VOICE: Values and Options in Cancer Care: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for VOICE: Values and Options in Cancer Care (NCT01485627), a randomized na study. The primary outcome is Mean Patient-centered Communication in Advanced Cancer Score.

2. Clinical Question

The registry’s brief description also identifies quality of life and quality of care as areas of interest. The primary endpoint, however, specifically concerns communication during the first physician visit after coaching for the intervention group or after study entry for the control group. Its timeframe is listed as 3 years. The comparison therefore addresses a communication outcome rather than a direct measure of survival or a clinical treatment effect. The registry record does not report further eligibility details or a more specific definition of the advanced-cancer population.

3. Trial Design

VOICE, or Values and Options in Cancer Care, is a completed randomized controlled trial. The registry lists a parallel design, no masking, and an enrollment of 485. The study began in 2011-04, and primary completion is listed as 2017-12-31. The sponsor was the University of Rochester. The phase is listed as not applicable.

The two groups were the communication intervention and usual care. In the intervention group, oncologists received communication training and patients received coaching for making the most of the visit. Patients in the control group received usual care. The study’s stated purpose was to assess whether a combined intervention for patients, caregivers, and oncologists improves communication, quality of life, and quality of care.

The primary outcome was assessed in patients only. The posted arm summaries include 130 intervention participants and 135 control participants for that outcome. These outcome-specific counts are not the same as the overall enrollment; the registry record does not explain in this entry why fewer participants appear in the primary outcome summaries. The primary measure was based on the first recorded physician visit after the relevant study entry point, and the registry lists a 3-year timeframe.

The registry lists 6 outcome measures and 5 analyses. Five secondary outcomes are named, covering caregiver prolonged grief symptoms, patient-physician differences in reported expectations of survival, patient well-being, caregiver mental health, and health-care utilization related to aggressive care at the end of life. The original sample-size justification, detailed recruitment and follow-up procedures, and the reasons for the difference between enrollment and outcome-specific counts 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.

Before assignment: 233 caregivers and 316 patients were consented. 204 caregivers were allocated and 127 of them were bereaved. 24 caregivers failed screening and 5 withdrew during screening. 26 patients failed screening and 9 withdrew during screening.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: This outcome measure was assessed in patients only.

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: Patients: 130 for the intervention and 135 for the control arms, who completed the audio-recorded office visit. Caregivers: 50 for the intervention and 53 for the control arms, who were bereaved and completed the month 2 and/or month 7 assessment.

6. Primary Endpoint

The registry lists the primary outcome as Mean Patient-centered Communication in Advanced Cancer Score, measured over 3 years.

The primary endpoint is the mean patient-centered communication score in advanced cancer. It is a continuous composite measure built from four communication domains: engaging patients in consultations, responding to patients’ emotions, informing patients about prognosis and treatment choices, and balanced framing of decisions. Trained university students coded the measures from audio recordings. The registry says the coding teams were audited continuously and blinded to study hypotheses and group assignment.

Each component was transformed to a standardized z score using means and standard deviations from the pre-randomization phase sample. The four standardized component scores were then averaged. A higher score indicates better communication. The measure is useful for bringing distinct observed communication behaviors onto a common standardized scale and combining them into one summary outcome. The registry reports that the maximum possible z score ranged from -0.69 to 20.08.

The outcome was based on the first physician visit recorded after coaching in the intervention group or after study entry in the control group. That timing links the measure to an observed consultation rather than relying only on a participant’s retrospective impression. At the same time, a composite can conceal different patterns across its component behaviors: an average score does not establish that every communication domain changed in the same way. The registry record does not give the separate component results in the primary outcome description.

7. Results Posted in the Registry

• Intervention: mean 0.16 versus -0.01 on control; difference 0.17 (95% CI -0.01 to 0.35), standardized effect d = 0.23; z-test p = 0.062.

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.

Caregiver Mean Prolonged Grief Symptoms as Measured by PG-13

• 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:
  • Intervention: Communication: mean 19.30 versus 20.60 on control; difference -1.30 (95% CI -3.83 to 1.23), standardized effect d = -0.20; z-test p = 0.314.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intervention: Communication and the control. The interval gives the range of effects that remain compatible with the data.

