Reducing Potentially Inappropriate Medication Prescribing for Older Patients in the ED: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Reducing Potentially Inappropriate Medication Prescribing for Older Patients in the ED (NCT04004936), a randomized na study. The primary outcome is Percentage of PIMs Prescribed.

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel-group trial with no masking. Its title identifies the EQUIPPED program and its focus on reducing potentially inappropriate medication prescribing for older patients in emergency departments. The study’s stated purpose was to examine implementation strategies and factors affecting implementation, as well as to micro-cost the active and passive feedback versions.

The two intervention groups were active feedback and passive feedback. Active feedback included one-to-one, in-person academic detailing from a professional colleague, with audit, feedback, peer benchmarking, and on-site expertise. Passive feedback used a monthly electronic dashboard with audit, feedback, and peer benchmarking. Thus, the central randomized contrast concerned how feedback was delivered within EQUIPPED.

Enrollment was 74. The registry reports the group-level counts as 36 for active feedback and 38 for passive feedback. The randomization unit is not reported in the registry record, so these counts should not be treated as evidence about whether providers, sites, or another unit were randomized. The record identifies the primary outcome population as discharge prescriptions for Veterans aged 65 and older at implementation sites within each group.

The primary outcome timeframe was 12 months after implementation. The registry also lists a perceived-control survey outcome measured 12 months after delivery of the first EQUIPPED report, and a micro-costing outcome. The trial status is completed. The record does not report the method of sequence generation, allocation concealment, details of the implementation sites, or whether outcome assessment was blinded. It also does not report a protocol-level sample-size justification or a detailed statistical analysis plan.

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.

Before assignment: Participants were EQUIPPED implementation team members or providers at EQUIPPED implementation sites.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Discharge prescriptions for Veterans 65 and older at EQUIPPED implementation sites within each group

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.

Population: Participants were EQUIPPED implementation team members or providers at EQUIPPED implementation sites.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of PIMs Prescribed, measured over 12-Months Post Implementation of EQUIPPED.

The primary endpoint is titled “Percentage of PIMs Prescribed.” Its description defines it as the percentage of prescriptions that are potentially inappropriate medications under Beers criteria, prescribed to adults aged 65 and older and discharged from the ED. The registry specifies discharge prescriptions for Veterans aged 65 and older at EQUIPPED implementation sites, assessed over the 12 months after implementation.

A percentage would describe the share of prescriptions meeting the PIM definition, rather than the raw number alone. That distinction matters because a prescription count has no denominator: without the total prescriptions assessed in each group, a count cannot show what proportion were PIMs. A comparison of percentages would also need an appropriate uncertainty estimate and analysis that accounts for how the data were collected.

The posted primary arm values are recorded as counts of units, with 1672 for active feedback and 2979 for passive feedback. The record does not provide the corresponding denominators or an estimable two-group comparison for the percentage endpoint. Therefore, these counts cannot be interpreted as group percentages or as evidence that one feedback strategy had a lower PIM prescribing rate. The stated endpoint is relevant to the study’s implementation question, but the posted numerical information does not permit the primary rate contrast to be assessed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help distinguish what the registry reports from what can be inferred. The first is the difference between a count and a rate. A count records the number of observed units, while a percentage relates a numerator to a denominator. The primary outcome is described as a percentage of prescriptions that were PIMs. Yet the posted primary arm summaries are classified as counts of units, and the denominators needed to interpret them as percentages are not reported in the registry record. No transformation of those counts into rates is justified by the posted information.

What does a confidence interval express?

General explanation: a confidence interval is a range of values compatible with an estimate under the statistical model and assumptions used to calculate it. It communicates precision as well as direction. A narrower interval indicates greater precision than a wider interval, while an interval that includes the value representing no difference is compatible with no difference under that calculation. A confidence interval is not a guarantee that the true value falls inside its particular limits, nor does it address biases or limitations not represented in the model.

For the perceived-control survey, the posted recalculation gives a mean difference of -0.8, with a 95% confidence interval from -1.517 to -0.083. The direction is passive feedback minus active feedback, so the negative estimate corresponds to a lower reported mean in the passive-feedback group. The interval is for this secondary survey comparison only. It cannot be transferred to the primary PIM percentage outcome, for which the registry does not provide a comparable estimate.

What does the mean difference compare?

General explanation: a difference between group means expresses how far apart the average observed scores are on the outcome scale. Its interpretation depends on the scale, the direction of subtraction, and the study design. Here, the comparison is defined as passive feedback minus active feedback, with reported means of 3.3 and 4.1, respectively. The recalculated difference is unadjusted: it does not account for covariates or other factors through a reported adjusted model. The registry does not list a sponsor-reported analysis for this secondary outcome.

