Behavioral Economics Applications to Geriatrics Leveraging EHRs: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Behavioral Economics Applications to Geriatrics Leveraging EHRs (NCT03704389), a na na study. The primary outcome is Misuse of Urinalysis or Urine Culture in Older Women.

2. Clinical Question

The study’s broader rationale concerns clinical decisions in older adults, with topics including testing and treatment for asymptomatic bacteriuria, prostate-specific antigen testing, and use of insulin or oral hypoglycemics for type 2 diabetes. The described primary outcome is narrower: it addresses urinalysis or urine culture in older women, among those who received a test. It is not a measure of all clinical decision support activity, all potentially inappropriate care, or patient health outcomes.

Because the record describes one group and no concurrent control, the primary question is chiefly descriptive in the available registry entry. A measured frequency could characterize testing patterns under the intervention period, but a causal comparison would require an appropriate contrast and analysis. Neither a primary outcome value nor such a comparison is reported in the registry record.

3. Trial Design

The study, titled Behavioral Economics Applications to Geriatrics Leveraging EHRs, is listed as completed and sponsored by Northwestern University. The recorded study start date is 2019-01-17, and the primary completion date is 2019-07-17. Enrollment is listed as 664. The phase and allocation are recorded as not applicable, masking is listed as none, and the model is single group.

The intervention is clinical decision support delivered through the electronic health record. Participating clinicians could receive any of three nudges when eligibility criteria were met within a patient’s chart. The registry describes an approach intended to address three topics of potential misuse in older adults. The description emphasizes the use of EHR-delivered prompts informed by social psychology, but the precise content of each nudge and its delivery frequency are not reported in the registry record.

The primary outcome timeframe is the 6 months intervention period. Its population is female patients aged 65 years and older with a primary care encounter and a urinalysis or urine culture, with the outcome focused on testing for non-specific reasons. The registry also lists prostate-specific antigen, asymptomatic bacteriuria, and type 2 diabetes mellitus as conditions. Those conditions provide context for the study’s broad scope; the primary outcome described in the record is specifically about urine testing in older women.

This is not a randomized comparison: the record lists a single group and no allocation. The enrollment figure describes the study’s recorded enrollment, not the number of patients who necessarily met the primary outcome definition or had complete outcome data. Those numbers are not reported in the registry record. Likewise, the registry record does not report a separate control group, a comparison estimate, or details sufficient to characterize how exposure to the nudges varied across clinicians or patients.

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: This is a small non-randomized pilot study. This trial enrolled patients with a waiver of informed consent. Clinicians were not enrolled. All outcome data were collected via the electronic health record. This explains why the number of participants started is the same as the number of participants completed.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Proportion of female patients aged 65 years and older with a primary care encounter who had a urinalysis or urine culture done for non-specific reasons among those with a test done

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: We report age and sex for PSA and UA/UC outcome measures. We have not collected sex for patients eligible for diabetes outcome measure.

6. Primary Endpoint

The registry lists the primary outcome as Misuse of Urinalysis or Urine Culture in Older Women, measured over 6 months intervention period.

The primary endpoint is the percentage of female patients aged 65 years and older who had a urinalysis and/or urine culture for suspected urinary tract infection without genitourinary-specific symptoms. The registry’s population description frames the denominator as older female patients with a primary care encounter who had a urine test, and the numerator as those whose test was performed for non-specific reasons. This makes the endpoint a measure of potentially non-symptom-specific testing among patients who were tested, rather than a measure of testing among all older women seen in primary care.

The endpoint is relevant to the study’s stated concern about misuse of services in older adults. The registry summary identifies testing and treatment for asymptomatic bacteriuria as an example of care where practice may not align with guidance for older adults. A symptom-based measure can describe whether urine testing occurred in the absence of genitourinary-specific symptoms, which is the behavior the endpoint is designed to capture.

The endpoint does not itself establish why a test was ordered, whether a clinician received a nudge, whether the test was clinically justified in an individual case, or what happened to the patient afterward. It is also not a direct measure of adverse health outcomes or of the overall quality of care. The registry record does not report the observed endpoint value, the number of eligible patients, or the operational details used to classify symptoms and testing reasons. Without those results and definitions, the endpoint’s observed frequency and precision cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical explanation: a binary outcome classifies each eligible observation into one of two categories, such as whether testing occurred without genitourinary-specific symptoms or not. A proportion summarizes the number in one category relative to the number eligible for that particular outcome. For this study, the relevant denominator is not automatically the full enrollment; it is the group described in the primary outcome definition. The registry record does not report that denominator or the resulting proportion.

What a proportion can describe

General statistical explanation: a proportion gives a measure of frequency within a defined population. Its meaning depends on how the population, outcome, and observation period are specified. Here, the outcome concerns older female patients who had a primary care encounter and received urinalysis or urine culture. It is therefore not a measure of the frequency of urine testing among all patients, nor a count of all instances of care considered potentially unnecessary. Clear eligibility and classification rules are important because a change in who is included can change the proportion even when the underlying clinical behavior has not changed.

