been recognized (1,2). Primary care and general medical settings usually serve as the first point of contact for depression care and are often the only sources of care (3,4). Depression is also highly comorbid with general medical conditions, and available evidence suggests that untreated or poorly managed depression complicates care for individuals with a number of chronic health conditions, especially cardiovascular disease and diabetes (5,6). More recent studies have further highlighted the increasing role of the general medical sector in the provision of depression treatment (7). General practitioners are also the most common source of antidepressant medications, prescribing approximately 62% of all retail antidepressants (8). Understanding the extent to which depression is routinely treated in general medical settings is important because health care providers have varying degrees of expertise in treating mental health conditions, with implications for adequacy and quality of care (9), including continuation of antidepressant medication use (10) and future use of mental health services (11).

Few past studies have compared individuals who receive care from both primary care and specialty mental health providers with those who receive care from only one of these provider types. This gap in knowledge is especially notable because there is a growing move toward greater coordination and integration of behavioral health care and primary health care in the United States. Since the original 2002 U.S. Preventive Task Force recommendation for depression screening in primary care (12), there have been many such efforts, including the Depression Improvement Across Minnesota, Offering a New Direction (DIAMOND) initiative, which provides collaborative care for depression in primary care settings (13), and the increased role of mental health care in patient-centered medical home projects.

In light of the Affordable Care Act and the Mental Health Parity and Addiction Equity Act, further characterization of persons who receive treatment from both general medical providers (GMPs) and specialty mental health providers (SMHPs) is becoming especially important. One type of care coordination between GMPs and SMHPs, the collaborative care model, has been shown to improve treatment outcomes such as depressive symptoms and remission (14,15) and to reduce health care costs (16).

Despite these promising initiatives, barriers such as regional mental health workforce shortages and persistent stigma toward mental health treatment may impede efforts to facilitate integration (3,17). A better understanding of the clinical and sociodemographic characteristics of individuals who receive depression care from both the primary and specialty mental health systems may inform public health infrastructure and future plans for care integration across services.

Using recent nationally representative data from community-dwelling adults ages 18–64 with a major depressive episode in the past year, we aimed to examine sociodemographic characteristics and health conditions of adults who received treatment from both GMPs and SMHPs and compare them with their counterparts who received treatment from either a GMP or an SMHP only. Our hypothesis was that adults who received treatment from both GMPs and SMHPs would be more likely to be functionally impaired by the major depressive episode and to live in counties with a greater number of SMHPs compared with adults who received treatment from only one type of provider.

Methods

Data Sources and Sample

This cross-sectional study was conducted with the National Survey on Drug Use and Health (NSDUH) restricted-use data collected from the civilian, noninstitutionalized U.S. population ages 12 years and older from all 50 states plus the District of Columbia. To increase the precision of the estimates by obtaining a larger sample, we combined five years of NSDUH data (2008–2012) for this study. Overall, the weighted response rates for 2008 through 2012 ranged from 62.9% to 66.8%. Data were collected by using a computer-assisted questionnaire and audio computer-assisted self-interview at the respondent’s home. Description of the NSDUH is available on the Substance Abuse and Mental Health Services Administration (SAMHSA) Web site (18,19). The NSDUH data were merged with the Area Health Resources Files (AHRF) (20) data to obtain county-level information on the psychiatrist workforce. An effort was made to merge the AHRF and NSDUH files from the corresponding year; for 2009 and 2012, years in which AHRF data were not available, the AHRF from the prior year were used.

The study focused on a sample of 17,700 adults ages 18–64 who met the DSM-IV-TR criteria for a major depressive episode in the past year. Of these adults, 9,100 received treatment for depression from either GMPs or SMHPs in the past year. After excluding individuals with missing information about the provider from whom they received depression treatment (<1% of the sample) and those with missing data on study covariates, our final sample size was 8,900 adults ages 18–64 with a past-year major depressive episode.

