JOURNAL OF DUAL DIAGNOSIS, 11(2), 128–135, 2015 C Taylor & Francis Group, LLC Copyright ISSN: 1550-4263 print / 1550-4271 online DOI: 10.1080/15504263.2015.1025024
PSYCHOPHARMACOLOGY & NEUROBIOLOGY
Psychiatric Services and Prescription Fills Among Veterans With Serious Mental Illness in Methadone Maintenance Treatment Carla Marienfeld, MD,1 and Robert A. Rosenheck, MD1,2
Objective: Comorbidity and co-prescription patterns of people with serious mental illness in methadone maintenance may complicate their treatment and have not been studied. The goal of this study was to examine the care and characteristics of people with serious mental illness in methadone maintenance treatment nationally in the Veterans Health Administration (VHA). Methods: Using national VHA data from FY2012, bivariate and multiple logistic regression analyses were used to compare veterans in methadone maintenance treatment who had a serious mental illness (schizophrenia, bipolar disorder, or major affective disorder) to patients in methadone maintenance treatment without serious mental illness and patients with serious mental illness who were not in methadone maintenance treatment. Results: Only a small fraction of patients with serious mental illness were receiving methadone maintenance treatment (0.65%), but a relatively large proportion in methadone maintenance treatment had a serious mental illness (33.2%). Compared to patients without serious mental illness, patients with serious mental illness in methadone maintenance treatment were more likely to have been homeless, to have had a recent psychiatric hospitalization, to be over 50% disabled, and to have had more fills for more classes of psychotropic drugs. Compared to other patients with serious mental illness, patients with serious mental illness in methadone maintenance treatment were more likely to have a drug abuse diagnosis and to reside in large urban areas. Conclusions: One-third of patients in methadone maintenance treatment have serious mental illness and more frequent psychiatric comorbidity, and they are more likely to use psychiatric and general health services and fill more types of psychiatric prescriptions. Further study and clinical awareness of potential drug-drug interactions in this high medication and service using population are needed. (Journal of Dual Diagnosis, 11:128–135, 2015)
Keywords methadone, severe mental illness, Veterans’ Health Administration
Methadone maintenance treatment has been shown to reduce the harms of illicit opioid use including hepatitis C and HIV viral transmission, improve health outcomes including psychiatric comorbidities, reduce criminal activity and possibly incarceration, and improve the likelihood of employment (Anglin, Speckart, Booth, & Ryan, 1989; Dole et al., 1969; Evans et al., 2012; Fullerton et al., 2013; Hser, Hoffman, Grella, & Anglin, 2001; Macarthur et al., 2012; Mattick, Breen, Kimber, & Davoli, 2009; Van Den Berg et al., 2007). Medical and psychiatric comorbidity are high in patients in methadone maintenance treatment (Brooner, King, Kidorf, Schmidt, & Bigelow, 1997; Ryan & White, 1996),
1Yale
University School of Medicine, New Haven, Connecticut, USA New England Mental Illness, Research, Education and Clinical Center, West Haven, Connecticut, USA Address correspondence to Carla Marienfeld, MD, Yale University School of Medicine, 300 George St., Suite 901, New Haven, CT 06511, USA. E-mail: [email protected] 2VA
and patients who receive integrated psychiatric and medical care with their addiction treatment can have better outcomes with these more cost-effective services (Weisner, Mertens, Parthasarathy, Moore, & Lu, 2001). Patients with dual diagnosis of serious mental illness and substance use disorders often have greater difficulty attending appointments and adhering to prescription medications (Bogenschutz & Siegfreid, 1998; Centorrino et al., 2001; Olfson et al., 2000; Teter et al., 2011). Integrated treatment of psychiatric and substance use disorders improves health service use and adherence to treatment, thus benefiting this population (Brunette & Mueser, 2006). A 2005 review found that major predictors of poor adherence to medications included the presence of psychological problems, particularly depression, cognitive impairment, lack of insight into the illness, and lack of belief in the benefits of treatment, all common problems in patients with serious mental illness and especially those with a dual diagnosis of serious mental illness and a substance use disorder (Osterberg & Blaschke, 2005). However, empirical
Veterans With SMI in Methadone Maintenance
research has less often addressed adherence in actually filling prescriptions within a specific time frame, a dimension that is especially amenable to evaluation in large clinical populations using administrative data, and offers a further opportunity to characterize the distinctive and potential vulnerabilities of patients with serious mental illness in methadone maintenance treatment. In this study, we use national administrative data from the Veterans’ Health Administration (VHA) to identify the proportion of veterans in methadone maintenance treatment with diagnosed serious mental illness and to describe their psychiatric and medical comorbidities as well as their use of diverse health services and psychiatric medications, as compared to those in methadone maintenance treatment who do not have serious mental illness. We also make these comparisons between patients with serious mental illness in methadone maintenance treatment and those with serious mental illness who do not receive methadone maintenance treatment. Through this broad descriptive exploration we hope to identify potential clinical problem areas deserving of further consideration in this population.
METHODS Sample and Data Source National VHA administrative records were used to identify all veterans who were enrolled in methadone maintenance treatment and/or who had a diagnosed serious mental illness in fiscal year (FY) 2012 in the VHA nationally. Data were de-identified by removing all personally identifiable information prior to analysis. Serious mental illness was defined, for these purposes, as schizophrenia and other psychotic disorders (International Classification of Diseases, 9th edition [ICD-9] codes 295.xx), bipolar spectrum disorders (ICD-9 codes 296.0x, 296.1x, 296.40–296.89), and major depressive disorders (ICD-9 codes 296.2–296.39), each of which requires extensive non–addiction-related treatment. The primary study population included veterans who were both in methadone maintenance treatment and had one or more diagnosed serious mental illnesses and was compared to veterans in methadone maintenance treatment without a serious mental illness diagnosis and to the population of veterans with one or more serious mental illnesses not participating in methadone maintenance treatment. The study was approved by the institutional review board of the VA Connecticut Healthcare System.
Measures Data were obtained on patient characteristics such as age, gender, income, marital status, receipt of Veterans’ Affairs
129
(VA) disability compensation or pension, and a designation of residence based on rural-urban commuting area (RUCA) codes (depts.washington.edu/uwruca). Information on ethnicity was not informative because of extensive missing data. Recent homelessness was identified by participation in a VHA specialized homeless service program or a V-60 code indicating housing problems. Clinical data included psychiatric and medical diagnoses (based on the ICD-9 coding system used by the VHA in FY2012) that were assigned to each patient at least once during the study year. The Charlson Index (Charlson, Pompei, Ales, & MacKenzie, 1987), an aggregate measure of medical comorbidity, has been shown to predict the 10-year mortality for patients based on the presence of a broad range of comorbid medical conditions and was used to assess medical comorbidity. Data on comorbid psychiatric diagnoses included all ICD-9 codes 290 through 319 (coded into 11 classes; available on request). VHA inpatient and outpatient service utilization was documented. Outpatient mental health specialty care and substance abuse clinic visits, and methadone maintenance treatment clinic visits in particular, were identified by standard VHA clinic codes. Inpatient psychiatric care or substance use hospitalizations were identified through bed section codes in computerized discharge summaries. Pharmacy benefit records documenting all VHA prescriptions filled were used to identify the total number of prescriptions filled by these patients in five psychiatric medication classes: antidepressants; antipsychotics; sedative, hypnotics, and anxiolytics; mood stabilizers (anticonvulsants); and lithium. The total number of psychotropic prescriptions filled by each veteran during the year was determined by summing the numbers of prescriptions in each separate class as well as the total number of classes from which a prescription was filled. Medications in these classes that were not filled in a VHA pharmacy were not included in the analysis.