Aim 1b&c Mean Difference in Reported Expectation of Survival in 2 Years Between Patients and Physicians

• 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:
  • Intervention: mean 2.20 versus 2.50 on control; difference -0.30 (95% CI -0.67 to 0.07), standardized effect d = -0.19; z-test p = 0.115.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Aim 2 Patient Well-being

• 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:
  • Intervention: Communication: mean -0.01 versus -0.07 on control; difference 0.06 (95% CI -0.12 to 0.24), standardized effect d = 0.08; z-test p = 0.523.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intervention: Communication and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover several distinct domains, including caregiver grief symptoms and mental health, patient well-being, expectations about survival, and health-care utilization related to aggressive care at the end of life. They broaden the trial’s scope beyond the primary communication score, but they are not interchangeable measures and do not necessarily share a common scale or target population.

Across the posted secondary results, the simple mean comparisons are generally imprecise, with confidence intervals that include zero. The registry also includes some model-based secondary analysis results, but not all listed secondary outcomes have a reported model result. These outcomes therefore do not provide a uniform pattern of corroboration for the primary result. Their relationship to the primary endpoint is conceptual: they concern possible effects of communication on patient or caregiver experience and care. A secondary outcome does not directly validate the primary communication measure.

General statistical explanation: when many outcomes or comparisons are examined, the chance of encountering a result that appears notable by chance can increase. Unadjusted secondary results do not account for that multiplicity. Unless a hierarchy or multiplicity procedure is specified, individual secondary p-values warrant less weight than a prespecified primary analysis. The registry record does not report whether the secondary analyses were part of a formal testing sequence or whether multiplicity was addressed.

Secondary findings are also constrained by outcome-specific counts and by the lack of reported details on missing data and analysis populations. The registry lists five secondary outcomes, while the record reports five analyses overall and does not give results for every outcome in the same form. This means the secondary results are best treated as exploratory context rather than as a single confirmatory body of evidence. The registry record does not report a plan for prioritizing these outcomes or interpreting them jointly.

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

• Caregiver Mean Prolonged Grief Symptoms as Measured by PG-13
• Aim 1b&c Mean Difference in Reported Expectation of Survival in 2 Years Between Patients and Physicians
• Aim 2 Patient Well-being
• Caregiver Mean Overall Mental Health as Measured by the SF-12 Assessment
• Health Care Utilization- Mean Index Score of Aggressive Care at the End of Life

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a continuous measure summarized by a mean and standard deviation. General statistical explanation: a mean describes the center of observed scores, while a standard deviation describes how dispersed individual scores are around that mean. In the registry summaries, the intervention mean is 0.16 with standard deviation 0.81, and the control mean is -0.01 with standard deviation 0.66. These summaries describe the outcome-specific groups listed in the record.

The recalculated comparison is an unadjusted difference between the two posted means. The estimate is 0.17. General statistical explanation: an unadjusted comparison uses the group summaries directly and does not account for other variables or repeated observations unless those are built into the calculation. It is therefore a straightforward descriptive contrast, but it need not reproduce a model-based estimate that accounts for additional structure or covariates. The registry record does not report enough detail to specify the sponsor model’s full structure.

The sponsor reports a mixed models analysis. General statistical explanation: mixed models can represent correlation among observations, such as repeated measurements or observations grouped within people or settings, and can incorporate fixed and random effects according to the model specification. The registry labels the reported parameter as an adjusted mean difference in difference. The exact model terms, covariance structure, adjustment variables, and handling of incomplete observations are not reported in the registry record. Those omissions prevent a full independent reconstruction of the sponsor analysis from the registry entry alone.

A confidence interval expresses statistical uncertainty around an estimate under the assumptions of its method. For the simple recalculation, the reported 95% interval includes zero. The sponsor reports a 95% interval from 0.06 to 0.62 for its adjusted estimate. General statistical explanation: an interval that includes zero is compatible with no difference under the corresponding model and confidence procedure, while an interval entirely on one side of zero is not. That statement concerns statistical compatibility, not whether an effect is important to patients or clinicians.

The p-value measures how unusual data at least as inconsistent with a specified null hypothesis would be if that hypothesis and the analysis assumptions were true. It is not the probability that the null hypothesis is true, nor a measure of the size or importance of an effect. The recalculated unadjusted p-value is 0.062; the reported model p-value is 0.05. Because the estimates come from different approaches, their p-values should not be treated as a contradiction that can be resolved without more information about the model and analysis plan.

