Improving Communication Between Cancer Patients & Oncologists: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Improving Communication Between Cancer Patients & Oncologists (NCT02969031), a randomized na study. The primary outcome is Provider to Patient Communication Score.

2. Clinical Question

This is a question about measured communication scores, not a direct measure of every aspect of the patient experience or of cancer outcomes. The registry’s description frames the project’s goal as improving communication and the experience of people living with cancer. Whether the training approaches produced a clear difference on the posted primary score is a separate statistical question from whether the goal is important. The posted estimate is imprecise, and the registry record does not establish that the groups differed on this outcome.

3. Trial Design

The registry identifies NCT02969031 as a completed, randomized, parallel-group trial with no masking. It lists Cancer as the condition and behavioral interventions in both arms. The phase is recorded as NA. Dana-Farber Cancer Institute is listed as the sponsor. The registry records a start date of 2017-02-14 and primary completion on 2018-08-31.

The two approaches were Enhanced SCOPE Training and Standard Communication Training. In the enhanced arm, the description includes a baseline questionnaire, anonymous patient satisfaction surveys, and audio recording of eight clinic visits with eight different patients. The project summary describes feedback from professional research-assistant coders and trained patient reviewers listening to recordings. In the standard arm, the description includes a baseline questionnaire, patient satisfaction surveys, receipt of their results, and a quality-improvement activity designed by the oncologist. The full registry description of the enhanced arm’s feedback process is not reported in the registry record.

Registry enrollment is 148. For the posted primary outcome, the arm summaries list 15 for Enhanced SCOPE Training and 12 for Standard Communication Training. The registry record does not explain the relationship between these outcome counts and the enrollment figure, or report why the outcome counts are smaller. These figures should therefore be read as the counts attached to the posted outcome summaries, not as a complete account of participant flow.

The primary outcome timeframe is 6 months. The registry also lists a secondary outcome at 6 months: Empathic Response to Patient Expressions of Negative Emotions. Although the record lists a randomized design, it does not report the randomization procedure, allocation concealment, or the original design assumptions. These details are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Enrolled and Randomized

Collection of Baseline Surveys

Collection of Baseline Recordings

Completed Intervention

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 Provider to Patient Communication Score, measured over 6 months.

The primary endpoint is the Provider to Patient Communication Score, a mean of a continuous measure. It is intended to reflect patients’ perceptions of how attentively the medical oncologist responded during recalled office encounters over 12 months. Patients answered the CG-CAHPS Version 1.0 questionnaire by computer or interactive voice response phone system.

The score uses responses about whether the provider explained things in an understandable way, listened carefully, showed respect for what the patient had to say, and spent enough time with the patient. Responses are coded as 1 for “Yes, definitely,” 2 for “Yes, somewhat,” 3 for “No,” and 4 for “I prefer not to answer.” The registry describes the aggregate score as the average of the scores for the first three questions. It states that lower scores indicate a more desirable outcome.

This endpoint is useful for the project’s question because it operationalizes patients’ views of several communication behaviors as a score that can be compared between training groups. It is a patient-reported measure, rather than a direct clinical outcome. The registry says averages were modeled with a linear mixed-effect model and that the data were analyzed both with missing responses dropped and with missing responses kept in the dataset. It does not report separate outcome estimates for those two missing-response approaches.

7. Results Posted in the Registry

• Enhanced SCOPE Training: mean 1.04 versus 1.07 on control; difference -0.03 (95% CI -0.17 to 0.11), standardized effect d = -0.18; z-test p = 0.664.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Empathic Response to Patient Expressions of Negative Emotions

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Enhanced SCOPE Training: mean 0.21 versus 0.12 on control; difference 0.09 (95% CI -0.12 to 0.29), standardized effect d = 0.31; z-test p = 0.417.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Enhanced SCOPE Training and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists one secondary outcome, Empathic Response to Patient Expressions of Negative Emotions. Its measure is the mean number of empathic responses relative to empathic opportunities. It addresses a related but distinct aspect of communication from the primary patient-reported score: the primary measure reflects patients’ perceptions of attentive communication, while the secondary measure concerns responses to expressions of negative emotion.

The secondary comparison does not provide a clear contrast between the training groups: its interval includes zero, and its unadjusted p-value is 0.417. Thus the secondary result does not add a statistically clear signal that reinforces the primary comparison. At the same time, the two outcomes measure different constructs, and the secondary result cannot be treated as a direct replication of the primary measure.

Unadjusted secondary results generally carry less weight than a primary result because examining multiple outcomes creates more opportunities for chance findings. This concern can arise even when each individual analysis is conducted correctly. The registry lists one secondary endpoint, but does not report whether additional outcomes were considered, whether a multiplicity strategy was planned, or whether any adjustment was used. The posted secondary p-value is explicitly unadjusted, so it should be read as one descriptive piece of evidence rather than as confirmatory evidence on its own.