Measures

The presence of a major depressive episode was assessed with nine questions about the symptoms of a major depressive episode according to the DSM-IV-TR criteria (21). Respondents met criteria for lifetime major depressive episode if they reported five or more symptoms nearly every day in the same two-week period during their worst or most recent period and at least one of the symptoms was depressed mood or “loss of interest or pleasure in daily activities.” The survey does not distinguish among major depressive episodes related to substance use, grief, or general medical comorbidity. Respondents who experienced these symptoms within the past 12 months were classified as those with a past-year major depressive episode.

Receipt of depression treatment was assessed by asking respondents if they had seen or talked to a medical doctor or other health care provider about their depression at any time in the past 12 months. Respondents were then asked a series of follow-up questions about the type of provider they had seen or talked to. On the basis of their answers, respondents were categorized into one of the following three groups: first, the GMP-only group, who received care only from GMPs, including general practitioners, family doctors, physicians other than psychiatrists, nurses, occupational therapists, or other health professionals; second, the SMHP-only group, who received care only from SMHPs, including psychiatrists, psychologists, counselors, social workers, or other mental health professionals, such as mental health nurses; and third, the GMP-SMHP group, who received care from at least one GMP and at least one SMHP.

The sociodemographic variables used for this study were age, gender, race-ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, or other), education (less than high school, high school completion, some college, or college graduate or higher), marital status (married, never married, or no longer married), employment status (full-time, part-time, unemployed, or other), family income as a percentage of federal poverty level defined by the U.S. Census Bureau (<100%, 100%−199%, ≥200%, or unknown for adults ages 18–22 living in a college dormitory), and health insurance status (private only, public [Medicaid, Medicare, or both], other [includes military insurance], or uninsured). Region was categorized as Northeast, Midwest, South, and West according to the U.S. Census region code, and metropolitan area was based on the U.S. Department of Agriculture rural-urban continuum codes and further categorized as large metropolitan area consisting of counties with a population of one million or more, small metropolitan area consisting of counties with less than one million population, and nonmetropolitan area consisting of urban and rural populations outside of the metropolitan area. Self-rated health, presence of past-year comorbid substance use disorder (meeting DSM-IV-TR criteria for substance abuse or dependence), the number of past-year comorbid general medical conditions, the number of past-year emergency room visits, presence of past-year suicidal ideation, presence of functional impairment caused by the major depressive episode per the Sheehan Disability Scale (22), and use of prescription medication for depression were also examined. The number of psychiatrists providing patient care within the county in which the respondent resided per 100,000 county residents was assessed through AHRF data.

Statistical Analysis

The sociodemographic characteristics of adult respondents were compared by provider type, and weighted population counts were estimated. Because the focus of the study was on factors that distinguish adults who received GMP-SMHP treatment from those who received care from only one type of provider, we limited comparisons to these three groups. These comparisons were accomplished with two logistic regression models. Both models adjusted for covariates to identify factors that were independently associated with receiving depression treatment from GMPs-SMHPs compared with the two sole-provider groups. Multicollinearity was assessed during multivariate modeling with the variance inflation factors (23). None of the selected independent variables assessed were collinear in the final multivariate models. Survey year was not included in the final models because no significant trend differences over time were noted. All analyses were conducted with Stata, version 13 (24), to account for NSDUH’s complex sample design and sampling weights. All percentages and estimates presented were weighted to the population. According to NSDUH guidelines for using restricted-use data, any description of the sample size must be rounded to the nearest 100 to minimize potential disclosure risk.

Results

Prevalence of Treatment Receipt for Depression, by Provider Type

Of the community-dwelling adults ages 18–64 who met the criteria for past-year major depressive episode (estimated 14.4 million), 38.5% (estimated 5.5 million) did not receive treatment from any provider for their depression in the past year. The remaining 61.5% comprised 21.0% (estimated 3.0 million) who received depression treatment from a GMP only, 18.5% (estimated 2.6 million) from an SMHP only, 19.2% (estimated 2.8 million) from both, and 2.9% (estimated .4 million) from alternative sources only, which included religious or spiritual advisors (such as ministers, priests, or rabbis) or another type of healer (such as herbalists, chiropractors, acupuncturists, or massage therapists).