The continuous endpoint’s standardized construction allows four communication components to contribute to one composite. General statistical explanation: standardization expresses a component relative to a reference mean and standard deviation, which supports combining measures that may originally have different scales. The registry says the reference quantities came from the pre-randomization phase sample. Averaging standardized components produces a composite, but an overall average can mask variation among its parts. Component-level results are not reported in the primary outcome description.

The registry identifies randomization and a parallel-group design, but it does not report the randomization method, allocation concealment procedures, stratification variables, or details of masking beyond listing masking as none. It says outcome coders were blinded to study hypotheses and group assignment. General statistical explanation: blinded outcome coding can reduce the risk that knowledge of group assignment influences subjective scoring, although it does not make the participants or clinicians masked.

Planning illustrations are based on the observed effect and stated assumptions, not on the original trial design. General statistical explanation: a sample-size calculation estimates how many participants would be needed to detect a specified effect with a chosen power under a statistical model. The registry’s original sample-size justification, target effect, and assumptions are not reported in the registry record. The planning illustrations should not be read as evidence about the trial’s achieved power.

Other important analytical details are also absent. The registry record does not report how missing outcome data were handled, whether the primary analysis followed an intention-to-treat principle, how the analysis population was defined, whether interim analyses were planned or performed, or whether multiplicity was controlled. These details matter when interpreting model results and a set of outcomes, but their absence does not justify inferring a particular procedure.

11. Confidence Intervals and Precision

The recalculated unadjusted mean difference is 0.17. Its 95% confidence interval runs from a value slightly below zero to a positive value, so the unadjusted summary leaves uncertainty about whether the underlying mean difference is zero or positive. The unadjusted p-value is 0.062. These results reflect the posted arm means and standard deviations, rather than the sponsor’s mixed-model adjustment.

The sponsor’s reported adjusted mean difference in difference is 0.34, with a 95% confidence interval from 0.06 to 0.62. This interval lies above zero, but its width still allows a range of positive estimates. The sponsor reports p = 0.05. Because the registry does not describe the model specification in enough detail to reproduce it, the interval is best interpreted as the uncertainty reported for that model rather than as a re-expression of the simple unadjusted comparison.

General statistical explanation: a confidence interval’s width reflects both the variability in the data and the amount of information available, as well as the method used to calculate it. A wider interval indicates less precision about the size of the effect. An interval crossing zero does not prove no effect; it indicates that the analysis does not exclude zero at that confidence level. An interval on one side of zero does not establish practical importance or rule out model limitations.

The outcome-specific group counts are 130 and 135, while total enrollment is 485. The registry record does not explain the relationship between these counts or report missing-data handling. Precision should therefore be read in the context of the posted outcome summaries and the sponsor’s stated analysis, without assuming that all enrolled participants contributed to each result. The registry does not report a minimally important difference for the standardized communication score, so statistical precision cannot be translated into a threshold for meaningful change.

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 figures are participants per arm, not total events. This is a continuous outcome, and the listed figures estimate participants in each of two equally allocated arms. The planning assumptions are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The registry’s original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the estimated participants per arm for a design with an 80% chance of detecting the specified effect under the stated assumptions, if that effect is the true one and the assumptions hold. The 90% power column gives the corresponding per-arm estimate for a 90% chance. These are planning illustrations, not observed power.

Planning for 100% of the observed effect requires 296 participants per arm for 80% power and 396 participants per arm for 90% power. Planning for 75% of the observed effect requires 526 participants per arm for 80% power and 703 participants per arm for 90% power. Planning for 50% of the observed effect requires 1,182 participants per arm for 80% power and 1,582 participants per arm for 90% power.

The curve represents the same planning relationship across effect sizes. Smaller effects need more participants to be detected with the same power, because a smaller difference is harder to distinguish from random variation. These estimates are based on the observed effect in this trial and should not be mistaken for the enrollment target used to design the trial.

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 = 0.170; pooled standard deviation σ = 0.737.
2. n = 2 × (1.96 + 0.842)² × 0.737² / 0.170² = 295.4, rounded up to 296 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 <- 0.17                      # observed difference in means
sigma <- 0.7373967672803                     # 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(1485627)
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 = 296 (80% power), 396 (90% power)
 75% of observed effect: n per arm = 526 (80% power), 703 (90% power)
 50% of observed effect: n per arm = 1,182 (80% power), 1,582 (90% power)
power.t.test: 296.3 
Simulated power at n = 296 per arm: 0.799 (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. Analyses Reported by the Sponsor

The sponsor’s own primary analysis is reported as a mixed models analysis, with an adjusted mean difference in difference of 0.34, a 95% confidence interval from 0.06 to 0.62, and p = 0.05. The registry labels the analysis as superiority. This is the trial’s posted model-based result.

