Analysis of longitudinal substance use outcomes using ordinal random-effects regression models

2012

The goal of this research project was to produce a model of the effects of drug dependence on general self-rated health. Due to power issues, two additional models, one for cocaine and one for heroin, were required. The models used data from the 2005-2009 National Survey on Drug Use and Health. The result of this effort was a ranking of the effects of drug dependence on general health for individuals and for the study population. The model controlled for infectious, chronic and mental illness as well as sociodemographic variables. Significantly increased odds ratios were found for alcohol, marijuana, analgesics, and cocaine at p < .001, and for heroin at p < .01. A ranking of odds ratios was constructed, but wide confidence intervals make the scale difficult to interpret and thus less useful for guiding policy.

The journal of mental health policy and economics, 2001

BACKGROUND: Health services researchers have increasingly used hazard functions to examine illness or treatment episode lengths and related treatment utilization and treatment costs. There has been little systematic hazard analysis, however, of mental health/substance abuse (MH/SA) treatment episodes. AIMS OF THE STUDY: This article uses proportional hazard functions to characterize multiple treatment episodes for a sample of insured clients with at least one alcohol or drug treatment diagnosis over a three-year period. It addresses the lengths and timing of treatment episodes, and the relationships of episode lengths to the types and locations of earlier episodes. It also identifies a problem that occurs when a portion of the sample observations is ǣpossibly censored. Failure to account for sample censoring will generate biased hazard function estimates, but treating all potentially censored observations as censored will overcompensate for the censoring bias. METHODS: Using insuran...

American Journal of Drug and Alcohol Abuse, 2015

Substance use disorders (addiction; abuse or dependence in DSM-IV terms) are chronic, relapsing conditions. Our current state of knowledge does not allow for accurate prediction of the trajectory of substance use patterns or development of the disorder in the general population, in a small subgroup (e.g. participants in a clinical trial), or in individual patients. At the population level, this limits our ability to efficiently allocate limited clinical, public health, and law enforcement resources to prevent the growth in use and abuse of existing and novel psychoactive substances (NPS). At the clinical trial level, this limits our ability to accurately identify factors influencing treatment outcome in subgroups of patients. At the patient level, this limits the clinician's ability to identify and promptly intervene during periods of heightened risk of relapse. Four articles in this issue of The American Journal of Drug and Alcohol Abuse by Wagner et al. (1), Westover et al. (2), Wakeland et al. (3) and Stogner ( ) describe statistical approaches that may improve the current state of the art, ranging from how to handle the raw data that is collected in order to extract the maximum information (in terms of level of aggregation and the best outcome distribution for modeling) (1) to the optimal model to use when evaluating (retrospectively or prospectively) substance use outcomes, whether in a circumscribed group of users (e.g. clinical trial participants) (2) or in the population as a whole (3,4). Overall, the examples in these papers show how sensitive the results of statistical analysis can be to the choice of data aggregation and modeling, highlighting the importance of using the best possible statistical approaches. (1) make several recommendations for improved handling of data on selfreported substance use to extract the maximum information. They illustrate their recommendations with data collected using the timeline follow-back (TLFB) method ( ), a semi-structured, calendar-based interview widely used in clinical studies. Their recommendations address two common issues with such data: treatment of substance use as a single variable, ignoring which specific substance(s) was used, and which mathematical

