Collaborative Care for Alcohol Use Disorders in the Patient-centered Medical Home: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Collaborative Care for Alcohol Use Disorders in the Patient-centered Medical Home (NCT02885311), a non-randomized phase 1, phase 2 study. The primary outcome is Weekly Change in Drinking.

2. Clinical Question

3. Trial Design

The registry record describes a completed Phase 1 and Phase 2 study sponsored by Northwell Health. The listed start was 2017-01, and primary completion was 2018-10. Enrollment was 27. The study concerned developing and testing a care model intended to increase identification and treatment of problem drinking in primary care.

Allocation was nonrandomized, masking was none, and the design model was listed as factorial. The record does not explain the factorial structure or give enough detail to identify specific factors, their levels, or how they entered an analysis. It describes care pathways for distinct clinical categories: At-Risk Drinkers, Problem Drinkers, people with physiological withdrawal, and people with a complex presentation. Because group membership was not randomized, differences between categories could reflect their underlying clinical characteristics as well as their care pathways.

The interventions listed include behavioral approaches, naltrexone, brief advice, and referral to specialty substance use disorder treatment. The arm descriptions add details about follow-up assessments and, for some groups, medication management or outpatient detoxification referral. These are descriptions of the care offered, not evidence that every participant received or completed every component.

The registry lists one primary outcome measure, Weekly Change in Drinking, and no analyses. It does not report outcome estimates, uncertainty intervals, hypothesis tests, or a statistical analysis plan. Details such as eligibility criteria beyond the screening description, the numbers enrolled in each group, adherence, and the handling of departures from the intended care pathways are not reported in the registry record. The design and care descriptions therefore support understanding the study question, but not an estimate of comparative treatment effects.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: We examined the change in average number of standard drinks consumed weekly during the baseline (8 weeks prior to study enrollment) and the 12-week treatment period.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Weekly Change in Drinking, measured over Baseline (8 weeks prior to study enrollment) and the 12-week treatment period.

The primary endpoint, Weekly Change in Drinking, concerns the average number of standard drinks consumed each week. The registry says drinking was assessed with the Timeline Followback Interview, using baseline covering the 8 weeks before study enrollment and the 12-week treatment period. A standard drink is described as about 9 grams of alcohol. The scale has a minimum of zero and no maximum.

The measure is intended to capture change in drinking over the study period. The registry specifies that higher negative scores reflect a greater reduction in drinking and better outcomes. This direction matters when reading a reported estimate: the sign and magnitude should be interpreted using that stated definition, rather than assuming that a higher score always means more drinking.

Weekly drinking is a continuous measure: participants can have a range of values rather than being classified only into outcome categories. In general, a continuous measure can retain differences in amount and change that would be lost if results were reduced to a simple yes-or-no outcome. Its interpretation still depends on how change was calculated and summarized. The registry identifies the outcome as a mean with standard deviation as the dispersion measure, but does not give the calculation details, group means, standard deviations, or analysis results. Thus, the endpoint is defined, but the size and precision of any change cannot be evaluated from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies Weekly Change in Drinking as a continuous outcome summarized by a mean, with standard deviation as the stated measure of dispersion. General statistical context: a mean represents the arithmetic average of measured values, while a standard deviation describes how spread out those values are around the average. These summaries answer different questions. The mean locates the group average; the standard deviation conveys variation among observations. Neither value is reported for the outcome in the registry entry.

What does the continuous outcome represent?

The outcome is based on the average weekly number of standard drinks, measured at baseline and during the 12-week treatment period. The baseline window is the 8 weeks before enrollment. The registry describes the Timeline Followback Interview as the collection method and defines a standard drink as about 9 grams of alcohol. The measure has a floor of zero and no maximum. The record says higher negative scores mean a greater reduction in drinking. It does not explain the precise computation used to turn the baseline and treatment-period measures into a change score, or how weekly observations within the treatment period were combined.

General statistical context: continuous outcomes can preserve information about the amount of change, but their interpretation depends on the scale, direction, and calculation. A group mean does not describe every participant's experience. The standard deviation can indicate how much observations vary, but it does not by itself show the uncertainty around the mean. A standard error or confidence interval is commonly used for uncertainty about an estimate; neither is reported for this outcome.

How should a group comparison be understood?

The record describes nonrandom allocation and care pathways for clinical categories. It does not state a statistical model for comparing those categories, nor does it report an adjusted or unadjusted treatment contrast. General statistical context: a comparison of group averages can be influenced by baseline differences, especially when group membership is not randomized. Adjustment can account for measured characteristics under a specified model, but it does not automatically remove all bias, and the registry does not report whether adjustment was performed.

