Validation of a Community Pharmacy-Based Prescription Drug Monitoring Program Risk Screening Tool: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Validation of a Community Pharmacy-Based Prescription Drug Monitoring Program Risk Screening Tool (NCT03936985), a registered phase not stated study. The primary outcome is Validity of the PDMP Narcotic Score Versus the WHO ASSIST.

2. Clinical Question

3. Trial Design

The registry describes a validation study of a community pharmacy-based PDMP risk screening tool. Its stated goal is to assess whether a PDMP-based opioid risk metric can discriminate between low, moderate, and high-risk opioid use disorder, using the WHO ASSIST to define patient risk levels. This is an evaluation of a screening metric against a reference instrument, not a test of a treatment effect. The registry explicitly states that no intervention or hypothesis will be tested.

The listed enrollment is 1464. The study status is COMPLETED, with a start date of 2019-11-04 and a primary completion date of 2021-08-30. The ASSIST was administered once at Day 1, according to the primary outcome timeframe. The record does not report how participants were recruited, the study sites, eligibility criteria, or participant characteristics.

Random allocation and treatment arms are not described. The registry does not state a formal design label or provide details about how the PDMP Narcotic Score was obtained, whether assessors were blinded, or how the comparison was operationalized. These details are not reported in the registry record. The absence of treatment arms means that concepts such as treatment allocation, a placebo control, and between-arm treatment effects do not describe the primary question.

The primary outcome is validity of the PDMP Narcotic Score versus the WHO ASSIST, assessed with an AUC and a 95% confidence interval. Although the record lists one outcome measure, it lists no analysis results. Thus, the study objective and planned type of summary are described, but the observed discrimination and its precision cannot be assessed from the registry record.

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: Participants who complete the World Health Organization Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST).

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 Validity of the PDMP Narcotic Score Versus the WHO ASSIST, measured over The ASSIST was administered once at Day 1..

The primary endpoint is the validity of the PDMP Narcotic Score against the WHO ASSIST, summarized by the area under the curve (AUC). The WHO ASSIST is described as the gold standard instrument for defining participant risk levels. The endpoint population is participants who complete that instrument, and the ASSIST was administered once at Day 1.

In general, an AUC summarizes how well a score separates people in different reference-standard categories across possible score thresholds. The registry explains that the AUC ranges from 0 to 1: 0.0 represents predictions that are 100% wrong, 1.0 represents predictions that are 100% correct, and 0.5 means the model has no means of separating the groups. For this study, the record says that values closer to 1.0 indicate greater similarity between the PDMP Narcotic Score and the WHO ASSIST.

An AUC is a measure of discrimination. It does not, by itself, establish that predicted risks are well calibrated, that a particular score threshold is useful, or that using the score improves outcomes. The record does not specify the exact AUC formulation for the low, moderate, and high-risk categories, nor does it report a numerical AUC or confidence interval. Those results are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary statistical method named in the registry is the area under the curve (AUC), with a 95% confidence interval. This method addresses discrimination: whether the PDMP Narcotic Score can distinguish the risk groups defined by the WHO ASSIST. The registry describes the WHO ASSIST as the gold standard instrument for assigning those risk levels. No numerical estimate or interval is reported, so the method can be explained, but its result cannot be interpreted for this study.

What does the AUC describe?

In general, a receiver operating characteristic curve considers how sensitivity and specificity vary as a score threshold changes. The AUC condenses that curve into a single summary of discrimination. A value closer to 1.0 indicates stronger separation under the interpretation stated in the registry. The registry describes 0.5 as having no means of separating the groups, 0.0 as predictions that are 100% wrong, and 1.0 as predictions that are 100% correct.

For a simple two-category problem, the AUC can also be understood generally as the chance that a randomly selected member of one category receives a more favorable score ranking than a randomly selected member of the other category, subject to the direction of the score. The study describes low, moderate, and high-risk groups, so the exact approach to combining or comparing categories matters. The registry does not report whether the analysis used pairwise comparisons, a multicategory method, or another formulation. That method is not reported in the registry record.

What does a confidence interval add?