Drug and Alcohol Dependence, 2004

Data from a longitudinal cohort study were used to directly compare the concurrent and predictive validity of four univariate typologic approaches with a multivariate approach in subtyping drug dependence. The four univariate typologies were based upon: (a) age-of-onset of drug abuse/dependence, (b) presence of drug abuse in first-degree relatives, (c) presence of antisocial personality disorder, and (d) sex. The multivariate typologic approach was based on indices of vulnerability, chronicity, consequences, and psychopathology, yielding the Type A/B dichotomy first demonstrated in alcohol dependence. Subtypes generated from the univariate typologies were then each compared with the multivariate typology on measures of concurrent and predictive validity, and the strength of association was compared statistically. There was evidence of significantly greater concurrent validity of the Type A/B typology compared with the univariate typologies across all the domains of validation (risk, substance use, psychopathology, personality, and overall functioning). The multivariate typology also fared better than the univariate ones in all three domains on which predictive validity was evaluated: substance use, psychopathology, and overall functioning, as well as the degree of change in several composite scores (drug, medical, legal, and psychiatric) and the global psychiatric symptom index. This direct method of comparison seemed to demonstrate the superior validity of the multivariate cluster-analytic approach over the univariate approaches to classifying subjects with drug dependence.

The paper describes advances in statistical methods for prevention research with a particular focus on substance abuse prevention. Standard analysis methods are extended to the typical research designs and characteristics of the data collected in prevention research. Prevention research often includes longitudinal measurement, clustering of data in units such as schools or clinics, missing data, and categorical as well as continuous outcome variables. Statistical methods to handle these features of prevention data are outlined. Developments in mediation, moderation, and implementation analysis allow for the extraction of more detailed information from a prevention study. Advancements in the interpretation of prevention research results include more widespread calculation of effect size and statistical power, the use of confidence intervals as well as hypothesis testing, detailed causal analysis of research findings, and meta-analysis. The increased availability of statistical software has contributed greatly to the use of new methods in prevention research. It is likely that the Internet will continue to stimulate the development and application of new methods.

Developmental Psychology, 2009

Analyzing problem-behavior trajectories can be difficult. The data are generally categorical and often quite skewed, violating distributional assumptions of standard normal-theory statistical models. In this paper, we present several currently-available modeling options, all of which make appropriate distributional assumptions for the observed categorical data. Three are based on the generalized linear model: a hierarchical generalized linear model (HGLM), a growth mixture model (GMM), and a latent class growth analysis (LCGA). We also describe a longitudinal latent class analysis (LLCA), which requires fewer assumptions than the first three. Finally, we illustrate all of the models using actual longitudinal adolescent alcohol-use data. We guide the reader through the model-selection process, comparing the results in terms of convergence properties, fit and residuals, parsimony, and interpretability. Advances in computing and statistical software have made the tools for these types of analyses readily accessible to most researchers. Using appropriate models for categorical data will lead to more accurate and reliable results, and their application in real data settings could contribute to substantive advancements in the field of development and the science of prevention.

Journal of Data Science, 2021

Alcohol and drug uses are common in today's society and it is well-known that they can lead to serious consequences. Studies have been conducted in order, for example, to understand short-or long-term temporal processes of alcohol and drug uses. This paper discusses statistical modeling for joint analysis of alcohol and drug uses and several models and the corresponding estimation approaches are presented. The methods are applied to a prospective study of alcohol and drug uses on college freshmen, which motivated this investigation. The analysis results suggest that female subjects seem to have much less consequences of alcohol and drug uses than male subjects and the consequences of alcohol and drug uses decrease along with ages.

American journal of epidemiology, 1989

Longitudinal studies aimed at assessing the impact of interventions on disease risk factors often confront several statistical problems. These problems include 1) dependent variables measured by ordered categories, 2) numerous potentially relevant patterns of transition between outcome levels, 3) mixed units of analysis (e.g., assignment by social unit while theorizing in terms of individuals), 4) incomplete randomization, and 5) correlated estimates for successive occasions of longitudinal measurement. Longitudinal data on use of cigarettes, alcohol, and marijuana among adolescents (n = 1,244, complete data) from the Midwestern Prevention Project are used to demonstrate solutions to each of these problems: 1) a proportional odds regression model, 2) conditional logistic models of transitions with interactions between baseline level and intervention effect, 3) a logistic model estimated with linear regression methods on measures aggregated by social unit, 4) conditional and uncondit...

