Improving Patient Satisfaction Improving Patient Satisfaction: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Improving Patient Satisfaction Improving Patient Satisfaction (NCT01263639), a randomized na study. The primary outcome is Patient Satisfaction as Measured by Giving an "Excellent" Score on a 5-point Rating.

2. Clinical Question

3. Trial Design

The registry describes a prospective quality-improvement initiative using randomized intervention assignment. It lists randomized allocation, a parallel-group model, and triple masking. The registry does not explain which people were masked or how masking was maintained, so the label alone does not establish that participants, clinicians, survey callers, or analysts were all unaware of assignment. The phase is listed as not applicable.

The setting was a Level 1 academic trauma center. The planned contrast was an attending surgeon biosketch card versus standard care without that card. The intervention description identifies a photograph and brief information about the surgeon’s background and interests; the arm description places delivery within 24 hours of admission. Recruitment procedures, eligibility criteria in detail, and the dates of individual enrollment are not reported in the registry record.

The registry lists an enrollment of 228 and a study start of 2011-01, primary completion of 2011-12, and status of completed. The study summary separately states that 212 patients were eligible, with 100 randomized to the intervention group and 112 randomized to control. It also reports that 76 patients could be reached for the satisfaction survey, including 34 in the intervention group and 42 in control. These figures describe different stages or registry fields; the record does not explain the discrepancy between the enrollment field and the eligible count.

The outcome was a satisfaction survey asking about the overall quality of inpatient care delivered by the attending surgeon. The survey was to be conducted within 2 weeks after discharge and before the first clinic appointment. The report describes calls by the Professional Resource Group as part of the medical center’s regular quality-improvement activity. The registry does not report the survey wording, the full five response categories, or detailed procedures for handling unanswered questions.

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

Reasons for not completing this period:

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Patient Satisfaction as Measured by Giving an "Excellent" Score on a 5-point Rating, measured over within 2 weeks of discharge and before first clinic appointment.

The primary endpoint was the number of participants giving an Excellent score on a 5-point rating of the overall quality of inpatient care delivered by the attending orthopaedic surgeon. The survey focused on the patient’s satisfaction with interactions with that surgeon, rather than a clinical outcome such as complications, recovery, or length of stay. The intended assessment window was within 2 weeks of discharge and before the patient’s first clinic appointment.

Using an Excellent rating as the endpoint turns a multi-category satisfaction response into a binary outcome: a response either meets the Excellent threshold or it does not. This can make a specific high rating straightforward to compare between groups, but it does not retain distinctions among the other possible ratings. The registry does not provide the complete response distribution, so the posted outcome cannot show whether patients who did not give an Excellent rating selected adjacent or substantially different ratings.

The outcome is relevant to the study’s stated aim because the intervention was intended to improve patients’ understanding of their surgeon and the satisfaction rating directly asks about the attending surgeon’s inpatient care. It is nevertheless a patient-reported rating, not an independent measure of technical care or a direct assessment of the patient-physician relationship. The registry reports the count giving an Excellent score in each group and the associated comparison; it does not report a separate endpoint for understanding the surgeon.

7. Results Posted in the Registry

• Intervention Group, Biosketch Card: 73.5% (25/34) versus 52.4% (22/42) on control; risk difference 21.1 percentage points (95% CI -0.0 to 42.3), risk ratio 1.40 (0.99–2.00); unadjusted two-proportion z-test p = 0.059.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary for analysis: each reached participant either gave an Excellent rating or did not. For a binary outcome, a group’s event count and denominator describe how often the event occurred in that group. The registry reports 25 Excellent ratings among 34 reached intervention participants and 22 among 42 reached control participants. These are respondent-level results; they should not be confused with the larger group counts described for eligibility and randomization.

What does a risk difference express?

A risk difference subtracts the event proportion in one group from the event proportion in the comparison group. It is an absolute measure: it describes the separation between the groups in the proportion experiencing the defined outcome. The posted estimate is 0.211, with the comparison directed from control to intervention. The positive direction means the observed event proportion was higher in the intervention respondents. An absolute measure is often useful for communicating how much the outcome frequency differs between groups, but its estimate remains dependent on the outcome definition and the population represented by the denominators.

How do relative risk and odds ratio differ?

