Treating Depression Among Low-Income Patients With Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treating Depression Among Low-Income Patients With Cancer (NCT00565110), a randomized na study. The primary outcome is Reduced Depression Symptoms.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group trial with single masking, listed as completed. The registry record gives the phase as NA, enrollment as 472, a start date of 2004-06, and primary completion in 2009-01. It identifies the University of Southern California as sponsor. The identity of the party or parties masked is not reported in the registry record.

The two groups were ADAPt-C intervention and enhanced usual care. The intervention combined depression treatment choices and care coordination using a stepped care algorithm. Enhanced usual care retained standard oncology services and supportive services, with depression education materials and a community and financial resource listing added. The registry summary describes the target setting as an urban public sector care system and the patient group as low-income and predominantly Hispanic.

For the primary endpoint, the posted group counts are 57 of 114 for enhanced usual care and 91 of 144 for ADAPt-C intervention. Those endpoint denominators do not cover the full enrollment listed in the registry. The reason for the difference, including whether it reflects eligibility for the endpoint, follow-up, or another analysis-population decision, is not reported in the registry record. The serious adverse event record uses separate at-risk denominators of 230 and 242; these should not be treated as the primary endpoint denominators.

The record posts a comparison for the primary outcome and a comparison for one secondary outcome. It does not report the original sample-size justification, a detailed statistical analysis plan, the identity of the masked party, or the analysis populations. These omissions limit how fully the posted results can be connected to the trial's original design decisions.

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

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 Reduced Depression Symptoms, measured over 12 months.

The primary endpoint counts participants with a 50% PHQ-9 score reduction since baseline at 12 months. It is a binary response measure: each participant either meets the specified reduction threshold or does not. The endpoint therefore describes the number and proportion meeting that criterion, rather than the full range of symptom scores or the size of each participant's score change.

A threshold-based response measure can make a group comparison easier to interpret: it asks how many participants crossed a defined level of improvement. The PHQ-9 score change supplies the basis for the response classification, while the registry outcome itself is reported as a count of participants. In general, turning a continuous score into a binary outcome also discards information. Participants just on opposite sides of the threshold are placed in different categories, while participants with very different changes can fall into the same category. The posted count alone cannot describe the distribution of symptom changes within either group.

The registry specifies the assessment time as 12 months and the threshold as 50% reduction from baseline. It does not report details such as the rule for handling a missing baseline or follow-up score, or whether the endpoint was assessed in all randomized participants. Those details are relevant to interpreting the count and are not reported in the registry record.

7. Results Posted in the Registry

• ADAPt-C Intervention: 63.2% (91/144) versus 50.0% (57/114) on control; risk difference 13.2 percentage points (95% CI 1.1 to 25.3), risk ratio 1.26 (1.01–1.58); unadjusted two-proportion z-test p = 0.033.

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.

Physical Composite Summary Score (PCS) Derived From the 12-item Short Form (SF-12) Health Survey

• 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 errors.
• Result:
  • ADAPt-C Intervention: mean 41.48 versus 38.68 on control; difference 2.80 (95% CI 0.37 to 5.23), standardized effect d = 0.28; z-test p = 0.024.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for ADAPt-C Intervention than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists a physical health secondary outcome based on the Physical Composite Summary Score from the 12-item Short Form Health Survey. The posted group comparison points in the same direction as the primary depression-response comparison: the intervention group has the higher reported group-level result. That directional consistency can be informative, because the outcomes capture different aspects of participants' health. It does not show that one outcome caused the other or that the two measures reflect the same underlying change.

Only one secondary outcome comparison is posted, although the registry lists two outcome measures. The record therefore does not give a complete set of secondary results corresponding to every measure listed. Nor does it report a prespecified hierarchy, a multiplicity correction, or a plan for interpreting the secondary outcomes alongside the primary endpoint.

When several outcomes are examined, unadjusted results carry less weight as confirmatory evidence. In general, looking at multiple outcomes increases the chance that at least one result appears notable by chance, even when no adjustment is made. A secondary result can still help describe a broader pattern, but without the analysis plan and multiplicity approach it is difficult to know whether it was intended as confirmatory or exploratory.

The secondary result should therefore be read as supportive descriptive evidence rather than as an independent confirmation of the primary finding. It concerns a different construct, has its own uncertainty, and does not fill the gaps in the primary endpoint's analysis-population and missing-data reporting. The registry record does not report whether the secondary comparison was adjusted for baseline values or other characteristics.

