|Home | About | Journals | Submit | Contact Us | Français|
To determine whether Medicaid-enrolled depressed adults receive adequate treatment for depression and to identify the characteristics of those receiving inadequate treatment.
Claims data from a Medicaid-enrolled population in a large mid-Atlantic state between July 2006 and January 2008.
We examined rates and predictors of minimally adequate psychotherapy and pharmacotherapy among adults with a new depression treatment episode during the study period (N=1,098).
Many depressed adults received either minimally adequate psychotherapy or pharmacotherapy. Black individuals and individuals who began their depression treatment episode with an inpatient psychiatric stay for depression were markedly less likely to receive minimally adequate psychotherapy and more likely to receive inadequate treatment.
Racial minorities and individuals discharged from inpatient treatment for depression are at risk for receiving inadequate depression treatment.
Depression treatment guidelines suggest that persons diagnosed with major depressive disorder be treated with a 4- to 8-week trial of an antidepressant medication, a 4- to 8-week trial of psychotherapy, or a combination of the two modalities (Agency for Health Care Policy and Research 1993; American Psychiatric Association 2000;). Recent research in primary care settings (Goldman, Nielsen, and Champion 1999; Unutzer et al. 2003;) and in community surveys (Wang, Berglund, and Kessler 2000; Young et al. 2001;) indicates that depressed patients do not, on average, receive treatment concordant with guideline recommendations.
Individuals enrolled in Medicaid are disproportionately poor, racially diverse, and have a greater prevalence of depression than other populations (Thomas et al. 2005). In addition, Medicaid-enrolled adults face significant barriers to depression care and have less access to mental health treatment (Melfi, Croghan, and Hanna 1999; Melfi et al. 2000;) than privately insured individuals. Medicaid enrollees also have more severe depression and greater depression-related medical costs than other depressed populations (Frank, Goldman, and Hogan 2003). Ensuring high-quality depression treatment for this vulnerable population is essential, yet little data exist on the extent to which Medicaid-enrolled adults receive depression care that reflects treatment guidelines.
In this study, we examine the quality of psychotherapy and pharmacotherapy received by Medicaid-enrolled adults undergoing treatment for depression in outpatient mental health specialty care settings. As quality of care across a variety of health conditions has been found to be worse for Medicaid-enrolled individuals than commercially insured individuals (Landon et al. 2007), we hypothesized that rates of receipt of guideline concordant depression treatment in this population would be lower than those reported in non-Medicaid populations (Wang et al. 2000). In addition, because significant disparities in the quality of mental health treatment have been found elsewhere (Rost et al. 2002; Miranda and Cooper 2004; Hinton et al. 2006;), we also hypothesized that men, racial minorities, and those living in rural areas would be less likely than their comparison groups to receive guideline concordant depression treatment.
Using specialty mental health care administrative data provided by a large nonprofit managed behavioral health organization and Medicaid pharmacy claims data provided by a large Mid-Atlantic state, we identified 1,098 adults aged 18–64 who initiated a new episode of depression treatment between October 28, 2006 and October 31, 2007. This study was conducted in compliance with the University of Pittsburgh IRB.
Individuals initiated a depression treatment episode when they (1) received two or more outpatient behavioral health services on different days in a 12-week period with a primary diagnosis of major depression (ICD-9 codes 296.2–296.36) or (2) were discharged from a psychiatric hospitalization with a diagnosis of major depression. The first outpatient service or the day of discharge was considered the index visit. Depression treatment episodes ended 144 days after the index visit or with any event (such as an inpatient stay) that would have prevented the individual from receiving outpatient care. Depression treatment was categorized as a new episode if an individual had not received any behavioral health services or antidepressant medication in the 4 months before the first service with a diagnosis of major depression (Figure 1). If individuals had multiple new treatment episodes for depression, we used the first one observed. To allow enough time for the clean period and the treatment period, we used data from July 1, 2006 to January 31, 2008 for depression treatment episodes where the index visit occurred between October 28, 2006 and October 31, 2007.
We excluded individuals who had received a service with a primary diagnosis of schizophrenia, schizoaffective disorder, or bipolar I disorder in the year before the index visit or at any point during the depression treatment episode. We excluded dually eligible individuals who had Supplemental Security Income with Medicare because Medicaid claims would not reflect all of the treatment they received. We also excluded individuals who had fewer than 90 days of Medicaid eligibility in the 144 days following the index visit because these individuals would not be eligible to receive mental health services during the majority of the study timeframe.
