A Clinician-Focused Nudging Intervention to Optimize Post-Surgical Prescribing: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Clinician-Focused Nudging Intervention to Optimize Post-Surgical Prescribing (NCT05299528), a randomized na study. The primary outcome is Percentage of Patients Contacted Who Agree to Participate in Study.

2. Clinical Question

3. Trial Design

This completed pilot was described as a single-site randomized controlled trial. The registry lists a randomized allocation, a parallel-group model, and no masking. The listed phase is NA. Enrollment is 465. The registry reports a start date of 2022-12-01 and a primary completion date of 2023-08-11.

The intervention was behavioral and named Direct Feedback. Surgeon participants assigned to that group were to receive procedure-specific feedback about patients’ opioid consumption-to-prescription ratio through electronic and written communication. The No Direct Feedback description says surgeons in that group would not be contacted directly during the study and would not receive the feedback. The registry describes the study as assessing feasibility and acceptability; it does not establish from these results whether feedback changed prescribing.

The primary outcome counts patients who agreed to participate among those contacted. The registry lists Phase 1 Patient Participants, Phase 2 Patient Participants in the Direct Feedback Group, and Phase 2 Patient Participants in the No Direct Feedback Group. Their reported enrollment counts for this outcome are 228, 42, and 66, respectively. The registry also states that patients whose surgery was performed by a surgeon who received the nudge intervention were analyzed only for safety, not contacted about study participation, and not included in the primary outcome population.

The registry reports 3 outcome measures and no posted secondary outcomes. The original protocol’s sample-size justification, its planned statistical analysis, and any interim analysis plan are not reported in the registry record. Consequently, the posted estimates can be described, but the record does not establish what analyses were prespecified or how the reported phase comparisons were intended to relate to randomization.

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: 94 surgeons were identified as potential participants in the study. 27 did not meet eligibility criteria. 20 did not agree to participate. 47 were consented and enrolled into the study. 12 surgeons met criteria to be randomized following Phase 1 Patient Recruitment period. Patients who underwent surgery performed by a surgeon who received the nudge intervention were patients who were analyzed only for safety and were not included in other groups. There was a waiver of consent for this group.

Only the first 6 of 7 groups are shown.

Surgeon Recruitment

Phase 1 Patient Recruitment

Reasons for not completing this period:

Surgeon Randomization

Phase 2 Patient Recruitment

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Total number of potentially eligible surgeons who were eligible

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.

Only the first 6 of 8 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Patients Contacted Who Agree to Participate in Study, measured over At consent (single time during study days 15-74 (group 1) or 122-181 (group 2)).

The primary endpoint is the percentage of patients contacted who agreed to participate in the study. It is reported as counts: 135 of 228 Phase 1 patients, 27 of 42 Phase 2 patients in the Direct Feedback group, and 31 of 66 Phase 2 patients in the No Direct Feedback group. The corresponding reported proportions are 0.592, 0.643, and 0.470.

This endpoint measures recruitment feasibility: whether people approached about participation agreed to join. For a pilot study, recruitment and consent can help characterize whether it is feasible to conduct the study as intended. The endpoint does not measure the amount of opioid prescribed, the amount consumed, or a change in the prescribing-to-consumption relationship. Those outcomes are not the primary endpoint.

The endpoint’s population definition matters for interpretation. The registry says that patients whose surgeons received the nudge intervention were analyzed only for safety and were not contacted about the study. Those patients were therefore not included in the primary participation outcome. The outcome concerns contacted patients, rather than every patient treated by a participating surgeon.

7. Results Posted in the Registry

• Phase 2 Patient Participants in the Direct Feedback Group: 64.3% (27/42) versus 59.2% (135/228) on control; risk difference 5.1 percentage points (95% CI -10.8 to 20.9), risk ratio 1.09 (0.85–1.39); unadjusted two-proportion z-test p = 0.537.
• Phase 2 Patient Participants in the No Direct Feedback Group: 47.0% (31/66) versus 59.2% (135/228) on control; risk difference -12.2 percentage points (95% CI -25.9 to 1.4), risk ratio 0.79 (0.60–1.05); unadjusted two-proportion z-test p = 0.077.

8. Statistical Interpretation

9. Statistical Methods Explained

These results concern a binary outcome: each contacted patient either agreed to participate or did not. The registry presents the number who agreed and the number contacted in each listed group. A proportion expresses the count agreeing relative to the contacted group; the reported proportions are 0.592 for Phase 1, 0.643 for Phase 2 Direct Feedback, and 0.470 for Phase 2 No Direct Feedback.

What does the risk difference describe?