Analysis Given the sample size, almost all comparisons revealed differences that were significant by conventional standards using chi-square tests for categorical data and t-tests for continuous data. Given the large sample size, effect sizes (i.e., risk ratios for dichotomous measures and Cohen’s d for continuous measures) were used rather than p values to identify meaningful differences. Risk ratios around greater than 1.5 or less than 0.67 were considered substantial. The difference between means divided by the pooled standard deviation was used to calculate Cohen’s d for continuous variables. A stepwise logistic regression model (odds ratios are reported) included all categorical variables with the risk ratios above 1.5 or less than 0.5 and all continuous variables with a Cohen’s d absolute effect size of 0.3 or greater (small to moderate effect size). The first logistic regression compared patients in methadone main-
2015, Volume 11, Number 2
130
C. Marienfeld and R. A. Rosenheck
TABLE 1 Veterans With Serious Mental Illness (SMI) and/or in Methadone Maintenance Treatment (MMT): Categorical Data1,2 MMT + SMI (n = 3319) Demographics Male Homeless Black White Hispanic Service in Iraq or Afghanistan
MMT Only (n = 6682)
SMI Only (n = 507403)
Risk Ratio MMT + SMI/MMT
Risk Ratio MMT + SMI/SMI
n 3100 1519 1037 2127 403 388
% 93.4% 45.8% 31.2% 64.1% 12.14% 11.69%
n 6468 2006 2314 4132 880 874
% 96.80% 30.02% 34.62% 61.84% 13.18% 13.08%
n 442827 58725 106258 378200 72364 52611
% 87.27% 11.57% 20.94% 74.54% 14.26% 10.37%
0.96 1.52 0.90 1.04 0.92 0.89
1.07 3.95 1.49 0.86 0.85 1.13
674 424 1871 165 12 1
20.31% 12.77% 56.37% 4.97% 0.36% 0.03%
1359 615 3719 554 51 13
20.34% 9.20% 55.66% 8.29% 0.76% 0.19%
74782 65860 227621 66344 23834 10345
14.74% 12.98% 44.86% 13.08% 4.70% 2.04%
1.00 1.39 1.01 0.60 0.47 0.15
1.38 0.98 1.26 0.38 0.08 0.01
1017 1087 440 775 400
30.64% 32.75% 13.26% 23.35% 12.05%
2455 2062 851 1314 810
36.74% 30.86% 12.74% 19.66% 12.12%
113172 120684 81781 191766 25465
22.30% 23.78% 16.12% 37.79% 5.02%
0.83 1.06 1.04 1.19 0.99
1.37 1.38 0.82 0.62 2.40
VHA service connection status > 50% < 50%
908 506
27.36% 15.25%
1203 1078
18.00% 16.13%
202124 83774
39.84% 16.51%
1.52 0.95
0.69 0.92
Rural-urban commuting area3 Large urban Urban Small rural Isolated small rural
153 2990 96 80
4.62% 90.08% 2.88% 2.42%
278 6047 218 139
4.16% 90.49% 3.26% 2.08%
58671 374722 42429 31581
11.56% 73.85% 8.36% 6.22%
1.11 1.00 0.88 1.16
0.40 1.22 0.34 0.39
Psychiatric diagnoses Dementia Schizophrenia Other psychotic disorders Bipolar disorder Depression PTSD Alcohol use disorder Other drug use disorder Dysthymia Anxiety disorder Adjustment disorder Personality disorder Any substance use disorder Any psychiatric disorder (not substance use) Any dual diagnosis
11 513 486 912 2017 1314 1772 3137 1753 1133 372 410 3229 3319 3319
0.33% 15.46% 14.64% 27.48% 60.77% 39.59% 53.39% 94.52% 52.82% 34.14% 11.21% 12.35% 97.29% 100.00% 100.00%
7 0 0 0 0 1739 2638 6360 2657 1464 578 295 6570 6653 6657
0.10% 0.00% 0.00% 0.00% 0.00% 26.03% 39.48% 95.18% 39.76% 21.91% 8.65% 4.41% 98.32% 99.57% 99.63%
7999 88669 52605 113191 307357 176490 94597 84396 197822 134197 37147 28410 133060 507403 507403
1.58% 17.48% 10.37% 22.31% 60.57% 34.78% 18.64% 16.63% 38.99% 26.45% 7.32% 5.60% 26.22% 100.00% 100.00%
3.16 n/a n/a n/a n/a 1.52 1.35 0.99 1.33 1.56 1.30 2.80 0.99 1.00 1.00
0.21 0.88 1.41 1.23 1.00 1.14 2.86 5.68 1.35 1.29 1.53 2.21 3.71 1.00 1.00
Medical diagnoses Nausea/vomiting Liver disease HIV All pain diagnoses
151 607 80 2309
4.55% 18.29% 2.41% 69.57%
229 1228 143 3927
3.43% 18.38% 2.14% 58.77%
12951 22751 4338 300308
2.55% 4.48% 0.85% 59.19%
1.33 1.00 1.13 1.18
1.78 4.08 2.82 1.18
Service use4 Substance use outpatient Psychiatric outpatient
3284 2808
98.95% 84.60%
6654 4392
99.58% 65.73%
45631 453306
8.99% 89.34%
0.99 1.29
Age < 40 40–49 50–64 65–74 75–85 > 85 Income < $7,000 $7–15,000 $15–25,000 > $25,000 VHA pension
Journal of Dual Diagnosis
11.00 0.95 (Continued on next page)
131
Veterans With SMI in Methadone Maintenance
TABLE 1 (Continued) Veterans With Serious Mental Illness (SMI) and/or in Methadone Maintenance Treatment (MMT): Categorical Data1,2 MMT + SMI (n = 3319) Any mental health outpatient visit Substance use inpatient or outpatient Any mental health service inpatient or outpatient Mental health inpatient admission
3314 3303 3316 1035
99.85% 99.52% 99.91% 31.18%
MMT Only (n = 6682) 6668 6674 6669 702
99.79% 99.88% 99.81% 10.51%
SMI Only (n = 507403) 456177 137572 456705 44506
89.90% 27.11% 90.01% 8.77%
Risk Ratio MMT + SMI/MMT
Risk Ratio MMT + SMI/SMI
1.00 1.00 1.00 2.97
1.11 3.67 1.11 3.56
Note. PTSD = posttraumatic stress disorder; HIV = human immunodeficiency virus. Given the large sample size, effect sizes (i.e., risk ratios) were used rather than p values to identify meaningful differences. Risk ratios greater than 1.5 or less than .67 were considered meaningful here. 1Patients in FY2012 from Veterans’ Health Administration (VHA) sites that have a methadone program and had diagnosed schizophrenia or other psychosis, bipolar disorder, or major depressive disorder. 2Chi-squared analysis (two degrees of freedom) was significant with p < .0001 for all categorical variables except VHA service connection less than 50%. 3Rural-urban commuting area (RUCA) codes (depts.washington.edu/uwruca). 4Among those with any service use in FY2012.
tenance treatment with and without serious mental illness, and the second compared patients with serious mental illness in or not in methadone maintenance treatment. For nested variables (e.g., substance abuse visits within the category of all outpatient visits), the broader variable encompassing the smaller ones was used to avoid large amounts of noninformative overlap. For example, since substance abuse outpatient visits are a subset of all outpatient visits, which itself showed substantial differences across groups, we used only the total number of outpatient visits in our models. A c-statistic was calculated to assess the goodness-of-fit of the logistic regression model in predicting the value of the dichotomous dependent variable.
RESULTS Demographics and Baseline Rates In the VHA nationally in FY2012, there were 510,722 veterans with diagnosed serious mental illness and 10,001 veterans who participated in methadone maintenance treatment. One-third (33.2%, n = 3319) of patients in methadone maintenance treatment received a serious mental illness diagnosis, but this represents only 0.65% of veterans with serious mental illness. The patients were, on average, in their mid-50s (M = 55.4, SD = 13.5). Patients in methadone maintenance treatment with serious mental illness were 1.5 times more likely to be or have been homeless as compared to the general methadone maintenance treatment population and almost 4 times more likely to be homeless than the general serious mental illness population (Table 1). Of patients in methadone maintenance treatment with serious mental illness, the most common comorbid psychiatric diagnoses were major depression (60.8%), other depression (52.8%), posttraumatic stress disorder (39.6%), anxiety disorder (34.1%), and bipolar disorder (27.5%). Schizophrenia and other psychotic disorders together totaled 30.1%. Approximately 12% of those with se-
rious mental illness in methadone maintenance treatment also had a personality disorder (Table 1).