The recalculation uses only the posted arm-level means and standard deviations. It gives an unadjusted mean difference of 0.17, with a 95% confidence interval that crosses zero and an unadjusted p-value of 0.062. It is a simple comparison rather than an attempt to reproduce the mixed model. The two estimates answer related but not necessarily identical statistical questions, since adjustment and model structure can affect both the estimate and its uncertainty.

The registry record does not give the mixed model’s full specification, including its adjustment variables or covariance structure. As a result, the sponsor’s reported estimate cannot be independently reconstructed from the summary statistics alone. The recalculation is useful as a transparent check on the posted group means, but it does not replace the sponsor’s analysis or explain why the model-based estimate differs. The record does not report further information that would resolve the difference.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary communication outcome and five secondary outcomes, covering different aspects of patient, caregiver, and care experience. It does not report whether the investigators adjusted for multiple comparisons, specified an outcome hierarchy, or controlled a family-wise error rate or false discovery rate. When several outcomes and analyses are considered, unadjusted results can be more likely to appear notable by chance. General statistical explanation: multiplicity procedures address this issue by changing how evidence is assessed across a family of tests. No particular procedure can be assumed from the registry entry.

Missing-data methods are not reported in the registry record. The primary outcome summaries include 130 intervention participants and 135 control participants, while the total enrollment is 485. These counts alone do not indicate how many participants were eligible for the outcome, how many had a usable recorded visit, why other participants are absent from the summaries, or whether missingness differed by group. No imputation, complete-case approach, sensitivity analysis, or other handling method is identified.

The analysis population is also not reported. The registry record does not say whether the primary analysis followed an intention-to-treat principle, used a per-protocol population, or applied another definition. Although the sponsor reports a mixed models analysis, the model details and its approach to incomplete observations are not described. This limits assessment of the assumptions behind the reported estimate.

Interim analyses and stopping rules are not reported in the registry record. The original sample-size justification and planned statistical analysis are also not reported. Without these details, it is not possible to establish from this record alone how the posted analyses relate to the initial analysis plan or whether any adaptations occurred. These are reporting gaps; they do not demonstrate that a particular method was or was not used.

16. 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 lists serious adverse events and other adverse events by group. For intervention patients, 139 were at risk and 0 had a serious adverse event; for control patients, 142 were at risk and 0 had a serious adverse event. The listed other adverse event counts are also 0 for both patient groups.

For caregivers, the registry lists two intervention groups with slightly different labels. The first has 105 at risk and 0 serious or other adverse events; the second has 99 at risk and 0 serious or other adverse events. These entries are not presented as a caregiver intervention-versus-control comparison, so they should not be interpreted as such. The registry lists 1 death in the first caregiver group and 0 in the second. It lists 91 deaths among intervention patients and 96 among control patients.

General statistical explanation: a zero event count in a posted group does not establish that an event is impossible or that groups are equivalent in safety. Interpretation depends on follow-up, event definitions, completeness of reporting, and the number of participants observed. The registry record does not provide further event descriptions or details about how these counts were collected. These figures should be read as registry-reported counts, not as a standalone assessment of safety.

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

The primary result concerns a standardized communication score observed during a recorded physician visit. The intervention group’s posted mean is higher than the control group’s mean, and the sponsor’s adjusted model reports a positive estimate. These results describe the communication measure and the analysis methods used; they do not directly measure survival, tumor response, or the effectiveness of cancer treatment.

The score combines several behaviors into one average, so an overall difference does not show that each component changed to the same extent. The registry record does not report a minimally important difference for the score. Without that benchmark, the numerical estimate cannot be translated into a claim about how meaningful the change would be to patients or clinicians. The standardized scale also does not have a direct unit such as minutes or a count of conversations.

The sponsor’s mixed-model estimate and the simple unadjusted recalculation have different values and uncertainty summaries. They should be understood in light of their distinct methods, with the sponsor’s result representing the reported model-based analysis and the recalculation representing a direct comparison of posted summary statistics. The registry record does not provide enough model detail to determine precisely why they differ.