The risk difference compares the proportion in a Phase 2 group with the proportion in Phase 1. A positive estimate means the Phase 2 proportion is higher at the point estimate; a negative estimate means it is lower. The reported risk difference is 0.051 for Direct Feedback versus Phase 1 and -0.122 for No Direct Feedback versus Phase 1. These are estimates from the posted counts, not proof that the intervention caused a change. The registry record does not report an adjusted analysis.

How do relative measures differ?

The posted results also give risk ratios and odds ratios. A risk ratio compares the proportions directly, while an odds ratio compares the odds of agreement. These measures use different scales, so their numerical values should not be read as interchangeable with a risk difference. For Direct Feedback versus Phase 1, the risk ratio is 1.086, with an interval from 0.846 to 1.394; the odds ratio is 1.24, with an interval from 0.626 to 2.458. For No Direct Feedback versus Phase 1, the risk ratio is 0.793, with an interval from 0.601 to 1.048; the odds ratio is 0.610, with an interval from 0.352 to 1.058.

What do the p-values contribute?

The unadjusted p-values are 0.537 for Direct Feedback versus Phase 1 and 0.077 for No Direct Feedback versus Phase 1. In general, a p-value measures how compatible observed 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 does it describe the size or practical importance of a difference. The record does not report the test specification or the threshold used in the trial’s own analysis plan.

“Unadjusted” indicates that these p-values are reported without a multiplicity adjustment. The registry record does not say whether additional analyses were planned, whether there were interim looks at the data, or how multiple outcomes or comparisons were to be handled. The comparison p-values therefore should not be treated as a complete account of the trial’s statistical plan.

What does an interval add?

In general, a confidence interval conveys the uncertainty around an estimate under a statistical method and its assumptions. It gives a range of values compatible with the observed data at the interval’s confidence level. The registry does not report the interval level or detailed calculation method. The risk-difference intervals extend from -0.108 to 0.209 for Direct Feedback versus Phase 1 and from -0.259 to 0.014 for No Direct Feedback versus Phase 1. Each crosses zero, so the posted ranges include no difference and also include differences in one or both directions.

Counts and proportions also depend on the population included. The primary outcome applies to patients who were contacted. The registry explicitly excludes from that endpoint patients whose surgeon received the nudge intervention and who were analyzed only for safety. A participation proportion should not be generalized to that excluded population without further information.

Finally, the registry’s trial design is described as randomized and parallel, but the primary results compare Phase 2 groups with Phase 1. The record does not explain the relationship between phase and randomization in enough detail to establish that these contrasts are direct randomized treatment comparisons. Nor does it report covariate adjustment, missing-data procedures, or an analysis population definition beyond the endpoint description. These omissions limit what can be concluded about the statistical method behind the posted estimates.

10. Confidence Intervals and Precision

The interval around the Direct Feedback risk difference extends from -0.108 to 0.209. The interval around the No Direct Feedback risk difference extends from -0.259 to 0.014. Both include zero, the value corresponding to no difference on the risk-difference scale. The posted ranges therefore do not isolate a single direction of difference.

In general, a confidence interval is a way to express uncertainty about an estimate. A narrower interval indicates greater precision under the method used; a wider interval leaves a broader range of values compatible with the data. The No Direct Feedback interval spans a broad range, including zero and negative values. The Direct Feedback interval also allows a range that includes zero and values on either side. The confidence level and interval calculation method are not reported in the registry record, so the intervals should be described by their posted bounds rather than assigned an unstated confidence level.

The group counts are not equal: the primary outcome lists 228 Phase 1 participants, 42 Phase 2 Direct Feedback participants, and 66 Phase 2 No Direct Feedback participants. The smaller Phase 2 groups contribute less information to their comparisons than a larger sample might, which is one general reason estimates based on smaller groups can be imprecise. The intervals reinforce that caution: the point estimates alone should not be treated as precise estimates of the underlying difference.

The risk-ratio and odds-ratio intervals are also broad and include 1 for both comparisons. In general, 1 represents no relative difference on these scales. The record does not provide enough detail to assess the interval method or the full analysis population, so precision should be judged from the posted estimates and bounds without assuming additional modeling choices.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

These planning illustrations are for participants per arm, not total events. The outcome is binary, and the sample-size assumptions state that all non-time-to-event illustrations are expressed as participants per arm. The calculation assumes two-sided significance level 0.05, equal allocation between two arms, and a normal approximation. It uses a stated fraction of the observed effect as the effect to detect. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column means the number of participants per arm that the planning calculation estimates would be needed to have an 80% chance of detecting the specified effect under those assumptions, if that effect and the model assumptions were correct. The 90% column gives the corresponding planning estimate for a 90% chance. These are design illustrations, not a statement about what happened in the trial; this is not observed power.