The registry record does not report secondary analyses beyond the posted comparison. There is therefore no basis for evaluating consistency across subgroups, time points, alternative outcome definitions, or adjusted models. The overall pattern is limited: the primary score and the single listed secondary measure both have uncertain comparisons, and neither establishes a clear between-group difference in the posted results.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Empathic Response to Patient Expressions of Negative Emotions

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a continuous score, and the registry reports arm means and standard deviations. A mean describes the central value of observations on a numeric scale. A standard deviation describes how spread out individual observations are around the mean; it is not the same as the uncertainty in the estimated mean or in the difference between group means.

What does a mean difference represent?

A mean difference compares the average score in one group with the average in the other. Here the posted comparison is Enhanced SCOPE Training minus Standard Communication Training, with a mean difference of -0.03. Since lower scores indicate a more desirable outcome, a negative difference points toward a lower average in the enhanced arm. This direction describes the observed contrast; it does not by itself show that the training caused a difference.

The comparison also includes an interval estimate and an unadjusted p-value. An interval estimate conveys uncertainty about the difference, under the assumptions used to construct it. In general, wider intervals indicate less precision, while narrower intervals indicate greater precision. The primary interval crosses zero, so the posted result does not distinguish a difference in one direction from no difference or a difference in the other direction at the stated interval level. The registry does not report the confidence level for this interval in the outcome entry; the planning assumptions separately specify a two-sided significance level of 0.05.

What does the p-value tell us?

The primary unadjusted p-value is 0.664. In general, a p-value measures how incompatible the observed data, or data more extreme, are with a specified null model, assuming that model and the analysis assumptions hold. It is not the probability that the null is true. It does not measure the size or importance of an effect, and it is not a direct measure of whether the intervention is useful. The term unadjusted indicates that this value is not adjusted for multiple comparisons. The registry record does not describe a multiplicity procedure.

The registry says averages were modeled with a linear mixed-effect model. In general, mixed-effect models can account for data that are grouped or correlated, such as repeated observations or observations associated with the same clinician. This can matter because observations within a group may be more alike than observations from different groups. The registry does not report the fitted model, the random or fixed effects used, the correlation structure, or the model-based estimate. The posted group summaries do not establish which features of the data were included in the model.

Why can over-dispersion matter?

In general, over-dispersion means that observed variability is greater than a statistical model expects. When a model understates variability, its uncertainty estimates can be too small and its intervals too narrow. A suitable model should reflect the structure and spread of the data. The primary endpoint is described as continuous and analyzed with a linear mixed-effect model; the registry does not report an over-dispersion assessment or any model diagnostic results. Over-dispersion is therefore a general consideration, not a reported finding for this trial.

The secondary measure is a mean proportion-like quantity, described as the number of empathic responses divided by the number of empathic opportunities. The registry posts arm means and a comparison for that measure, but does not report a fitted model for it. For outcomes based on counts or proportions, statistical methods should reflect the possible range of values and the data-generating structure. That is a general principle; the registry record does not report which modeling choices were used for this secondary endpoint.

The registry describes two approaches to missing responses for the primary score: dropping missing responses and retaining them in the dataset. In general, handling missing data can affect both the analysis population and the estimate, depending on why data are missing and how the analysis treats them. Merely naming two approaches does not show how many responses were missing, how retained missing values were handled, whether the approaches gave similar results, or what assumptions supported either analysis. Those details are not reported in the registry record.

The posted primary comparison is a simple unadjusted recalculation from the reported summary statistics. It should be distinguished from the linear mixed-effect model mentioned in the endpoint description. When the registry’s own analysis results are available, those results take precedence over a simple recalculation. In this record, the discrepancy indicator is false, but the model-based output itself is not reported. The p-value and interval attached to the posted comparison do not supply the missing model details.

11. Confidence Intervals and Precision

Precision concerns how tightly an estimate is determined by the available data. For a group difference, the confidence interval is more informative about precision than the point estimate alone: it shows a range of values compatible with the data and the interval procedure. The posted primary interval ranges from a negative to a positive difference and includes zero. Thus the data do not pin down whether the underlying mean score is lower or higher in the enhanced arm. The point estimate is -0.03, but the interval indicates substantial uncertainty around that estimate.

The primary group means are 1.042 and 1.072, with posted outcome counts of 15 and 12. These summaries are based on limited counts, and the interval’s span is consistent with limited precision. The registry record does not give a model-based interval, so the interval accompanying the simple comparison should not be treated as a description of uncertainty from the stated linear mixed-effect model.