Comparison of Groups by Sociodemographic Characteristics

Adults with a major depressive episode who received depression treatment from GMPs only, SMHPs only, or both types differed in regard to several sociodemographic characteristics, health conditions, and functional impairment caused by a major depressive episode (Table 1). For example, adults with higher education and those who reported having functional impairment were more likely to receive treatment from both types of provider rather than from either GMPs or SMHPs.

After we adjusted for other covariates, many of these differences remained. For example, adults with past-year suicidal ideation were almost twice as likely to receive depression treatment from the combination of practitioners (GMPs-SMHPs) as from GMPs only. Adults with a major depressive episode who resided in a county with a greater concentration of psychiatrists providing patient care than found in other counties were also more likely to receive depression treatment from both GMPs-SMHPs than from GMPs alone. Adults who resided in the South compared with those in the Northeast and those ages 35–64 compared with younger adults ages 18–25 were significantly less likely to receive depression care from both provider types versus from GMPs only.

In the adjusted model, women, married adults, residents in a nonmetropolitan urban or rural area compared with large metropolitan areas, those reporting fair or poor self-rated health compared with excellent or very good health, and those with more comorbid general medical conditions were more likely to received depression treatment from GMPs-SMHPs versus from SMHPs only. Compared with depressed adults who were non-Hispanic white, depressed adults who were non-Hispanic black or Hispanic were less likely to receive treatment from both types of providers than from SMHPs only.

Symptoms, Functional Impairment, and Prescription Medication Use

Compared with those who received care from SMHPs only or GMPs only, a larger percentage of participants who received care from GMPs-SMHPs met all nine criteria for major depressive episode (39.4% versus 33.2% for those seeing SMHPs only and 27.1% for those seeing GMPs only; p<.05 [Figure 1]). Compared with the SMHP-only and the GMP-only groups, the GMP-SMHP group included both a higher proportion of adults prescribed medication for their depression (88.1% versus 69.0% and 74.7%, respectively) and those with functional impairment (78.2% versus 72.3% and 63.4%, respectively; all comparisons p<.05 [Figure 2]). Over half of the adults from GMP-SMHP and GMP-only groups who received care had one or more general medical comorbid conditions (56.8% and 56.6%, respectively), a higher proportion than for adults who received care from an SMHP only (40.6%; both comparisons p<.001 [Figure 2]).

Discussion

Using recent data from nationally representative surveys of mental health status and service utilization, we found that more than three of every five community-dwelling adults ages 18–64 who met the diagnostic criteria for past-year major depressive episode reported receiving depression treatment from at least one health care provider. Approximately 21% of adults with a major depressive episode received depression care from GMPs only, 19% from SMHPs only, and 19% from at least one of both types of provider. Individuals who received depression treatment from both types of provider were distinct from those who received treatment from GMPs alone in having more suicidal ideation and a greater level of functional impairment caused by major depressive episode and for SMHPs alone in a larger number of general medical conditions and functional impairment. Even after controlling for these factors, we found that several sociodemographic and contextual factors, such as education and the number of psychiatrists at the county level who were seeing patients, remained significantly associated with using both types of provider compared with seeing GMPs alone.

Few past studies have compared individuals who receive care from both primary care and specialty mental health providers with those who receive care from one type of provider. Some of these studies have found that patients who received treatment from psychiatrists were more likely to have a diagnosis of mood or anxiety disorder and to have a greater number of depressive symptoms than those who sought treatment from primary care physicians (10,25). Other studies did not find differences in depression severity, psychiatric comorbidity, and depressive symptoms among individuals receiving care in primary and specialty care settings (26–28). Findings from this study are consistent with those of these other studies (10,25), and we found differences in depressive symptoms according to type of provider the patient saw.