Patients With Serious Mental Illness on Methadone Maintenance Compared to Those Without Serious Mental Illness On bivariate analysis, patients in methadone maintenance treatment with serious mental illness were more likely to have been homeless than other patients in methadone maintenance treatment and to have VHA service–connected disability benefits rated greater than 50%, and they were less likely to be older than 65. In addition to having schizophrenia, bipolar, or major depressive disorders, they were more likely to have other psychiatric diagnoses of dementia, posttraumatic stress disorder, anxiety disorder, and personality disorder. They were also more likely to have had a mental health inpatient admission (Table 1). There was a small effect size (0.3–0.4) in the difference between groups for the number of general psychiatric outpatient visits, the number of emergency room visits, and the number of antipsychotic, anxiolytic, and anticonvulsant prescriptions. There was a moderate effect size (0.5–0.7) between groups for the number of substance abuse, psychiatric, or any outpatient visits as well as for the number of any psychotropic medication or antidepressant prescriptions. There was a large effect size (0.8 or greater) for the number of classes of psychiatric medications filled (Table 2). Logistic regression analysis showed that veterans in methadone maintenance treatment with serious mental illness were independently more likely to have been homeless, to have a VA service connected disability greater than 50%, to have a diagnosis of personality disorder, to have a mental health inpatient admission, and to have prescription fills in a larger number of psychiatric medication classes than veterans in methadone maintenance treatment without an serious mental illness diagnosis (Table 3). 2015, Volume 11, Number 2
132
C. Marienfeld and R. A. Rosenheck
TABLE 2 Veterans With a Serious Mental Illness (SMI) and/or in Methadone Maintenance Treatment (MMT): Continuous Data1 MMT + SMI (n = 3319) Demographics Age (in years) Income Medical comorbidity Charlson Index Score Service use2 Substance use outpatient clinic visits General psychiatric outpatient visits Psychiatric or substance use outpatient visits Methadone visits Primary care visits Medical or surgical outpatient visits Number of emergency room visits Number of days hospitalized3 Any outpatient visits Prescriptions filled4 Any psychotropic prescriptions Antidepressant prescriptions Antipsychotic prescriptions Anxiolytic/sedative/hypnotic prescriptions Stimulant prescriptions Anticonvulsant prescriptions Lithium prescriptions Opiate prescriptions Number of classes of psychiatric meds filled
MMT Only (n = 6682)
SMI Only (n = 507403)
MMT + SMI vs. MMT
MMT + SMI vs. SMI
Effect size −0.09 0.01
Effect size −0.34 −0.22
Mean 50.90 15406.10
SD 11.99 17956.93
Mean 52.14 14944.62
SD 12.86 24459.53
Mean 55.55 23278.47
SD 13.88 36134.26
Pooled SD 13.86 35919.93
2.54
2.24
2.58
2.38
2.89
2.41
2.41
−0.02
−0.14
61.18 22.24
77.70 35.23
51.78 10.16
72.06 21.17
2.06 12.01
12.02 26.11
15.73 26.12
0.60 0.46
3.76 0.39
83.42
88.01
61.90
77.72
14.07
30.68
32.42
0.66
2.14
35.31 4.68 12.79
65.06 5.25 13.89
35.34 3.72 10.11
63.83 4.30 12.30
0.00 3.78 10.71
0.00 3.95 11.99
8.93 3.96 12.01
0.00 0.24 0.22
3.95 0.23 0.17
2.72 16.36 96.21
4.47 21.92 90.53
1.59 9.50 72.04
3.09 10.50 80.10
0.98 17.65 24.78
2.37 74.49 34.20
2.40 73.80 35.81
0.47 0.09 0.67
0.73 −0.02 1.99
32.22 13.49 8.31 4.06
55.03 21.82 25.91 8.95
10.88 5.23 1.35 1.77
21.36 11.18 6.49 5.24
20.79 8.06 4.94 3.61
38.29 15.29 17.87 6.93
38.25 15.29 17.83 6.92
0.56 0.54 0.39 0.33
0.30 0.36 0.19 0.07
0.16 5.28 0.73 11.69 2.20
1.58 16.19 4.77 12.10 1.18
0.14 1.83 0.03 11.26 1.14
1.46 7.09 0.62 10.71 1.10
0.17 3.05 0.04 7.79 1.94
1.62 11.68 4.17 8.02 1.13
1.62 11.66 4.14 8.09 1.13
0.01 0.30 0.17 0.05 0.94
−0.01 0.19 0.17 0.48 0.23
Note. Given the large sample size, effect sizes (i.e., Cohen’s d) were used rather than p values to identify meaningful differences. Cohen’s d effect sizes greater than .30 were considered meaningful here. 1Patients in FY2012 from Veterans’ Health Administration sites that have a methadone program and had diagnosed schizophrenia or other psychosis, bipolar disorder, or major depressive disorder. 2Among those with any service use in FY2012. 3Among those with a hospitalization. 4Among those with prescriptions filled in FY2012, mean reflects the average number of prescriptions filled in the year. TABLE 3 Logistic Regression Analysis of Patients in Methadone Maintenance Treatment (MMT) With Serious Mental Illness (SMI) (n = 3227) Compared to Those Without SMI (n = 6531) Variable
Estimate
Standardized Estimate
Odds Ratio
95% Confidence Interval
Age Homeless VHA service connection greater than 50% Personality disorder Mental health inpatient admission Any outpatient visit Antidepressant prescriptions Antipsychotic prescriptions Anxiolytic/sedative/hypnotic prescriptions Anticonvulsant prescriptions Number of classes of psychiatric meds filled
0.00997 0.2903 0.3262 0.2696 0.4106 0.00119 0.00873 0.0384 −0.00927 −1.0105 0.6003
0.693 0.0766 0.731 0.0376 0.0849 0.0558 0.0759 0.3385 −0.0344 −0.0642 0.406
1.01 1.337 1.386 1.309 1.508 1.001 1.009 1.039 0.991 0.99 1.823
[1.006–1.014] [1.204–1.484] [1.235–1.555] [1.084–1.582] [1.315–1.728] [1.001–1.002] [1.005–1.013] [1.031–1.047] [0.983–0.999] [0.984–0.995] [1.729–1.922]
Note. VHA = Veterans’ Health Administration. The goodness-of-fit (C-Statistic) for this model = 0.777.
Journal of Dual Diagnosis
133
Veterans With SMI in Methadone Maintenance
Patients With Serious Mental Illness in Methadone Maintenance Treatment Compared to Those Not in Methadone Maintenance Treatment
disorder, an adjustment disorder, a personality disorder, and filled prescriptions in fewer classes of psychiatric medications (Table 4).
On bivariate analysis, veterans with serious mental illness who were in methadone maintenance treatment were more likely to have been homeless and to receive a VA pension than veterans with serious mental illness who did not receive methadone maintenance treatment. Those who received methadone maintenance treatment were more likely to have a drug, alcohol, or any substance use disorder as well as to have an adjustment disorder or a personality disorder. They were also more likely to have diagnosed liver disease and HIV. They used more outpatient substance use services and were more likely to have had a mental health inpatient admission (Table 1). There was a small effect size (0.3–0.4) in the difference between patients with serious mental illness in methadone maintenance treatment and other patients with serious mental illness for being younger, the number of general psychiatric outpatient visits, and the number of any psychiatric prescriptions filled, antidepressant prescriptions filled, and opiate prescriptions filled. There was a moderate effect size (0.5–0.7) between groups for the number of emergency room visits. There was a large effect size (0.8 or greater) for the substance use and psychiatric outpatient visits and for any outpatient visits (Table 2). Logistic regression analysis showed that patients with serious mental illness receiving methadone maintenance treatment were independently more likely to receive a VA pension, to live in a large urban area, to have a diagnosis of drug use disorder or liver disease, or to have had a mental health inpatient stay. Veterans with serious mental illness veterans in methadone maintenance treatment, as compared to those not in methadone maintenance treatment, had less likelihood of an alcohol use
DISCUSSION This is the first national study of which we are aware to explore the distinctive characteristics, clinical diagnoses, and service use and psychiatric prescription fills among veterans with serious mental illness in methadone maintenance treatment. While only a small fraction (0.65%) of patients with serious mental illness received methadone maintenance treatment, a relatively large proportion (33.2%) of patients receiving methadone maintenance treatment also had a diagnosed serious mental illness. Significant predictors of having a serious mental illness while in methadone maintenance treatment included social factors (homelessness, VA service connection benefit status), psychiatric comorbidity (personality disorder), service use (having a mental health inpatient admission), and greater psychiatric medication usage (the total number of classes of psychiatric medications filled). However, factors associated with being in methadone maintenance treatment among patients with diagnosed serious mental illness were somewhat different, the most prominent of which—not surprisingly—was having a drug use diagnosis. The psychiatric comorbid diagnoses (alcohol use disorder, adjustment disorder, personality disorder) and the number of classes of psychiatric medications filled were negatively associated with receiving methadone maintenance treatment. A medical comorbidity, liver disease, was associated with receiving methadone maintenance treatment, although this is not surprising given the risk of hepatitis C in intravenous drug users. Interestingly, having a mental health
TABLE 4 Logistic Regression Analysis of Patients With Serious Mental Illness (SMI) in Methadone Maintenance Treatment (MMT) (n = 3227) Compared to Those Not in MMT (n = 486080) Variable
Estimate
Standardized Estimate
Odds Ratio
95% Confidence Interval
Age Homeless VHA pension RUCA large urban Alcohol substance use disorder Any drug use disorder Anxiety disorder Adjustment disorder Personality disorder Liver disease Mental health inpatient admission Any outpatient visit Antipsychotic prescriptions Number of classes of psychiatric meds filled
−0.0122 0.0874 0.3807 0.7568 −0.3472 4.1425 0.0946 −0.2477 −0.1874 0.6434 0.1899 0.00928 −0.00385 −0.1165
−0.0928 0.0157 0.0461 0.1831 −0.0753 0.8669 0.023 −0.0356 −0.024 0.0741 0.0299 0.1822 −0.0384 −0.0724
0.988 1.091 1.463 2.132 0.707 62.958 1.099 0.781 0.829 1.903 1.209 1.009 0.996 0.89
[0.985–0.991] [1.008–1.182] [1.304–1.642] [1.893–2.400] [0.653–0.765] [53.636–73.9] [1.016–1.189] [0.694–0.878] [0.740–0.929] [1.726–2.098] [1.106–1.321] [1.009–1.010] [0.994–0.998] [0.860–0.922]
Note. VHA = Veterans’ Health Administration; RUCA = rural-urban commuting area. The goodness-of-fit (C-Statistic) for this model = 0.927.