The registry record identifies a 6 months intervention period and describes the endpoint in terms of testing without genitourinary-specific symptoms. It does not report the numerical outcome, the number meeting the outcome criteria, or detailed classification procedures. It also does not report the handling of repeated encounters or tests for the same patient. These details affect how an outcome estimate should be interpreted, and are not reported in the registry record.

Uncertainty and confidence intervals

General statistical explanation: a confidence interval expresses uncertainty around an estimate under a specified statistical model. A narrower interval generally indicates greater precision, while a wider interval indicates less precision. An interval is not a guarantee about the true value, and its interpretation depends on the analysis assumptions and the way observations were selected and measured. No confidence interval for the primary endpoint is reported in the registry record, so its precision cannot be assessed.

For a proportion, statistical methods commonly account for the number of eligible observations and the observed frequency. The enrollment total of 664 is not enough to reconstruct an endpoint interval because the number of eligible older women with testing and the number meeting the endpoint definition are not reported. No recalculated estimate can therefore be obtained from the registry information described.

Single-group interpretation

General statistical explanation: a single-group study can describe outcomes observed among participants exposed to an intervention, but without a concurrent comparator it cannot directly estimate the difference between intervention and control conditions. A causal effect asks what would have happened to comparable patients under an alternative condition. That counterfactual is not observed merely by recording an outcome in one group. The registry lists the design as single group, with allocation not applicable, and does not report a control group or a comparative effect estimate.

Even a change from a pre-intervention measure to a later measure would need careful interpretation. General considerations include changes in the population, clinical practice, record documentation, and the availability of tests over time. The registry record does not report whether such a comparison was planned or conducted, and no before-and-after result is reported. It would therefore be inappropriate to infer an intervention effect from the design description alone.

Planning illustrations and analysis choices

The record describes 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. It also states that the effect to detect is expressed as a fraction of the observed effect. These are planning assumptions, not evidence that the study used a two-arm design: the registry records a single group and no allocation. The illustrations are not observed power and cannot replace an analysis of trial outcomes. Since the registry record does not report the primary outcome result or an observed effect estimate, the basis for interpreting effect-dependent planning values is not available in the registry record.

General statistical explanation: over-dispersion occurs when outcome variability is greater than a model assumes. In count or proportion analyses, ignoring extra variability can make uncertainty appear smaller than it is. Whether over-dispersion is relevant depends on the outcome structure and model. The registry record does not report an analysis model or diagnostics, so it is not possible to say whether over-dispersion was assessed or whether it affected an analysis.

The registry record lists no analyses. It does not report a statistical analysis plan, model specifications, covariate adjustments, hypothesis tests, or methods for estimating uncertainty. These omissions mean that the statistical methods used for the primary outcome cannot be reconstructed from the registry entry. General methodological considerations can explain what such choices mean, but cannot establish which choices were made in this study.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Clinical Decision Support Intervention")
n      <- c(230)
events <- c(25)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Clinical Decision Support Intervention  230  25  10.9  7.2  15.6  7.5  15.6  7.5  15.6

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists three outcome measures, but the primary outcome description is the only outcome detail available in the record. No analysis results are listed. The record does not report how multiple outcomes or multiple statistical tests were handled. In general, evaluating multiple outcomes or tests can increase the chance of a chance finding if each is interpreted without accounting for the number of analyses. Whether and how that issue was addressed in this study is not reported in the registry record.

Missing-data methods and analysis populations are also not reported in the registry record. It does not state how missing symptom information, incomplete encounters, or unavailable test records were handled. It does not report whether the primary analysis included all enrolled participants, only those meeting the endpoint eligibility criteria, or another defined population. These details matter because exclusions and missing observations can affect both the estimated frequency and its precision.

The enrollment of 664 does not identify the analysis denominator for the primary endpoint. The endpoint is restricted to older female patients with a primary care encounter and a urine test, and the record does not report the number satisfying these criteria. It also does not give a flow of participants or encounters from enrollment to outcome assessment. Consequently, the extent and potential impact of missing or excluded observations cannot be assessed from the registry record.

No imputation, sensitivity analysis, multiplicity adjustment, or prespecified analysis population is described. These methods may or may not have been used; the registry record does not say. Without a reported analysis plan and outcome results, it is not possible to evaluate whether the reported statistical analysis, if any, aligned with prespecified choices.

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’s safety entries list the clinical decision support intervention group with 664 participants at risk for serious adverse events and 0 affected. For other adverse events, 664 are listed at risk and 0 affected. Deaths are listed as 0. No serious adverse event terms are listed.

These are counts in the registry record, not estimates of the primary urine-testing outcome. The entries report no affected participants in the listed categories; they do not establish that every possible event was systematically detected or that no unlisted event occurred. The record does not describe the event collection procedures, follow-up completeness, or whether events were judged related to the intervention. Causality should therefore not be inferred from the counts alone.

The 664 listed at risk is a safety denominator and should not be treated as the denominator for the primary outcome. The primary endpoint concerns a more restricted group of older female patients who had a primary care encounter and a urine test, and its denominator is not reported. Zero reported safety events also cannot fill the gap in the missing primary endpoint results or establish an effect on clinical practice.

