ORIGINAL RESEARCH

Comparison of Health Service Use Among Veterans With Methamphetamine Versus Alcohol Use Disorders Benjamin J. Morasco, PhD, Maya E. O’Neil, PhD, Jonathan P. Duckart, MPS, and Linda Ganzini, MD, MPH

Objectives: Methamphetamine use disorders (MUD) are associated with severe health effects and psychiatric comorbidities, but little is known about the health care utilization of patients with MUD. The goal of this study was to describe health service use among veterans with MUD relative to a group of veterans with an alcohol use disorder (AUD). Methods: Using Veterans Affairs (VA) administrative data, we identified 718 patients who were diagnosed with MUD and had confirmatory drug testing. Data were compared with those of 744 patients who had diagnoses of an AUD also with confirmatory testing. We examined diagnoses and medical utilization for 5 years after their index date. Results: Patients with MUD and laboratory-confirmed recent use were younger and more likely to be diagnosed with a mood disorder, posttraumatic stress disorder, and a psychotic-spectrum disorder (all P values < 0.05). After statistical controls, patients with MUD were more likely to have an inpatient hospitalization (80% vs 70%, odds ratio [OR] = 1.8; 95% confidence interval [CI] = 1.4-2.3), discharge from an inpatient admission against medical advice (23.4% vs 8.3%, OR = 2.6, 95% CI = 1.9-3.7), receive care at 3 or more VA medical centers (13.1% vs 5.4%, OR = 2.3, 95% CI = 1.5-3.5), have a behavioral flag in the medical record (5.6% vs 1.1%, OR = 4.6, 95%

From the Mental Health and Clinical Neurosciences Division (BJM, MEO, LG), Portland VA Medical Center, Portland, Oregon ; Department of Psychiatry (BJM, MEO, LG), Oregon Health & Science University, Portland, Oregon; Portland Center for the Study of Chronic Comorbid Mental and Physical Disorders (BJM, MEO, JPD, LG), Health Services Research and Development, Portland VA Medical Center, Portland, Oregon. Received for publication May 24, 2013; accepted October 12, 2013. Supported in part by award 018165 from the National Institute on Drug Abuse to Drs. Ganzini and Morasco, as part of the Methamphetamine Research Center at the Portland VA Medical Center and Oregon Health & Science University. Dr Ganzini, Dr O’Neil, and Mr Duckart were supported in part by grant REA 06-174 from the VA Health Services Research and Development service. Dr Morasco received funding from grant K23DA023467 from the National Institute on Drug Abuse. The authors declare no conflicts of interest. The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the Department of Veterans Affairs or the National Institute on Drug Abuse. Send correspondence and reprint requests to Benjamin Morasco, PhD, Portland VA Medical Center (R&D99), 3710 SW US Veterans Hospital Road, Portland, OR 97239. E-mail: [email protected]. C 2014 American Society of Addiction Medicine Copyright  ISSN: 1932-0620/14/0801-0047 DOI: 10.1097/ADM.0000000000000005

CI = 2.1-10.6), and have more total missed appointments in the 5-year study period (M = 33.1 vs M = 23.5, P < 0.001). Conclusions: Among veterans with substance use disorders, those with MUD and laboratory-confirmed recent use have additional behavioral, health care utilization, and psychiatric characteristics that need to be considered in developing programs of care. Key Words: alcohol abuse, comorbidity, health service utilization, medical care, methamphetamine abuse (J Addict Med 2014;8: 47–52)

R

ecent years have witnessed a modest decrease in the use of methamphetamine. The frequency of past-month users is 0.2% of the US population, with higher rates in western and rural states (Substance Abuse and Mental Health Services Administration, 2013). Despite the decreased use of methamphetamine, it is marked by pernicious outcomes. Long-term use of high doses of methamphetamine can affect motor coordination, lead to significant weight loss, stroke, seizures, toxicity of the kidneys and liver, and prenatal complications, and may complicate care of many other medical illnesses such as hypertension, cardiac disease, and cognitive impairment (Darke et al., 2008). In addition, methamphetamine use disorder (MUD) may result in negative psychological states that include anxiety, confusion, aggression, depression, paranoia, psychosis, and suicidal ideation (Rawson et al., 2001; Zweben et al., 2004; McKetin et al., 2006; Salo et al., 2011); up to 50% of people with MUD have comorbid depression, personality disorders, and/or psychosis (Committee on Opportunities in Drug Abuse Research, 1996; Huang et al., 2006). Despite the high medical and psychiatric comorbidity associated with MUD, empirical data on the health care utilization of patients with MUD are limited. Some data suggest that medical inpatients testing positive for methamphetamine have higher direct costs and utilize more resources than patients who test negative for methamphetamine (Richards et al., 1999; London et al., 2009). A recent study of individuals with current methamphetamine use and comorbid opioid dependence found that patients with comorbid methamphetamine and opioid use had increased likelihood of hospital admissions, relative to patients with opioid dependence alone (Pilowsky et al., 2011). Clinical experience working with MUD patients on inpatient medical and surgical wards, and in traditional outpatient substance abuse treatment programs, suggests that this

J Addict Med r Volume 8, Number 1, January/February 2014

Copyright © 2014 American Society of Addiction Medicine. Unauthorized reproduction of this article is prohibited.