2015, Volume 11, Number 2
134
C. Marienfeld and R. A. Rosenheck
inpatient admission predicted both having serious mental illness among those in methadone maintenance treatment and being in methadone maintenance treatment among those with serious mental illness. The high rate of additional psychiatric and substance use diagnoses in veterans with serious mental illness in methadone maintenance treatment was notable, as was increased service utilization on measures of psychiatric hospitalization and emergency department visits. There was also a greater use of primary care services, suggesting that this pattern of increased VHA service use applies to both acute and chronic health services. Ideally, increased primary care visits would be linked to decreased use of acute care. However, this pattern is consistent with another recent analysis of the VHA health care system that showed increased emergency department use even with readily available and almost free access to outpatient care in patients with high psychosocial and medical needs (Doran, Raven, & Rosenheck, 2013). Not surprisingly, veterans with serious mental illness in methadone maintenance treatment were overall more likely to fill prescriptions for psychiatric medications than those in methadone maintenance treatment without serious mental illness. However, being in methadone maintenance treatment, with its more intensive level of supervision, may confer an advantage in that those in methadone maintenance treatment are more likely to fill prescriptions than people with serious mental illness in general. This runs contrary to the generally held view that patients with dual diagnosis adhere less to medication regimens, but is consistent with recent data from the VHA showing increased prescription fills in veterans with diagnosed schizophrenia or bipolar disorders and either comorbid alcohol or drug use disorders (Marienfeld & Rosenheck, 2013). This could imply that the structure of methadone maintenance treatment leads to improved adherence to medications prescribed and possibly improved health outcomes. This is particularly relevant given that the rates of psychiatric diagnoses in the methadone maintenance treatment population in this study were much higher than general population rates of psychiatric disorders (Grant et al., 2004). Prior work in patients with dual diagnosis has looked at ways to address non-compliance with aftercare in this population, along with lack of engagement in treatment, poor adherence to recommended psychiatric visits, and barriers to implementation of quality treatment (Bogenschutz & Siegfreid, 1998; Centorrino et al., 2001; Drake et al., 2001; Wolpe, Gorton, Serota, & Sanford, 1993). This study expands on prior work showing that, in fact, the patients with serious mental illness in methadone maintenance treatment used more health services and filled more psychotropic medication prescriptions both than other patients in methadone maintenance treatment and than other patients with serious mental illness who did not use methadone maintenance treatment in FY2012. These results raise concerns about drug-drug interactions among clinically indicated but sedating medications that can
Journal of Dual Diagnosis
increase the risk of respiratory depression and falls. These concerns about drug-drug interactions and related risks of oversedation, falls, and even depressed respiration in patients in methadone maintenance treatment with serious mental illness receiving additional psychiatric and opiate medication deserve greater attention both clinically and in services research studies (McCance-Katz, Sullivan, & Nallani, 2010). In addition, there are risks of interactions between methadone and psychiatric medications, which we have shown to be filled in greater numbers in this population (Gruber & McCance-Katz, 2010). Overall, this study of veterans in methadone maintenance treatment with serious mental illness is, to our knowledge, the first to study nationally the intersection of serious mental illness and methadone maintenance treatment. Several methodological limitations require comment. First, this study is limited to the VHA population, which is predominantly male and older, and thus may not be representative of other methadone maintenance treatment programs. Second, this study relied on administrative diagnoses, which are of uncertain completeness and validity, although they can be taken as accurate representations of clinician judgments and thus are of special relevance to the study of real-world practice. Third, we did not have urine toxicology data or other clinical outcomes measures available to use in comparing groups. In addition, we choose to include major affective disorders as criterion for serious mental illness rather than limiting our classification to psychotic disorders. We included patients with major depressive disorder in the serious mental illness group because they are a relatively large group that often requires extensive psychiatric treatment in addition to their substance use disorder treatment. Finally, our data only represent prescriptions filled in VHA pharmacies.
Conclusion Veterans with comorbid serious mental illness in methadone maintenance treatment have high additional psychiatric comorbidity and are more likely to use psychiatric and general health services and to fill psychiatric medications than either patients in methadone maintenance treatment without serious mental illness or patients with serious mental illness who do not use methadone maintenance treatment. These data call attention to the need for further study and clinical awareness of potential drug-drug interactions in this high medication and service using population.
DISCLOSURES Dr. Marienfeld and Dr. Rosenheck report no financial relationships with commercial interests or any additional income to report.
Veterans With SMI in Methadone Maintenance
FUNDING There was no financial support or grants to conduct this study. REFERENCES Anglin, M. D., Speckart, G. R., Booth, M. W., & Ryan, T. M. (1989). Consequences and costs of shutting off methadone. Addictive Behaviors, 14(3), 307–326. Bogenschutz, M. P., & Siegfreid, S. L. (1998). Factors affecting engagement of dual diagnosis patients in outpatient treatment. Psychiatric Services, 49(10), 1350–1352. Brooner, R. K., King, V. L., Kidorf, M., Schmidt, C. W., & Bigelow, G. E. (1997). Psychiatric and substance use comorbidity among treatmentseeking opioid abusers. Archives of General Psychiatry, 54(1), 71–80. Brunette, M. F., & Mueser, K. T. (2006). Psychosocial interventions for the long-term management of patients with severe mental illness and cooccurring substance use disorder. Journal of Clinical Psychiatry, 67(Suppl 7), 10–17. Centorrino, F., Hern´an, M. A., Drago-Ferrante, G., Rendall, M., Apicella, A., L¨angar, G., & Baldessarini, R. J. (2001). Factors associated with noncompliance with psychiatric outpatient visits. Psychiatric Services, 52(3), 378–380. doi:10.1176/appi.ps.52.3.378 Charlson, M. E., Pompei, P., Ales, K. L., & MacKenzie, C. R. (1987). A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. Journal of Chronic Diseases, 40(5), 373– 383. Dole, V. P., Robinson, J. W., Orraca, J., Towns, E., Searcy, P., & Caine, E. (1969). Methadone treatment of randomly selected criminal addicts. New England Journal of Medicine, 280(25), 1372–1375. doi:10.1056/NEJM196906192802502 Doran, K. M., Raven, M. C., & Rosenheck, R. A. (2013). What drives frequent emergency department use in an integrated health system? National data from the Veterans Health Administration. Annals of Emergency Medicine, 62(2), 151–159. doi:10.1016/j.annemergmed.2013.02.016 Drake, R. E., Essock, S. M., Shaner, A., Carey, K. B., Minkoff, K., Kola, L., . . . Rickards, L. (2001). Implementing dual diagnosis services for clients with severe mental illness. Psychiatric Services, 52(4), 469–476. doi:10.1176/appi.ps.52.4.469 Evans, J. L., Tsui, J. I., Hahn, J. A., Davidson, P. J., Lum, P. J., & Page, K. (2012). Mortality among young injection drug users in San Francisco: A 10-year follow-up of the UFO study. American Journal of Epidemiology, 175(4), 302–308. doi:10.1093/aje/kwr318 Fullerton, C. A., Kim, M., Thomas, C. P., Lyman, D. R., Montejano, L. B., Dougherty, R. H., . . . Delphin-Rittmon, M. E. (2014). Medication-assisted treatment with methadone: Assessing the evidence. Psychiatric Services, 65(2), 146–157. doi:10.1176/appi.ps.201300235 Grant, B. F., Stinson, F. S., Dawson, D. A., Chou, S. P., Dufour, M. C., Compton, W., . . . Kaplan, K. (2004). Prevalence and co-occurrence
135
of substance use disorders and independent mood and anxiety disorders: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry, 61(8), 807–816. doi:10.1001/archpsyc.61.8.807 Gruber, V. A., & McCance-Katz, E. F. (2010). Methadone, buprenorphine, and street drug interactions with antiretroviral medications. Current HIV/AIDS Reports, 7(3), 152–160. doi:10.1007/s11904-010-0048-2 Hser, Y. I., Hoffman, V., Grella, C. E., & Anglin, M. D. (2001). A 33-year follow-up of narcotics addicts. Archives of General Psychiatry, 58(5), 503–508. doi:10.1001/archpsyc.58.5.503 Macarthur, G. J., Minozzi, S., Martin, N., Vickerman, P., Deren, S., Bruneau, J., . . . Hickman, M. (2012). Opiate substitution treatment and HIV transmission in people who inject drugs: Systematic review and meta-analysis. BMJ, 345, e5945. doi:10.1136/bmj.e5945 Marienfeld, C., & Rosenheck, R. (2013). National psychotropic prescription dispensation among dually diagnosed patients. Advances in Dual Diagnosis, 6(1), 34–44. doi:10.1108/17570971311309015 Mattick, R. P., Breen, C., Kimber, J., & Davoli, M. (2009). Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. The Cochrane Database of Systematic Reviews, 3, CD002209. doi:10.1002/14651858.CD002209.pub2 McCance-Katz, E. F., Sullivan, L. E., & Nallani, S. (2010). Drug interactions of clinical importance among the opioids, methadone and buprenorphine, and other frequently prescribed medications: A review. The American Journal on Addictions, 19(1), 4–16. doi:10.1111/j.1521-0391.2009.00005.x Olfson, M., Mechanic, D., Hansell, S., Boyer, C. A., Walkup, J., & Weiden, P. J. (2000). Predicting medication noncompliance after hospital discharge among patients with schizophrenia. Psychiatric Services, 51(2), 216–222. doi:10.1176/appi.ps.51.2.216 Osterberg, L., & Blaschke, T. (2005). Adherence to medication. New England Journal of Medicine, 353(5), 487–497. doi:/NEJMra050100 Ryan, C. F., & White, J. M. (1996). Health status at entry to methadone maintenance treatment using the SF-36 health survey questionnaire. Addiction, 91(1), 39–45. Teter, C. J., Falone, A. E., Bakaian, A. M., Tu, C., Ongur, D., & Weiss, R. D. (2011). Medication adherence and attitudes in patients with bipolar disorder and current versus past substance use disorder. Psychiatry Research, 190(2–3), 253–258. doi:10.1016/j.psychres.2011.05.042 Van Den Berg, C., Smit, C., Van Brussel, G., Coutinho, R., Prins, M., & Cohort, A. (2007). Full participation in harm reduction programmes is associated with decreased risk for human immunodeficiency virus and hepatitis C virus: Evidence from the Amsterdam Cohort Studies among drug users. Addiction, 102(9), 1454–1462. doi:10.1111/j.1360-0443.2007. 01912.x Weisner, C., Mertens, J., Parthasarathy, S., Moore, C., & Lu, Y. (2001). Integrating primary medical care with addiction treatment: A randomized controlled trial. JAMA: The Journal of the American Medical Association, 286(14), 1715–1723. doi:10.1001/jama.286.14.1715 Wolpe, P. R., Gorton, G., Serota, R., & Sanford, B. (1993). Predicting compliance of dual diagnosis inpatients with aftercare treatment. Hospital & Community Psychiatry, 44(1), 45–49.