For example, planning for 100% of the observed Direct Feedback comparison gives 1,439 participants per arm for 80% power and 1,925 participants per arm for 90% power. Planning for 75% of that observed effect gives 2,574 and 3,445 participants per arm, respectively. For the No Direct Feedback comparison, planning for 100% of the observed effect gives 260 participants per arm for 80% power and 348 for 90% power.

In general, a smaller effect is harder to distinguish from random variation, so detecting a smaller effect requires more participants under the same assumptions. The planning estimates rise as the assumed effect fraction decreases for each comparison. These figures are based on the observed effects in this pilot, not on a design justification documented by the trial investigators, and should be interpreted as illustrative rather than as the trial’s original target.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.592; treatment proportion p₁ = 0.643; average p̄ = 0.617.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6873 = 1.3471.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6864 = 0.5777.
4. n = (1.3471 + 0.5777)² / (0.0508)² = 1,438.3, rounded up to 1,439 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.592105263157895                          # control proportion
p1 <- 0.642857142857143                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(5299528)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 1,439 (80% power), 1,925 (90% power)
 75% of observed effect: n per arm = 2,574 (80% power), 3,445 (90% power)
 50% of observed effect: n per arm = 5,826 (80% power), 7,799 (90% power)
power.prop.test: 1438.3 
Simulated power at n = 1,439 per arm: 0.799 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists 3 outcome measures, while no secondary outcomes are posted. It reports two Phase 2 comparisons for the primary participation outcome: Direct Feedback versus Phase 1 and No Direct Feedback versus Phase 1. The posted p-values are labeled unadjusted. The registry record does not report whether these comparisons were prespecified as separate primary contrasts, whether a multiplicity procedure was planned, or how any additional outcome analyses were to be interpreted.

In general, testing multiple outcomes or comparisons can increase the chance of observing an apparently unusual result by chance. A multiplicity adjustment is one possible way to account for a family of tests, but the record does not report one for these estimates. The unadjusted p-values should therefore be read as individual comparison results, not as evidence that the full set of trial questions was controlled for multiplicity.

Missing-data methods are not reported in the registry record. The posted endpoint counts identify patients contacted and patients who agreed, but do not describe whether contact attempts with no recorded response were counted as non-agreement, excluded, or handled in another way. The registry’s population description also excludes patients whose surgeon received the intervention from the primary endpoint because they were not contacted. No further analysis-population definitions are reported.

The record does not report an intention-to-treat analysis, per-protocol analysis, covariate-adjusted model, or interim analysis plan. It also does not explain in detail how Phase 1 relates to the randomized Phase 2 groups. These gaps mean that the posted summary estimates can be described as reported, but the analysis population, handling of missing outcomes, and relationship to randomization cannot be independently reconstructed from the registry record.

14. 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 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: vomiting (3.7%), abdominal pain (1.3%), gastritis (0.4%), clostridioides difficile infection (0.4%), chest pain (0.4%).

The registry reports serious adverse event counts for three groups. In the Phase 2 Direct Feedback group, 1 of 27 patients at risk had a serious adverse event; in the Phase 2 No Direct Feedback group, 0 of 31 did. Among patients who underwent surgery performed by a surgeon who received the nudge intervention, 5 of 225 had a serious adverse event. Deaths were 0 in each of the three groups.

For other adverse events, the reported counts are 2 of 27 in Direct Feedback, 1 of 31 in No Direct Feedback, and 5 of 225 among patients whose surgeon received the nudge intervention. These counts describe the registry’s safety groups; the record does not report a comparative statistical analysis for them. The safety-only group is distinct from the contacted patients included in the primary participation endpoint.

The registry lists vomiting at 3.7%, abdominal pain at 1.3%, gastritis at 0.4%, Clostridioides difficile infection at 0.4%, and chest pain at 0.4% among the top serious adverse event terms. The record does not specify in this listing how those terms map to each group. Event counts alone do not establish a cause, and the groups and denominators differ.

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

The primary results show that 135 of 228 Phase 1 patients agreed to participate, compared with 27 of 42 Phase 2 patients in the Direct Feedback group and 31 of 66 Phase 2 patients in the No Direct Feedback group. The reported risk differences are 0.051 for Direct Feedback versus Phase 1 and -0.122 for No Direct Feedback versus Phase 1. Their intervals both include zero, and each comparison has an unadjusted p-value: 0.537 and 0.077, respectively.

These findings describe agreement among patients who were contacted. They do not establish that surgeon feedback changed opioid prescribing, postoperative consumption, or the relationship between prescriptions and consumption. The primary outcome is a participation measure, not a prescribing or clinical outcome. The registry describes the project as a pilot intended to assess feasibility and acceptability, and the posted participation results speak to that limited question.