Health Services Research, 2010

Objective. To analyze the relationships between illicit drug use and three types of health services utilization: emergency room utilization, hospitalization, and medical attention required due to injury(s). Data. Waves 1 and 2 (11,253 males and 13,059 females) from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). Study Design. We derive benchmark estimates by employing standard cross-sectional data models to pooled waves of NESARC data. To control for potential bias due to time-invariant unobserved individual heterogeneity, we reestimate the relationships with fixed-effects models. Principal Findings. The cross-sectional data models suggest that illicit drug use is positively and significantly related to health services utilization in almost all specifications. Conversely, the only significant (po.05) relationships in the fixed-effects models are the odds of receiving medical attention for an injury and the number of injuries requiring medical attention for men, and the number of times hospitalized for men and women. Conclusions. Failing to control for time-invariant individual heterogeneity could lead to biased coefficients when estimating the effects of illicit drug use on health services utilization. Moreover, it is important to distinguish between types of drug user (casual versus heavy) and estimate gender-specific models.

Drug and Alcohol Dependence, 2007

Background: Severity measures for clients in substance abuse treatment programs are becoming increasingly important as funders adopt payment systems linked to agency performance. Recently, two severity measures based on administrative data have been developed. This study validated these measures using prospective data. Methods: Subjects were participants in the Drug Abuse Treatment Outcomes Study (adult or adolescent components) or the Substance Abuse and Mental Health Services Administration Medicaid Managed Behavioral Healthcare and Vulnerable Populations project (adult or adolescent chemical dependency components). Severity measures were calculated based on data obtained at entry into substance abuse treatment. The baseline severity measures were included along with age, gender, and race/ethnicity in logistic regression models predicting abstinence at follow-up for alcohol use, marijuana use, cocaine use, or heroin use. Results: For adults, the severity measures were highly statistically significant (p < 0.001) for all models in both data sets, indicating that adults with higher severity were more likely (and much more likely in many cases) to use alcohol, marijuana, cocaine, or heroin at the follow-up interview than were those with lower severity. For adolescents, the severity measure was highly statistically significant (p < 0.001) for marijuana in both data sets and for alcohol in the Medicaid data set. Conclusions: Baseline severity measures were powerful predictors of abstinence at follow-up. These measures, derived from routinely available electronic records, appear to have noteworthy predictive validity. The severity indicators can be used for administrative purposes such as riskadjustment when examining treatment agency performance.

JAMA Psychiatry, 2019

IMPORTANCE Limited empirical research has examined the extent to which cohort-level prevalence of substance use is associated with the onset of drug use and transitioning into greater involvement with drug use. OBJECTIVE To use cross-national data to examine time-space variation in cohort-level drug use to assess its associations with onset and transitions across stages of drug use, abuse, dependence, and remission. DESIGN, SETTING, AND PARTICIPANTS The World Health Organization World Mental Health Surveys carried out cross-sectional general population surveys in 25 countries using a consistent research protocol and assessment instrument. Adults from representative household samples were interviewed face-to-face in the community in relation to drug use disorders. The surveys were conducted between 2001 and 2015. Data analysis was performed from July 2017 to July 2018. MAIN OUTCOMES AND MEASURES Data on timing of onset of lifetime drug use, DSM-IV drug use disorders, and remission from these disorders was assessed using the Composite International Diagnostic Interview. Associations of cohort-level alcohol prevalence and drug use prevalence were examined as factors associated with these transitions. RESULTS Among the 90 027 respondents (48.1% [SE, 0.2%] men; mean [SE] age, 42.1 [0.1] years), 1 in 4 (24.8% [SE, 0.2%]) reported either illicit drug use or extramedical use of prescription drugs at some point in their lifetime, but with substantial time-space variation in this prevalence. Among users, 9.1% (SE, 0.2%) met lifetime criteria for abuse, and 5.0% (SE, 0.2%) met criteria for dependence. Individuals who used 2 or more drugs had an increased risk of both abuse (odds ratio, 5.17 [95% CI, 4.66-5.73]; P < .001) and dependence (odds ratio, 5.99 [95% CI, 5.02-7.16]; P < .001) and reduced probability of remission from abuse (odds ratio, 0.86 [95% CI, 0.76-0.98]; P = .02). Birth cohort prevalence of drug use was also significantly associated with both initiation and illicit drug use transitions; for example, after controlling for individuals' experience of substance use and demographics, for each additional 10% of an individual's cohort using alcohol, a person's odds of initiating drug use increased by 28% (odds ratio, 1.28 [95% CI, 1.26-1.31]). Each 10% increase in a cohort's use of drug increased individual risk by 12% (1.12 [95% CI, 1.11-1.14]). CONCLUSIONS AND RELEVANCE Birth cohort substance use is associated with drug use involvement beyond the outcomes of individual histories of alcohol and other drug use. This has important implications for understanding pathways into and out of problematic drug use.