2015, Volume 11, Number 2
Copyright of Journal of Dual Diagnosis is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
PSYCHOPHARMACOLOGY & NEUROBIOLOGY
Psychiatric Services and Prescription Fills Among Veterans With Serious Mental Illness in Methadone Maintenance Treatment Carla Marienfeld, MD,1 and Robert A. Rosenheck, MD1,2
Objective: Comorbidity and co-prescription patterns of people with serious mental illness in methadone maintenance may complicate their treatment and have not been studied. The goal of this study was to examine the care and characteristics of people with serious mental illness in methadone maintenance treatment nationally in the Veterans Health Administration (VHA). Methods: Using national VHA data from FY2012, bivariate and multiple logistic regression analyses were used to compare veterans in methadone maintenance treatment who had a serious mental illness (schizophrenia, bipolar disorder, or major affective disorder) to patients in methadone maintenance treatment without serious mental illness and patients with serious mental illness who were not in methadone maintenance treatment. Results: Only a small fraction of patients with serious mental illness were receiving methadone maintenance treatment (0.65%), but a relatively large proportion in methadone maintenance treatment had a serious mental illness (33.2%). Compared to patients without serious mental illness, patients with serious mental illness in methadone maintenance treatment were more likely to have been homeless, to have had a recent psychiatric hospitalization, to be over 50% disabled, and to have had more fills for more classes of psychotropic drugs. Compared to other patients with serious mental illness, patients with serious mental illness in methadone maintenance treatment were more likely to have a drug abuse diagnosis and to reside in large urban areas. Conclusions: One-third of patients in methadone maintenance treatment have serious mental illness and more frequent psychiatric comorbidity, and they are more likely to use psychiatric and general health services and fill more types of psychiatric prescriptions. Further study and clinical awareness of potential drug-drug interactions in this high medication and service using population are needed. (Journal of Dual Diagnosis, 11:128–135, 2015)
Keywords methadone, severe mental illness, Veterans’ Health Administration
Methadone maintenance treatment has been shown to reduce the harms of illicit opioid use including hepatitis C and HIV viral transmission, improve health outcomes including psychiatric comorbidities, reduce criminal activity and possibly incarceration, and improve the likelihood of employment (Anglin, Speckart, Booth, & Ryan, 1989; Dole et al., 1969; Evans et al., 2012; Fullerton et al., 2013; Hser, Hoffman, Grella, & Anglin, 2001; Macarthur et al., 2012; Mattick, Breen, Kimber, & Davoli, 2009; Van Den Berg et al., 2007). Medical and psychiatric comorbidity are high in patients in methadone maintenance treatment (Brooner, King, Kidorf, Schmidt, & Bigelow, 1997; Ryan & White, 1996),
1Yale
University School of Medicine, New Haven, Connecticut, USA New England Mental Illness, Research, Education and Clinical Center, West Haven, Connecticut, USA Address correspondence to Carla Marienfeld, MD, Yale University School of Medicine, 300 George St., Suite 901, New Haven, CT 06511, USA. E-mail: [email protected] 2VA
and patients who receive integrated psychiatric and medical care with their addiction treatment can have better outcomes with these more cost-effective services (Weisner, Mertens, Parthasarathy, Moore, & Lu, 2001). Patients with dual diagnosis of serious mental illness and substance use disorders often have greater difficulty attending appointments and adhering to prescription medications (Bogenschutz & Siegfreid, 1998; Centorrino et al., 2001; Olfson et al., 2000; Teter et al., 2011). Integrated treatment of psychiatric and substance use disorders improves health service use and adherence to treatment, thus benefiting this population (Brunette & Mueser, 2006). A 2005 review found that major predictors of poor adherence to medications included the presence of psychological problems, particularly depression, cognitive impairment, lack of insight into the illness, and lack of belief in the benefits of treatment, all common problems in patients with serious mental illness and especially those with a dual diagnosis of serious mental illness and a substance use disorder (Osterberg & Blaschke, 2005). However, empirical
Veterans With SMI in Methadone Maintenance
research has less often addressed adherence in actually filling prescriptions within a specific time frame, a dimension that is especially amenable to evaluation in large clinical populations using administrative data, and offers a further opportunity to characterize the distinctive and potential vulnerabilities of patients with serious mental illness in methadone maintenance treatment. In this study, we use national administrative data from the Veterans’ Health Administration (VHA) to identify the proportion of veterans in methadone maintenance treatment with diagnosed serious mental illness and to describe their psychiatric and medical comorbidities as well as their use of diverse health services and psychiatric medications, as compared to those in methadone maintenance treatment who do not have serious mental illness. We also make these comparisons between patients with serious mental illness in methadone maintenance treatment and those with serious mental illness who do not receive methadone maintenance treatment. Through this broad descriptive exploration we hope to identify potential clinical problem areas deserving of further consideration in this population.
METHODS Sample and Data Source National VHA administrative records were used to identify all veterans who were enrolled in methadone maintenance treatment and/or who had a diagnosed serious mental illness in fiscal year (FY) 2012 in the VHA nationally. Data were de-identified by removing all personally identifiable information prior to analysis. Serious mental illness was defined, for these purposes, as schizophrenia and other psychotic disorders (International Classification of Diseases, 9th edition [ICD-9] codes 295.xx), bipolar spectrum disorders (ICD-9 codes 296.0x, 296.1x, 296.40–296.89), and major depressive disorders (ICD-9 codes 296.2–296.39), each of which requires extensive non–addiction-related treatment. The primary study population included veterans who were both in methadone maintenance treatment and had one or more diagnosed serious mental illnesses and was compared to veterans in methadone maintenance treatment without a serious mental illness diagnosis and to the population of veterans with one or more serious mental illnesses not participating in methadone maintenance treatment. The study was approved by the institutional review board of the VA Connecticut Healthcare System.
Measures Data were obtained on patient characteristics such as age, gender, income, marital status, receipt of Veterans’ Affairs
129
(VA) disability compensation or pension, and a designation of residence based on rural-urban commuting area (RUCA) codes (depts.washington.edu/uwruca). Information on ethnicity was not informative because of extensive missing data. Recent homelessness was identified by participation in a VHA specialized homeless service program or a V-60 code indicating housing problems. Clinical data included psychiatric and medical diagnoses (based on the ICD-9 coding system used by the VHA in FY2012) that were assigned to each patient at least once during the study year. The Charlson Index (Charlson, Pompei, Ales, & MacKenzie, 1987), an aggregate measure of medical comorbidity, has been shown to predict the 10-year mortality for patients based on the presence of a broad range of comorbid medical conditions and was used to assess medical comorbidity. Data on comorbid psychiatric diagnoses included all ICD-9 codes 290 through 319 (coded into 11 classes; available on request). VHA inpatient and outpatient service utilization was documented. Outpatient mental health specialty care and substance abuse clinic visits, and methadone maintenance treatment clinic visits in particular, were identified by standard VHA clinic codes. Inpatient psychiatric care or substance use hospitalizations were identified through bed section codes in computerized discharge summaries. Pharmacy benefit records documenting all VHA prescriptions filled were used to identify the total number of prescriptions filled by these patients in five psychiatric medication classes: antidepressants; antipsychotics; sedative, hypnotics, and anxiolytics; mood stabilizers (anticonvulsants); and lithium. The total number of psychotropic prescriptions filled by each veteran during the year was determined by summing the numbers of prescriptions in each separate class as well as the total number of classes from which a prescription was filled. Medications in these classes that were not filled in a VHA pharmacy were not included in the analysis.