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

• Physical Composite Summary Score (PCS) Derived From the 12-item Short Form (SF-12) Health Survey

9. Statistical Interpretation

10. Statistical Methods Explained

For a count outcome, the main comparison can be expressed using the proportion meeting the response criterion in each group. The risk difference describes the absolute separation between those proportions. The risk ratio compares the proportions as a ratio, while the odds ratio compares the odds of response. Each measure presents the same underlying two-group data in a different scale. Absolute measures are often more direct for understanding the size of a difference; relative measures can help describe how the event frequency compares across groups. None should be interpreted without considering its uncertainty and the population to which it applies.

What does a confidence interval describe?

In general, a confidence interval expresses the sampling uncertainty around an estimate under the method used to calculate it. A narrower interval indicates greater statistical precision under that method; a wider interval indicates less precision. An interval is not a range that guarantees the true effect lies within its endpoints, and it does not account automatically for biases, missing data, or choices made in the analysis. The risk-difference interval for the primary result has positive limits, and the reported intervals for the relative measures are also above their respective no-effect values. Their breadth still allows uncertainty about the magnitude.

What does the p-value add?

The primary comparison has an unadjusted p-value of 0.033. In general, a p-value measures how unusual data at least as discrepant as those observed would be if a specified null model held and the test assumptions were satisfied. It is not an effect-size measure, a measure of clinical importance, or a substitute for a confidence interval. The registry does not report whether this was the prespecified analysis, which test was used, or whether any adjustment was made for other comparisons. The word unadjusted is important: the p-value does not account for a family of tests unless a relevant adjustment is applied.

Why distinguish counts from continuous scores?

The primary endpoint is binary because a PHQ-9 change is classified according to a response threshold. The secondary physical health endpoint is continuous and reported as a mean with a standard error. A mean difference describes the separation of group averages on that scale; it does not mean every participant experienced that amount of change. The standard error describes uncertainty in the estimated mean, not the spread of individual participant scores. In general, the standard deviation describes individual variation, while the standard error describes precision of an estimated mean.

The posted secondary comparison reports a pooled standard deviation and a standardized effect estimate as well as the mean difference. Standardizing expresses a mean difference relative to a measure of variation, which can aid comparisons across scales; it does not itself establish clinical importance. The registry record does not report an adjusted secondary analysis. As with the primary outcome, a posted summary comparison cannot reveal the full distribution of participant outcomes or the handling of incomplete measurements.

Why do design and analysis details matter?

Randomization is intended, in general, to balance prognostic factors between groups on average. It does not eliminate uncertainty, guarantee balance in a particular sample, or correct analysis choices made after randomization. A complete interpretation would benefit from details on the analysis population, missing-data approach, any covariate adjustment, and prespecified outcome hierarchy. Those details are not reported in the registry record.

Dispersion also matters for continuous outcomes. In general, greater variation among participants makes a given mean difference harder to estimate precisely, while less variation can improve precision. For counts and proportions, precision depends on the number of observations and the frequency of the outcome. These considerations explain why a point estimate is not enough: the associated interval and the underlying analysis population are also needed to judge how stable a comparison may be.

The registry includes an unadjusted p-value and interval-based comparisons, but it does not give a full account of model assumptions or diagnostic checks. The sample-size illustrations use a normal approximation, but those illustrations are not a description of the trial's original analysis plan. Whether the original trial used a particular model, adjustment, or interim analysis is not reported in the registry record.

11. Confidence Intervals and Precision

The primary response comparison is estimated with a positive risk-difference interval and relative-effect intervals that are also on the side of more responses in the intervention group. For the risk ratio, the posted estimate is 1.264 and its confidence interval runs from 1.012 to 1.578. For the odds ratio, the estimate is 1.717 and the interval runs from 1.042 to 2.829. These intervals indicate that the point estimates are not precise to a single narrow value: effects closer to the no-difference point and more pronounced relative differences remain compatible with the reported uncertainty.

Precision is not the same as correctness. A confidence interval reflects sampling uncertainty under its calculation method, but it does not resolve the unreported details about endpoint denominators, missing data, or analysis populations. It also does not address how well the trial's population represents other settings. The registry record does not report adjusted intervals or enough methodological detail to assess whether the posted estimates incorporate other sources of uncertainty.

The secondary physical health comparison also has a confidence interval around its mean difference. In general, that interval indicates uncertainty about the group-average separation on the scale, not the range of individual outcomes. The standard errors reported for the group means are not confidence intervals for the mean difference. The secondary outcome's posted interval and unadjusted p-value should be considered in light of the fact that it is one of multiple outcomes listed and no multiplicity approach is reported.

Interpretation should therefore focus on both direction and uncertainty. The primary intervals have limits on the intervention-favoring side of the no-effect value, but the interval span means the exact size of the difference is not pinned down. Neither the p-value nor the interval alone establishes that the estimated difference will reproduce in another population or under another analysis.