Secondary outcomes address caregiver and patient experiences and aspects of care, but they are not a substitute for the primary endpoint. Their interpretation is limited by the absence of a reported multiplicity plan and by incomplete detail on analysis populations and missing data. No medical advice follows from these statistical summaries. The registry record supports describing the posted estimates and their uncertainty, not making a broader claim about clinical benefit or lack of benefit.

18. Why This Trial Matters Statistically

This trial raises a statistical challenge common in communication research: the main outcome is a composite measure of observed behaviors rather than a single, directly tangible clinical event. Combining multiple domains can give a concise summary, but it also requires careful attention to how components are standardized, scored, and interpreted. The registry describes the construction of the composite and the use of trained coders, which helps readers understand what the primary score represents.

It also illustrates why analyses based on arm summaries and adjusted models may not produce identical estimates. A simple comparison is transparent, while a model-based result can account for features of the data that are not represented by the two means alone. The registry reports the sponsor’s mixed-model estimate but not enough detail to reproduce the model. That contrast makes clear why statistical estimates should be read together with their analysis methods.

The trial also spans multiple outcomes involving patients, caregivers, and care. This makes the distinction between a primary endpoint and secondary outcomes important. Without a reported multiplicity plan, secondary results should be treated with restraint. Similarly, outcome-specific counts that differ from total enrollment emphasize the value of reporting analysis populations and missing-data handling.

For biostatisticians and clinical researchers, the record offers a useful example of interpreting a standardized patient-centered endpoint alongside a reported adjusted analysis, a transparent unadjusted re-calculation, and planning illustrations. The estimates describe statistical uncertainty, but the registry does not report the clinical threshold needed to judge the importance of a score change. A careful reading therefore separates measurement, model-based evidence, and practical meaning.

19. 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 does not report the original sample-size justification or the assumptions used when the trial was designed. The planning illustrations based on the observed effect are not a substitute for that information. It is therefore not possible from this entry to assess the planned detectable effect or determine whether the trial’s enrollment target was based on the primary endpoint.

The primary outcome summaries include 130 intervention participants and 135 control participants, compared with total enrollment of 485. The reason for this difference is not reported in the registry record. It also does not report the analysis population, missing-data methods, or sensitivity analyses. Those gaps limit assessment of selection and missing-outcome concerns.

The sponsor reports a mixed models analysis, but the full model specification is not reported. Details such as adjustment variables, model structure, and handling of incomplete observations are not available in the registry entry. The adjusted estimate therefore cannot be independently reproduced from the arm-level summary statistics. The unadjusted recalculation offers a useful contrast but does not recover the sponsor’s analysis.

The primary score averages four standardized communication components. The component-level outcome estimates are not reported in the primary outcome description, and the registry does not provide a minimally important difference for the composite. This limits interpretation of which behaviors contributed to the overall result and whether the numerical change is practically meaningful.

The record also does not report a multiplicity plan for the secondary outcomes, an interim-analysis plan, detailed eligibility criteria, or a full statistical analysis plan. Masking is listed as none, although coders are described as blinded to group assignment and study hypotheses. These features and missing details should be considered when interpreting the posted results, without assuming procedures that the registry does not describe.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: VOICE: Values and Options in Cancer Care — NCT01485627.
• Registry record last updated 2019-01-10; status last verified 2019-01.
• Results first posted 2019-01-10.
• 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.

23. Record Summary

VOICE: Values and Options in Cancer Care is a completed randomized, parallel-group trial sponsored by the University of Rochester. The registry lists enrollment of 485, no masking, and a communication training and coaching intervention compared with usual care. The study began in 2011-04, with primary completion listed as 2017-12-31.

The primary endpoint was mean patient-centered communication in advanced cancer, a standardized composite based on four communication measures from recorded physician visits. The sponsor reported a mixed-model adjusted mean difference in difference of 0.34, with a 95% confidence interval from 0.06 to 0.62 and p = 0.05. A simple unadjusted recalculation gives a mean difference of 0.17 and p = 0.062.

The record lists five secondary outcomes, but does not report a multiplicity plan, detailed model specification, missing-data methods, or the original sample-size justification. The statistical results should be interpreted in light of those reporting limits and the distinction between the sponsor’s adjusted analysis and the unadjusted comparison.