The standard deviation describes variation among observations within a group; it is not the uncertainty interval for the group mean. The reported standard deviations are 2.1 for passive feedback and 0.8 for active feedback. The difference in observed variability is important context for the comparison, but does not by itself identify its cause. The registry’s recalculated comparison uses the posted summary statistics and a pooled standard deviation of 1.605. These are summary-statistic calculations, not a substitute for a fully specified analysis of the trial’s original data.

Why is over-dispersion relevant?

General explanation: count outcomes are often analyzed with models that relate the expected count to an exposure or denominator. If the observed variability in counts is greater than a model expects, over-dispersion may make standard errors too small and confidence intervals too narrow unless the model accounts for it. For prescribing outcomes, the volume of prescriptions can differ across groups or sites, so the denominator and the structure of the observations matter. The registry record does not report a count model, exposure adjustment, over-dispersion assessment, or other primary-outcome analysis. No such method should be assumed to have been used.

What do the planning illustrations mean?

General explanation: sample-size planning calculations describe how many participants per arm might be required under stated assumptions to detect a chosen effect with a target power. They are not observed power and do not describe what the completed trial achieved. The posted illustration uses a two-sided significance level of 0.05, equal allocation, and a normal approximation. It uses fractions of the observed perceived-control effect as the effects to detect. For planning for 100% of the observed effect, the illustration gives 64 participants per arm for 80% power and 85 participants per arm for 90% power. For planning for 75% of the observed effect, it gives 113 and 151 participants per arm for those power targets. For planning for 50% of the observed effect, it gives 253 and 339 participants per arm. These are hypothetical planning quantities based on the observed secondary effect, not the trial’s original design justification. The original sample-size justification is not reported in the registry record.

Finally, a p-value is a measure of how incompatible the data are with a specified null model, under the assumptions of the test. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The unadjusted p-value of 0.029 belongs to the recalculated secondary survey comparison. The record does not report adjustment for multiple comparisons, so this value should not be treated as a multiplicity-adjusted confirmatory result.

10. Multiplicity, Missing Data and Analysis Populations

The registry record does not report how missing outcome data were handled. It does not state whether all randomized units contributed data to each outcome, whether observations were excluded, or whether any imputation or sensitivity analyses were conducted. Consequently, the potential effect of missingness on either the prescribing endpoint or the survey result cannot be assessed from the posted information.

Analysis populations are also not reported in the registry record. There is no stated intention-to-treat, per-protocol, or other analysis population for the primary endpoint, and the record does not explain how the analyzed populations relate to the randomization. The reported group counts of 36 and 38 should therefore be read as registry arm summaries, not as proof of the analysis denominator for every outcome.

The record does not report whether the primary analysis adjusted for site, baseline prescribing, clustering, or other covariates. This is relevant because the outcome is defined across implementation sites, but the randomization unit and analysis structure are not specified. No particular adjustment or design feature should be inferred from the trial description alone.

Multiplicity procedures are not reported. The study description includes multiple aims and outcomes, and a recalculated secondary comparison has an unadjusted p-value. Without a reported plan for managing multiple outcomes or tests, the chance of at least one apparently notable result across a set of comparisons is difficult to evaluate. The survey comparison should not be interpreted as if it were adjusted for all analyses.

The registry also does not report a detailed statistical analysis plan, interim analyses, the handling of protocol deviations, or a sample-size justification. These omissions limit assessment of whether analyses followed a prespecified plan and how much weight to place on the posted results. They do not establish that such methods were absent from the conduct of the study; these details are simply not reported in the registry record.

11. 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 reports zero serious adverse events and zero other adverse events in the active-feedback group, with 36 listed as at risk in each adverse-event category. It reports zero serious adverse events and zero other adverse events in the passive-feedback group, with 38 listed as at risk in each category. Deaths are reported as zero in both groups.

These figures mean that no events in those categories were recorded in the posted registry results. They should not be read as proof that no events occurred outside the recorded categories or observation process. The registry does not provide additional detail about event definitions, collection methods, duration of safety follow-up, or completeness of reporting. Zero observed events also do not establish that the approaches have identical safety profiles, particularly when the record does not report an inferential comparison for safety.

The adverse-event figures are separate from the prescribing outcome. They do not provide denominators for calculating PIM percentages and do not resolve the primary endpoint’s missing rate comparison. The appropriate reading is descriptive: the posted record lists no serious events, other events, or deaths in either group.

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

The trial addresses a practical implementation question: whether active, in-person provider feedback differs from monthly electronic feedback when both are used with EQUIPPED. The registry describes a randomized parallel-group comparison and identifies PIM prescribing among ED discharge prescriptions for older Veterans as the primary outcome. These features make the intended contrast clear, but the posted primary numbers do not allow the central percentage outcome to be compared directly.

The primary values are counts, not reported percentages. Without denominators, readers cannot determine what share of prescriptions were PIMs in either group or estimate the difference in those shares. The record therefore does not show that active feedback reduced PIM prescribing relative to passive feedback, nor does it show that passive feedback was better. It also does not show that the strategies were equivalent. Those conclusions require an analyzable comparison of the stated endpoint.