The contrasts should also be interpreted in light of the record’s description of Phase 1 and Phase 2. It reports the study as randomized and parallel, but presents Phase 1 as the reference group for both Phase 2 comparisons without explaining the relationship between phase and assignment. It is therefore not possible from the registry record alone to attribute the observed differences to the feedback intervention.

The interval estimates leave substantial uncertainty about the size and direction of the group differences. The record does not report the confidence level or the original analysis plan, and the p-values are unadjusted. These results should not be treated as a definitive estimate of how feedback would affect participation in other settings or populations.

The safety data list adverse event counts in separate groups, including a group analyzed only for safety. They do not provide a reported causal comparison. Taken together, the posted results are best understood as descriptive pilot findings about participation and safety reporting, not as proof of a treatment effect or a basis for clinical decision-making.

16. Why This Trial Matters Statistically

This trial raises a statistical question that is easy to overlook: when an intervention is directed at clinicians, a patient-level participation outcome may describe a different part of the study process than the intervention’s intended effect on clinician behavior. The reported primary endpoint is agreement to participate, which can inform recruitment feasibility. It does not directly quantify prescribing behavior or opioid consumption.

The record also illustrates why study structure matters when interpreting estimates. The trial is described as randomized and parallel, but the primary results use Phase 1 as the reference for both Phase 2 groups. Without a fuller account of how these phases relate to randomization, it is difficult to know whether the contrasts capture only group assignment or also reflect other differences between phases. This is a design and reporting issue, not something that can be resolved by the point estimates alone.

Precision is another important lesson. The posted risk-difference intervals cross zero, and the Phase 2 groups have fewer participants than Phase 1. The estimates therefore warrant cautious interpretation rather than a simple reading of the point estimate as a stable effect. Planning illustrations based on a pilot’s observed effects can show how sample-size needs change with the assumed effect size, but they do not replace the original design justification or answer whether the pilot was adequately powered for a different endpoint.

Finally, the record reports no secondary outcomes and does not describe multiplicity, missing-data handling, or the original statistical plan. Those details are needed to place isolated comparisons within the intended analysis framework. The example underscores the value of separating feasibility endpoints, clinician behavior endpoints, patient outcomes, and safety reporting, and of making the analysis populations and planned contrasts explicit.

17. 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 registry record leaves important details of the trial design and analysis unclear. Although the study is listed as randomized and parallel, the primary outcome comparisons use Phase 1 participants as the reference for each Phase 2 group. The relationship between those phases and randomized assignment is not reported in detail. This limits interpretation of the contrasts as randomized comparisons.

The primary endpoint is agreement among contacted patients, not a direct measure of whether the intervention altered prescribing or opioid consumption. Patients whose surgeon received the nudge intervention were described as analyzed only for safety and not contacted about the study, so they were not included in the primary endpoint. The endpoint therefore does not represent participation among all patients whose care was potentially affected by the surgeon intervention.

The Phase 2 groups contributing to the primary endpoint include 42 Direct Feedback participants and 66 No Direct Feedback participants, compared with 228 Phase 1 participants. The reported intervals are broad and include zero. The record does not report a confidence level, missing-data procedures, covariate adjustment, or an analysis population beyond the endpoint description.

The registry does not report the original sample-size justification, prespecified analysis plan, interim analyses, or multiplicity strategy. The planning illustrations are based on observed pilot effects and do not document the investigators’ original design assumptions. The safety results are counts in groups with different denominators, and no comparative safety analysis is reported. These limitations mean the posted results are informative as pilot descriptions but do not establish a causal effect on prescribing or a precise estimate of participation differences.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Clinician-Focused Nudging Intervention to Optimize Post-Surgical Prescribing — NCT05299528.
• Registry record last updated 2024-10-08; status last verified 2024-09.
• Results first posted 2024-09-19.
• 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.

21. Record Summary

NCT05299528 is a completed, single-site pilot randomized controlled trial of a clinician-focused nudging intervention intended to optimize post-surgical prescribing. The registry lists random allocation, a parallel model, no masking, and enrollment of 465. The intervention was procedure-specific direct feedback to surgeons about patients’ opioid consumption-to-prescription ratio.

The primary endpoint was the percentage of contacted patients who agreed to participate. The posted counts were 135 of 228 in Phase 1, 27 of 42 in Phase 2 Direct Feedback, and 31 of 66 in Phase 2 No Direct Feedback. The reported risk differences versus Phase 1 were 0.051 and -0.122, with intervals that include zero.

The results address participation and safety reporting, not an effect on prescribing or opioid consumption. The registry record does not report the trial’s original sample-size justification, detailed analysis plan, missing-data procedures, or the relationship between the Phase 1 reference group and randomized assignment.