Analysis Given the sample size, almost all comparisons revealed differences that were significant by conventional standards using chi-square tests for categorical data and t-tests for continuous data. Given the large sample size, effect sizes (i.e., risk ratios for dichotomous measures and Cohen’s d for continuous measures) were used rather than p values to identify meaningful differences. Risk ratios around greater than 1.5 or less than 0.67 were considered substantial. The difference between means divided by the pooled standard deviation was used to calculate Cohen’s d for continuous variables. A stepwise logistic regression model (odds ratios are reported) included all categorical variables with the risk ratios above 1.5 or less than 0.5 and all continuous variables with a Cohen’s d absolute effect size of 0.3 or greater (small to moderate effect size). The first logistic regression compared patients in methadone main-
2015, Volume 11, Number 2
130
C. Marienfeld and R. A. Rosenheck
TABLE 1 Veterans With Serious Mental Illness (SMI) and/or in Methadone Maintenance Treatment (MMT): Categorical Data1,2 MMT + SMI (n = 3319) Demographics Male Homeless Black White Hispanic Service in Iraq or Afghanistan
MMT Only (n = 6682)
SMI Only (n = 507403)
Risk Ratio MMT + SMI/MMT
Risk Ratio MMT + SMI/SMI
n 3100 1519 1037 2127 403 388
% 93.4% 45.8% 31.2% 64.1% 12.14% 11.69%
n 6468 2006 2314 4132 880 874
% 96.80% 30.02% 34.62% 61.84% 13.18% 13.08%
n 442827 58725 106258 378200 72364 52611
% 87.27% 11.57% 20.94% 74.54% 14.26% 10.37%
0.96 1.52 0.90 1.04 0.92 0.89
1.07 3.95 1.49 0.86 0.85 1.13
674 424 1871 165 12 1
20.31% 12.77% 56.37% 4.97% 0.36% 0.03%
1359 615 3719 554 51 13
20.34% 9.20% 55.66% 8.29% 0.76% 0.19%
74782 65860 227621 66344 23834 10345
14.74% 12.98% 44.86% 13.08% 4.70% 2.04%
1.00 1.39 1.01 0.60 0.47 0.15
1.38 0.98 1.26 0.38 0.08 0.01
1017 1087 440 775 400
30.64% 32.75% 13.26% 23.35% 12.05%
2455 2062 851 1314 810
36.74% 30.86% 12.74% 19.66% 12.12%
113172 120684 81781 191766 25465
22.30% 23.78% 16.12% 37.79% 5.02%
0.83 1.06 1.04 1.19 0.99
1.37 1.38 0.82 0.62 2.40
VHA service connection status > 50% < 50%
908 506
27.36% 15.25%
1203 1078
18.00% 16.13%
202124 83774
39.84% 16.51%
1.52 0.95
0.69 0.92
Rural-urban commuting area3 Large urban Urban Small rural Isolated small rural
153 2990 96 80
4.62% 90.08% 2.88% 2.42%
278 6047 218 139
4.16% 90.49% 3.26% 2.08%
58671 374722 42429 31581
11.56% 73.85% 8.36% 6.22%
1.11 1.00 0.88 1.16
0.40 1.22 0.34 0.39
Psychiatric diagnoses Dementia Schizophrenia Other psychotic disorders Bipolar disorder Depression PTSD Alcohol use disorder Other drug use disorder Dysthymia Anxiety disorder Adjustment disorder Personality disorder Any substance use disorder Any psychiatric disorder (not substance use) Any dual diagnosis
11 513 486 912 2017 1314 1772 3137 1753 1133 372 410 3229 3319 3319
0.33% 15.46% 14.64% 27.48% 60.77% 39.59% 53.39% 94.52% 52.82% 34.14% 11.21% 12.35% 97.29% 100.00% 100.00%
7 0 0 0 0 1739 2638 6360 2657 1464 578 295 6570 6653 6657
0.10% 0.00% 0.00% 0.00% 0.00% 26.03% 39.48% 95.18% 39.76% 21.91% 8.65% 4.41% 98.32% 99.57% 99.63%
7999 88669 52605 113191 307357 176490 94597 84396 197822 134197 37147 28410 133060 507403 507403
1.58% 17.48% 10.37% 22.31% 60.57% 34.78% 18.64% 16.63% 38.99% 26.45% 7.32% 5.60% 26.22% 100.00% 100.00%
3.16 n/a n/a n/a n/a 1.52 1.35 0.99 1.33 1.56 1.30 2.80 0.99 1.00 1.00
0.21 0.88 1.41 1.23 1.00 1.14 2.86 5.68 1.35 1.29 1.53 2.21 3.71 1.00 1.00
Medical diagnoses Nausea/vomiting Liver disease HIV All pain diagnoses
151 607 80 2309
4.55% 18.29% 2.41% 69.57%
229 1228 143 3927
3.43% 18.38% 2.14% 58.77%
12951 22751 4338 300308
2.55% 4.48% 0.85% 59.19%
1.33 1.00 1.13 1.18
1.78 4.08 2.82 1.18
Service use4 Substance use outpatient Psychiatric outpatient
3284 2808
98.95% 84.60%
6654 4392
99.58% 65.73%
45631 453306
8.99% 89.34%
0.99 1.29
Age < 40 40–49 50–64 65–74 75–85 > 85 Income < $7,000 $7–15,000 $15–25,000 > $25,000 VHA pension
Journal of Dual Diagnosis
11.00 0.95 (Continued on next page)
131
Veterans With SMI in Methadone Maintenance
TABLE 1 (Continued) Veterans With Serious Mental Illness (SMI) and/or in Methadone Maintenance Treatment (MMT): Categorical Data1,2 MMT + SMI (n = 3319) Any mental health outpatient visit Substance use inpatient or outpatient Any mental health service inpatient or outpatient Mental health inpatient admission
3314 3303 3316 1035
99.85% 99.52% 99.91% 31.18%
MMT Only (n = 6682) 6668 6674 6669 702
99.79% 99.88% 99.81% 10.51%
SMI Only (n = 507403) 456177 137572 456705 44506
89.90% 27.11% 90.01% 8.77%
Risk Ratio MMT + SMI/MMT
Risk Ratio MMT + SMI/SMI
1.00 1.00 1.00 2.97
1.11 3.67 1.11 3.56
Note. PTSD = posttraumatic stress disorder; HIV = human immunodeficiency virus. Given the large sample size, effect sizes (i.e., risk ratios) were used rather than p values to identify meaningful differences. Risk ratios greater than 1.5 or less than .67 were considered meaningful here. 1Patients in FY2012 from Veterans’ Health Administration (VHA) sites that have a methadone program and had diagnosed schizophrenia or other psychosis, bipolar disorder, or major depressive disorder. 2Chi-squared analysis (two degrees of freedom) was significant with p < .0001 for all categorical variables except VHA service connection less than 50%. 3Rural-urban commuting area (RUCA) codes (depts.washington.edu/uwruca). 4Among those with any service use in FY2012.
tenance treatment with and without serious mental illness, and the second compared patients with serious mental illness in or not in methadone maintenance treatment. For nested variables (e.g., substance abuse visits within the category of all outpatient visits), the broader variable encompassing the smaller ones was used to avoid large amounts of noninformative overlap. For example, since substance abuse outpatient visits are a subset of all outpatient visits, which itself showed substantial differences across groups, we used only the total number of outpatient visits in our models. A c-statistic was calculated to assess the goodness-of-fit of the logistic regression model in predicting the value of the dichotomous dependent variable.
RESULTS Demographics and Baseline Rates In the VHA nationally in FY2012, there were 510,722 veterans with diagnosed serious mental illness and 10,001 veterans who participated in methadone maintenance treatment. One-third (33.2%, n = 3319) of patients in methadone maintenance treatment received a serious mental illness diagnosis, but this represents only 0.65% of veterans with serious mental illness. The patients were, on average, in their mid-50s (M = 55.4, SD = 13.5). Patients in methadone maintenance treatment with serious mental illness were 1.5 times more likely to be or have been homeless as compared to the general methadone maintenance treatment population and almost 4 times more likely to be homeless than the general serious mental illness population (Table 1). Of patients in methadone maintenance treatment with serious mental illness, the most common comorbid psychiatric diagnoses were major depression (60.8%), other depression (52.8%), posttraumatic stress disorder (39.6%), anxiety disorder (34.1%), and bipolar disorder (27.5%). Schizophrenia and other psychotic disorders together totaled 30.1%. Approximately 12% of those with se-
rious mental illness in methadone maintenance treatment also had a personality disorder (Table 1).
Patients With Serious Mental Illness on Methadone Maintenance Compared to Those Without Serious Mental Illness On bivariate analysis, patients in methadone maintenance treatment with serious mental illness were more likely to have been homeless than other patients in methadone maintenance treatment and to have VHA service–connected disability benefits rated greater than 50%, and they were less likely to be older than 65. In addition to having schizophrenia, bipolar, or major depressive disorders, they were more likely to have other psychiatric diagnoses of dementia, posttraumatic stress disorder, anxiety disorder, and personality disorder. They were also more likely to have had a mental health inpatient admission (Table 1). There was a small effect size (0.3–0.4) in the difference between groups for the number of general psychiatric outpatient visits, the number of emergency room visits, and the number of antipsychotic, anxiolytic, and anticonvulsant prescriptions. There was a moderate effect size (0.5–0.7) between groups for the number of substance abuse, psychiatric, or any outpatient visits as well as for the number of any psychotropic medication or antidepressant prescriptions. There was a large effect size (0.8 or greater) for the number of classes of psychiatric medications filled (Table 2). Logistic regression analysis showed that veterans in methadone maintenance treatment with serious mental illness were independently more likely to have been homeless, to have a VA service connected disability greater than 50%, to have a diagnosis of personality disorder, to have a mental health inpatient admission, and to have prescription fills in a larger number of psychiatric medication classes than veterans in methadone maintenance treatment without an serious mental illness diagnosis (Table 3). 2015, Volume 11, Number 2