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 sample-size figures are planning illustrations, not observed trial results. For these outcomes, the figures are participants per arm, not total events. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The trial's own original design assumptions are not reported in the registry record.

The 80% power column gives the planned number of participants per arm for a design with an 80% chance of detecting the specified effect under those assumptions, if that effect is the true effect and the model assumptions hold. The 90% power column uses a higher probability of detecting that same specified effect, and therefore requires more participants per arm. These are design calculations, not the probability that this completed trial found a real effect. This is not observed power.

For the primary response endpoint, planning for 100% of the observed effect gives 221 participants per arm for 80% power and 295 for 90% power. Planning for 75% gives 396 and 530 per arm, while planning for 50% gives 897 and 1,200 per arm. The secondary outcome has its own corresponding illustrations: 198 and 264 per arm at 100%, 351 and 470 at 75%, and 789 and 1,056 at 50%.

The central pattern is that smaller effects need more participants to detect with the same power. The values depend on the stated assumptions and the observed effect used as a planning reference; they do not explain why the original trial enrolled its actual number of participants.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.500; treatment proportion p₁ = 0.632; average p̄ = 0.566.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7009 = 1.3738.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6947 = 0.5847.
4. n = (1.3738 + 0.5847)² / (0.1319)² = 220.3, rounded up to 221 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 <- 0.5                          # control proportion
p1 <- 0.631944444444444                          # 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(565110)
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 = 221 (80% power), 295 (90% power)
 75% of observed effect: n per arm = 396 (80% power), 530 (90% power)
 50% of observed effect: n per arm = 897 (80% power), 1,200 (90% power)
power.prop.test: 220.3 
Simulated power at n = 221 per arm: 0.795 (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.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures and posts one secondary outcome comparison, in addition to the primary endpoint comparison. It does not report a multiplicity strategy, such as a prespecified hierarchy or an adjustment across outcomes. Without that information, the unadjusted p-values should not be treated as though they necessarily preserve a single family-wide error rate. This matters particularly for secondary outcomes, which often have a more exploratory role unless the protocol states otherwise.

The registry record does not report how missing primary or secondary outcome data were handled. It does not specify whether participants without a 12-month measure were excluded, whether any imputation was used, or whether sensitivity analyses were performed. The primary endpoint counts use 114 participants in enhanced usual care and 144 in ADAPt-C, whereas total enrollment is 472. The record does not explain the difference between those figures, so the reason for the endpoint denominators cannot be determined from the posted information.

The record also does not identify the analysis population used for the primary comparison. It does not state whether analyses followed assigned groups regardless of treatment received, or whether another population was used. That distinction is important in randomized trials because post-randomization exclusions can affect the comparability created by assignment.

Other details that are not reported in the registry record include the full statistical analysis plan, any covariate adjustment, interim analyses, and the original sample-size justification. The planning illustrations in the registry use stated assumptions, but they are not evidence of the assumptions used to design this trial. These omissions do not invalidate the posted comparison; they limit independent assessment of how it was produced and how robust it may be.

15. 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.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: death (23.5%), in hospice or palliative care (0.8%).

The serious adverse event figures use at-risk groups of 230 participants for enhanced usual care and 242 for ADAPt-C intervention. Serious events affected 55 participants in enhanced usual care and 47 in ADAPt-C. The record lists zero other adverse events among 230 at risk in enhanced usual care and zero among 242 at risk in ADAPt-C. Deaths are listed separately as 54 and 45, respectively.

The registry also identifies Death and In hospice or palliative care among the leading serious event terms, with listed percentages of 23.5% and 0.8%. The record does not specify the denominator for those term percentages. The event-term percentages should not be assumed to use the same denominator as the affected-participant counts. Nor should the separate death counts be treated as equivalent to the serious-event counts without information about coding and overlap.

These figures describe what was posted, not the cause of an event or a comparison adjusted for differences between groups. The record does not provide further detail on event definitions, timing, relatedness to the intervention, or statistical testing of safety outcomes. Serious adverse event data should therefore be read as descriptive counts with the stated denominators.

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

The registry results show a higher count and proportion meeting the 12-month PHQ-9 response criterion in the ADAPt-C group than in enhanced usual care. The posted comparison estimates and confidence intervals describe the size and uncertainty of that difference. The secondary physical health comparison is directionally consistent at the group level, but it measures a distinct outcome and is not a substitute for the primary endpoint.

These findings do not establish that every participant in the intervention group improved, or that every participant in enhanced usual care did not. The primary measure classifies participants using a threshold; it does not display all individual score trajectories. Group averages and response counts also do not describe the experiences of each participant.