The study model is listed as factorial, but the record does not explain the factors or analysis structure. A factorial design, in general, studies more than one intervention component or factor and may assess both individual effects and interactions. That general meaning does not establish which factors were tested in this study. Without the factor definitions, allocation details, and analysis plan, no particular factorial contrast can be inferred.

What do the planning assumptions say?

The stated planning framework uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The target effect is expressed as a fraction of the observed effect. General statistical context: a significance level is a threshold used in a hypothesis-testing procedure; a two-sided procedure allows departures in either direction. A normal approximation uses a mathematical distribution to estimate how a statistic behaves under specified assumptions. Equal allocation means the planning setup assumes the same number of participants in each comparison arm.

Those assumptions describe planning calculations, not the analysis actually conducted. The registry record says the sample-size illustration is computed from the observed effect, but does not report the trial's own sample-size justification. Such an illustration should not be interpreted as observed power. Power is a design property under specified assumptions, and a post-study calculation based on an observed effect does not establish that the study could reliably detect an effect. No illustration values are stated in the registry record available for this analysis.

What can be learned from absent analysis details?

The registry lists one outcome measure and no analyses. It does not report a test statistic, p-value, confidence interval, model, covariates, or analysis population. General statistical context: a p-value, when reported, is interpreted in relation to a specified null hypothesis and testing procedure; it is not a measure of the size or clinical importance of an effect. An interval estimate is needed to assess precision, while the effect estimate itself is needed to describe direction and magnitude. Here, neither is available.

The registry also does not state how repeated weekly observations, if used, were handled, or whether any distributional features such as over-dispersion were assessed. General statistical context: over-dispersion occurs when variability exceeds what a statistical model assumes; ignoring it can make uncertainty estimates unreliable. No claim can be made about whether it is relevant to the reported outcome, because the analysis model and outcome results are not reported.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("At-Risk Drinkers (AR)", "Problem Drinkers (PD)", "AD With Physiological Withdrawal (AD-W)", "AD With Complex Presentation (AD-CMPLX)")
n   <- c(6, 8, 9, 2)
m   <- c(-11.4, -22.2, -39.6, -29.3)
sp  <- c(4.1, 8, 33.5, 9.9)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  At-Risk Drinkers (AR)  6  -11.400  4.100  -15.703  -7.097
CARM  Problem Drinkers (PD)  8  -22.200  8.000  -28.888  -15.512
CARM  AD With Physiological Withdrawal (AD-W)  9  -39.600  33.500  -65.350  -13.850
CARM  AD With Complex Presentation (AD-CMPLX)  2  -29.300  9.900  -118.248  59.648

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and no analyses. A single listed outcome does not establish that only one outcome or one statistical test was planned; it means the record does not provide a set of reported analyses from which multiplicity handling can be assessed. The statistical analysis plan, any prespecified contrasts, and any correction for multiple testing are not reported in the registry record.

Missing-data handling is also not described. The record does not give the number of participants with an observed primary outcome, the amount or pattern of missing outcome data, reasons for missingness, or any imputation or sensitivity analysis. General statistical context: missing observations can affect estimates when the likelihood of being observed is related to participants' characteristics or outcomes. Different assumptions and methods can lead to different interpretations, so transparent reporting is needed to assess robustness.

The analysis population is not reported. It is not possible to determine from the registry entry whether analyses included all enrolled participants, only participants who received a particular intervention, or a different defined group. The record also does not report exclusions, adherence, or handling of participants who changed or did not complete their offered care. These omissions prevent assessment of how closely any analysis would reflect enrollment and the intended care pathways.

Because neither the analysis population nor outcome completeness is described, the enrollment total should not be treated as the number contributing to the primary result. No statistical approach to multiplicity or missingness can be attributed to the study from the information in the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry lists serious and other adverse-event counts by care group, along with deaths. For At-Risk Drinkers, it lists 6 at risk and 0 affected for serious events, and 6 at risk and 0 affected for other events. For Problem Drinkers, the corresponding entries are 10 and 0 for serious events, and 10 and 0 for other events. For participants with physiological withdrawal, they are 9 and 0 for each category. For participants with a complex presentation, they are 2 and 0 for each category. Deaths are listed as 0 in every group.

These are registry counts, not a comparison of safety between randomized groups. The study was nonrandomized, and the groups represent different clinical categories and care pathways. The registry lists no serious adverse-event terms. The figures should be read as the recorded counts and categories, not as evidence that an intervention caused or prevented an event. The registry record does not provide further detail on event definitions, timing, severity assessment, or how the at-risk counts relate to participants exposed to each care component. The absence of affected counts in this entry does not by itself establish that no adverse experiences occurred outside the registry's reported categories or reporting period.

