INtegrating DEPrEssioN and Diabetes treatmENT (INDEPENDENT) Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for INtegrating DEPrEssioN and Diabetes treatmENT (INDEPENDENT) Study (NCT02022111), a randomized na study. The primary outcome is Percentage of Combined Improvement of Depressive Symptoms and CVD Risk Factors.

2. Clinical Question

The study was designed around the challenge of addressing depression alongside diabetes care. Its stated aim was to develop a culturally adapted model of combined care and evaluate its effectiveness and costs in routine clinic settings in India. The registry describes patient focus groups and individual interviews as part of the adaptation work, followed by a trial across four clinics.

The primary outcome is a combined, participant-level measure rather than a single depression or diabetes measure. To qualify, a participant needed a reduction of at least 50% in SCL-20 score and at least one of the specified cardiovascular risk-factor improvements: a reduction of at least 0.5% in HbA1c, at least 5 mmHg in systolic blood pressure, or at least 10mg/dl in LDL-c. The endpoint therefore represents joint improvement across depression and one or more risk factors, not improvement in every component.

The posted group percentages are 71.6% for the intervention program and 54.7% for the control arm. The registry also posts an unadjusted risk difference and confidence interval. These results address the defined combined endpoint at the stated follow-up, not every outcome of interest in depression, diabetes, cardiovascular risk, or care costs.

3. Trial Design

The INtegrating DEPrEssioN and Diabetes treatmENT (INDEPENDENT) Study was a randomized, parallel-group trial sponsored by Emory University. Its status is completed, and the registry records total enrollment as 404. The intervention and control arms had posted group sizes of 196 and 208, respectively. The study began in 2014-03, with primary completion recorded as 2018-07-14.

The trial took place at four diabetes clinics in India, according to the registry summary. The intervention program included patient education and behavioral activation delivered by a care coordinator, support for self-care, psychiatrist and diabetologist reviews, and a decision-support electronic health record system. The control arm received the existing standard of care and routine diabetes treatment at the clinic, with the care provider notified about depressive symptoms. The registry’s control-arm description ends mid-sentence, so additional details about that arm are not reported in the registry record.

Allocation is recorded as randomized, the model as parallel, and masking as single. Who was masked and how masking was implemented are not reported in the registry record. The trial phase is listed as NA. The randomization procedure, allocation concealment, and the original design assumptions are not reported in the registry record.

The primary endpoint was the percentage of participants meeting a combined depression and cardiovascular risk-factor improvement definition 24 months post-intervention. The registry also lists secondary outcomes concerning common effects, achievement of cardiovascular risk targets, changes in SCL-20, HbA1c, systolic blood pressure and LDL-c, and target achievement or significant reductions for some individual factors. Although the registry lists outcome measures, it does not post an analysis count beyond recording zero analyses. Details of the analysis population, handling of missing outcomes, and adjustment strategy 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.

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.

Population: At four public and private urban diabetes clinics in India, we randomized diabetes patients with moderate-to-severe depressive symptoms and poorly-controlled cardio-metabolic indices to the INDEPENDENT collaborative care model or usual care in a 1:1 ratio.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Combined Improvement of Depressive Symptoms and CVD Risk Factors, measured over 24-months post-intervention.

The primary endpoint is the percentage of participants in each arm who met a combined improvement definition at 24 months post-intervention. Meeting the definition required a reduction of at least 50% in SCL-20 score and at least one of three cardiovascular risk-factor changes: a reduction of at least 0.5% in HbA1c, at least 5 mmHg in systolic blood pressure, or at least 10mg/dl in LDL-c.

This composite uses a conjunctive rule for depression and a disjunctive rule for the risk factors: improvement in depression was required, while improvement in any one or more of the listed cardiovascular measures could satisfy the second part. It does not require improvement in all three cardiovascular measures. A participant who improved on a single component but did not meet the full combined definition would not count as achieving this endpoint.

A combined endpoint can summarize whether improvement spans more than one clinical domain, which fits the study’s aim of integrating depression and diabetes care. It also makes the result dependent on every part of the definition, including measurement and threshold choices. The overall percentage does not show which component drove the result or how many participants met each individual criterion. Those component-level details are not reported in the registry record as primary-endpoint results.

The posted values are 71.6% in the intervention program arm and 54.7% in the control arm. These are percentages meeting the combined definition, not mean changes in symptoms or risk factors. They should be read as outcome proportions at the stated follow-up rather than as evidence about all aspects of care or longer-term outcomes.

7. Results Posted in the Registry

• Intervention Program of Care: 71.4% (140/196) versus 54.8% (114/208) on control; risk difference 16.6 percentage points (95% CI 7.4 to 25.9), risk ratio 1.30 (1.12–1.52); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: a participant either met the combined improvement definition or did not. A percentage in each arm describes the share meeting that definition. The main contrast is a risk difference, which expresses the difference between the two arm proportions on the absolute probability scale. The registry also reports a relative risk and an odds ratio, which describe the comparison on ratio scales.