A confidence interval expresses uncertainty around an estimated quantity under the assumptions of the method used. Its width is informative about precision: a narrower interval generally indicates a more precise estimate, while a wider interval indicates more uncertainty. A 95% confidence interval is specified for the primary outcome, but neither its endpoints nor the AUC estimate are reported. It is therefore not possible to judge the observed precision or whether the estimate is compatible with substantially different levels of discrimination.

A confidence interval does not remove design limitations or establish clinical usefulness. It reflects sampling uncertainty under the analysis assumptions, not every possible source of bias or measurement error. In general, the validity of the estimate also depends on how participants were selected, how the score and reference assessment were measured, and how incomplete observations were handled. The registry does not describe these details.

What analysis details are missing?

The record does not report the score calculation, the statistical software or model, the direction of the score, the handling of tied values, or the method used to construct the confidence interval. It also does not describe whether any covariates were included or whether analyses were adjusted. These methods are not reported in the registry record. Without such information, the analysis cannot be independently reproduced from the registry entry.

The record identifies the ASSIST as the reference standard, but does not describe how its categories were operationalized for the AUC analysis or whether all participants received the same assessment procedures. It states that the ASSIST was administered once at Day 1. The timing and handling of the PDMP Narcotic Score measurement are not specified. Those details are not reported in the registry record.

In general, an AUC evaluates ranking or separation, not calibration. Calibration concerns agreement between predicted risk and observed risk; a score can rank participants well without its numerical values corresponding closely to observed probabilities. The registry does not report a calibration analysis. Nor does it identify a score threshold, report threshold-specific sensitivity or specificity, or describe decision-analytic methods. These are not reported in the registry record.

The record's sample-size assumptions refer to planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The study itself is described as having no intervention or hypothesis test, and no like-for-like randomized arm comparison is established. The registry does not provide a planning illustration applicable to the AUC validation endpoint. Those assumptions should not be mistaken for the validation analysis or for observed power.

The endpoint is a probability-type measure, while the registry categorizes the outcome as count other. The record does not explain that categorization or describe a count model. No count-model inference can be inferred from the endpoint description alone. The primary analysis result remains unavailable in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and zero analyses. It identifies the primary endpoint as AUC for the PDMP Narcotic Score versus the WHO ASSIST, but does not list a result. The number of analyses listed in the record does not establish whether analyses were conducted elsewhere; it means that the registry entry does not provide analysis details or results to evaluate.

Multiplicity concerns arise when several outcomes, comparisons, subgroup analyses, or analytic choices are considered. The record does not describe secondary outcomes, subgroup analyses, multiple AUC formulations, or an adjustment for multiple testing. It also states that no hypothesis will be tested. A multiplicity strategy is therefore not reported in the registry record, and it is not possible to assess whether any multiplicity adjustment was used.

The endpoint applies to participants who complete the WHO ASSIST. The record does not report how many participants completed the assessment, whether any data were missing, or how missing score or reference-standard information was handled. It does not specify an analysis population beyond the endpoint's completion criterion. The number included in the analysis and the treatment of incomplete records are not reported in the registry record.

There is also no description of exclusions after enrollment, protocol deviations, or sensitivity analyses. These details matter because exclusions or missing assessments can affect how representative the analyzed participants are of those enrolled. Without those details and a reported analysis, the degree to which such issues could affect the result cannot be assessed.

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

The study addresses whether a PDMP Narcotic Score can distinguish categories defined by the WHO ASSIST. Its chosen summary, the AUC, is designed to describe discrimination, and the registry explains how values across its range are interpreted. Because no numerical AUC or confidence interval is reported, the posted results do not establish the degree of discrimination observed in the study. They do not support a conclusion that the tool was validated successfully or unsuccessfully.

Even a reported AUC would answer a limited statistical question. In general, it describes separation across possible thresholds; it does not, on its own, determine an appropriate threshold for screening or quantify the consequences of decisions made using that threshold. It also does not measure calibration, prove that scores reflect an individual's actual risk, or establish that using the score changes behavior or outcomes. Those outcomes and analyses are not reported in the registry record.

The WHO ASSIST serves as the reference standard in the stated design. The comparison therefore concerns agreement in risk discrimination relative to that instrument, not a randomized comparison between treatments. The record's statement that no intervention or hypothesis will be tested is consistent with that distinction. No causal treatment effect can be inferred from this validation objective.