13. What the Result Does, and Does Not, Mean

The registry record establishes that this study examined a primary-care care model for alcohol use disorders and defined weekly change in drinking as its primary outcome. It also describes nonrandomized care pathways for people categorized by drinking-related clinical presentation. Those details clarify the intended question and the outcome being tracked.

The record does not report the primary outcome analyses. No group estimates, measure of variability, uncertainty interval, or statistical test is listed. Therefore, the available registry information cannot establish whether average weekly drinking increased, decreased, or stayed similar during the treatment period. It cannot quantify a change or indicate how precisely a change was estimated.

The record also cannot support a comparative conclusion about the care pathways. Participants were not randomly allocated, and no common control group is identified. General statistical context: when treatment pathways correspond to different clinical categories, outcome differences can reflect both the care received and differences in the participants entering each category. A causal claim would require an appropriate design and analysis, together with sufficient detail to assess assumptions and potential confounding.

The enrollment total and the adverse-event entries provide context about the study record, but they do not substitute for primary-outcome results. The listed safety counts should likewise be interpreted descriptively, with attention to the nonrandomized group structure and the limited event detail.

In short, the registry describes a study question, care pathways, and an outcome measure, but not the outcome findings needed to judge the magnitude, direction, precision, or comparability of drinking changes. This is a limit of what the registry record reports, not evidence for either benefit or lack of benefit.

14. Why This Trial Matters Statistically

This study is statistically notable because the care model was evaluated in a primary-care setting while pathways differed according to participants' drinking-related categories. The record identifies the practical aim of improving identification and treatment of problem drinking, and it makes clear that the outcome of interest was change in average weekly drinking. That combination illustrates why a well-defined outcome and transparent description of assignment are both necessary for interpreting clinical research.

The nonrandomized allocation is central to statistical interpretation. General statistical context: randomized comparisons are designed to reduce systematic differences between groups, while nonrandomized comparisons require careful consideration of why participants entered each pathway. In this study, group descriptions indicate different clinical presentations and different care, so a simple difference between group averages would not automatically isolate the effect of a treatment component.

The record also highlights the distinction between registering an outcome and reporting its analysis. It identifies a continuous measure and its intended direction, but reports no analyses. Without estimates and uncertainty, readers cannot assess the amount of change or the strength of evidence. Stating the outcome definition is useful, yet it does not replace results or methods for handling missing data, group differences, and repeated measurement.

Finally, the listed planning framework is not the same as the trial's own sample-size justification or an account of observed precision. General statistical context: design assumptions can help explain how a study might plan to detect a specified effect, but interpretation after completion depends on the estimates and their uncertainty. The registry entry leaves those results and details unreported. The statistical importance of the record is therefore as much about the limits of inference as about the study question itself.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The main trial-specific limitation for interpretation is the nonrandomized allocation. The registry describes separate clinical categories with different care pathways, so the groups are not a like-for-like control comparison. Differences in outcomes, if reported elsewhere, could reflect differences in the people in each category, the care offered, or other factors. The registry does not report an adjusted comparison that would allow those explanations to be assessed.

Enrollment was 27, but the registry does not give enrollment by group or the number contributing data to the primary outcome. It also does not describe outcome completion, missing data, adherence, or the analysis population. Those gaps limit assessment of precision, potential selection effects, and how representative any estimate would be of all enrolled participants.

The endpoint is described as an average weekly drinking measure based on baseline and a 12-week treatment period, but the precise change calculation and treatment of repeated measurements are not reported. The factorial model label is not accompanied by the factors or analysis structure. The record also does not report the statistical analysis plan, sample-size justification, or primary-outcome analyses. Thus, estimates, uncertainty, and analytic choices cannot be checked against a prespecified approach.

The adverse-event entry provides counts but no event terms, timing, or additional definitions. With no reported primary outcome results and limited design and analysis detail, the registry record supports a description of the study but not a conclusion about the magnitude or causal effect of any care pathway.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Collaborative Care for Alcohol Use Disorders in the Patient-centered Medical Home — NCT02885311.
• Registry record last updated 2023-10-30; status last verified 2023-10.
• Results first posted 2023-10-30.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The completed Northwell Health study, NCT02885311, examined collaborative care for alcohol use disorders in a patient-centered medical home. It was listed as Phase 1 and Phase 2, nonrandomized, with no masking and a factorial model. Enrollment was 27.

The registry describes care pathways for At-Risk Drinkers, Problem Drinkers, participants with physiological withdrawal, and participants with a complex presentation. Its primary outcome was Weekly Change in Drinking, based on average weekly standard drinks at baseline and during the 12-week treatment period. The registry reports no analyses for that outcome and does not provide outcome estimates or uncertainty measures. Adverse-event counts are listed by group, with 0 affected and 0 deaths in each group.