The secondary perceived-control comparison provides a separate, unadjusted result. It describes a mean-score difference between groups, with a confidence interval that indicates uncertainty around that estimate. It is not a measure of prescribing and should not be used to infer the primary result. Since the record does not report multiplicity handling or sponsor analyses for that comparison, the secondary finding has limited confirmatory weight.

The posted data also do not explain the unit of randomization, analysis population, missing-data handling, or statistical model for prescribing counts. Without these details, it is not possible to assess whether site structure, prescription volume, or other features were addressed in the primary analysis. The posted safety counts are descriptive and do not supply that missing information.

Overall, the record documents the intervention contrast and some group-level outcomes, but it does not establish a between-group difference in the primary PIM percentage. This is a limit of what can be concluded from the registry results, not a claim about what analyses may have been conducted outside the record.

13. Why This Trial Matters Statistically

This trial is statistically important because implementation strategies are not simply drugs or procedures applied to individual patients. The comparison asks whether the way providers receive feedback—personal academic detailing or a monthly dashboard—relates to prescribing practice across ED implementation sites. The endpoint is intended to quantify a meaningful prescribing pattern using the proportion of discharge prescriptions classified as PIMs.

For such an outcome, a count alone is insufficient when the research question concerns a percentage. The number of PIM prescriptions must be interpreted in relation to the total prescriptions assessed. If prescription volumes differ between groups, raw counts can give a misleading impression of the underlying rate. The registry record does not include the denominators or an estimable rate comparison, illustrating why endpoint definitions and posted data structures need to align.

The trial also highlights the importance of design and analysis details in implementation research. The record reports randomization and parallel groups, but not the randomization unit or how the site structure was handled statistically. Those details can affect how uncertainty should be quantified. Likewise, the record’s lack of reported missing-data methods, analysis populations, and multiplicity procedures constrains interpretation of secondary findings.

The perceived-control outcome offers a distinct example of why effect estimates need context. A mean difference and confidence interval describe a secondary survey contrast, but neither answers whether the PIM share differed. Keeping outcome-specific conclusions separate protects against treating a result on one construct as proof about another.

The central statistical lesson is that a study’s design question, endpoint definition, and reported analysis must match. The registry makes the implementation contrast and intended primary measure understandable, while the missing denominator and primary comparison prevent a direct conclusion about the targeted prescribing percentage.

14. 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.

A key limitation specific to this record is the mismatch between the primary endpoint’s stated form and the posted arm summaries. The endpoint is a percentage of PIM prescriptions, while the numerical values are recorded as counts of units. The denominators and an estimable between-group comparison are not reported in the registry record. This prevents assessment of the principal rate contrast and means the raw counts should not be read as rates.

The registry reports the trial as randomized and parallel, but the unit of randomization is not reported. The study involves implementation sites, yet the record does not state how site-level structure was incorporated into the analysis or whether observations were clustered. No model details are reported for the primary outcome. As a result, the appropriateness and precision of any count-based analysis cannot be evaluated from the registry entry.

Enrollment was 74, and the arm summaries list 36 for active feedback and 38 for passive feedback. The record does not identify the analysis population for each outcome or explain whether these counts correspond to all observations used. It also does not report missing-data procedures, exclusions, or sensitivity analyses. These gaps make it difficult to judge the impact of incomplete or unavailable observations.

The available secondary comparison is unadjusted and has no reported sponsor analysis. The registry does not report a multiplicity plan, and the micro-costing outcome has no numerical results in the record. It also does not report the original sample-size justification, a detailed analysis plan, or whether interim analyses were conducted. These are limitations of the information available for interpreting the posted record; they do not prove that the trial team did not address these matters elsewhere.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Reducing Potentially Inappropriate Medication Prescribing for Older Patients in the ED — NCT04004936.
• Registry record last updated 2026-02-25; status last verified 2026-02.
• Results first posted 2024-08-06.
• 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.

18. Record Summary

The registry describes a completed randomized, parallel-group trial with no masking and enrollment of 74. It compared EQUIPPED with active provider feedback against EQUIPPED with passive provider feedback. The primary outcome was the percentage of PIM prescriptions among ED discharge prescriptions for Veterans aged 65 and older, assessed 12 months after implementation.

The posted primary arm values are counts of units, 1672 for active feedback and 2979 for passive feedback; denominators and an estimable comparison of the stated percentage outcome are not reported. A secondary perceived-control comparison reports group means of 4.1 and 3.3 and an unadjusted recalculated mean difference of -0.8. The registry reports no serious adverse events, other adverse events, or deaths in either group. Details on analysis populations, missing data, multiplicity, and the primary statistical model are not reported in the registry record.