What does a risk difference describe?

A risk difference subtracts one group’s outcome proportion from the other group’s proportion. It is an absolute measure: it describes how far apart the groups are in the proportion meeting the endpoint. The registry reports the risk difference as 0.166. No conversion to another scale is needed to interpret the posted estimate. Its confidence interval ranges from 0.074 to 0.259.

In the posted comparison, the intervention program is compared with the control arm. The event counts are 140 of 196 and 114 of 208, respectively. The group percentages in the arm summaries are 71.6% and 54.7%. The comparison’s unadjusted estimates use the event counts and denominators recorded for that analysis. The registry does not report a method explaining the difference in rounding between those summaries, and it does not report an adjusted comparison.

How do relative risk and odds ratio differ?

The relative risk compares the probability of meeting the endpoint between arms. The odds ratio compares the odds of meeting it. Probabilities and odds are different quantities; as an outcome becomes more common, the odds ratio can be farther from 1 than the relative risk. Therefore, the odds ratio of 2.061 should not be described as a doubling of the probability. The posted relative risk is 1.303. Both estimates have confidence intervals recorded in the registry.

These measures each summarize a group contrast, but none identifies why the contrast arose. Because the primary endpoint is composite, the aggregate comparison does not separately estimate the contrast for SCL-20, HbA1c, systolic blood pressure, or LDL-c. Secondary outcomes are listed, but their detailed results are not reported in the registry record.

What does the p-value tell the reader?

The posted p-value is unadjusted and reported as p < 0.001. In general, a p-value assesses how unusual data at least as incompatible with a null model would be if the null model and the test assumptions held. It is not the probability that the null is correct, and it does not quantify the clinical importance of the result. A small p-value alone does not provide the effect size or its precision; those are conveyed by the estimate and confidence interval.

The registry does not describe the exact test used for the primary comparison, the handling of any clustering, or whether covariates were included. Randomized parallel allocation is recorded, but the randomization method and analysis details are not. Thus, the reported comparison should be understood as the unadjusted estimate identified in the record, rather than as an adjusted or model-based estimate of a different kind.

Why can a composite endpoint be useful and challenging?

Combining measures can represent a goal that spans depression and cardiovascular risk factors. Here, meeting the endpoint requires improvement in SCL-20 and in at least one listed risk factor. This can give a concise measure of joint progress, but its meaning depends on the component thresholds and the rule for combining them. The overall proportion does not tell readers whether all components changed in the same direction or which component most often qualified participants.

General statistical consideration: composite endpoints can be influenced by components that occur more often or are easier to meet. Component-specific results help readers understand the makeup of the overall outcome. The registry lists several secondary measures, but component-specific outcome estimates and their uncertainty are not reported in the registry record in the posted primary results.

What is not described about the analysis?

The registry records zero analyses and does not report the prespecified statistical analysis plan, the analysis population, missing-data methods, or any adjustment for multiple comparisons. It does not report whether the primary result used an intention-to-treat population or another population. These omissions limit assessment of how the analysis was conducted and whether it followed its original plan. The original trial design assumptions, including the sample-size justification, are also not reported in the registry record.

General statistical consideration: randomization supports a fair comparison when allocation and follow-up are handled appropriately, but it does not remove the need to understand exclusions, missing outcomes, and analysis decisions. Without those details, readers can describe the posted unadjusted result and its precision, but cannot fully audit the analysis or assess sensitivity to alternative assumptions.

10. Confidence Intervals and Precision

The risk-difference estimate is 0.166, with a 95% confidence interval from 0.074 to 0.259. A confidence interval describes the statistical uncertainty around an estimate under the method used to construct it. In general, a 95% confidence interval is produced by a procedure that would cover the true parameter in 95% of repeated samples when its assumptions hold. It is not a guarantee that a particular interval contains the true value.

For this comparison, the entire reported interval lies above zero. The lower and upper limits indicate a range of positive risk differences compatible with the data and analysis method. The interval does not capture every uncertainty relevant to interpretation: it reflects sampling uncertainty under the calculation, not possible bias from missing outcomes, measurement choices, or analysis decisions that are not described in the registry record.

The relative risk is 1.303, with a confidence interval from 1.12 to 1.517. The odds ratio is 2.061, with a confidence interval from 1.364 to 3.116. Those intervals are on different scales from the risk-difference interval, so their numerical widths should not be compared directly. Each interval is best interpreted alongside its corresponding effect measure.

The registry posts no confidence intervals for the arm percentages themselves and no uncertainty estimates for the individual components of the combined endpoint. It also does not report an adjusted interval or a sensitivity analysis. Thus, the posted intervals describe the reported unadjusted comparisons, not all possible sources of uncertainty or all possible analysis approaches.

11. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total participants and not total events. They are illustrations calculated from the observed effect, rather than the trial’s own original design assumptions; those original assumptions are not reported in the registry record. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect.

For planning at 100% of the observed effect, the figures are 132 participants per arm for 80% power and 175 participants per arm for 90% power. At 75% of the observed effect, they are 240 participants per arm for 80% power and 320 participants per arm for 90% power. At 50% of the observed effect, they are 549 participants per arm for 80% power and 735 participants per arm for 90% power.

In general, power is the chance that a specified study design will detect an effect of the size assumed when the null hypothesis is false, given the chosen testing procedure and its assumptions. The 80% and 90% columns represent different target chances of detecting the assumed effect. The higher target generally requires more participants under the stated planning setup. Smaller assumed effects also require more participants, as the listed scenarios illustrate.

These figures are planning illustrations based on the observed result, not observed power. They do not describe the actual enrollment or prove that the completed trial had a particular probability of detecting its effect. The registry does not report the trial’s original sample-size justification or design assumptions.

12. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Control proportion p₀ = 0.548; treatment proportion p₁ = 0.714; average p̄ = 0.631.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6823 = 1.3374.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6721 = 0.5657.
4. n = (1.3374 + 0.5657)² / (0.1662)² = 131.1, rounded up to 132 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 <- 0.548076923076923                          # control proportion
p1 <- 0.714285714285714                          # 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(2022111)
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 = 132 (80% power), 175 (90% power)
 75% of observed effect: n per arm = 240 (80% power), 320 (90% power)
 50% of observed effect: n per arm = 549 (80% power), 735 (90% power)
power.prop.test: 131.1 
Simulated power at n = 132 per arm: 0.803 (20000 replicates)

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

13. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and multiple secondary outcome measures, including common effects, cardiovascular risk-target achievement, and changes or target achievement for SCL-20, HbA1c, systolic blood pressure, and LDL-c. It does not report an analysis count beyond recording zero analyses, and detailed secondary-outcome results are not included in the posted record. The order of testing, if any, is not reported in the registry record.

When many hypotheses are tested, the chance of at least one small p-value can increase. General statistical consideration: multiplicity procedures, such as prespecified testing sequences or adjusted significance thresholds, can address this issue. Whether any such procedure was planned or used for this trial is not reported in the registry record. The primary comparison is reported with an unadjusted p-value, so it should not be described as multiplicity-adjusted.

The registry does not report the amount or pattern of missing primary outcomes, how missing data were handled, or whether participants with missing results were included in a particular analysis population. It also does not report whether the primary comparison followed an intention-to-treat approach. These details matter because missing outcomes and exclusions can affect the estimate, particularly when their occurrence differs between arms or relates to the outcome.

The record provides arm denominators and event counts for the reported primary comparison, but those numbers alone do not explain the treatment of all randomized participants or the reasons for any differences between enrollment and the comparison denominators. The randomization procedure and allocation concealment are also not reported. Readers can interpret the posted group comparison, but cannot establish from the record whether every analysis decision was prespecified or assess the robustness of the result to missing-data assumptions.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

The registry reports serious adverse events in zero of 196 participants at risk in the intervention program arm and zero of 208 participants at risk in the control arm. Other adverse events are also reported in zero of 196 and zero of 208 participants at risk, respectively. No serious adverse-event terms are listed.

Deaths are reported separately as 2 in the intervention program arm and 7 in the control arm. These counts should be read as the registry’s reported figures, not as evidence that the intervention caused or prevented deaths. The record does not report timing, causes, relatedness assessments, or a statistical comparison for deaths. It also does not provide further detail about how serious and other adverse events were defined, collected, or attributed.

Zero reported events in a category does not establish that an event is impossible or that no unrecorded events occurred. The figures describe what the registry reports for the listed groups and categories. They do not, by themselves, support a comparative safety conclusion. Interpretation is limited by the lack of event descriptions, follow-up detail, and analyses in the registry record.

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

The posted primary result is a comparison of the percentages meeting a combined outcome definition at 24 months post-intervention. The intervention program arm’s posted percentage is 71.6%, and the control arm’s is 54.7%. The registry’s unadjusted risk difference is 0.166, with a 95% confidence interval from 0.074 to 0.259 and p < 0.001. In statistical terms, the observed comparison favors a higher percentage meeting the combined endpoint in the intervention program arm, with the reported interval above zero.

The endpoint requires improvement in depressive symptoms and at least one of three specified cardiovascular risk factors. It does not require improvement in every risk factor, and it does not measure each domain separately. The overall finding therefore cannot identify which endpoint component accounts for the group contrast. The registry lists secondary measures, but detailed component results are not reported in the registry record.

