Abstract
Aims:
To identify factors associated with early and sustained cessation of heavy drinking.
Design:
Retrospective cohort study over 5 years.
Setting:
Kaiser Permanente Northern California, USA.
Participants:
Adults reporting heavy drinking during primary care-based alcohol screening between 06/01/2013 and 05/31/2014. The sample (n=85,434) was 41% female and 34% non-White; mean age was 50.3 years (SD=18.1).
Measurements:
Following U.S. guidelines, early and sustained cessation of heavy drinking was defined as reporting lower-risk drinking or abstinence at 1 year and through 5 years after achieving early cessation, respectively. Associations between patient characteristics and service use and cessation outcomes were examined using logistic regression with inverse probability weights addressing attrition.
Findings:
Nearly two-thirds of participants achieved early cessation of heavy drinking. Women (odds ratio (OR)=1.39, 95% confidence interval (CI)=1.35, 1.44), middle-age (35-64 years: ORs=1.16 to 1.19), non-White race/ethnicity (ORs=1.31 to 1.57), medical conditions (OR=1.05, 95% CI=1.04, 1.06), psychiatric (OR=1.10, 95% CI=1.06, 1.15) and drug use disorders (OR=1.35, 95% CI=1.17, 1.56), and addiction treatment (OR=1.19, 95% CI=1.09, 1.30) were associated with higher odds of early cessation, while older age (≥65 years: OR=0.91, 95% CI=0.86, 0.96), smoking (OR=0.81, 95% CI=0.77, 0.84), higher index drinking levels (exceeding both daily and weekly limits: OR=0.30, 95% CI=0.29, 0.32), and psychiatric treatment (OR=0.91, 95% CI=0.84, 0.99) were associated with lower odds. Among those who achieved early cessation (n=19,200), about 60% sustained cessation. Associations between patient factors and sustained cessation paralleled those observed in analyses of early cessation. Additionally, routine primary care (OR=1.57, 95% CI=1.44, 1.71) and addiction treatment post-1 year (OR=1.41, 95% CI=1.19, 1.66) were associated with higher odds of sustained cessation. Lower-risk drinking versus abstinence at 1-year was associated with lower odds of sustained cessation (OR=0.62, 95% CI=0.57, 0.66).
Conclusions:
Nearly two-thirds of a large, diverse sample of patients who reported heavy drinking in a California USA healthcare system achieved early and sustained cessation of heavy drinking. Vulnerable subgroups (i.e., non-White patients and those with psychiatric disorders), patients who received routine primary care, and those who received addiction treatment were more likely to sustain cessation of heavy drinking than other participants.
Introduction
Heavy alcohol use is a major risk factor for chronic illness and mortality[1,2]. While heavy drinking in the United States (U.S.) is common [prevalence of 6% to 28%, depending on the definition and whether alcohol use disorder (AUD) is included][3–5], cessation of heavy drinking and recovery are possible. From U.S. general population surveys, the prevalence of recovery from AUD was 34% in 2012-2013[6], and the prevalence of heavy drinking cessation was 27% in 2004-2005[7].
Relatively little is known about who achieves heavy drinking cessation and for how long. Moreover, most studies have focused on AUD recovery among alcohol treatment populations, making it difficult to examine the role of treatment and generalize findings to the wider population that drinks heavily but does not meet criteria for an AUD. Additionally, studies have differed in recovery definitions, with more recent studies including asymptomatic, lower-risk drinking in addition to abstinence, given evidence that it can be a suitable goal for some individuals[8–10]. Consistent predictors of AUD recovery include female sex, older age, and lower severity drinking and AUD symptoms[6,11–18]. Other predictors of recovery include health service use (alcohol treatment[14,19], continuing care[20]), not smoking[14,15,21], better self-control[22], and having supportive relationships such as marriage[6,11].
Studies evaluating predictors of heavy drinking cessation among individuals with and without AUDs are sparse, though findings are consistent with those on AUD recovery. A recent U.S. population-based study found that older age, female sex, Black or Hispanic race/ethnicity, no AUD, no tobacco or drug use, alcohol treatment, and having children were associated with higher odds of heavy drinking cessation[7]. Additional evidence comes from latent class or trajectory analyses, which describe longitudinal drinking patterns, but are difficult for inferring categorical changes in drinking, such as heavy drinking cessation or recovery[23–27]. More studies of heavy drinking cessation are needed to understand service needs and inform targeted interventions.
The aim of this exploratory study was to identify predictors of early and sustained cessation of heavy drinking over 5 years among an adult primary care population within an integrated health care system where comprehensive data are collected in the electronic health record (EHR). Key factors of interest included sociodemographic characteristics, alcohol and tobacco use, medical and psychiatric diagnoses, and health service utilization, including routine primary care and specialty treatment. The ability to examine a multitude of diagnoses and different types of health service in association with heavy drinking cessation is novel. Findings may help inform future research on mechanisms of heavy drinking cessation (e.g., examining how routine primary care services can influence recovery), which would help clinicians and health systems design new (or tailor existing) interventions.
Methods
This manuscript followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cohort studies[28].
Setting
Kaiser Permanente Northern California (KPNC) serves 4.3 million members, approximately one-third of the Northern Californian population. The membership is racially diverse and reflects the U.S. population with access to care[29]. Membership includes enrollees from Medicaid, Medicare, employer-based plans, and health insurance exchanges. The KPNC Institutional Review Board reviewed the study and granted a waiver of informed consent to examine EHR data.
Since June 2013, KPNC has systematically screened for heavy drinking in adult primary care[30], and has maintained an >80% average screening rate. Medical assistants ask patients 3 questions about their alcohol use, including a modified version of the evidence-based National Institute on Alcohol Abuse and Alcoholism (NIAAA) single-item screening question to determine the number of heavy drinking days: 1) “How many times in the past three months have you had 5 or more drinks containing alcohol in a day?” (for men aged 18-65 years), or “4 or more drinks” (for women ≥18 years and men ≥66 years)[3]; and two questions to calculate average drinks consumed per week: 2) “On average, how many days per week do you have an alcoholic drink?”; and 3) “On a typical drinking day, how many drinks do you have?”
Study Design and Sample
We conducted a retrospective cohort study using observational, longitudinal EHR-based data from the KPNC Adult Alcohol Registry[31]. We identified 200,458 adults (≥18 years) who screened positive for heavy drinking between June 1, 2013 and May 31, 2014; the index date was the first date during this period where the patient screened positive for heavy drinking. For each patient, all alcohol screenings after the index date through 63 months post-index were extracted. The screening closest to 12 months post-index (between 6 and 15 months) was chosen as the 1-year follow-up screening (1Y). The screening closest to 60 months post-index (between 54 and 63 months) was chosen as the 5-year follow-up screening (5Y).
Cessation Outcomes
Definitions of heavy drinking cessation were based on the alcohol consumption component of the recent NIAAA Recovery from AUD definition[32]. Heavy drinking was defined as exceeding either the daily (≥5 drinks/day for men 18-65 years, or ≥4 drinks/day for men ≥66 years and women) or weekly limits (>14 drinks/week for men 18-65 years, or >7 drinks/day for men ≥66 years and women) from the NIAAA drinking guidelines[3]. Early cessation of heavy drinking was defined as reporting lower-risk drinking (no heavy drinking days in the past 3 months and drinking within weekly limits) or abstinence at 1Y. Sustained cessation of heavy drinking was defined as achieving early cessation and reporting lower-risk drinking or abstinence at all screenings through 5Y.
Predictors
We extracted EHR-based data from the KPNC Adult Alcohol Registry to create the following variables of interest[31]:
Sociodemographic characteristics.
Variables included sex, age, race/ethnicity, and household income. Household income was estimated by geocoding patients’ most recent residential address in the year before index to 2010 U.S. Census blocks and using the median household income of the census block[33]. Household income was categorized into tertiles of the overall distribution (low: ≤$61,765, middle: $61,766-$89,952, and high: ≥$89,953).