47

J Addict Med r Volume 8, Number 1, January/February 2014

Morasco et al.

patient population has high rates of severe and complex medical comorbidities, high rates of recidivism, and poor treatment compliance. Information on health care utilization can be used to design more appropriate and effective health services for this patient population. The goal of this study was to describe patterns of health care use for veterans in the Pacific Northwest, a section of the United States with relatively high rates of methamphetamine use, who are diagnosed with laboratory-confirmed MUD. We predicted that patients with MUD would demonstrate high rates of psychiatric comorbidity and poor outpatient adherence. A secondary objective was to compare this group with a sample of veterans with alcohol use disorders (AUD), also restricted to those with laboratory confirmation indicating intoxication during the medical encounter. We hypothesized that patients with MUD and laboratory-confirmed recent use, compared with this sample of patients with AUD, would have medical care that is more episodic, chaotic, spread across many different locations, and characterized by overutilization of inpatient care and underutilization of outpatient care. We chose for comparison a cohort of patients with AUD and laboratory-confirmed recent use because intoxication could be determined, it is a patient population with high rates of medical utilization (Cryer et al., 1999; Alexandre et al., 2001; Polen et al., 2001), and traditional Veterans Affairs (VA) substance abuse treatment services were originally developed to meet the needs of persons with problematic alcohol use. This comparison may assist in outlining how substance abuse treatment may transform to more effectively care for patients with MUD, and specialized approaches to health care delivery for this patient population.

METHODS Participants All data were obtained for veterans receiving medical care at a VA facility in Veterans Integrated Service Network 20 (VISN-20) in the Pacific Northwest (Washington, Oregon, Idaho, and Alaska). Data were downloaded from the VISN20 Data Warehouse after approval by the institutional review board at the Portland VA Medical Center. Two groups of patients were included: (1) all veterans from VISN-20 who had been diagnosed with MUD anytime in a 3-year period, between January 1, 2003, and December 31, 2005, and (2) veterans from VISN-20 diagnosed with an AUD from the same time period. To be included, patients in the MUD group had both a diagnosis of MUD recorded in the electronic medical record (EMR) and a urine drug screen positive for methamphetamine metabolites. Patients in the AUD group had a diagnosis recorded in the EMR and with at least 1 blood alcohol level ≥ 80 mg/dl (the legal level of intoxication). Confirmatory testing was included to ensure that the use of methamphetamine was recent. Confirmatory testing was required for the alcohol group to develop a comparison cohort with similar severity of illness. Each patient was assigned an index date, for the purpose of tracking comorbid diagnoses and subsequent medical utilization. The specific index date was the date of the laboratory test in which a patient screened positive for methamphetamine or alcohol; these tests were conducted for standard

48

clinical purposes and may have occurred in inpatient, emergency, or ambulatory settings. Patients were excluded from the AUD comparison group if they were diagnosed with another illicit substance use disorder, including marijuana, cocaine, amphetamines, or illicit opioid use. Because polysubstance is almost universal in persons with MUD, other drug use was not an exclusion criterion for this group.

Demographic, Diagnostic, and Medical Utilization Data Demographic data included age, gender, race, marital status, and VA service-connected status. Inpatient and outpatient diagnoses were based on the International Classification of Diseases, Clinical Modification–9th Revision. Patients were coded as having a disorder if it was documented in the medical record within 1 year before the index date. Past-year inpatient hospitalization and outpatient diagnoses were used to calculate the Charlson Comorbidity score (Charlson et al., 1987), a standardized technique that uses diagnostic codes to provide a measure of illness severity (Quan et al., 2002, 2005). The Charlson Comorbidity score predicts 1-year mortality for patients with 1 of 17 medical conditions associated with increased risk of dying such as heart disease, acquired immunodeficiency syndrome, or cancer. Higher scores are associated with greater medical comorbidity. We extracted VA medical utilization data on each patient for 5 years after the index date. VA medical service utilization was assessed using the following variables: the number of inpatient stays overall, the number of emergency department visits, and the number of days hospitalized on medical units, surgical units, and psychiatric units. We also collected data on the number of VA hospitals in the area where the patient received care and episodes of discharge against medical advice (AMA) (inpatient admissions with discharge dispositions coded as “irregular” were classified as an AMA discharge). Patients with AMA discharges are given limited supplies of medications, are more likely to leave without scheduled follow-up care, and often do not have the opportunity to receive patient education around illness, medication, and treatment. Data were collected from the Veterans’ Health Administration Vital Status File on subsequent patient deaths. Data were obtained on whether the patient had a behavioral flag in their medical record. Behavioral flags are alerts in the EMR that are displayed immediately upon opening a patient’s record (before seeing demographic or clinical data). Their purpose is to alert clinical staff of immediate critical information that directly relates to safety of the patient or others and would not otherwise be known. Examples of reasons for behavioral flags include a patient with a history of violence in the medical center, repeated violence against others, credible threats of harm against patients or staff, history of repeated disruptive behavior in the medical center, or a history of sexual harassment toward patients or staff. Patients with a behavioral flag often require a police escort during medical appointments.