Sociodemographic variables, including age, gender, race, and Medicaid eligibility category, were obtained from the state's membership and eligibility files. Race/ethnicity was categorized as Caucasian, African American, or other. Consistent with other analyses of Medicaid-enrolled individuals (Zito et al. 2005), we categorized individuals into Medicaid eligibility categories according to whether they were Medicaid eligible as a result of general assistance, medical or mental health disability (e.g., Supplemental Security Income [SSI]), or income (e.g., Temporary Assistance to Needy Families [TANF]). Individuals were categorized as living in an urban area if their county of residence had a population density of greater than 1,000 individuals/square mile.
Variables measuring prior inpatient psychiatric admissions and use of behavioral services for substance abuse were developed using behavioral health claims data. Individuals were categorized as having a prior inpatient psychiatric admission if they had an inpatient psychiatric admission in the 12 months before the “clean period” (i.e., 3–15 months before the start of the depression treatment episode). Individuals were categorized as having prior substance abuse treatment if they had received any substance abuse treatment services in the 12 months before the start of the depression treatment episode. In addition, we categorized individuals based on the type of service use that qualified them for a current depression treatment episode: inpatient, outpatient with the same provider, and outpatient with different providers for the two episode-initiating visits.
Consistent with other studies of depression quality of care, we defined minimally adequate psychotherapy as four or more individual, group, or family psychotherapy visits during the first 84 days (12 weeks) of a depression treatment episode (Bao, Sturm, and Croghan 2003; Schoenbaum et al. 2002; Teh et al. 2008;).
Minimally adequate pharmacotherapy was defined as having a filled prescription for an antidepressant for 84 of the 144 days following the index visit, a modification of the HEDIS acute antidepressant treatment, which measures appropriate pharmacotherapy as a filled prescription for an antidepressant medication on 84 out of the first 114 days of treatment (HEDIS 2008). Because depression treatment episodes could begin with either an outpatient visit or discharge from an inpatient stay, individuals initiating treatment for depression may not initially see a prescribing physician (i.e., the index visit may be with a clinical social worker). To account for any lag time between having a visit with a nonphysician mental health worker, being referred to a prescribing physician (e.g., psychiatrist or primary care physician), and filling a prescription, we extended the timeframe by 30 days (from 114 days, as per the HEDIS measure, to 144 days).
Inadequate treatment was defined as having neither adequate psychotherapy nor adequate pharmacotherapy during the treatment episode.
To determine patterns of behavioral health care and antidepressant medication use during a new episode of depression treatment, we conducted frequency counts of dependent variables for the adult Medicaid population and examined differences between groups of particular interest using t-tests and χ2-tests as appropriate. We used multivariate logistic regression models to determine which factors were associated with receiving (1) minimally adequate psychotherapy, (2) minimally adequate pharmacotherapy, and (3) inadequate treatment. For each outcome, the final model included all independent variables, such that assessment of the effect of each individual covariate controlled for the effects of all other measured covariates. Results were considered significant at the p<.05 level. Since we were concerned that the odds ratios resulting from this analysis may exaggerate a risk association or treatment effect, we report instead the approximated risk ratios (Zhang and Yu 1998). All analyses were performed using the Statistical Analysis System (SAS), version 9.1 (SAS Institute, Cary, NC).
We identified 1,098 Medicaid-enrolled adults who initiated new depression treatment episodes between July 1, 2006 and January 31, 2008 (Table 1). The mean age was 36.7 years (SD=11.9) and nearly 70 percent were women. Almost half were nonwhite. Approximately 80 percent (n=634) were Medicaid-eligible through general assistance and state expansion programs and as TANF beneficiaries (n=634; 80 percent), while 20 percent (n=464) were eligible as SSI beneficiaries. Very few had either had an inpatient psychiatric admission (3 percent) or had received treatment for substance abuse (8 percent). The majority of depression episodes began with two outpatient mental health visits with the same provider (78 percent), though 16 percent began with an inpatient psychiatric stay.