The registry also posts a relative risk of 1.404 and an odds ratio of 2.525. Relative risk compares event proportions directly. An odds ratio compares the odds of the event, where odds relate the event count to the count without the event. Because odds and probabilities are not the same quantity, an odds ratio should not be described as a direct multiplication of the event probability. Relative measures can communicate the direction and scale of an association, while the risk difference conveys the absolute group contrast. Each measure answers a related but distinct question.

What does the p-value say?

The unadjusted p-value is 0.059. In general, a p-value is calculated under a null model that specifies no group association and summarizes how compatible the observed data are with that model. A smaller p-value indicates that data at least as incompatible with the null model would be less common if the model held, but it does not measure the probability that the null is true. It also does not describe the size of the group difference or whether that difference matters to patients. The registry does not identify the exact test used, so the test procedure should not be inferred from the p-value alone.

The sample-size planning assumptions list a two-sided significance level of 0.05. That value is an assumption for the planning illustration; the registry does not report the trial’s original analysis plan or confirm that the same threshold governed the primary analysis. The p-value of 0.059 is close to that planning value, but the interpretation should not be reduced to a threshold label. The effect estimate, its confidence interval, the follow-up data available, and the endpoint’s meaning all matter.

Why are confidence intervals useful?

A confidence interval gives a range of effect estimates compatible with the data under the method used to construct it. It conveys uncertainty around an estimate, rather than only whether a test crossed a chosen threshold. For the risk difference, the reported interval includes zero. For the relative risk and odds ratio, the reported intervals include 1, their respective no-association values. The confidence level and interval-construction method are not reported in the registry record, so those details cannot be evaluated from the posted summary alone.

What does unadjusted mean?

An unadjusted comparison is not reported as controlling for other patient or clinical characteristics. Random allocation can support a causal comparison when assignment, outcome collection, follow-up, and analysis are appropriately conducted. It does not by itself remove concern about incomplete follow-up or guarantee that a respondent-only analysis preserves the intended comparison. The registry provides no adjusted primary result, covariate list, or analysis-population definition. It also reports no analyses in the analysis listing, so the exact statistical procedure and any model diagnostics are not described.

For binary outcomes, general methods may assume that observations are independent and that the model adequately represents the outcome distribution. If outcomes are more variable than a model expects, confidence intervals can be too narrow and tests can be too optimistic unless the method accounts for that extra variation. That is a general statistical consideration, not a finding about this study: the registry does not report a model-based analysis, an over-dispersion assessment, or diagnostics. The posted aggregate counts permit the stated summary comparison, but do not reveal how individual-level data were analyzed.

10. Confidence Intervals and Precision

Precision concerns how much uncertainty surrounds an estimated group contrast. The posted risk difference is 0.211, and its reported confidence interval extends from -< 0.001 to 0.423. Since that interval includes zero, the posted data do not tightly distinguish a positive difference from no difference; the interval also allows a range of positive effects. Its breadth indicates that the estimate is uncertain. The confidence level used for this interval is not reported in the registry record.

The relative measures tell a similar story about uncertainty in a different scale. The relative risk estimate is 1.404, and the odds ratio estimate is 2.525, but the confidence interval for each includes 1. For both measures, 1 represents no relative difference between groups. The estimates point in the same direction as the risk difference, while their intervals indicate that the magnitude is not precisely established by the posted respondent data.

Precision is affected by the amount of outcome information available, not only by the apparent size of the estimate. The primary comparison uses the reached survey respondents: 34 in the intervention group and 42 in control. The summary reports larger eligibility and randomized-group counts, but those counts are not the denominators for the posted satisfaction responses. Without outcomes for patients who could not be reached, the uncertainty intervals do not resolve possible selection related to follow-up.

Confidence intervals are not a guarantee that the true effect falls within the displayed range, and they do not account automatically for every design or missing-data concern. They summarize sampling uncertainty under the statistical method used. Because the registry record does not specify the confidence level, interval method, or full analysis procedure, precision can be interpreted from the posted interval’s relation to the no-difference value and its breadth, but not fully audited.

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.

The sample-size planning values are participants per arm, not total events. The registry’s planning illustration is based on the observed effect and assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It considers planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The original trial design assumptions and any original sample-size justification are not reported in the registry record.

The 80% power column gives the number of participants per arm in the planning illustration for a future study designed to have an 80% chance of detecting the specified effect under its assumptions. The 90% power column gives the corresponding per-arm planning number for a 90% chance. These are prospective design calculations, not descriptions of what happened in this trial. The entries are 81 and 108 participants per arm for planning around 100% of the observed effect; 149 and 198 for 75% of the observed effect; and 342 and 457 for 50% of the observed effect, respectively.