This result does not establish that every participant benefited, that every component improved, or that the same result would occur in other healthcare settings. It does not show the effect of each intervention component separately, because the program combines education and behavioral activation, self-care support, specialist reviews, and an electronic decision-support system. The registry does not report a comparison that isolates any one of those elements.

The posted estimate is unadjusted. The analysis population, missing-data handling, and adjustment strategy are not reported in the registry record. These omissions limit assessment of how the estimate might depend on analytic choices. The confidence interval describes sampling uncertainty under the method used; it does not address every possible source of bias or uncertainty.

The findings describe this trial’s recorded outcome and follow-up. The registry identifies four diabetes clinics in India and an integrated-care model intended for routine clinics. The result should not be treated as a guarantee of the same outcomes in other populations or care systems. Nor does the combined endpoint alone establish effects on other clinical outcomes or costs, for which detailed results are not reported in the registry record.

16. Why This Trial Matters Statistically

This trial poses a statistical question that spans more than one domain: whether participants met a joint threshold for improvement in depressive symptoms and cardiovascular risk. That framing makes the endpoint relevant to an integrated-care model, while also requiring careful interpretation of the composite definition. A favorable result on the combined outcome does not reveal which component contributed most or whether each component changed in the same way.

Randomized allocation and a parallel design are recorded, offering a structure for comparing the intervention program with routine care. The reported arm sizes and unadjusted comparison allow readers to examine both absolute and relative measures of association. Reporting a risk difference alongside relative risk and odds ratio makes clear that a treatment contrast can be expressed on several scales, each with a distinct interpretation.

The confidence interval is important because the p-value alone would not describe how precisely the difference was estimated. The reported interval for the risk difference lies above zero, while the record’s lack of adjusted analyses, component-specific primary results, missing-data methods, and multiplicity details limits a complete appraisal. This illustrates why statistical interpretation depends not only on a headline comparison but also on endpoint construction and analysis transparency.

The planning illustrations also show a general feature of study design: targeting detection of a smaller effect requires more participants under the listed assumptions. These are not the original design assumptions and are not observed power. Keeping planning estimates separate from the completed trial’s observed result prevents a retrospective illustration from being mistaken for the study’s actual design justification.

Overall, the trial is statistically informative as a comparison of a defined integrated-care program with standard care on a combined endpoint. The registry record also shows why estimates must be read alongside uncertainty, endpoint rules, and reporting gaps before drawing broader conclusions.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry provides an abbreviated account of the trial’s methods and results. It records randomized allocation, a parallel model, and single masking, but does not identify who was masked or describe allocation concealment, randomization procedures, or the original statistical analysis plan. The control-arm description is incomplete, ending mid-sentence. These gaps restrict interpretation of how the comparison was implemented.

The primary endpoint combines SCL-20 improvement with at least one cardiovascular risk-factor improvement. The posted primary comparison does not separate the components, and detailed component-specific results are not reported in the registry record. As a result, the overall percentage cannot show which measures most influenced endpoint achievement. The endpoint’s thresholds and combination rule also mean that it captures one defined form of joint improvement, not all possible patterns of change.

The registry does not report the primary analysis population, missing-outcome methods, adjusted estimates, or sensitivity analyses. It does not explain the original sample-size justification or design assumptions. The posted planning illustrations are based on the observed effect and should not be confused with those missing original assumptions. The record also provides no detailed results for the listed secondary outcomes and no multiplicity strategy.

The trial is described as taking place at four diabetes clinics in India, but further information about participant eligibility and clinic-level variation is not reported in the registry record. The duration is specified for the primary endpoint as 24 months post-intervention; other follow-up details are not reported. These limits mean the posted comparison can be characterized, but its broader applicability and sensitivity to analysis choices cannot be fully assessed from the registry record alone.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: INtegrating DEPrEssioN and Diabetes treatmENT (INDEPENDENT) Study — NCT02022111.
• Registry record last updated 2021-05-03; status last verified 2021-04.
• Results first posted 2020-05-14.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

The INtegrating DEPrEssioN and Diabetes treatmENT (INDEPENDENT) Study was a completed, randomized, parallel trial sponsored by Emory University. It enrolled 404 participants and was conducted at four diabetes clinics in India. The intervention program combined patient education and behavioral activation, self-care support, psychiatrist and diabetologist reviews, and an electronic decision-support system; the control arm received existing standard care and routine diabetes treatment.

The primary endpoint was the percentage meeting a combined depression and cardiovascular risk-factor improvement definition 24 months post-intervention. The posted percentages were 71.6% for the intervention program and 54.7% for control. The unadjusted risk difference was 0.166, with a 95% confidence interval from 0.074 to 0.259 and p < 0.001. Analysis details, missing-data methods, and the original design assumptions are not reported in the registry record.