Sixty-three percent of depressed Medicaid-enrolled adults initiating a depression treatment episode had four or more psychotherapy visits in the first 12 weeks of the episode. In our multivariate analyses, men and individuals living in rural counties were more likely to receive minimally adequate psychotherapy, while younger adults (ages 18–29, corrected relative risk [RR]=0.87), African Americans (corrected RR=0.88), and those whose depression treatment episode began with an inpatient stay (corrected RR=0.44) were significantly less likely than their comparison groups to receive minimally adequate psychotherapy during the depression treatment episode (Table 2).
Thirty percent of depressed Medicaid-enrolled adults initiating a depression treatment episode had a filled prescription for an antidepressant medication for at least 84 days (12 weeks) of the 144 days following the index visit. Rates were even lower among individuals discharged from a psychiatric hospitalization, of whom 65 percent initiated pharmacotherapy, but only 26 percent received minimally adequate pharmacotherapy. Multivariate analyses revealed that older adults (ages 45–64) and those whose depression episode began with two outpatient visits with different providers were more likely to receive minimally adequate pharmacotherapy. African Americans were significantly less likely to receive minimally adequate pharmacotherapy than whites or individuals of other races (corrected RR=0.079) (Table 2).
Thirty percent of individuals received inadequate depression treatment, defined as receiving neither adequate psychotherapy nor adequate pharmacotherapy. African Americans received inadequate treatment more frequently (43 percent) than whites (25 percent) and those of other races (27 percent); in multivariate analyses, they were more likely than whites to receive inadequate treatment (corrected RR=1.14).
Almost two-thirds (65 percent) of individuals whose depression treatment episode began with an inpatient stay (N=172) received inadequate treatment; this group was more likely to receive inadequate treatment than those whose episodes began with outpatient visits (corrected RR=1.15).
We found that approximately 70 percent of Medicaid-enrolled adults initiating treatment for major depression in specialty mental health care settings received minimally adequate care. Rates of minimally adequate psychotherapy treatment appear to be reasonably high—almost two-thirds of depressed Medicaid-enrolled adults received four or more psychotherapy visits during the first 12 weeks of treatment. However, rates of minimally adequate pharmacotherapy are substantially lower with only 30 percent of individuals receiving adequate care. Our findings that 70 percent of individuals in a high risk, low SES Medicaid population are receiving minimally adequate depression care are consistent with other studies reporting that 60 percent of all individuals being treated for mood disorders in specialty mental health settings are receiving minimally adequate care (Wang et al. 2007), and they are encouraging given more recent reports about modest rates of quality care across a range of disorders and settings (McGlynn et al. 2003).
We found substantial racial disparities in the rates of minimally adequate depression care. African Americans consistently received lower rates of minimally adequate treatment than other individuals, even after controlling for other numerous factors that could explain these differences. This finding is consistent with other studies documenting racial/ethnic disparities in mental health treatment quality (DHHS 1999; Wang et al. 2007;). These disparities may be due to differences in access to care (IOM 2002), patient preferences, including the use of individuals outside the health care system to address mental health issues (Young, Griffith, and Williams 2003), stigma (Corrigan 2004), or provider bias or discrimination (Atdjian and Vega 2005). Policy-level interventions to improve the quality of mental health care overall and to improve access to care for racial minorities (McGuire and Miranda 2008) and patient-level educational interventions to reduce stigma in minority populations may help to reduce these disparities in the quality of depression treatment.
We also found low rates of minimally adequate care for individuals initiating depression treatment with a psychiatric hospitalization, a concern since such individuals likely have more severe depression than those initiating treatment in outpatient settings. Hospitalized individuals were markedly less likely to receive minimally adequate psychotherapy or pharmacotherapy after discharge; only a third of individuals discharged from a hospital received any adequate treatment for depression. Low rates of clinical engagement following discharge from psychiatric hospitalization have been documented previously (Stein et al. 2007), but given appropriate follow-up care's association with lower rates of readmission (Nelson, Maruish, and Axler 2000), attention to improving the adequacy of outpatient depression treatment for this population is essential.
We found that individuals whose episode-initiating outpatient visits were with different providers were slightly more likely to receive adequate pharmacotherapy than those whose visits were with the same provider. Although this finding seems to conflict with research showing that continuity of care is an essential element of quality care (Teh et al. 2008), we suspect that in this case, many individuals in this group had a first visit with a nonphysician mental health provider (e.g., clinical social worker) and were then appropriately referred to a psychiatrist for pharmacotherapy. We were unable to test this hypothesis, however, since our data did not allow identification of individual provider types.