Confidence intervals are not guarantees that the true effect lies inside their endpoints. In general, their interpretation depends on the statistical method and assumptions used. Nor does an interval crossing zero demonstrate equivalence: establishing equivalence would require an appropriate design and a prespecified margin, neither of which is reported in the registry record. The posted interval instead indicates that the available comparison is not precise enough to rule out effects on either side of zero.

The secondary endpoint’s posted interval also spans values on both sides of zero. This indicates uncertainty about the direction and size of its group difference as well. The registry does not report multiplicity-adjusted intervals. Taken together, the precision information supports cautious interpretation: the point estimates describe the observed summaries, while the intervals show that considerably different underlying contrasts remain compatible with those observations.

12. 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 planning values are participants per arm, not total events. They use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect being planned for is a stated fraction of the observed effect: 100%, 75%, or 50%. These calculations are planning illustrations based on the observed effect, not the trial’s original design assumptions; those original assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation associates with an 80% chance of detecting the specified effect under its assumptions. The 90% power column gives the corresponding planning figure for a 90% chance. These are prospective design calculations, not measures of what happened in this trial. They are not observed power.

For the primary outcome, planning for 100% of the observed effect gives 509 participants per arm for 80% power and 681 participants per arm for 90% power. Planning for 75% of the observed effect gives 904 participants per arm for 80% power and 1,210 participants per arm for 90% power. Planning for 50% of the observed effect gives 2,033 participants per arm for 80% power and 2,722 participants per arm for 90% power.

The pattern reflects a general feature of power calculations: smaller effects are harder to distinguish from random variation, so more participants are needed to achieve the same planned power. These values are conditional on the calculation’s assumptions and the observed effect used as its starting point. They should not be interpreted as a recommendation or as evidence that the original trial was designed to detect those effects. The registry does not report the original sample-size justification or its planned target for the primary comparison.

13. 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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.030; pooled standard deviation σ = 0.171.
2. n = 2 × (1.96 + 0.842)² × 0.171² / -0.030² = 508.2, rounded up to 509 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
delta <- -0.03                      # observed difference in means
sigma <- 0.170697405955685                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(2969031)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 509 (80% power), 681 (90% power)
 75% of observed effect: n per arm = 904 (80% power), 1,210 (90% power)
 50% of observed effect: n per arm = 2,033 (80% power), 2,722 (90% power)
power.t.test: 509.2 
Simulated power at n = 509 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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.

14. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several statistical questions are tested. Even if each test uses the same nominal threshold, carrying out multiple tests increases the chance of at least one apparently notable result occurring by chance. A prespecified strategy may address this, for example by identifying a primary outcome or adjusting inference across a family of comparisons. The registry identifies a primary outcome and one secondary outcome, but a multiplicity plan is not reported in the registry record.

The primary comparison is labeled unadjusted, and the secondary comparison also has an unadjusted p-value. No adjustment method, testing hierarchy, or family of outcomes is described. The registry lists two outcome measures and reports no analyses in the analysis listing, while posted outcome summaries include comparisons. The record does not explain this difference in reporting. As a result, the posted p-values should not be assumed to reflect a formal multiplicity procedure or a complete prespecified analysis plan.

Missing data are especially relevant to the primary endpoint. Its description states that missing responses were handled in two ways: dropped from the data, and kept in the dataset. The registry does not report the number or pattern of missing responses, the results under each approach, or the assumptions used to justify retaining missing responses. It also does not report whether missingness differed between the arms. Without these details, the robustness of the primary comparison to missing-data decisions cannot be assessed.

The registry record does not report defined analysis populations, such as an intention-to-treat population, a per-protocol population, or a population defined by available outcome data. It does not explain the relationship between enrollment of 148 and the primary outcome counts of 15 and 12. Participant flow, exclusions, adherence, and the handling of participants without outcome data are not reported in the registry record.

These omissions do not establish that analyses were improperly conducted. They limit what can be assessed from the registry record: whether analyses followed a prespecified plan, how missing observations were handled in practice, and how strongly the reported comparisons should be interpreted in light of multiple testing.

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

The posted results describe average communication scores and an unadjusted comparison between two behavioral training approaches. The primary group means are 1.042 for Enhanced SCOPE Training and 1.072 for Standard Communication Training. Since lower scores are considered more desirable, the observed mean is lower in the enhanced arm. The comparison’s uncertainty includes zero, however, so the posted data do not establish a clear difference between the groups on the primary measure.

This result does not demonstrate that the approaches are equivalent. Equivalence requires a specific design and a prespecified range of differences considered practically unimportant; such a margin and analysis are not reported in the registry record. Nor does the result prove that neither approach affects communication. An uncertain estimate can arise when the underlying difference is small, when outcome data are limited, or for other reasons; the registry results do not determine which explanation applies.