Smoking status.
Based on the most recent tobacco screening in the year before the index date, patients were classified as whether they currently or never/formerly smoked.
Index drinking level.
We categorized patients into 3 mutually exclusive groups based on NIAAA guidelines: exceeding only daily limits, exceeding only weekly limits, and exceeding both daily and weekly limits[3].
Baseline diagnoses.
Using International Classification of Diseases, 9th Edition (ICD-9) codes in the year before the index date, we identified patients with chronic medical conditions prevalent in primary care[34], psychiatric disorders, non-remission AUD, and non-remission drug use disorder diagnoses[31]. More details can be found in Palzes et al. (2020).
Health service utilization.
Based on prior research[20], we evaluated health service utilization in the year (1Y) after index, and for those who achieved early cessation, also over 1Y to 5Y. An indicator for receiving addiction treatment was defined as receiving an AUD diagnosis and either having ≥1 outpatient visits in addiction medicine or receiving pharmacotherapy for AUD (i.e., acamprosate, disulfiram, naltrexone, or topiramate)[35,36] from an outpatient KPNC pharmacy, during the respective follow-up period (1=AUD and received treatment, or no AUD; 0=AUD and did not receive treatment). An indicator for receiving psychiatric treatment was defined as receiving a psychiatric diagnosis and either having ≥1 outpatient visits in psychiatry or filling a prescription for an antidepressant, mood stabilizer, sedative, benzodiazepine, other anxiolytic, or antipsychotic (1=psychiatric disorder and received treatment, or no psychiatric disorder; 0=psychiatric disorder and did not receive treatment)[37–39]. An indicator for routine primary care was defined as having an annual outpatient office or telephone visit in the post-1Y period.
Type of early cessation.
Alcohol use reported at 1Y was categorized into lower-risk drinking or abstinence based on NIAAA drinking guidelines[3].
Control variables.
To account for different follow-up times, we included the number of alcohol screenings and follow-up days (<2 years or ≥2 years) between 1Y to 5Y as control variables.
Missing Data
Proportions of patients missing covariate data were low, with 0.1% (n=285) having unknown household income and 1.6% (n=3,249) having unknown smoking status. While there are many possible reasons why patients could be missing household income and smoking status data (e.g., geocoding failure, medical assistants forgetting to screen for smoking), we expect that data is missing at random and is not associated with the outcomes. Since the proportion of incomplete cases was low (<5% [43]), we imputed missing household income values as the mean ($79,400) and missing smoking status as the mode (never or former) among the full sample of patients identified with heavy drinking (n=200,458).
Analytical Samples
Two analytical samples were generated (Figure S1):
Early cessation of heavy drinking.
Among the 200,458 patients identified with heavy drinking, we excluded patients with incomplete index screenings (n=12,136) and patients with <70% KPNC membership in the year before the index date (n=31,133), resulting in 157,189 patients. Among these patients, 71,755 (45.6%) were lost to follow-up due to death (n=370), disenrollment from KPNC (n=8,159), or not having a 1Y screening (n=63,226), resulting in an analytical sample of 85,434 (54.4%) patients with observed outcome data.
Sustained cessation of heavy drinking.
Since sustained cessation of heavy drinking is conditional on achieving early cessation of heavy drinking[32], all patients that achieved early cessation were included with no further exclusion criteria applied (n=53,916). Among this sample, 34,716 (64.4%) patients were lost to follow-up due to death (n=2,267), disenrollment from KPNC (n=11,572), not having a 5Y screening (n=19,707), or having either <70% membership coverage between the index date and 5Y screening or no follow-up screenings between the 1Y and 5Y screening (n=1,170), resulting in an analytical sample of 19,200 (35.6%) patients with observed outcome data.
Inverse Probability of Attrition Weights (IPAW)
Given the large proportions of patients lost to follow-up, we examined potential selection bias due to attrition by calculating standardized differences for patient characteristics between the subgroups retained and lost to follow-up, and found that age, index drinking level, and the number of chronic medical conditions meaningfully differed (standardized difference >0.2 [41,42]) between the subgroups retained and lost to follow-up in both analytical samples (Tables S1–S2).
We next estimated and applied inverse probability of attrition weights (IPAW) for each observation in models of early and sustained cessation. In both, the IPAW were estimated based on predicted probabilities of no attrition through logistic regression models, conditional on a set of covariates that were hypothesized to be associated with attrition and the outcome. For the analysis on early cessation (n=157,189), the IPAW model was fit using baseline patient characteristics. For the analysis on sustained cessation (n=53,916), the IPAW model was fit on the same set of covariates, plus receiving addiction and psychiatric treatment in 1Y after index and the type of early cessation. The distribution of the weights was reasonable for both models of early cessation (mean=1.84, median=1.84, min=1.04, max=3.41) and sustained cessation (mean=2.81, median=2.46, min=1.82, max=6.32), and they were not unacceptably unstable according to guidelines by Seaman & White (2013)[43].
Statistical Analysis
For analyses of early cessation of heaving drinking, we fit a multivariable logistic regression model among the analytical sample with observed outcome data (n=85,434), weighted by IPAW, including all baseline patient characteristics (sex, age group, race/ethnicity, household income, smoking status, index drinking level, AUD, number of chronic medical conditions, any psychiatric disorder, and any drug use disorder) and health service utilization (addiction and psychiatric treatment) in 1Y after index. Variance was estimated using the conservative robust sandwich standard error estimator approach[44]. We also examined whether there was effect measure modification of addiction treatment (in 1Y after index) by sex, age, race/ethnicity, and index drinking level by including an interaction term in separate models and evaluating whether the Type 3 test was statistically significant at p<0.05.
For analyses of sustained cessation of heavy drinking, we fit a multivariable logistic regression model among the analytical sample who achieved early cessation and had observed outcome data (n=19,200), weighted by IPAW, including all baseline patient characteristics, addiction and psychiatric treatment in 1Y after index, and post-1Y variables (type of early cessation, number of alcohol screenings during follow-up, number of days of follow-up, routine primary care, addiction and psychiatric treatment). We also examined whether there was effect measure modification of addiction treatment (post-1Y) by sex, age, race/ethnicity, and the type of early cessation (e.g., abstinence, lower-risk drinking), and whether AUD was an effect measure modifier for type of early cessation. We utilized the same methods for estimating variance and testing for effect measure modification as described above.
All statistical analyses were conducted using SAS software, Version 9.4 (SAS Institute Inc., Cary, NC). The analysis plan was not pre-registered, thus results should be considered exploratory.
Results
Early Cessation of Heavy Drinking
The analytical sample with observed outcome data consisted of 85,434 patients with heavy drinking. About two-fifths (n=34,761) were female; 56,581 (66.2%) were White, 7,186 (8.4%) were Asian, Native Hawaiian or Pacific Islander, 5,096 (6.0%) were Black, 14,184 (16.6%) were Latinx/Hispanic, and 1,008 (1.2%) were Native American (Table 1). The mean age was 50.3 years (SD=18.1) and 13.9% (n=11,914) reported current smoking. Based on the index alcohol screening, 50,193 (58.8%) patients exceeded only daily limits, 20,835 (24.4%) exceeded only weekly limits, and 14,406 (16.9%) exceeded both daily and weekly limits. There were 4,057 (4.7%) patients with an AUD.
Table 1.