Statistical Analyses All data were analyzed using IBM SPSS Statistics v. 19.0 software. To examine MUD and AUD differences in health care utilization and health outcomes, we used analysis of  C

2014 American Society of Addiction Medicine

Copyright © 2014 American Society of Addiction Medicine. Unauthorized reproduction of this article is prohibited.

J Addict Med r Volume 8, Number 1, January/February 2014 Health Service Use of Veterans with Methamphetamine Use Disorders

covariance (ANCOVA), logistic regression, and survival analysis techniques. Because patients who lived longer would have had more opportunity to experience certain health care utilization variables, and because age was potentially related to our outcomes of interest, we included both age and time until death as covariates in all relevant analyses, and adjusting for race and gender. We used ANCOVA to examine mean differences between groups on continuous outcome variables while accounting for the potentially confounding variables of gender, race, age, and time until death; these analyses included comparisons of frequency of medication utilization variables. We used logistic regression analysis adjusting for gender, race, age, and time until death to examine group differences on dichotomous outcome variables; these analyses included whether patients had a discharge from an inpatient AMA admission, received care at 3 or more different VA medical centers in the region, or had a behavioral flag in the medical record. Survival analysis was used to evaluate time until death between the 2 groups. Twenty patients in the MUD group also had clinically significant alcohol laboratory test results. These 20 patients did not significantly differ from the larger MUD group on any demographic or clinical variable. All analyses were run including the 20 participants and again with them excluded. The removal of these 20 participants did not impact the significance of the results or direction of effect in the analyses, and they were retained in the MUD group.

RESULTS Between January 2003 and December 2005, 2300 veterans who received health care at a VA Medical Center in the Pacific Northwest carried a current or past diagnosis of methamphetamine abuse or dependence. In 718 veterans, recent methamphetamine use was confirmed by urine drug screen—these veterans are the focus of our investigation. Veterans with MUD and laboratory-confirmed recent use were, in general, white, unmarried, middle-aged men, had modest levels of medical illness, with 41.5% having a service connected disability (a medical or mental illness causing impaired function that was identified originally during military service, or resulted from military service), and were mostly from the Vietnam War era (Table 1). Common comorbid medical diagnoses in veterans with MUD included hypertension (35.7%), hepatitis C (34.5%), and diabetes mellitus (12.3%). The prevalence of psychiatric comorbidity was substantial: 71% were diagnosed with a mood disorder, 33% with posttraumatic stress disorder, 26% with a psychotic spectrum disorder, and almost 6% with a suicide attempt (Table 2). In the same study time period, there were 25,042 veterans in this region with a diagnosis of alcohol abuse or dependence. We excluded 8846 patients who had medical record documentation of a current comorbid substance use disorder and 15,452 patients who did not have blood alcohol testing during the 5-year time period or whose blood alcohol testing showed a level of less than 80 mg/dl. The remaining 744 patients with an AUD who were legally intoxicated during the index medical encounter were retained for comparison with patients in the MUD group. In unadjusted analyses, patients with MUD were younger than patients with AUD, more likely to be female,  C

TABLE 1. Demographic Characteristics by Group*

Age Male sex Race Asian or Pacific Islander Black or African American White Unknown Ethnicity Not of Hispanic origin Hispanic Unknown Marital status Married Divorced Widowed Separated Never married Period of military service World War II Korean War era Vietnam War era Persian Gulf era Other VA service-connected Charlson Comorbidity Index

Methamphetamine Use Disorder (n = 718)

Alcohol Use Disorder (n = 744)

46.2 (8.2) 93.7% (673)

53.9 (9.9) 96.9% (721)

1.0% (7) 6.0% (43) 78.4% (563) 13.9% (100)

1.3% (10) 5.6% (42) 72.0% (536) 18.4% (137)

85.8% (616) 0.3% (2) 13.9% (100)

81.2% (604) 0.4% (3) 18.4% (137)

15.2% (109) 46.5% (334) 3.1% (22) 7.1% (51) 24.5% (176)

19.0% (141) 50.0% (372) 6.6% (49) 5.9% (44) 16.9% (126)

0

1.9% (14) 11.4% (85) 76.9% (572) 9.7% (72) 0.1% (1) 34.7% (258) 2.1 (2.5)

P

Comparison of health service use among veterans with methamphetamine versus alcohol use disorders.

Methamphetamine use disorders (MUD) are associated with severe health effects and psychiatric comorbidities, but little is known about the health care...
97KB Sizes 0 Downloads 0 Views