The listed enrollment is 1464, while the endpoint is defined for participants who complete the ASSIST. The record does not report the analyzed population or the count of completed assessments, so enrollment should not be treated as the analysis denominator. Details on missing data and exclusions are also unavailable.

In practical terms, the registry record documents the question, reference instrument, endpoint, and intended summary statistic, but not the numerical evidence needed to judge the score's performance. Conclusions about precision, threshold performance, calibration, or generalizability would require information not reported in the registry record.

12. Why This Trial Matters Statistically

Risk screening tools are often judged not only by whether they produce a score, but by whether that score separates groups defined by a reference assessment. This study's stated statistical question is therefore an important validation question: how well a PDMP Narcotic Score discriminates among risk levels defined by the WHO ASSIST. AUC is a recognizable summary for that purpose because it evaluates separation across possible thresholds rather than at a single cutoff.

The registry also makes clear why the endpoint should not be conflated with a treatment outcome. The study is not testing an intervention or hypothesis; it is evaluating a metric against a reference instrument. That distinction helps keep interpretation focused on measurement performance rather than causal effects.

At the same time, a useful validation judgment requires the estimate and its uncertainty, along with enough detail to understand how the analysis was performed. The record specifies an AUC and a 95% confidence interval but reports neither numerical result. It also does not describe the AUC formulation for the three risk categories or provide threshold-specific or calibration results. As a result, the statistical importance of the question is clearer than the empirical answer available in the registry.

This record illustrates a broader point about diagnostic statistics: a named endpoint and an explanation of its scale are not substitutes for posted findings. Without the estimate, interval, analysis population, and method details, readers cannot evaluate the magnitude, precision, or reproducibility of the validation result. The registry record supports a description of the intended analysis, but not a conclusion about the tool's performance.

13. 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 primary limitation for interpreting this study from the registry record is the absence of numerical outcome results. Although the endpoint is defined as AUC with a 95% confidence interval, neither the estimate nor interval is reported. The degree and precision of discrimination therefore cannot be evaluated.

The record does not describe the exact AUC method for the low, moderate, and high-risk categories. It does not state how category comparisons were combined, how the score was oriented, or how the interval was calculated. These details are not reported in the registry record, limiting reproducibility and interpretation.

The listed enrollment does not establish the primary analysis population. The endpoint pertains to participants who complete the WHO ASSIST, but the registry does not report the number completing it, the number analyzed, missing assessments, exclusions, or the handling of incomplete data. These details are not reported in the registry record.

Information about recruitment, study sites, participant characteristics, eligibility criteria, and the timing or derivation of the PDMP Narcotic Score is also absent. The registry does not report whether the score and reference assessment were evaluated independently or whether assessors were blinded. These limitations prevent assessment of representativeness and potential measurement or assessment bias from the record alone.

Finally, AUC addresses discrimination but not calibration or performance at a selected threshold. The registry does not report calibration, threshold-specific performance, or decision consequences. These omissions limit the conclusions that could be drawn even if the AUC estimate were available.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Validation of a Community Pharmacy-Based Prescription Drug Monitoring Program Risk Screening Tool — NCT03936985.
• Registry record last updated 2024-07-11; status last verified 2024-07.
• Results first posted 2024-07-11.
• 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.

17. Record Summary

The registry record describes a completed study titled Validation of a Community Pharmacy-Based Prescription Drug Monitoring Program Risk Screening Tool, sponsored by the University of Utah. It lists enrollment of 1464, a start date of 2019-11-04, and a primary completion date of 2021-08-30.

The stated goal is to assess a PDMP-based opioid risk metric against risk levels defined by the WHO ASSIST. The primary outcome is validity of the PDMP Narcotic Score versus the WHO ASSIST, measured with an AUC and a 95% confidence interval. The ASSIST was administered once at Day 1. The record states that no intervention or hypothesis will be tested.

One outcome measure is listed, but no analysis results are listed. A numerical AUC, confidence interval, analysis population, and missing-data approach are not reported in the registry record. The adverse-event record lists zero serious events, other events, and deaths for the all-participants group; these entries do not provide evidence about the screening metric's discrimination.