Baseline characteristics and health service utilization of the analytical sample of patients who reported heavy drinking between 6/1/2013 and 5/31/2014 in Kaiser Permanente Northern California for examining predictors of early cessation of heavy drinking.a,b
| Achieved early cessation of heavy drinking |
|||
|---|---|---|---|
| Characteristic | Overall (n=85,434) |
Yes (n=53,916) |
No (n=31,518) |
| Baseline characteristics | |||
| Sex, n (%) | |||
| Male | 50,673 (59.3) | 30,704 (56.9) | 19,969 (63.4) |
| Female | 34,761 (40.7) | 23,212 (43.1) | 11,549 (36.6) |
| Age, mean (SD) | 50.3 (18.1) | 48.9 (17.8) | 52.7 (18.3) |
| Age group, n (%) | |||
| 18-34 | 21,249 (24.9) | 14,267 (26.5) | 6,982 (22.2) |
| 35-49 | 19,303 (22.6) | 13,183 (24.5) | 6,120 (19.4) |
| 50-64 | 20,969 (24.5) | 13,776 (25.6) | 7,193 (22.8) |
| ≥65 | 23,913 (28.0) | 12,690 (23.5) | 11,223 (35.6) |
| Race/ethnicity, n (%) | |||
| White | 56,581 (66.2) | 33,520 (62.2) | 23,061 (73.2) |
| Asian, Native Hawaiian, or Pacific Islander | 7,186 (8.4) | 5,173 (9.6) | 2,013 (6.4) |
| Black | 5,096 (6.0) | 3,762 (7.0) | 1,334 (4.2) |
| Latinx/Hispanic | 14,184 (16.6) | 9,913 (18.4) | 4,271 (13.6) |
| Native American | 1,008 (1.2) | 664 (1.2) | 344 (1.1) |
| Other/Unknown | 1,379 (1.6) | 884 (1.6) | 495 (1.6) |
| Household income, n (%) | |||
| Low (≤$61,765) | 26,882 (31.5) | 17,567 (32.6) | 9,315 (29.6) |
| Middle ($61,766-$89,952) | 28,672 (33.6) | 18,264 (33.9) | 10,408 (33.0) |
| High (≥$89,953) | 29,880 (35.0) | 18,085 (33.5) | 11,795 (37.4) |
| Smoking status, n (%) | |||
| Never or former | 73,520 (86.1) | 46,918 (87.0) | 26,602 (84.4) |
| Current | 11,914 (13.9) | 6,998 (13.0) | 4,916 (15.6) |
| Index drinking level, n (%) | |||
| Exceeding only daily limits | 50,193 (58.8) | 36,682 (68.0) | 13,511 (42.9) |
| Exceeding only weekly limits | 20,835 (24.4) | 11,060 (20.5) | 9,775 (31.0) |
| Exceeding both limits | 14,406 (16.9) | 6,174 (11.5) | 8,232 (26.1) |
| Alcohol use disorder (AUD), n (%) | 4,057 (4.7) | 2,194 (4.1) | 1,863 (5.9) |
| Number of chronic medical conditions, n (%) | |||
| 0 | 29,835 (34.9) | 19,583 (36.3) | 10,252 (32.5) |
| 1 | 17,095 (20.0) | 11,018 (20.4) | 6,077 (19.3) |
| 2 | 11,815 (13.8) | 7,131 (13.2) | 4,684 (14.9) |
| ≥3 | 26,689 (31.2) | 16,184 (30.0) | 10,505 (33.3) |
| Any psychiatric disorder, n (%) | 15,078 (17.6) | 9,927 (18.4) | 5,151 (16.3) |
| Anxiety disorder | 8,219 (9.6) | 5,519 (10.2) | 2,700 (8.6) |
| Mood disorder | 9,730 (11.4) | 6,419 (11.9) | 3,311 (10.5) |
| Psychoses | 281 (0.3) | 205 (0.4) | 76 (0.2) |
| Any drug use disorder | 1,148 (1.3) | 801 (1.5) | 347 (1.1) |
| Cannabis | 551 (0.6) | 368 (0.7) | 183 (0.6) |
| Cocaine | 96 (0.1) | 75 (0.1) | 21 (0.1) |
| Opioid | 236 (0.3) | 188 (0.3) | 48 (0.2) |
| Other drug | 373 (0.4) | 273 (0.5) | 100 (0.3) |
| Stimulant | 118 (0.1) | 97 (0.2) | 21 (0.1) |
| Health service utilization in 1Y after index | |||
| Addiction treatment, n (%) | |||
| AUD and did not receive treatment | 3,051 (3.6) | 1,552 (2.9) | 1,499 (4.8) |
| AUD and received treatment, or no AUD | 82,383 (96.4) | 52,364 (97.1) | 30,019 (95.2) |
| AUD and received outpatient care | 1,130 (1.3) | 791 (1.5) | 339 (1.1) |
| AUD and received pharmacotherapy | 144 (0.2) | 111 (0.2) | 33 (0.1) |
| No AUD | 81,224 (95.1) | 51,559 (95.6) | 29,665 (94.1) |
| Psychiatric treatment, n (%) | |||
| Psychiatric disorder and did not receive treatment | 3,116 (3.6) | 2,094 (3.9) | 1,022 (3.2) |
| Psychiatric disorder and received treatment, or no psychiatric disorder | 82,318 (96.4) | 51,822 (96.1) | 30,496 (96.8) |
| Psychiatric disorder and received outpatient care | 4,816 (5.6) | 3,388 (6.3) | 1,428 (4.5) |
| Psychiatric disorder and received pharmacotherapy | 12,864 (15.1) | 8,391 (15.6) | 4,473 (14.2) |
| No psychiatric disorder | 68,577 (80.3) | 42,815 (79.4) | 25,762 (81.7) |
Heavy drinking was defined as exceeding either daily or weekly drinking limits recommended by the NIAAA (≥5 drinks/day or >14 drinks/week for men 18-65 years, or ≥4 drinks/day or >7 drinks/week for men >65 years).
Early cessation of heavy drinking was defined as reporting lower-risk drinking (no heavy drinking days in the past 3 months and drinking within weekly limits) or abstinence at 1Y.
A total of 53,916 (63.1%) patients achieved early cessation of heavy drinking (Table 1); of them, 25,289 (46.9%) reported abstinence and 28,627 (53.1%) reported lower-risk drinking. From the main effects model weighted by IPAW (Table 2), baseline characteristics associated with higher odds of early cessation were female versus male, middle age (35-64 years) versus younger age (18-34 years), and low versus middle household income. All non-White races/ethnicities were associated with higher odds of early cessation, with Black patients having the highest relative likelihood compared to White patients (OR=1.57, 95% CI=1.46, 1.68). Having more chronic medical conditions or a psychiatric or drug use disorder were associated with higher odds of early cessation. Receiving addiction treatment in 1Y after index was associated with higher odds of early cessation (OR=1.19, 95% CI=1.09, 1.30). Older (≥65 years) versus younger age (18-34 years), high versus middle household incomes, smoking, higher index drinking levels, and receiving psychiatric treatment in 1Y after index were associated with lower odds of early cessation. Having an AUD was not associated with early cessation.
Table 2.