Evaluation Review, 2008

Why do some people abuse drugs? Why are some people able to stop abusing drugs while others progress to an addiction that renders them unable to discontinue drug use in spite of dire consequences? What interventions can help people stop abusing drugs? Why do many people treated for drug abuse relapse? What interventions can reduce or prevent relapse? Addiction researchers struggle with these important questions-all of which require a longitudinal perspective. Researchers at the University of California at Los Angeles introduced the term "career" in reference to observed patterns in drug addiction behaviors and became one of the first groups to view addiction from a longitudinal perspective (McGlothlin, Anglin, & Wilson, 1978). Since then, the notion of a drug use pattern or career has been extended to observable patterns in treatment relapse (Hser, Anglin, Grella, Longshore, & Prendergast, 1997; Anglin, Hser., & Grella, 1997). Almost three decades later, the field has come to view addiction from the perspective of a chronic disease model (McLellan,

Background—This study aims to estimate the odds and predictors of Cannabis Use Disorders (CUD) relapse among individuals in remission. Methods—Analyses were done on the subsample of individuals with lifetime history of a CUD (abuse or dependence) who were in full remission at baseline (Wave 1) of the National Epidemiological Survey of Alcohol and Related Conditions (NESARC) (n=2350). Univariate logistic regression models and hierarchical logistic regression model were implemented to estimate odds of relapse and identify predictors of relapse at 3 years follow up (Wave 2). Results—The relapse rate of CUD was 6.63% over an average of 3.6 year follow-up period. In the multivariable model, the odds of relapse were inversely related to time in remission, whereas having a history of conduct disorder or a major depressive disorder after Wave 1 increased the risk of relapse.

BMC medical research methodology, 2014

The reduction of crime is an important outcome of opioid maintenance treatment (OMT). Criminal intensity and treatment regimes vary among OMT patients, but this is rarely adjusted for in statistical analyses, which tend to focus on cohort incidence rates and rate ratios. The purpose of this work was to estimate the relationship between treatment and criminal convictions among OMT patients, adjusting for individual covariate information and timing of events, fitting time-to-event regression models of increasing complexity. National criminal records were cross linked with treatment data on 3221 patients starting OMT in Norway 1997-2003. In addition to calculating cohort incidence rates, criminal convictions was modelled as a recurrent event dependent variable, and treatment a time-dependent covariate, in Cox proportional hazards, Aalen&#39;s additive hazards, and semi-parametric additive hazards regression models. Both fixed and dynamic covariates were included. During OMT, the number...

Alcohol and Alcoholism, 2007

Aims-Individuals in treatment for alcohol use disorders are more likely to die from cigarette use than from alcohol consumption. Advanced statistical methodologies that increase study power and clinical relevance have been advocated to examine the timevarying nature of substance use relapse and abstinence, including drinking and smoking. The purpose of this investigation was to examine timevarying factors that are associated with smoking cessation among smokers in the general population, including alcohol use, self-efficacy, and depression, to determine if they were also related to smoking cessation during and after treatment for alcohol use disorders.