Women are more likely to seek treatment for depression than men (Galbaud du Fort et al. 1999). We found, however, that depressed women initiating treatment for major depression received lower rates of minimally adequate psychotherapy than men. Medicaid-enrolled women may face additional barriers to treatment than men, such as lack of adequate child care that could differentially reduce their ability to attend routine psychotherapy visits. Better understanding these barriers and implementing interventions to increase access to mental health care for Medicaid-enrolled women must be a goal of future public policy. Surprisingly, we also found that individuals living in rural counties received higher rates of minimally adequate psychotherapy than those in urban counties, despite the limited availability, accessibility, and acceptability of mental health care services common in rural areas (Sawyer, Gale, and Lambert 2006). Further research exploring the causes and generalizability of this finding is needed.
Our findings are based on mental health specialty and pharmacy claims data from a Medicaid-enrolled population in a large mid-Atlantic state. A limitation of this analysis is lack of generalizability: we do not know to what extent our findings generalize to populations from other regions, or to populations in areas with different Medicaid mental health benefits. Because of the racial composition of our study state and limitations of the way race was recorded in the data, we were not able to report on differences in depression care quality by race and ethnicity beyond a simple comparison between white and black. We could not observe depression treatment services provided outside of the behavioral health system, such as pastoral counseling or psychotherapy services provided by a physical health provider. While such services may benefit individuals with depression, we note, however, that they are not commonly considered in assessing the adequacy of psychotherapy provided (Schoenbaum et al. 2002; Teh et al. 2008;). We are unable to assess the content or quality of psychotherapy visits, and given that most mental health visits do not contain evidence-based treatment, it is likely that the number of individuals receiving effective psychotherapy is low (Tai-Seale et al. 2007). Our measure of minimally adequate acute-phase psychotherapy also provides for a very low level of treatment intensity; however, even such minimal depression treatment has been associated with improved depression outcomes (Schoenbaum et al. 2002).
Better understanding of the factors associated with receipt of quality care for depression among Medicaid-enrolled individuals who initiate depression treatment will allow for more targeted efforts to improve treatment and outcomes for this population. Findings in the current study suggest that targeting efforts to increase the adequacy of antidepressant prescribing and adherence as a good place to start. Moreover, given that racial minorities and individuals discharged from inpatient treatment are at highest risk for receiving inadequate care for depression, specific interventions might target these individuals as well. Quality improvement efforts can build upon established models of depression care management (e.g., care coordination, communication, patient education) that have documented success in improving outcomes such as rates of appropriate use of antidepressant medication for individuals with depression (Williams et al. 2007). System-level interventions might include efforts to increase payment rates for psychotherapy and medication management and to provide clinicians with information about their performance on key outcomes versus their peers. Medicaid enrollees face significant barriers to receiving quality behavioral health care. It is essential that systems and policy makers use available information about factors that are related to quality of care and respond in a manner that ensures high-quality depression treatment for this vulnerable population.
Joint Acknowledgment/Disclosure Statement: The authors would like to gratefully acknowledge support for this work from the National Institute of Mental Health (T32 MH19986) and the Community Care/ University of Pittsburgh Department of Psychiatry Community Academic Partnership Award. We would like to thank Community Care Behavioral Health Organization's Performance Management Committee for feedback on prior versions of the manuscript. Mrs. Sorbero and Mihalyo and Drs. Kogan, Schuster, and Stein are affiliated with Community Care Behavioral Health.
Disclosures: An early version of the paper was presented at the National Association of State Mental Health Program Directors Research Institute Annual Conference on February 11, 2008 and at the NIMH Mental Health Services Research Conference on July 23, 2007.
This paper was made possible by funding from Community Care Behavioral Health Organization and the University of Pittsburgh Department of Psychiatry Community Academic Partnership award (Carrie Farmer Teh, PI) and from the National Institute of Mental Health grant no. T32 MH19986 (Charles F. Reynolds, III, PI). We are grateful to Community Care Behavioral Health Organization's Performance Management Committee for useful feedback and suggestions on an earlier version of the manuscript. An abstract of this report was presented in poster format at the NASMHPD Annual State Mental Health Services Research Conference in February 2008.
Additional supporting information may be found in the online version of this article:
Appendix SA1: Author Matrix.
Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.