Inverse probability of attrition weighted (IPAW) associations of baseline and follow-up patient characteristics and utilization with early cessation of heavy drinking, n=85,434.a
| Characteristic | OR (95% CI) | p-value |
|---|---|---|
| Baseline characteristics | ||
| Female (reference group: male) | 1.39 (1.35, 1.44) | <.001 |
| Age group (reference group: 18-34 years) | ||
| 35-49 years | 1.16 (1.11, 1.21) | <.001 |
| 50-64 years | 1.19 (1.14, 1.25) | <.001 |
| ≥65 years | 0.91 (0.86, 0.96) | <.001 |
| Race/ethnicity (reference group: White) | ||
| Asian, Native Hawaiian, or Pacific Islander | 1.46 (1.38, 1.55) | <.001 |
| Black | 1.57 (1.46, 1.68) | <.001 |
| Latinx/Hispanic | 1.32 (1.27, 1.38) | <.001 |
| Native American | 1.31 (1.14, 1.51) | <.001 |
| Other/Unknown | 1.03 (0.91, 1.16) | 0.642 |
| Household income (reference group: middle) | ||
| Low | 1.06 (1.02, 1.10) | 0.005 |
| High | 0.91 (0.88, 0.94) | <.001 |
| Smoking status (reference group: never or former) | 0.81 (0.77, 0.84) | <.001 |
| Index drinking level (reference group: exceeding only daily limits) | ||
| Exceeding only weekly limits | 0.45 (0.43, 0.46) | <.001 |
| Exceeding both daily and weekly limits | 0.30 (0.29, 0.32) | <.001 |
| Alcohol use disorder | 0.94 (0.87, 1.02) | 0.155 |
| Number of chronic medical conditions | 1.05 (1.04, 1.06) | <.001 |
| Any psychiatric disorder | 1.10 (1.06, 1.15) | <.001 |
| Any drug use disorder | 1.35 (1.17, 1.56) | <.001 |
| Utilization in 1Y after index | ||
| Addiction treatment | 1.19 (1.09, 1.30) | <.001 |
| Psychiatric treatment | 0.91 (0.84, 0.99) | 0.025 |
Early cessation of heavy drinking was defined as reporting lower-risk drinking (no heavy drinking days in the past 3 months and drinking within weekly limits) or abstinence at 1Y.
Index drinking level moderated the association between receiving addiction treatment in 1Y after index and early cessation of heavy drinking (Figure 1). Among patients who reported exceeding only daily limits, those who received addiction treatment in 1Y after index, compared to those who did not, had higher odds of early cessation (OR=1.50, 95% CI=1.29,1.75). Receiving addiction treatment in 1Y after index was not associated with early cessation among those who reported exceeding only weekly limits (OR=1.03, 95% CI=0.90,1.19) or both daily and weekly limits (OR=1.10, 95% CI=0.97, 1.26). We did not find evidence of effect measure modification of addiction treatment (in 1Y after index) by sex, age group, and race/ethnicity.
Figure 1.
ORs and 95% CIs of early cessation of heavy drinking, comparing patients who received addiction treatment in 1Y after index to those who did not, by index drinking level (>Daily limits: exceeding only daily limits, >Weekly limits: exceeding only weekly limits, >Both limits: exceeding both daily and weekly limits).
Sustained Cessation of Heavy Drinking
Sustained cessation of heavy drinking was examined among the subgroup that achieved early cessation and had observed outcome data (n=19,200). A total of 11,539 (60.0%) patients sustained cessation of heavy drinking over 5 years (Table 3). From the main effects model weighted by IPAW (Table 4), baseline characteristics associated with higher odds of sustained cessation of heavy drinking were female versus male and middle-to-older (≥35 years) versus younger age (18-34 years). Asian, Native Hawaiian, or Pacific Islander, Latinx/Hispanic, and Native American patients had higher odds of sustained cessation compared to White patients, but no association was observed among Black patients. Having more chronic medical conditions or a psychiatric disorder were associated with higher odds of sustained cessation. Routine primary care (OR=1.57, 95% CI=1.44, 1.71) and receiving addiction treatment post-1Y (OR=1.41, 95% CI=1.19,1.66) were associated with higher odds of sustained cessation. Higher index drinking levels, smoking, and lower-risk drinking (versus abstinence) at 1Y were associated with lower odds of sustained cessation. Household income, AUD, drug use disorders, receiving addiction treatment in 1Y after index, and receiving psychiatric treatment in 1Y after index were not associated with sustained cessation.
Table 3.
Baseline and follow-up patient characteristics and health service utilization of the analytical sample of patients who achieved early cessation of heavy drinking for examining predictors of sustained cessation of heavy drinking.a,b
| Sustained cessation of heavy drinking |
|||
|---|---|---|---|
| Characteristic | Overall (n=19,200) |
Yes (n=11,539) |
No (n=7,661) |
| Baseline characteristics | |||
| Sex, n (%) | |||
| Male | 11,039 (57.5) | 6,328 (54.8) | 4,711 (61.5) |
| Female | 8,161 (42.5) | 5,211 (45.2) | 2,950 (38.5) |
| Age, mean (SD) | 53.3 (16.8) | 52.0 (16.8) | 55.2 (16.8) |
| Age group, n (%) | |||
| 18-34 | 3,253 (16.9) | 2,106 (18.3) | 1,147 (15.0) |
| 35-49 | 4,345 (22.6) | 2,838 (24.6) | 1,507 (19.7) |
| 50-64 | 5,664 (29.5) | 3,456 (30.0) | 2,208 (28.8) |
| ≥65 | 5,938 (30.9) | 3,139 (27.2) | 2,799 (36.5) |
| Race/ethnicity, n (%) | |||
| White | 12,252 (63.8) | 6,865 (59.5) | 5,387 (70.3) |
| Asian, Native Hawaiian, or Pacific Islander | 1,749 (9.1) | 1,252 (10.9) | 497 (6.5) |
| Black | 1,332 (6.9) | 862 (7.5) | 470 (6.1) |
| Latinx/Hispanic | 3,387 (17.6) | 2,243 (19.4) | 1,144 (14.9) |
| Native American | 243 (1.3) | 159 (1.4) | 84 (1.1) |
| Other/Unknown | 237 (1.2) | 158 (1.4) | 79 (1.0) |
| Household income, n (%) | |||
| Low (≤$61,765) | 5,975 (31.1) | 3,695 (32.0) | 2,280 (29.8) |
| Middle ($61,766-$89,952) | 6,489 (33.8) | 3,861 (33.5) | 2,628 (34.3) |
| High (≥$89,953) | 6,736 (35.1) | 3,983 (34.5) | 2,753 (35.9) |
| Smoking status, n (%) | |||
| Never or former | 17,091 (89.0) | 10,313 (89.4) | 6,778 (88.5) |
| Current | 2,109 (11.0) | 1,226 (10.6) | 883 (11.5) |
| Index drinking level, n (%) | |||
| Exceeding only daily limit | 12,181 (63.4) | 8,462 (73.3) | 3,719 (48.5) |
| Exceeding only weekly limit | 4,674 (24.3) | 2,157 (18.7) | 2,517 (32.9) |
| Exceeding both limits | 2,345 (12.2) | 920 (8.0) | 1,425 (18.6) |
| Alcohol use disorder (AUD), n (%) | 736 (3.8) | 372 (3.2) | 364 (4.8) |
| Number of chronic medical conditions, n (%) | |||
| 0 | 5,309 (27.7) | 3,327 (28.8) | 1,982 (25.9) |
| 1 | 3,838 (20.0) | 2,357 (20.4) | 1,481 (19.3) |
| 2 | 2,973 (15.5) | 1,718 (14.9) | 1,255 (16.4) |
| ≥3 | 7,080 (36.9) | 4,137 (35.9) | 2,943 (38.4) |
| Any psychiatric disorder, n (%) | 3,577 (18.6) | 2,202 (19.1) | 1,375 (17.9) |
| Anxiety disorder | 2,002 (10.4) | 1,249 (10.8) | 753 (9.8) |
| Mood disorder | 2,354 (12.3) | 1,471 (12.7) | 883 (11.5) |
| Psychoses | 71 (0.4) | 37 (0.3) | 34 (0.4) |
| Any drug use disorder, n (%) | 238 (1.2) | 151 (1.3) | 87 (1.1) |
| Cannabis | 101 (0.5) | 60 (0.5) | 41 (0.5) |