Smaller target effects require more participants in this illustration because weaker differences are harder to distinguish from sampling variation. The curve should be read as the same planning relationship across effect sizes, not as a record of accrued enrollment or observed outcomes. These calculations use an effect estimated from the posted results and are sensitive to that estimate. They are not observed power: observed power is not a useful substitute for interpreting the estimated effect and its uncertainty after data collection.

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 = 0.524; treatment proportion p₁ = 0.735; average p̄ = 0.630.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6830 = 1.3386.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6664 = 0.5608.
4. n = (1.3386 + 0.5608)² / (0.2115)² = 80.7, rounded up to 81 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 <- 0.523809523809524                          # control proportion
p1 <- 0.735294117647059                          # 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(1263639)
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 = 81 (80% power), 108 (90% power)
 75% of observed effect: n per arm = 149 (80% power), 198 (90% power)
 50% of observed effect: n per arm = 342 (80% power), 457 (90% power)
power.prop.test: 80.7 
Simulated power at n = 81 per arm: 0.805 (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 one primary outcome measure and no posted secondary outcome measures. It also lists no analyses. This does not establish that only one analysis was conducted; it means that a detailed analysis listing is not present in the registry record. The statistical plan, the number of outcomes considered during study conduct, and any planned or unplanned subgroup analyses are not reported.

Multiplicity arises when several outcomes, comparisons, time points, or subgroups are tested. Repeated testing can increase the chance of at least one apparently unusual result, even when there is no underlying group effect. Statistical approaches can address this, but the registry does not report a multiplicity strategy or enough information to determine whether one was needed. The primary comparison should therefore be interpreted as the posted result, without assuming that it was the only analysis considered.

The summary says 212 patients were eligible and identifies 100 randomized to intervention and 112 randomized to control. It reports that 76 patients could be reached for the follow-up survey, with 34 intervention respondents and 42 control respondents. This difference between the randomized groups described and the outcome respondents is important for understanding the analysis population. The record does not say whether the analysis followed intention-to-treat, included only respondents, or used another defined population. It does not report the reasons patients could not be reached or compare their characteristics with those reached.

Missing outcome data can affect both the estimated group contrast and its uncertainty if follow-up is related to assignment, satisfaction, or other factors associated with the rating. The registry does not report imputation, weighting, sensitivity analyses, or another method for addressing unavailable survey responses. It also does not report the full rating distribution or a participant flow description sufficient to reconcile each stage. These omissions limit assessment of how representative the posted respondent results are of the randomized groups.

14. Safety Results

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

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 serious adverse events and other adverse events separately for the intervention and control groups. For serious adverse events, the reported figures are 0 among 100 at risk in the intervention group and 0 among 112 at risk in the control group. For other adverse events, the figures are also 0 among 100 at risk in the intervention group and 0 among 112 at risk in control. These are registry event counts, not results from the satisfaction survey denominators.

A zero count in a reported group means no event was recorded in that category for the listed group and risk count. It does not establish that an event is impossible, nor does it quantify the chance of an event beyond the observation and reporting captured in the record. The registry does not report the adverse-event collection period, definitions, ascertainment procedures, or whether the event reporting covered the same participants as the satisfaction outcome.

Death counts are not reported in the record. The zero adverse-event counts should not be interpreted as a comparison proving equivalent safety between groups; the record does not provide a powered safety analysis or a formal comparison for these outcomes. The intervention was a card, but the event figures remain registry-reported counts and do not by themselves support a broader safety conclusion.

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

The posted satisfaction results show that 25 of 34 reached patients assigned to the biosketch-card group gave an Excellent rating, while 22 of 42 reached control patients did so. The estimated risk difference is positive, and the reported relative risk and odds ratio also point toward a higher rate of Excellent ratings in the card group among survey respondents. The estimates are uncertain: the confidence intervals include their respective no-difference values, and the unadjusted p-value is 0.059.

This is not a conclusion that the card does or does not improve satisfaction. The result concerns a specific rating threshold, a specific care setting, and respondents reached for a survey. It does not describe all satisfaction ratings, and it is not a direct measure of whether patients understood the surgeon’s background. It also does not measure clinical outcomes or establish the quality of care beyond the patient-reported endpoint.