Adults who received treatment from both types of provider versus a GMP were more likely to have had past-year suicidal ideation; however, one-fifth of adults who received depression care from a GMP alone also reported suicidal ideation. This finding highlights the importance of continuing education and training for suicide prevention and mental health care in a primary care setting. The finding also reinforces the importance of depression screening and treatment provided in general medical settings, especially in counties in which psychiatrists are not readily available. These results further suggest the need for greater efforts to increase accessibility of mental health services across the United States through the development of health system infrastructure (29) and collaborative care models (14,15). Variation in perceived quality and perceived effectiveness of treatment by provider type also underscores the importance of understanding the population that receives treatment from both types of provider (30) and the barriers to such care, including the inability for GMPs and SMHPs to bill for services on the same day (31). A recent study also suggested that individuals seeing GMPs-SMHPs found depression treatment more helpful than that provided by GMPs, similar to findings with adults who received treatment from SMHPs only (32).

The GMP-SMHP group in this study was also significantly different from the SMHP-only group in sociodemographic characteristics. Persons from racial-ethnic minority groups, specifically non-Hispanic blacks and Hispanics, were more likely to receive care from SMHPs alone than from both types of provider. Prior studies have found racial-ethnic disparities in access to care and in adequacy of depression treatment (33,34). Another study found an increase in disparity of mental health care in general medical settings over time (35). To our knowledge, this study is the first to identify racial-ethnic differences among individuals receiving care from both types of provider or one type alone. Because people from racial-ethnic minorities, particularly African Americans with mental illness, may have a higher rate of multiple chronic medical conditions (36), the differences in seeking treatment from GMPs-SMHPs compared with treatment-seeking patterns by whites may represent a significant disparity in access. In addition to addressing behavioral health care needs in a general medical setting, a parallel movement to incorporate primary care needs in mental health care settings may help meet the needs of persons from racial-ethnic minority groups, who may be more likely receive care from SMHPs alone.

Several limitations of this study should be noted. First, NSDUH is a cross-sectional survey that limits drawing causal inference from the data. Second, NSDUH does not sample homeless persons not living in a shelter and the institutionalized population who may need mental health care. Third, NSDUH does not assess prior treatment history, preferred treatment type, and perceived need for care specifically for depression, and the survey uses a self-report instrument to assess diagnoses and service use and is therefore prone to recall biases. Fourth, the study could not examine the treatment intensity for major depressive episode across different types of provider because of the limitations of NSDUH questionnaires. Furthermore, talking to or seeing health professionals about depression does not necessarily constitute treatment and may be subject to recall bias. Fifth, NSDUH does not assess whether participants who report receiving care from both types of provider are referred by a GMP to an SMHP, referred by an SMHP to a GMP, receive care from both providers concomitantly, or receive collaborative care. Finally, the survey does not distinguish a primary major depressive episode from one that is related to substance use, grief, or general medical illness. A major depressive episode in the context of bipolar disorder also was not distinguishable from a major depressive episode in the context of major depressive disorder.

Conclusions

The study used nationally representative data, and the results are generalizable to a majority of the population of U.S. adults ages 18–64 with a major depressive episode. In addition, the integration of workforce characteristics from the AHRF enabled the examination of how differences in availability of psychiatrists relate to receipt of depression treatment for adults who have had a major depressive episode. The study results suggest that a sizeable proportion of adults with depression receive depression treatment from GMPs, either alone or along with care from SMHPs. These results further reinforce the vital role that primary care providers play in both screening for and treating depression. Adults who received care from both types of provider had more complicated health care needs. However, even after controlling for these factors, we found that sociodemographic and contextual factors, such as higher level of education and access to psychiatrists, may influence whether individuals receive care from both GMPs and SMHPs rather than from one provider type and may inform barriers to accessing potential integrated care. Continued monitoring of service use across different settings will be important as integrated models of care become more prevalent.