132
C. Marienfeld and R. A. Rosenheck
TABLE 2 Veterans With a Serious Mental Illness (SMI) and/or in Methadone Maintenance Treatment (MMT): Continuous Data1 MMT + SMI (n = 3319) Demographics Age (in years) Income Medical comorbidity Charlson Index Score Service use2 Substance use outpatient clinic visits General psychiatric outpatient visits Psychiatric or substance use outpatient visits Methadone visits Primary care visits Medical or surgical outpatient visits Number of emergency room visits Number of days hospitalized3 Any outpatient visits Prescriptions filled4 Any psychotropic prescriptions Antidepressant prescriptions Antipsychotic prescriptions Anxiolytic/sedative/hypnotic prescriptions Stimulant prescriptions Anticonvulsant prescriptions Lithium prescriptions Opiate prescriptions Number of classes of psychiatric meds filled
MMT Only (n = 6682)
SMI Only (n = 507403)
MMT + SMI vs. MMT
MMT + SMI vs. SMI
Effect size −0.09 0.01
Effect size −0.34 −0.22
Mean 50.90 15406.10
SD 11.99 17956.93
Mean 52.14 14944.62
SD 12.86 24459.53
Mean 55.55 23278.47
SD 13.88 36134.26
Pooled SD 13.86 35919.93
2.54
2.24
2.58
2.38
2.89
2.41
2.41
−0.02
−0.14
61.18 22.24
77.70 35.23
51.78 10.16
72.06 21.17
2.06 12.01
12.02 26.11
15.73 26.12
0.60 0.46
3.76 0.39
83.42
88.01
61.90
77.72
14.07
30.68
32.42
0.66
2.14
35.31 4.68 12.79
65.06 5.25 13.89
35.34 3.72 10.11
63.83 4.30 12.30
0.00 3.78 10.71
0.00 3.95 11.99
8.93 3.96 12.01
0.00 0.24 0.22
3.95 0.23 0.17
2.72 16.36 96.21
4.47 21.92 90.53
1.59 9.50 72.04
3.09 10.50 80.10
0.98 17.65 24.78
2.37 74.49 34.20
2.40 73.80 35.81
0.47 0.09 0.67
0.73 −0.02 1.99
32.22 13.49 8.31 4.06
55.03 21.82 25.91 8.95
10.88 5.23 1.35 1.77
21.36 11.18 6.49 5.24
20.79 8.06 4.94 3.61
38.29 15.29 17.87 6.93
38.25 15.29 17.83 6.92
0.56 0.54 0.39 0.33
0.30 0.36 0.19 0.07
0.16 5.28 0.73 11.69 2.20
1.58 16.19 4.77 12.10 1.18
0.14 1.83 0.03 11.26 1.14
1.46 7.09 0.62 10.71 1.10
0.17 3.05 0.04 7.79 1.94
1.62 11.68 4.17 8.02 1.13
1.62 11.66 4.14 8.09 1.13
0.01 0.30 0.17 0.05 0.94
−0.01 0.19 0.17 0.48 0.23
Note. Given the large sample size, effect sizes (i.e., Cohen’s d) were used rather than p values to identify meaningful differences. Cohen’s d effect sizes greater than .30 were considered meaningful here. 1Patients in FY2012 from Veterans’ Health Administration sites that have a methadone program and had diagnosed schizophrenia or other psychosis, bipolar disorder, or major depressive disorder. 2Among those with any service use in FY2012. 3Among those with a hospitalization. 4Among those with prescriptions filled in FY2012, mean reflects the average number of prescriptions filled in the year. TABLE 3 Logistic Regression Analysis of Patients in Methadone Maintenance Treatment (MMT) With Serious Mental Illness (SMI) (n = 3227) Compared to Those Without SMI (n = 6531) Variable
Estimate
Standardized Estimate
Odds Ratio
95% Confidence Interval
Age Homeless VHA service connection greater than 50% Personality disorder Mental health inpatient admission Any outpatient visit Antidepressant prescriptions Antipsychotic prescriptions Anxiolytic/sedative/hypnotic prescriptions Anticonvulsant prescriptions Number of classes of psychiatric meds filled
0.00997 0.2903 0.3262 0.2696 0.4106 0.00119 0.00873 0.0384 −0.00927 −1.0105 0.6003
0.693 0.0766 0.731 0.0376 0.0849 0.0558 0.0759 0.3385 −0.0344 −0.0642 0.406
1.01 1.337 1.386 1.309 1.508 1.001 1.009 1.039 0.991 0.99 1.823
[1.006–1.014] [1.204–1.484] [1.235–1.555] [1.084–1.582] [1.315–1.728] [1.001–1.002] [1.005–1.013] [1.031–1.047] [0.983–0.999] [0.984–0.995] [1.729–1.922]
Note. VHA = Veterans’ Health Administration. The goodness-of-fit (C-Statistic) for this model = 0.777.
Journal of Dual Diagnosis
133
Veterans With SMI in Methadone Maintenance
Patients With Serious Mental Illness in Methadone Maintenance Treatment Compared to Those Not in Methadone Maintenance Treatment
disorder, an adjustment disorder, a personality disorder, and filled prescriptions in fewer classes of psychiatric medications (Table 4).
On bivariate analysis, veterans with serious mental illness who were in methadone maintenance treatment were more likely to have been homeless and to receive a VA pension than veterans with serious mental illness who did not receive methadone maintenance treatment. Those who received methadone maintenance treatment were more likely to have a drug, alcohol, or any substance use disorder as well as to have an adjustment disorder or a personality disorder. They were also more likely to have diagnosed liver disease and HIV. They used more outpatient substance use services and were more likely to have had a mental health inpatient admission (Table 1). There was a small effect size (0.3–0.4) in the difference between patients with serious mental illness in methadone maintenance treatment and other patients with serious mental illness for being younger, the number of general psychiatric outpatient visits, and the number of any psychiatric prescriptions filled, antidepressant prescriptions filled, and opiate prescriptions filled. There was a moderate effect size (0.5–0.7) between groups for the number of emergency room visits. There was a large effect size (0.8 or greater) for the substance use and psychiatric outpatient visits and for any outpatient visits (Table 2). Logistic regression analysis showed that patients with serious mental illness receiving methadone maintenance treatment were independently more likely to receive a VA pension, to live in a large urban area, to have a diagnosis of drug use disorder or liver disease, or to have had a mental health inpatient stay. Veterans with serious mental illness veterans in methadone maintenance treatment, as compared to those not in methadone maintenance treatment, had less likelihood of an alcohol use
DISCUSSION This is the first national study of which we are aware to explore the distinctive characteristics, clinical diagnoses, and service use and psychiatric prescription fills among veterans with serious mental illness in methadone maintenance treatment. While only a small fraction (0.65%) of patients with serious mental illness received methadone maintenance treatment, a relatively large proportion (33.2%) of patients receiving methadone maintenance treatment also had a diagnosed serious mental illness. Significant predictors of having a serious mental illness while in methadone maintenance treatment included social factors (homelessness, VA service connection benefit status), psychiatric comorbidity (personality disorder), service use (having a mental health inpatient admission), and greater psychiatric medication usage (the total number of classes of psychiatric medications filled). However, factors associated with being in methadone maintenance treatment among patients with diagnosed serious mental illness were somewhat different, the most prominent of which—not surprisingly—was having a drug use diagnosis. The psychiatric comorbid diagnoses (alcohol use disorder, adjustment disorder, personality disorder) and the number of classes of psychiatric medications filled were negatively associated with receiving methadone maintenance treatment. A medical comorbidity, liver disease, was associated with receiving methadone maintenance treatment, although this is not surprising given the risk of hepatitis C in intravenous drug users. Interestingly, having a mental health
TABLE 4 Logistic Regression Analysis of Patients With Serious Mental Illness (SMI) in Methadone Maintenance Treatment (MMT) (n = 3227) Compared to Those Not in MMT (n = 486080) Variable
Estimate
Standardized Estimate
Odds Ratio
95% Confidence Interval
Age Homeless VHA pension RUCA large urban Alcohol substance use disorder Any drug use disorder Anxiety disorder Adjustment disorder Personality disorder Liver disease Mental health inpatient admission Any outpatient visit Antipsychotic prescriptions Number of classes of psychiatric meds filled
−0.0122 0.0874 0.3807 0.7568 −0.3472 4.1425 0.0946 −0.2477 −0.1874 0.6434 0.1899 0.00928 −0.00385 −0.1165
−0.0928 0.0157 0.0461 0.1831 −0.0753 0.8669 0.023 −0.0356 −0.024 0.0741 0.0299 0.1822 −0.0384 −0.0724
0.988 1.091 1.463 2.132 0.707 62.958 1.099 0.781 0.829 1.903 1.209 1.009 0.996 0.89
[0.985–0.991] [1.008–1.182] [1.304–1.642] [1.893–2.400] [0.653–0.765] [53.636–73.9] [1.016–1.189] [0.694–0.878] [0.740–0.929] [1.726–2.098] [1.106–1.321] [1.009–1.010] [0.994–0.998] [0.860–0.922]
Note. VHA = Veterans’ Health Administration; RUCA = rural-urban commuting area. The goodness-of-fit (C-Statistic) for this model = 0.927.