The difference between the randomized-group counts described in the summary and the smaller respondent groups matters. If the patients who could not be reached differed in their ratings or in ways associated with their ratings, results among respondents might not represent the randomized groups. No missing-data analysis or sensitivity analysis is reported to examine this possibility. The confidence intervals quantify uncertainty under their statistical method, but do not automatically account for this potential selection.

The trial’s setting and intervention description provide useful context for interpreting the question, but the registry record does not establish whether the same results would apply elsewhere. Details of the original statistical analysis plan, survey properties, and response patterns are not reported. The appropriate reading is therefore that the posted data suggest a direction of difference among respondents, while the magnitude and broader applicability remain uncertain.

16. Why This Trial Matters Statistically

This trial is statistically interesting because its primary outcome is a patient-reported binary endpoint derived from a rating scale. Turning the scale into an Excellent-versus-not-Excellent outcome makes the comparison easy to define, but it discards distinctions among all the other ratings. The interpretation consequently depends on the chosen cutoff, not simply on satisfaction as a broad concept.

The study also illustrates why the denominator matters. The registry summary describes eligible and randomized patients, while the posted outcome uses those reached for follow-up. Readers should base the event comparison on the respondents who actually contribute outcome data and separately consider what the unavailable responses may mean. A difference in follow-up can undermine the representativeness of the observed contrast even when assignment was randomized.

Several effect measures are posted for the same binary outcome. The risk difference communicates the absolute contrast, while relative risk and odds ratio express relative contrasts. Their estimates are all positive in direction, but the intervals include the relevant null values. Looking across these measures and their uncertainty is more informative than focusing only on the p-value.

Finally, the registry’s planning illustration makes clear that sample-size requirements depend on the effect a future study aims to detect. Smaller target effects require larger per-arm planning numbers under the stated assumptions. Because that illustration uses the observed effect rather than the original trial design assumptions, it should be read as a planning aid and not as evidence about the trial’s original power or its result.

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.

A central limitation is the gap between the randomized groups described and the smaller groups with posted survey outcomes. The registry says 76 patients were reached, with 34 intervention and 42 control respondents, but does not explain why other patients could not be reached or report their outcome data. If response or reachability was associated with satisfaction, the respondent comparison could be biased. The record gives no missing-data analysis to assess the direction or extent of that risk.

The registry enrollment field reports 228, whereas the study summary states that 212 patients were eligible. It does not explain how enrollment relates to eligibility, randomization, or follow-up. This limits reconstruction of participant flow. The outcome analysis population is also not formally defined, and no intention-to-treat or other population label is reported.

The endpoint is an Excellent rating on a 5-point scale. The full scale distribution and exact survey instrument are not reported, so the posted comparison cannot show how responses across the other rating levels differed. The survey was intended within 2 weeks of discharge and before the first clinic appointment, but actual assessment timing for respondents is not reported. The record does not report a separate measure of patient understanding, despite that being part of the intervention rationale.

The registry lists no detailed analysis plan and no analyses in its analysis listing. The test procedure, confidence level, adjustment strategy, and any checks of assumptions are not reported. The original sample-size rationale is also not reported; the available planning illustration is based on the observed effect. Finally, the findings come from a single academic trauma-center setting as described in the summary, and the registry does not report evidence establishing applicability to other settings.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Improving Patient Satisfaction Improving Patient Satisfaction — NCT01263639.
• Registry record last updated 2014-07-28; status last verified 2014-06.
• Results first posted 2014-07-28.
• 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

The registry record describes a completed randomized, parallel-group quality-improvement study at a Level 1 academic trauma center, comparing an orthopaedic attending biosketch card with standard care without the card. The listed allocation is randomized and masking is listed as triple, but details of masking are not reported. The registry enrollment field is 228; the summary states that 212 patients were eligible, with 100 randomized to intervention and 112 to control.

The primary outcome was the number of participants giving an Excellent score on a 5-point rating of inpatient care by the attending surgeon, intended to be collected within 2 weeks after discharge and before the first clinic appointment. The posted outcome includes 34 reached intervention respondents and 42 reached control respondents. The comparison reports a positive risk difference, an unadjusted p-value of 0.059, and confidence intervals that include the respective no-difference values. The record does not report secondary outcomes, a detailed analysis plan, or a missing-data analysis.