| Cocaine | 19 (0.1) | 8 (0.1) | 11 (0.1) |
| Opioid | 62 (0.3) | 50 (0.4) | 12 (0.2) |
| Other drug | 86 (0.4) | 65 (0.6) | 21 (0.3) |
| Stimulant | 32 (0.2) | 21 (0.2) | 11 (0.1) |
| Health service utilization in 1Y after index | |||
| Addiction treatment, n (%) | |||
| AUD and did not receive treatment | 542 (2.8) | 278 (2.4) | 264 (3.4) |
| AUD and received treatment, or no AUD | 18,658 (97.2) | 11,261 (97.6) | 7,397 (96.6) |
| AUD and received outpatient care | 243 (1.3) | 157 (1.4) | 86 (1.1) |
| AUD and received pharmacotherapy | 20 (0.1) | 11 (0.1) | 9 (0.1) |
| No AUD | 18,413 (95.9) | 11,104 (96.2) | 7,309 (95.4) |
| Psychiatric treatment, n (%) | |||
| Psychiatric disorder and did not receive treatment | 705 (3.7) | 442 (3.8) | 263 (3.4) |
| Psychiatric disorder and received treatment, or no psychiatric disorder | 18,495 (96.3) | 11,097 (96.2) | 7,398 (96.6) |
| Psychiatric disorder and received outpatient care | 1,125 (5.9) | 720 (6.2) | 405 (5.3) |
| Psychiatric disorder and received pharmacotherapy | 2,978 (15.5) | 1,824 (15.8) | 1,154 (15.1) |
| No psychiatric disorder | 15,325 (79.8) | 9,147 (79.2) | 6,178 (80.6) |
| Post-1Y characteristics and utilization | |||
| Type of early cessation, n (%) | |||
| Abstinence | 8,561 (44.6) | 5,759 (49.9) | 2,802 (36.6) |
| Lower-risk drinking | 10,639 (55.4) | 5,780 (50.1) | 4,859 (63.4) |
| Number of follow-up alcohol screenings | |||
| ≤3 | 14,070 (73.3) | 9,737 (84.4) | 4,333 (56.6) |
| >3 | 5,130 (26.7) | 1,802 (15.6) | 3,328 (43.4) |
| Follow-up duration | |||
| <2 years | 8,183 (42.6) | 5,128 (44.4) | 3,055 (39.9) |
| ≥2 years | 11,017 (57.4) | 6,411 (55.6) | 4,606 (60.1) |
| Routine primary care, n (%) | 15,038 (78.3) | 9,020 (78.2) | 6,018 (78.6) |
| Addiction treatment, n (%) | |||
| AUD and did not receive treatment | 909 (4.7) | 399 (3.5) | 510 (6.7) |
| AUD and received treatment, or no AUD | 18,291 (95.3) | 11,140 (96.5) | 7,151 (93.3) |
| AUD and received outpatient care | 541 (2.8) | 289 (2.5) | 252 (3.3) |
| AUD and received pharmacotherapy | 149 (0.8) | 82 (0.7) | 67 (0.9) |
| No AUD | 17,729 (92.3) | 10,839 (93.9) | 6,890 (89.9) |
| Psychiatric treatment, n (%) | |||
| Psychiatric disorder and did not receive treatment | 802 (4.2) | 489 (4.2) | 313 (4.1) |
| Psychiatric disorder and received treatment, or no psychiatric disorder | 18,398 (95.8) | 11,050 (95.8) | 7,348 (95.9) |
| Psychiatric disorder and received outpatient care | 2,937 (15.3) | 1,857 (16.1) | 1,080 (14.1) |
| Psychiatric disorder and received pharmacotherapy | 5,721 (29.8) | 3,474 (30.1) | 2,247 (29.3) |
| No psychiatric disorder | 12,295 (64.0) | 7,332 (63.5) | 4,963 (64.8) |
Heavy drinking was defined as exceeding either daily or weekly drinking limits recommended by the NIAAA.
Sustained cessation of heavy drinking was defined as achieving early cessation of heavy drinking and reporting lower-risk drinking or abstinence through 5Y.
Table 4.
Inverse probability of attrition weighted (IPAW) associations of baseline and follow-up patient characteristics with sustained cessation of heavy drinking, n=19,200.a
| Characteristic | OR (95% CI) | p-value |
|---|---|---|
| Baseline characteristics | ||
| Sex (female vs male) | 1.37 (1.27, 1.47) | <.001 |
| Age group (reference group: 18-34 years) | ||
| 35-49 years | 1.22 (1.10, 1.35) | <.001 |
| 50-64 years | 1.35 (1.21, 1.50) | <.001 |
| ≥65 years | 1.40 (1.23, 1.59) | <.001 |
| Race/ethnicity (reference group: White) | ||
| Asian, Native Hawaiian, or Pacific Islander | 1.56 (1.37, 1.77) | <.001 |
| Black | 1.11 (0.97, 1.28) | 0.125 |
| Latinx/Hispanic | 1.20 (1.09, 1.32) | <.001 |
| Native American | 1.35 (1.00, 1.82) | 0.050 |
| Other/Unknown | 1.18 (0.88, 1.59) | 0.266 |
| Household income (reference group: middle) | ||
| Low | 1.08 (0.99, 1.17) | 0.084 |
| High | 1.02 (0.94, 1.10) | 0.651 |
| Smoking status (reference group: never or former) | 0.88 (0.79, 0.98) | 0.021 |
| Index drinking level (reference group: exceeding only daily limits) | ||
| Exceeding only weekly limits | 0.41 (0.37, 0.44) | <.001 |
| Exceeding both daily and weekly limits | 0.32 (0.29, 0.36) | <.001 |
| Alcohol use disorder | 0.86 (0.70, 1.05) | 0.130 |
| Number of chronic medical conditions | 1.08 (1.06, 1.10) | <.001 |
| Any psychiatric disorder | 1.14 (1.04, 1.25) | 0.005 |
| Any drug use disorder | 1.17 (0.84, 1.64) | 0.349 |
| Utilization in 1Y after index | ||
| Addiction treatment | 0.92 (0.73, 1.15) | 0.453 |
| Psychiatric treatment | 1.00 (0.83, 1.20) | 0.969 |
| Post-1Y characteristics and utilization | ||
| Type of early cessation (lower-risk drinking vs abstinence) | 0.62 (0.57, 0.66) | <.001 |
| Number of alcohol screenings during follow-up | 0.42 (0.39, 0.44) | <.001 |
| Number of days of follow-up (reference group: <2 years) | 1.10 (1.03, 1.18) | 0.006 |
| Routine primary care | 1.57 (1.44, 1.71) | <.001 |
| Addiction treatment | 1.41 (1.19, 1.66) | <.001 |
| Psychiatric treatment | 1.02 (0.86, 1.21) | 0.804 |
Sustained cessation of heavy drinking was defined as achieving early cessation of heavy drinking and reporting lower-risk drinking or abstinence through 5Y.
AUD modified the findings for the type of early cessation. Overall, achieving lower-risk drinking was associated with lower odds of sustained cessation compared to abstinence (OR=0.62, 95% CI=0.57, 0.66; Table 4); however, the odds were lower among those with AUD (OR=0.44, 95% CI=0.31, 0.62; Figure 2) compared to those without (OR=0.62, 95% CI=0.58,0.67).
Figure 2.
ORs and 95% CIs of sustained cessation of heavy drinking, comparing patients who reported lower-risk drinking versus abstinence at 1Y, by whether patients had an AUD diagnosis.
Discussion
In a large, diverse sample of patients who reported heavy drinking in a Northern California healthcare system, we found that nearly two thirds achieved early and sustained cessation of heavy drinking, which was much higher than a prior U.S. general population study that found less than a third ceased heavy drinking[7]. This is of great clinical importance, since reducing drinking is associated with decreased risks of alcohol-related problems and health conditions, such as AUDs and liver disease, in addition to death[45,46]. Heavy drinking cessation is a critical component of recovery from AUD, as it is associated with reduction in AUD symptoms and improved quality of life[8,14]. In this study, we identified many patient characteristics (many of which were consistent with prior research, i.e., sex, age, and drinking severity[6,7,11–18]) and health service factors associated with early and sustained cessation of heavy drinking. Findings are largely generalizable to patients with access to care and can inform the development of strategies to reduce the significant harm caused by heavy drinking.