The outcome is a patient-reported score about communication behaviors, not a measure of disease progression, treatment effectiveness, or survival. The registry’s goal is to improve communication between oncologists and patients, but the posted outcome cannot establish all aspects of the patient experience. It also does not tell whether the training changes other clinical or organizational outcomes not reported in the record.

The record lists a secondary measure concerning empathic responses to negative emotions, but its comparison is also uncertain. Taken together, these posted results do not provide clear statistical evidence of separation between the groups on the reported outcomes. That is narrower than saying the training approaches have no value or no effect, and it does not support medical advice or a treatment recommendation.

The results should be understood alongside reporting limitations: the primary outcome counts are smaller than the enrollment figure, the relation between them is not explained, and the model-based result mentioned in the endpoint description is not reported. The findings are therefore best read as the observed summaries and comparisons recorded, rather than as a complete account of all trial analyses or effects.

16. Why This Trial Matters Statistically

This trial raises a statistical question that is common in behavioral and communication research: how to translate a broad goal, such as improving clinical conversations, into measurable outcomes. The primary score combines patient ratings of several communication behaviors. A secondary measure focuses on empathic responses to negative emotions. Keeping those constructs distinct helps prevent a score on one aspect of communication from being treated as a complete measure of the encounter.

The design also illustrates why an observed difference and evidence of a reliably estimated difference are not the same thing. The primary mean is lower in the enhanced arm, but the comparison interval includes zero. Interpretation therefore depends not only on the direction of the point estimate but also on uncertainty, the analysis method, and the amount of outcome data available. A numerical contrast alone cannot establish a clear group difference.

The registry gives a useful example of how missing-data choices and model descriptions affect reproducibility. It names a linear mixed-effect model and two approaches to missing responses, but does not report the model-based result or the results under each missing-data approach. Without those details, readers cannot assess whether the methods account for the structure of the observations or whether choices about missingness materially change the estimate.

Finally, the gap between recorded enrollment and the outcome counts emphasizes the importance of transparent participant flow. Clear reporting of who contributed to each outcome, how missing responses were handled, and how planned analyses relate to posted comparisons would make the statistical evidence easier to assess. These considerations matter even when the research question is about communication rather than a conventional clinical endpoint.

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 primary outcome summaries list 15 observations for Enhanced SCOPE Training and 12 for Standard Communication Training, while registry enrollment is 148. The record does not explain the relationship between these figures, identify the outcome analysis population, or report participant flow. This makes it difficult to judge how the posted summaries relate to the full enrollment.

The endpoint description refers to a linear mixed-effect model, but a model-based estimate, model specification, and model diagnostics are not reported. The posted comparison is an unadjusted summary-based contrast. The registry also says that missing primary responses were analyzed both after dropping them and while keeping them in the dataset, but does not report the results of those separate approaches or the amount and pattern of missingness.

The enhanced-arm description ends before fully specifying what feedback participants received. The project summary describes coder and patient-reviewer feedback, but the detailed arm description is incomplete. The registry record also does not report implementation fidelity, how consistently the training was delivered, or how the quality-improvement activity in the standard arm varied across clinicians.

The primary score concerns patients’ perceptions of communication over recalled office encounters spanning 12 months, with a 6-month outcome timeframe. The record does not provide additional detail about the recall or measurement schedule beyond that description. The outcome uses coded responses, including “I prefer not to answer,” and the registry describes the aggregate as the average of the first three questions. Further details on scoring and the treatment of missing or non-substantive responses are not reported.

Finally, the registry does not report the trial’s original sample-size assumptions, allocation procedure, or a multiplicity plan. These omissions limit assessment of design intent and analysis robustness, but do not by themselves establish that the trial methods were inadequate.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Improving Communication Between Cancer Patients & Oncologists — NCT02969031.
• Registry record last updated 2020-01-03; status last verified 2019-12.
• Results first posted 2020-01-03.
• 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

NCT02969031 is a completed randomized, parallel-group, unmasked behavioral trial sponsored by Dana-Farber Cancer Institute. It compares Enhanced SCOPE Training with Standard Communication Training in a project concerning communication between oncologists and patients with cancer. The registry records enrollment of 148, a start date of 2017-02-14, and primary completion on 2018-08-31.

The primary outcome is the Provider to Patient Communication Score at 6 months; lower scores are described as more desirable. The posted group means are 1.042 and 1.072, with a mean difference of -0.03 and an unadjusted p-value of 0.664. The comparison interval includes zero, so the posted primary result does not establish a clear group difference. The one listed secondary outcome concerns empathic responses to expressions of negative emotions and also has an uncertain comparison.

The registry describes a linear mixed-effect model and two approaches to missing responses, but does not report model-based results, the missing-data results separately, or the relationship between enrollment and the primary outcome counts. The original design assumptions and multiplicity plan are not reported in the registry record.