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

The registry record describes a behavioral clinical decision support intervention delivered through electronic health records and identifies a primary outcome concerning urine testing in older women. It also reports enrollment of 664 and records zero serious adverse events, zero other adverse events, and zero deaths in the intervention group’s safety entries. These registry entries characterize the registered study and its reported safety counts.

The record does not report the observed primary outcome, its denominator, a confidence interval, or a statistical test. It therefore does not show the frequency of testing without genitourinary-specific symptoms, how precisely that frequency was estimated, or whether it changed during the intervention period. The record also does not report a comparison group or an estimated intervention effect. The available information cannot establish whether the nudges changed clinical decisions.

The primary outcome is a process measure, not a direct measure of patient health. Even if a process measure had been reported, it would not by itself establish effects on morbidity, mortality, or other clinical outcomes. The study’s background describes concerns about care in older adults, but the registry record does not provide results for all the clinical topics in its broader rationale.

Nor does the record give enough information to determine whether the outcome classification was complete, how missing information was handled, or which patients contributed to an analysis. Enrollment should not be mistaken for the endpoint denominator. The absence of those details limits assessment of both the estimate and its generalizability.

In short, the registry entry documents a completed single-group study and its intended primary outcome, but does not provide the primary result needed to assess the observed testing pattern. The safety counts should be read as recorded counts rather than proof of no risk, and the single-group design does not provide a direct treatment-versus-control estimate. No conclusion about an intervention effect is supported by the registry record alone.

14. Why This Trial Matters Statistically

This study addresses a statistical challenge common in quality improvement and clinical decision support research: the outcome is a behavior embedded in routine care, while the intervention is delivered through the same clinical information system in which that behavior is documented. The primary measure is not simply whether a test occurred. It focuses on urine testing among older women who were tested and classifies the reason in relation to genitourinary-specific symptoms. The denominator and the accuracy of that classification are therefore central to the meaning of any result.

The single-group design is also important. A descriptive estimate can characterize the intervention period, but a comparison is needed to distinguish a change associated with an intervention from changes that might have occurred for other reasons. The registry record does not describe a concurrent control or report a comparative result. That makes it especially important not to interpret a hypothetical frequency as a causal effect.

The entry also illustrates why enrollment alone does not establish statistical precision. The primary endpoint applies to a subset defined by sex, age, encounter, testing, and symptom status. The number meeting that definition is not reported. Without it, readers cannot assess how much information underlies the endpoint or whether the safety denominator overlaps with the outcome population.

Finally, the record’s missing primary result limits what can be learned from a completed study entry. The study’s rationale and design identify a relevant question, but understanding the evidence requires outcome estimates, uncertainty, analysis methods, and clear accounting of eligible observations. Those elements are not reported in the registry record.

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 limitation for interpreting this study from the registry entry is the absence of a reported primary outcome value. No outcome denominator, estimate, confidence interval, or test result is included. This prevents assessment of the observed frequency and its precision, even though the study is listed as completed.

The single-group design limits causal interpretation. No concurrent comparator is described, and the registry record does not report a comparative analysis. It is not possible to distinguish an intervention-associated change from other influences using the information recorded. The allocation is listed as not applicable, and masking is listed as none.

The primary outcome applies to a specific subset: female patients aged 65 years and older with a primary care encounter and a urine test. The registry does not report the number who met these conditions or explain how symptoms and reasons for testing were classified. It also does not report how repeated encounters, incomplete symptom documentation, or missing outcome data were handled. These omissions make the analytic population and measurement process difficult to assess.

Although the study description discusses three topics of potential misuse, the detailed primary outcome available in the registry record concerns urine testing in older women. The record does not give results for the other topics in the broader study rationale. It also does not report detailed intervention implementation, statistical methods, analysis populations, or multiplicity and missing-data procedures. These are limitations of what can be concluded from the registry record and do not establish what may have been documented elsewhere.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Behavioral Economics Applications to Geriatrics Leveraging EHRs — NCT03704389.
• Registry record last updated 2025-10-07; status last verified 2025-09.
• Results first posted 2025-10-07.
• 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

Behavioral Economics Applications to Geriatrics Leveraging EHRs is listed as a completed, single-group study sponsored by Northwestern University. The recorded start date is 2019-01-17, the primary completion date is 2019-07-17, and enrollment is 664. Masking is listed as none, and allocation and phase are listed as not applicable.

The intervention consists of clinical decision support nudges delivered through the electronic health record when eligibility criteria are met. The primary outcome concerns urine testing among female patients aged 65 years and older during the 6 months intervention period, focusing on tests performed without genitourinary-specific symptoms. The primary outcome result, eligible outcome denominator, statistical analysis, and comparative effect are not reported in the registry record.

The safety entries list 0 serious adverse events among 664 at risk, 0 other adverse events among 664 at risk, and 0 deaths. These counts do not provide a result for the primary endpoint or a controlled estimate of intervention effect.