Routine primary care was one of the strongest predictors of sustained cessation of heavy drinking in our study. The role of primary care providers has become increasingly important in addressing alcohol problems, particularly because heavy drinking may lead to chronic illness and impact disease progression and severity[47–49]. Screening, brief intervention, and referral to treatment (SBIRT) workflows in primary care are essential for identifying alcohol problems and determining appropriate care, and are effective in reducing drinking in adults with less severe problems[50]. The prevalence of AUD diagnoses in our study was low, which speaks to the utility of primary care-based SBIRT in addressing heavy drinking widely, regardless of whether AUD criteria are met. For individuals with AUDs, routine primary care is associated with long-term remission and reduced overall costs of care[20,51]. Although understanding the mechanisms that underlie these associations was beyond the scope of this study, we were able to examine the important role of primary care, which has not been examined in other studies. Having annual primary care visits may help address medical needs, identify and treat relapse, and promote healthier lifestyles and better self-management skills, leading to recovery from AUD.
A key contribution of this study is the examination of treatment utilization as predictors of heavy drinking cessation. Overall, receiving addiction treatment was associated with a higher likelihood of early and sustained heavy drinking cessation, which underscore its importance in short-term and long-term recovery for individuals with AUD. Specifically, receiving addiction treatment in 1Y after index was associated with a higher likelihood of early cessation among patients who reported exceeding only daily limits, suggesting that early initiation and better retention in addiction treatment, higher intensity of treatment, or more emphasis on alcohol use reduction may be needed for those with higher baseline levels of drinking. Among patients who achieved early cessation, we found that receiving addiction treatment when needed was important for sustaining heavy drinking cessation long-term. Interestingly, patients who received psychiatric treatment were less likely to achieve early cessation of heavy drinking, yet patients with a psychiatric disorder were more likely. This finding is similar to our prior study among AUD patients and suggests that receiving psychiatric treatment may be a marker for more severe psychiatric disorders[13]. However, psychiatric treatment may not be as important for predicting sustained cessation of heavy drinking, as we did not find an association with this outcome. Factors such as treatment engagement and retention, and motivation and readiness to change, can influence treatment effectiveness and should be examined in future studies[16,30,52–54].
We did not find AUD to be associated with early cessation of heavy drinking but found that it was an effect measure modifier of the type of early cessation on sustained cessation. Prior studies suggest that lower-risk drinking is a suitable goal for some individuals[8], while others show that abstinence tends to lead to better outcomes[9,10]. In our study, patients who reported lower-risk drinking at 1Y had a lower likelihood of sustained cessation of heavy drinking compared to those who reported abstinence, but the difference was more pronounced among individuals with AUD. This finding supports the notion that while lower-risk drinking is a suitable goal for alcohol treatment, certain individuals, such as those with AUD, might require more frequent monitoring and support post-treatment in sustaining long-term heavy drinking cessation.
Compared to White patients, non-White patients were generally more likely to achieve heavy drinking cessation, which is consistent with a recent study of the general U.S. population[7]. Studies of individuals with AUD have reported mixed results, with some showing non-White patients are more likely to recover[13], and others showing they are less likely[6], which might be due to differences in populations (i.e., primary care versus general U.S. populations, respectively). The mechanism underlying why non-White patients may be more likely to stop drinking heavily is unclear, though lower cultural and social acceptance of heavy drinking, higher religiosity, and financial factors are possible explanations[55,56]. In this study, non-White patients had higher proportions of patients who reported abstinence at 1Y (47% to 58%) compared to White patients (38%). Additionally, Black and Hispanic/Latinx patients had the highest proportions who received routine primary care (83% and 85%, respectively), which could be driven by health disparities observed in these subgroups[57,58], but is promising in terms of increasing the likelihood of sustained cessation.
Strengths and Limitations
We demonstrate that data recorded in the EHR during patient care can be leveraged to identify patient-level factors, including service utilization, associated with different stages of heavy drinking cessation. As the aim of this study was to conduct an exploratory analysis of multiple predictors of heavy drinking cessation, we did not adjust for multiple hypothesis testing[59]. Future studies are needed to confirm these exploratory findings. Results concerning demographics and drinking levels were consistent with prior longitudinal survey studies and suggest the utility of EHR-based data in understanding heavy drinking cessation. However, the availability of EHR-based alcohol screening data depends on when patients receive care, making it difficult to infer exact durations of heavy drinking cessation. We were not able to examine reductions in total daily alcohol consumption (i.e., drinks or grams of alcohol per day), a continuous unit of measure that the World Health Organization (WHO) utilizes in defining drinking risk levels. Prior studies have shown that reductions in the WHO drinking risk levels were associated with fewer alcohol-related consequences and better mental health and quality of life[46,60–65]; thus, it is possible that patients in our study may have experienced good health outcomes without necessarily reducing consumption below the NIAAA recommended limits. Future research examining our findings in the context of other study settings using other valid alcohol use measures is warranted.
We could not exclude the possibility that the decreases in alcohol use were due to regression to the mean, a statistical phenomenon. Self-report alcohol use data collected during clinical care is also subject to social desirability bias; however, the healthcare system in this study underwent a trial to determine the best workflow for conducting alcohol screening in a way that would minimize stigma and optimize patient comfort in disclosing alcohol use[30], and its findings informed the subsequent system-wide implementation. However, future research, including collection of in-depth qualitative data, might help clarify the role and extent of this potential bias.
Attrition in our study was mainly due to missing outcome alcohol screening data. While screenings are conducted systematically, a patient may not be screened every year due to busy workflows, medical assistants forgetting, patients arriving late to appointments or not having a primary care visit every year. We found that patients with observed outcome data were older with more medical conditions compared to those lost to follow-up, so we adjusted for potential selection bias due to attrition using IPAW.
Conclusion
This study highlights patient characteristics and health service factors related to early and sustained cessation of heavy drinking. Vulnerable subgroups (i.e., non-White patients and those with psychiatric disorders), patients who received routine primary care, and those who received addiction treatment were more likely to sustain cessation of heavy drinking, an encouraging outcome; however, more research is needed to understand the mechanisms underlying these findings. Findings may inform primary care-based interventions that can be implemented within health systems (in the U.S. or internationally) to facilitate heavy drinking cessation. Future studies should examine whether these cessation durations are associated with improved physical and mental health, and overall functioning.
Supplementary Material
Acknowledgments
We gratefully acknowledge Dr. Raye Litten and Dr. Daniel Falk for their expertise in research on recovery from alcohol use disorders and reviewing this manuscript. We also would like to thank Dr. Romain Neugebauer for providing biostatistical consultation.
Conflict of interest declaration:
The authors declare no conflicts of interest. This study was supported by grants from the National Institute on Alcohol Abuse and Alcoholism (NIAAA) in the United States: 75N94019P00907 (C.W.) and R01AA025902 (S.S.).