2015, Volume 11, Number 2
134
C. Marienfeld and R. A. Rosenheck
inpatient admission predicted both having serious mental illness among those in methadone maintenance treatment and being in methadone maintenance treatment among those with serious mental illness. The high rate of additional psychiatric and substance use diagnoses in veterans with serious mental illness in methadone maintenance treatment was notable, as was increased service utilization on measures of psychiatric hospitalization and emergency department visits. There was also a greater use of primary care services, suggesting that this pattern of increased VHA service use applies to both acute and chronic health services. Ideally, increased primary care visits would be linked to decreased use of acute care. However, this pattern is consistent with another recent analysis of the VHA health care system that showed increased emergency department use even with readily available and almost free access to outpatient care in patients with high psychosocial and medical needs (Doran, Raven, & Rosenheck, 2013). Not surprisingly, veterans with serious mental illness in methadone maintenance treatment were overall more likely to fill prescriptions for psychiatric medications than those in methadone maintenance treatment without serious mental illness. However, being in methadone maintenance treatment, with its more intensive level of supervision, may confer an advantage in that those in methadone maintenance treatment are more likely to fill prescriptions than people with serious mental illness in general. This runs contrary to the generally held view that patients with dual diagnosis adhere less to medication regimens, but is consistent with recent data from the VHA showing increased prescription fills in veterans with diagnosed schizophrenia or bipolar disorders and either comorbid alcohol or drug use disorders (Marienfeld & Rosenheck, 2013). This could imply that the structure of methadone maintenance treatment leads to improved adherence to medications prescribed and possibly improved health outcomes. This is particularly relevant given that the rates of psychiatric diagnoses in the methadone maintenance treatment population in this study were much higher than general population rates of psychiatric disorders (Grant et al., 2004). Prior work in patients with dual diagnosis has looked at ways to address non-compliance with aftercare in this population, along with lack of engagement in treatment, poor adherence to recommended psychiatric visits, and barriers to implementation of quality treatment (Bogenschutz & Siegfreid, 1998; Centorrino et al., 2001; Drake et al., 2001; Wolpe, Gorton, Serota, & Sanford, 1993). This study expands on prior work showing that, in fact, the patients with serious mental illness in methadone maintenance treatment used more health services and filled more psychotropic medication prescriptions both than other patients in methadone maintenance treatment and than other patients with serious mental illness who did not use methadone maintenance treatment in FY2012. These results raise concerns about drug-drug interactions among clinically indicated but sedating medications that can
Journal of Dual Diagnosis
increase the risk of respiratory depression and falls. These concerns about drug-drug interactions and related risks of oversedation, falls, and even depressed respiration in patients in methadone maintenance treatment with serious mental illness receiving additional psychiatric and opiate medication deserve greater attention both clinically and in services research studies (McCance-Katz, Sullivan, & Nallani, 2010). In addition, there are risks of interactions between methadone and psychiatric medications, which we have shown to be filled in greater numbers in this population (Gruber & McCance-Katz, 2010). Overall, this study of veterans in methadone maintenance treatment with serious mental illness is, to our knowledge, the first to study nationally the intersection of serious mental illness and methadone maintenance treatment. Several methodological limitations require comment. First, this study is limited to the VHA population, which is predominantly male and older, and thus may not be representative of other methadone maintenance treatment programs. Second, this study relied on administrative diagnoses, which are of uncertain completeness and validity, although they can be taken as accurate representations of clinician judgments and thus are of special relevance to the study of real-world practice. Third, we did not have urine toxicology data or other clinical outcomes measures available to use in comparing groups. In addition, we choose to include major affective disorders as criterion for serious mental illness rather than limiting our classification to psychotic disorders. We included patients with major depressive disorder in the serious mental illness group because they are a relatively large group that often requires extensive psychiatric treatment in addition to their substance use disorder treatment. Finally, our data only represent prescriptions filled in VHA pharmacies.
Conclusion Veterans with comorbid serious mental illness in methadone maintenance treatment have high additional psychiatric comorbidity and are more likely to use psychiatric and general health services and to fill psychiatric medications than either patients in methadone maintenance treatment without serious mental illness or patients with serious mental illness who do not use methadone maintenance treatment. These data call attention to the need for further study and clinical awareness of potential drug-drug interactions in this high medication and service using population.
DISCLOSURES Dr. Marienfeld and Dr. Rosenheck report no financial relationships with commercial interests or any additional income to report.
Veterans With SMI in Methadone Maintenance
FUNDING There was no financial support or grants to conduct this study. REFERENCES Anglin, M. D., Speckart, G. R., Booth, M. W., & Ryan, T. M. (1989). Consequences and costs of shutting off methadone. Addictive Behaviors, 14(3), 307–326. Bogenschutz, M. P., & Siegfreid, S. L. (1998). Factors affecting engagement of dual diagnosis patients in outpatient treatment. Psychiatric Services, 49(10), 1350–1352. Brooner, R. K., King, V. L., Kidorf, M., Schmidt, C. W., & Bigelow, G. E. (1997). Psychiatric and substance use comorbidity among treatmentseeking opioid abusers. Archives of General Psychiatry, 54(1), 71–80. Brunette, M. F., & Mueser, K. T. (2006). Psychosocial interventions for the long-term management of patients with severe mental illness and cooccurring substance use disorder. Journal of Clinical Psychiatry, 67(Suppl 7), 10–17. Centorrino, F., Hern´an, M. A., Drago-Ferrante, G., Rendall, M., Apicella, A., L¨angar, G., & Baldessarini, R. J. (2001). Factors associated with noncompliance with psychiatric outpatient visits. Psychiatric Services, 52(3), 378–380. doi:10.1176/appi.ps.52.3.378 Charlson, M. E., Pompei, P., Ales, K. L., & MacKenzie, C. R. (1987). A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. Journal of Chronic Diseases, 40(5), 373– 383. Dole, V. P., Robinson, J. W., Orraca, J., Towns, E., Searcy, P., & Caine, E. (1969). Methadone treatment of randomly selected criminal addicts. New England Journal of Medicine, 280(25), 1372–1375. doi:10.1056/NEJM196906192802502 Doran, K. M., Raven, M. C., & Rosenheck, R. A. (2013). What drives frequent emergency department use in an integrated health system? National data from the Veterans Health Administration. Annals of Emergency Medicine, 62(2), 151–159. doi:10.1016/j.annemergmed.2013.02.016 Drake, R. E., Essock, S. M., Shaner, A., Carey, K. B., Minkoff, K., Kola, L., . . . Rickards, L. (2001). Implementing dual diagnosis services for clients with severe mental illness. Psychiatric Services, 52(4), 469–476. doi:10.1176/appi.ps.52.4.469 Evans, J. L., Tsui, J. I., Hahn, J. A., Davidson, P. J., Lum, P. J., & Page, K. (2012). Mortality among young injection drug users in San Francisco: A 10-year follow-up of the UFO study. American Journal of Epidemiology, 175(4), 302–308. doi:10.1093/aje/kwr318 Fullerton, C. A., Kim, M., Thomas, C. P., Lyman, D. R., Montejano, L. B., Dougherty, R. H., . . . Delphin-Rittmon, M. E. (2014). Medication-assisted treatment with methadone: Assessing the evidence. Psychiatric Services, 65(2), 146–157. doi:10.1176/appi.ps.201300235 Grant, B. F., Stinson, F. S., Dawson, D. A., Chou, S. P., Dufour, M. C., Compton, W., . . . Kaplan, K. (2004). Prevalence and co-occurrence
135
of substance use disorders and independent mood and anxiety disorders: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry, 61(8), 807–816. doi:10.1001/archpsyc.61.8.807 Gruber, V. A., & McCance-Katz, E. F. (2010). Methadone, buprenorphine, and street drug interactions with antiretroviral medications. Current HIV/AIDS Reports, 7(3), 152–160. doi:10.1007/s11904-010-0048-2 Hser, Y. I., Hoffman, V., Grella, C. E., & Anglin, M. D. (2001). A 33-year follow-up of narcotics addicts. Archives of General Psychiatry, 58(5), 503–508. doi:10.1001/archpsyc.58.5.503 Macarthur, G. J., Minozzi, S., Martin, N., Vickerman, P., Deren, S., Bruneau, J., . . . Hickman, M. (2012). Opiate substitution treatment and HIV transmission in people who inject drugs: Systematic review and meta-analysis. BMJ, 345, e5945. doi:10.1136/bmj.e5945 Marienfeld, C., & Rosenheck, R. (2013). National psychotropic prescription dispensation among dually diagnosed patients. Advances in Dual Diagnosis, 6(1), 34–44. doi:10.1108/17570971311309015 Mattick, R. P., Breen, C., Kimber, J., & Davoli, M. (2009). Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. The Cochrane Database of Systematic Reviews, 3, CD002209. doi:10.1002/14651858.CD002209.pub2 McCance-Katz, E. F., Sullivan, L. E., & Nallani, S. (2010). Drug interactions of clinical importance among the opioids, methadone and buprenorphine, and other frequently prescribed medications: A review. The American Journal on Addictions, 19(1), 4–16. doi:10.1111/j.1521-0391.2009.00005.x Olfson, M., Mechanic, D., Hansell, S., Boyer, C. A., Walkup, J., & Weiden, P. J. (2000). Predicting medication noncompliance after hospital discharge among patients with schizophrenia. Psychiatric Services, 51(2), 216–222. doi:10.1176/appi.ps.51.2.216 Osterberg, L., & Blaschke, T. (2005). Adherence to medication. New England Journal of Medicine, 353(5), 487–497. doi:/NEJMra050100 Ryan, C. F., & White, J. M. (1996). Health status at entry to methadone maintenance treatment using the SF-36 health survey questionnaire. Addiction, 91(1), 39–45. Teter, C. J., Falone, A. E., Bakaian, A. M., Tu, C., Ongur, D., & Weiss, R. D. (2011). Medication adherence and attitudes in patients with bipolar disorder and current versus past substance use disorder. Psychiatry Research, 190(2–3), 253–258. doi:10.1016/j.psychres.2011.05.042 Van Den Berg, C., Smit, C., Van Brussel, G., Coutinho, R., Prins, M., & Cohort, A. (2007). Full participation in harm reduction programmes is associated with decreased risk for human immunodeficiency virus and hepatitis C virus: Evidence from the Amsterdam Cohort Studies among drug users. Addiction, 102(9), 1454–1462. doi:10.1111/j.1360-0443.2007. 01912.x Weisner, C., Mertens, J., Parthasarathy, S., Moore, C., & Lu, Y. (2001). Integrating primary medical care with addiction treatment: A randomized controlled trial. JAMA: The Journal of the American Medical Association, 286(14), 1715–1723. doi:10.1001/jama.286.14.1715 Wolpe, P. R., Gorton, G., Serota, R., & Sanford, B. (1993). Predicting compliance of dual diagnosis inpatients with aftercare treatment. Hospital & Community Psychiatry, 44(1), 45–49.
2015, Volume 11, Number 2
Copyright of Journal of Dual Diagnosis is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