The results do not establish a mechanism for any group difference. ADAPt-C included multiple care components, and the registry record does not separate the effects of medication management, psychotherapy, education, follow-up, or other components. The comparison is between the overall assigned care approaches, not a test of each element individually.

The registry summary describes low-income, predominantly Hispanic patients with cancer receiving care in an urban public sector system. The findings should be interpreted in that context. The record does not establish that the results apply equally to other populations, healthcare systems, or care arrangements. The posted endpoint denominators and absent missing-data details also limit conclusions about the relationship between enrollment and the analyzed outcomes.

In general, statistical evidence is one part of interpreting a trial. The posted estimates do not by themselves determine clinical importance, durability beyond the stated assessment, or the balance of benefits and harms. Further details about the protocol, analysis plan, follow-up, and event reporting are not reported in the registry record. The appropriate reading is therefore a description of the observed group comparisons and their statistical uncertainty, not a universal claim about treatment effects.

17. Why This Trial Matters Statistically

This trial illustrates the difference between a trial's overall enrollment and the denominators used for a particular posted outcome. The record lists enrollment of 472, while the primary endpoint comparison uses 114 participants in one group and 144 in the other. The record does not explain that difference. Making the analysis population and missing-data rules explicit is important because denominators shape how readers understand both the estimates and their applicability.

It also illustrates why reporting several effect measures can aid interpretation while still requiring care. The response outcome is expressed as a count and compared using absolute and relative measures. Confidence intervals show that a point estimate should not be read as exact. Reporting the outcome definition, time point, group denominators, and interval together helps readers distinguish the observed data from the uncertainty around the comparison.

A further statistical issue is the relationship between primary and secondary outcomes. The secondary physical health measure points in the same direction as the primary result, but the record does not report multiplicity handling or a hierarchy. Readers therefore need to distinguish a consistent descriptive pattern from a set of prespecified confirmatory claims.

Finally, the sample-size illustrations show how planning requirements depend on the effect chosen for detection. Smaller target effects require more participants per arm under the stated assumptions. Because these are illustrations based on observed effects and not the original design justification, they cannot establish whether the completed study was designed with adequate power for its planned questions. Together, these features make the record useful for understanding effect measures, precision, planning, and the importance of complete statistical reporting.

18. 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.

Several limitations specific to this record affect interpretation. First, the primary outcome is a thresholded response measure. It gives the number meeting a 50% PHQ-9 reduction criterion at 12 months, but does not show the full distribution of score changes or whether outcomes differed among participants who did not meet that threshold.

Second, the primary endpoint denominators are smaller than the enrollment listed in the registry, and the record does not explain why. No missing-data approach or analysis population is reported. This prevents readers from determining how much incomplete follow-up, eligibility, or other analysis decisions may have influenced the posted comparison.

Third, the masking description is limited to single masking, without identifying who was masked. Details about outcome assessment procedures and the full statistical analysis plan are not reported. The record also does not report whether comparisons were adjusted for baseline characteristics or whether interim analyses took place.

The target population and setting are described as low-income, predominantly Hispanic patients with cancer in an urban public sector care system. This provides useful context, but the registry record does not establish how results would generalize to populations or settings outside that description. The primary outcome is reported at 12 months; longer-term outcomes are not included in the posted primary comparison.

Finally, one secondary outcome comparison is posted, while the record lists two outcome measures. There is no reported multiplicity approach, and the serious adverse event summary does not give further detail on event timing, relatedness, or the denominator for the leading event-term percentages. These reporting limitations constrain interpretation of the posted results; they do not explain the reasons for the missing information.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Treating Depression Among Low-Income Patients With Cancer — NCT00565110.
• Registry record last updated 2017-07-24; status last verified 2017-04.
• Results first posted 2017-07-24.
• 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.

22. Record Summary

NCT00565110 is a completed randomized, parallel-group trial comparing ADAPt-C stepped care with enhanced usual care among low-income patients with cancer and depression. The registry lists single masking, enrollment of 472, and the University of Southern California as sponsor.

The primary outcome was the number of participants with a 50% PHQ-9 score reduction since baseline at 12 months. The posted counts were 57 of 114 for enhanced usual care and 91 of 144 for ADAPt-C, with an unadjusted p-value of 0.033 and relative-effect confidence intervals favoring the intervention group. A physical health secondary comparison was also posted in the same direction.

The registry record does not report the analysis population, missing-data methods, multiplicity plan, or original sample-size justification. Those gaps, together with the unexplained primary endpoint denominators, limit how completely the posted results can be assessed.