References
- 1.Esser MB, Sherk A, Liu Y, Naimi TS, Stockwell T, Stahre M, et al. Deaths and years of potential life lost from excessive alcohol use - United States, 2011-2015. MMWR Morbidity and mortality weekly report. 2020. October 2;69(39):1428–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.GBD 2016 Alcohol Collaborators. Alcohol use and burden for 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016 - The Lancet. The Lancet. 2018;392:1015–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.National Institute on Alcohol Abuse and Alcoholism. Helping patients who drink too much: a clinician’s guide [Internet]. 2005. [cited 2018 Mar 13]. Available from: https://pubs.niaaa.nih.gov/publications/Practitioner/CliniciansGuide2005/guide.pdf
- 4.Azagba S, Shan L, Latham K, Manzione L. Trends in binge and heavy drinking among adults in the United States, 2011-2017. Subst Use Misuse. 2020. January 30;1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Grant BF, Chou SP, Saha TD, Pickering RP, Kerridge BT, Ruan WJ, et al. Prevalence of 12-month alcohol use, high-risk drinking, and DSM-IV alcohol use disorder in the United States, 2001-2002 to 2012-2013: results from the National Epidemiologic Survey on Alcohol and Related Conditions. JAMA Psychiatry. 2017. 01;74(9):911–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Fan AZ, Chou SP, Zhang H, Jung J, Grant BF. Prevalence and correlates of past-year recovery from DSM-5 alcohol use disorder: results from National Epidemiologic Survey on Alcohol and Related Conditions-III. Alcohol Clin Exp Res. 2019. November;43(11):2406–20. [DOI] [PubMed] [Google Scholar]
- 7.Saitz R, Heeren TC, Zha W, Hingson R. Transitions to and from at-risk alcohol use in adults in the United States. J Subst Use. 2019;24(1):41–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Witkiewitz K, Wilson AD, Pearson MR, Montes KS, Kirouac M, Roos CR, et al. Profiles of recovery from alcohol use disorder at three years following treatment: can the definition of recovery be extended to include high functioning heavy drinkers? Addiction. 2018;114(1):69–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kline-Simon AH, Litten RZ, Weisner CM, Falk DE. Posttreatment low-risk drinking as a predictor of future drinking and problem outcomes among individuals with alcohol use disorders: a 9-year follow-up. Alcohol Clin Exp Res. 2017;41(3):653–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kline-Simon AH, Falk DE, Litten RZ, Mertens JR, Fertig J, Ryan M, et al. Posttreatment low-risk drinking as a predictor of future drinking and problem outcomes among individuals with alcohol use disorders. Alcohol Clin Exp Res. 2013. January;37 Suppl 1:E373–380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Satre DD, Chi FW, Mertens JR, Weisner CM. Effects of age and life transitions on alcohol and drug treatment outcome over nine years. J Stud Alcohol Drugs. 2012. May;73(3):459–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Tuithof M, Ten Have M, van den Brink W, Vollebergh W, de Graaf R. The relationship between excessive alcohol consumption and alcohol use disorders according to DSM-IV and DSM-5. Alcohol Clin Exp Res. 2014. January;38(1):249–56. [DOI] [PubMed] [Google Scholar]
- 13.Palzes VA, Kline-Simon AH, Satre DD, Sterling S, Weisner C, Chi FW. Remission from unhealthy drinking among patients with an alcohol use disorder: a longitudinal study using systematic, primary care-based alcohol screening data. J Stud Alcohol Drugs. 2020;81(4):436–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Dawson DA, Goldstein RB, Grant BF. Rates and correlates of relapse among individuals in remission from DSM-IV alcohol dependence: a 3-year follow-up. Alcohol Clin Exp Res. 2007. December;31(12):2036–45. [DOI] [PubMed] [Google Scholar]
- 15.Dawson DA, Goldstein RB, Ruan WJ, Grant BF. Correlates of recovery from alcohol dependence: a prospective study over a 3-year follow-up interval. Alcohol Clin Exp Res. 2012. July;36(7):1268–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Weisner C, Ray GT, Mertens JR, Satre DD, Moore C. Short-term alcohol and drug treatment outcomes predict long-term outcome. Drug Alcohol Depend. 2003. September 10;71(3):281–94. [DOI] [PubMed] [Google Scholar]
- 17.Dawson DA, Grant BF, Stinson FS, Chou PS, Huang B, Ruan WJ. Recovery from DSM-IV alcohol dependence: United States, 2001-2002. Addiction. 2005. March;100(3):281–92. [DOI] [PubMed] [Google Scholar]
- 18.Boschloo L, Vogelzangs N, van den Brink W, Smit JH, Beekman ATF, Penninx BWJH. Predictors of the 2-year recurrence and persistence of alcohol dependence. Addiction. 2012. September;107(9):1639–40. [DOI] [PubMed] [Google Scholar]
- 19.Weisner C, Matzger H, Kaskutas LA. How important is treatment? One-year outcomes of treated and untreated alcohol-dependent individuals. Addiction (Abingdon, England). 2003. July;98(7):901–11. [DOI] [PubMed] [Google Scholar]
- 20.Chi FW, Parthasarathy S, Mertens JR, Weisner CM. Continuing care and long-term substance use outcomes in managed care: early evidence for a primary care-based model. Psychiatr Serv. 2011. October;62(10):1194–200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tsoh JY, Chi FW, Mertens JR, Weisner CM. Stopping smoking during first year of substance use treatment predicted 9-year alcohol and drug treatment outcomes. Drug Alcohol Depend. 2011. April 1;114(2–3):110–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Stein E, Witkiewitz K. Trait self-control predicts drinking patterns during treatment for alcohol use disorder and recovery up to three years following treatment. Addict Behav. 2019;99:106083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Soloski KL, Durtschi JA. Identifying different ways people change: a latent basis growth mixture model example identifying nonlinear trajectories of binge drinking. J Marital Fam Ther. 2020. October;46(4):638–60. [DOI] [PubMed] [Google Scholar]
- 24.Vickers Smith R, Kranzler HR, Justice AC, Tate JP, VA Million Veteran Program. Longitudinal drinking patterns and their clinical correlates in Million Veteran Program participants. Alcohol Clin Exp Res. 2019;43(3):465–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jackson KM, Sher KJ. Similarities and differences of longitudinal phenotypes across alternate indices of alcohol involvement: a methodologic comparison of trajectory approaches. Psychol Addict Behav. 2005. December;19(4):339–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cranford JA, Krentzman AR, Mowbray O, Robinson EAR. Trajectories of alcohol use over time among adults with alcohol dependence. Addict Behav. 2014. May;39(5):1006–11. [DOI] [PubMed] [Google Scholar]
- 27.Kline-Simon AH, Chi FW, Mertens JR, Weisner C. Trajectories of remission and mortality over 13 years after intake to substance use treatment. Am J Drug Alcohol Abuse. 2017;43(5):583–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Elm E von Altman DG, Egger M Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for reporting observational studies. PLOS Medicine. 2007. October 16;4(10):e296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Gordon NP. Similarity of the adult Kaiser Permanente membership in Northern California to the insured and general population in Northern California: statistics from the 2011 California Health Interview Survey [Internet]. 2015. [cited 2019 Mar 15]. Available from: https://divisionofresearch.kaiserpermanente.org/projects/memberhealthsurvey/SiteCollectionDocuments/chis_non_kp_2011.pdf
- 30.Mertens JR, Chi FW, Weisner CM, Satre DD, Ross TB, Allen S, et al. Physician versus non-physician delivery of alcohol screening, brief intervention and referral to treatment in adult primary care: the ADVISe cluster randomized controlled implementation trial. Addict Sci Clin Pract. 2015. November 19;10:26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Palzes VA, Weisner C, Chi FW, Kline-Simon AH, Satre DD, Hirschtritt ME, et al. The Kaiser Permanente Northern California Adult Alcohol Registry, an electronic health records-based registry of patients with alcohol problems: development and implementation. JMIR Med Inform. 2020. July 22;8(7):e19081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.National Institute on Alcohol Abuse and Alcoholism. NIAAA Recovery Research Definitions [Internet]. 2020. [cited 2021 Mar 2]. Available from: https://www.niaaa.nih.gov/research/niaaa-recovery-from-alcohol-use-disorder/definitions
- 33.U.S. Census Bureau. Explore Census Data [Internet]. 2020. [cited 2020 Jul 10]. Available from: https://data.census.gov/cedsci/
- 34.Ornstein SM, Nietert PJ, Jenkins RG, Litvin CB. The prevalence of chronic diseases and multimorbidity in primary care practice: a PPRNet report. J Am Board Fam Med 2013. October;26(5):518–24. [DOI] [PubMed] [Google Scholar]
- 35.U.S. Department of Health and Human Services. Medication for the Treatment of Alcohol Use Disorder: A Brief Guide [Internet]. HHS Publication No. (SMA) 15-4907. Rockville, MD; 2015. [cited 2019 May 2]. Available from: https://store.samhsa.gov/sites/default/files/d7/priv/sma15-4907.pdf [Google Scholar]
- 36.Blodgett JC, Del Re AC, Maisel NC, Finney JW. A meta-analysis of topiramate’s effects for individuals with alcohol use disorders. Alcohol Clin Exp Res. 2014. June;38(6):1481–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Hirschtritt ME, Palzes VA, Kline-Simon AH, Kroenke K, Campbell CI, Sterling SA. Benzodiazepine and unhealthy alcohol use among adult outpatients. Am J Manag Care. 2019;25(12):e358–65. [PMC free article] [PubMed] [Google Scholar]
- 38.Center for Medicare & Medicaid Services. Atypical Antipsychotics: U.S. Food and Drug Administration-Approved Indications and Dosages for Use in Adults [Internet]. [cited 2019 May 29]. Available from: https://www.cms.gov/Medicare-Medicaid-Coordination/Fraud-Prevention/Medicaid-Integrity-Education/Pharmacy-Education-Materials/Downloads/atyp-antipsych-adult-dosingchart.pdf
- 39.Public Resources from the Mental Health Research Network: MHResearchNetwork/MHRN-Central [Internet]. MHResearchNetwork; 2019. [cited 2019 Jul 24]. Available from: https://github.com/MHResearchNetwork/MHRN-Central [Google Scholar]
- 40.Cohen J Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 1988. [Google Scholar]
- 41.Yang D, Dalton JE. A unified approach to measuring the effect size between two groups using SAS. In: SAS Global Forum [Internet]. SAS Institute Inc.; 2012. [cited 2020 Oct 27]. Available from: https://support.sas.com/resources/papers/proceedings12/335-2012.pdf [Google Scholar]
- 42.Seaman SR, White IR. Review of inverse probability weighting for dealing with missing data. Stat Methods Med Res. 2013. June;22(3):278–95. [DOI] [PubMed] [Google Scholar]
- 43.Lee KJ, Tilling KM, Cornish RP, Little RJA, Bell ML, Goetghebeur E, et al. Framework for the treatment and reporting of missing data in observational studies: The Treatment And Reporting of Missing data in Observational Studies framework. J Clin Epidemiol. 2021. February 2;134:79–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Austin PC. Variance estimation when using inverse probability of treatment weighting (IPTW) with survival analysis. Stat Med. 2016. December 30;35(30):5642–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Charlet K, Heinz A. Harm reduction-a systematic review on effects of alcohol reduction on physical and mental symptoms. Addict Biol. 2017. September;22(5):1119–59. [DOI] [PubMed] [Google Scholar]
- 46.Knox J, Wall M, Witkiewitz K, Kranzler HR, Falk D, Litten R, et al. Reduction in nonabstinent WHO drinking risk levels and change in risk for liver disease and positive AUDIT-C scores: prospective 3-year follow-up results in the U.S. general population. Alcohol Clin Exp Res. 2018;42(11):2256–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Sullivan LE, Fiellin DA, O’Connor PG. The prevalence and impact of alcohol problems in major depression: a systematic review. Am J Med. 2005. April;118(4):330–41. [DOI] [PubMed] [Google Scholar]
- 48.Ostacher MJ. Comorbid alcohol and substance abuse dependence in depression: impact on the outcome of antidepressant treatment. Psychiatr Clin North Am. 2007. March;30(1):69–76. [DOI] [PubMed] [Google Scholar]
- 49.Praud D, Rota M, Rehm J, Shield K, Zatoński W, Hashibe M, et al. Cancer incidence and mortality attributable to alcohol consumption. Int J Cancer. 2016. March 15;138(6):1380–7. [DOI] [PubMed] [Google Scholar]
- 50.Kaner EFS, Beyer F, Dickinson HO, Pienaar E, Campbell F, Schlesinger C, et al. Effectiveness of brief alcohol interventions in primary care populations. Cochrane Database Syst Rev. 2007. April 18;(2):CD004148. [DOI] [PubMed] [Google Scholar]
- 51.Parthasarathy S, Chi FW, Mertens JR, Weisner C. The role of continuing care in 9-year cost trajectories of patients with intakes into an outpatient alcohol and drug treatment program. Med Care. 2012. June;50(6):540–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Rice SL, Hagler KJ, Tonigan JS. Longitudinal trajectories of readiness to change: alcohol use and help-seeking behavior. J Stud Alcohol Drugs. 2014. May;75(3):486–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Witbrodt J, Ye Y, Bond J, Chi F, Weisner C, Mertens J. Alcohol and drug treatment involvement, 12-step attendance and abstinence: 9-year cross-lagged analysis of adults in an integrated health plan. J Subst Abuse Treat. 2014. April;46(4):412–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Satre DD, Mertens JR, Areán PA, Weisner C. Five-year alcohol and drug treatment outcomes of older adults versus middle-aged and younger adults in a managed care program. Addiction. 2004. October;99(10):1286–97. [DOI] [PubMed] [Google Scholar]
- 55.Meyers JL, Brown Q, Grant BF, Hasin D. Religiosity, race/ethnicity, and alcohol use behaviors in the United States. Psychol Med. 2017;47(1):103–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Zapolski TCB, Pedersen SL, McCarthy DM, Smith GT. Less drinking, yet more problems: understanding African American drinking and related problems. Psychol Bull. 2014. January;140(1):188–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Horowitz B, Miskulin D, Zager P. Epidemiology of hypertension in CKD. Adv Chronic Kidney Dis. 2015. March;22(2):88–95. [DOI] [PubMed] [Google Scholar]
- 58.Clements JM, West BT, Yaker Z, Lauinger B, McCullers D, Haubert J, et al. Disparities in diabetes-related multiple chronic conditions and mortality: The influence of race. Diabetes Res Clin Pract. 2020. January;159:107984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Bender R, Lange S. Adjusting for multiple testing--when and how? J Clin Epidemiol. 2001. April;54(4):343–9. [DOI] [PubMed] [Google Scholar]
- 60.Knox J, Scodes J, Wall M, Witkiewitz K, Kranzler HR, Falk D, et al. Reduction in non-abstinent WHO drinking risk levels and depression/anxiety disorders: 3-year follow-up results in the US general population. Drug Alcohol Depend. 2019. April 1;197:228–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Witkiewitz K, Hallgren KA, Kranzler HR, Mann KF, Hasin DS, Falk DE, et al. Clinical validation of reduced alcohol consumption after treatment for alcohol dependence using the World Health Organization risk drinking levels. Alcohol Clin Exp Res. 2017. January;41(1):179–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Witkiewitz K, Kranzler HR, Hallgren KA, O’Malley SS, Falk DE, Litten RZ, et al. Drinking risk level reductions associated with improvements in physical health and quality of life among individuals with alcohol use disorder. Alcohol Clin Exp Res. 2018. December;42(12):2453–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63.Witkiewitz K, Falk DE, Litten RZ, Hasin DS, Kranzler HR, Mann KF, et al. Maintenance of World Health Organization risk drinking level reductions and posttreatment functioning following a large alcohol use disorder clinical trial. Alcohol Clin Exp Res. 2019. May;43(5):979–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Witkiewitz K, Heather N, Falk DE, Litten RZ, Hasin DS, Kranzler HR, et al. World Health Organization risk drinking level reductions are associated with improved functioning and are sustained among patients with mild, moderate and severe alcohol dependence in clinical trials in the United States and United Kingdom. Addiction. 2020. September;115(9):1668–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.Witkiewitz K, Kranzler HR, Hallgren KA, Hasin DS, Aldridge AP, Zarkin GA, et al. Stability of drinking reductions and long-term functioning among patients with alcohol use disorder. J Gen Intern Med. 2021. February;36(2):404–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.