Increasing Use of Mental Health Services: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Increasing Use of Mental Health Services (NCT00605358), a randomized na study. The primary outcome is The Primary Outcome is Engagement Defined as at Least One Visit With a Mental Health Provider Who Can Offer Treatment of Depression..

2. Clinical Question

3. Trial Design

This was a randomized, single-masked, parallel-group trial. The registry lists an enrollment of 161 and records 81 participants in the Open Door arm and 80 in the Services Referral control arm. The masking description does not identify who was masked, so the role of masking is not reported in the registry record. The method used to generate the allocation sequence and details of allocation concealment are also not reported in the registry record.

The trial is listed as completed, with a start date of 2007-08 and a primary completion date of 2014-01. Its phase is listed as NA. Weill Medical College of Cornell University is recorded as the sponsor. The intervention was behavioral, and the study model was parallel group.

Open Door participants worked with a counselor to identify barriers to participating in treatment, set goals, problem-solve, and receive a referral. The Services Referral arm received an evaluation and a local mental health provider referral, along with booklet information and an application for HEAP. The registry description of that arm is cut off after describing reduced rates, so further details about the service are not reported in the registry record.

The registry records one primary outcome measure and one analysis. It does not report the original sample-size justification, the recruitment target rationale, or whether the planned enrollment was changed. Additional details about site organization, the exact intervention schedule, and fidelity monitoring 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: A number of participants were excluded from the trial after screening but prior to randomization, due to the following reasons: Non-English speaking, active suicidality, starting new antidepressant or therapy, cognitive impairment, substance abuse, psychosis.

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 The Primary Outcome is Engagement Defined as at Least One Visit With a Mental Health Provider Who Can Offer Treatment of Depression., measured over 12 and 24 weeks.

The primary endpoint was engagement with a mental health provider who could offer treatment for depression. It was recorded as a binary outcome: a participant either had at least one qualifying visit or did not. The registry defines engagement as having a visit by either 12 or 24 weeks, so a participant could meet the outcome at either follow-up time point.

The registry explains that a first visit could include both an initial evaluation and the start of treatment in some settings. Counting at least one visit therefore captures initial contact with a provider without requiring a later treatment visit. The registry also states that older adults may take up to 6 months to follow through on a referral. Including follow-up at 12 and 24 weeks was intended to allow for that delay while retaining one primary outcome, rather than treating the two time points as separate primary endpoints.

This endpoint measures whether contact occurred, not the number of visits, the duration or quality of treatment, adherence after the initial visit, or a change in depressive symptoms. Those outcomes are not reported as secondary results in the registry record. Thus the endpoint addresses service engagement as defined by the trial, rather than clinical improvement or sustained use of care.

7. Results Posted in the Registry

• Open Door Intervention: 74.1% (60/81) versus 56.2% (45/80) on control; risk difference 17.8 percentage points (95% CI 3.4 to 32.3), risk ratio 1.32 (1.04–1.66); unadjusted two-proportion z-test p = 0.018.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: engagement was counted when a participant had at least one qualifying visit, and otherwise was not counted as engaged. For a binary endpoint, the arm-specific proportion describes how common the event was in each group. A comparison can then be expressed on more than one scale. The posted arm percentages, risk difference, risk ratio, and odds ratio all describe the same general contrast, but they answer it using different mathematical definitions.

What does a risk difference describe?

A risk difference subtracts one group’s event proportion from the other group’s event proportion. It is an absolute contrast on the proportion scale. The recalculated risk difference is 0.178, with a confidence interval from 0.034 to 0.323. Its interpretation stays on that scale; it should not be confused with an odds ratio or read as a count of additional participants. The arm percentages are reported separately as 74.1% and 56.3%.

What do a risk ratio and odds ratio describe?

A risk ratio divides one group’s event proportion by the other group’s event proportion. The recalculated risk ratio is 1.317, with a confidence interval from 1.044 to 1.661. An odds ratio instead compares the odds of the event between groups. The registry’s reported estimate is an odds ratio of 2.4, and the unadjusted recalculation gives 2.222. Odds and probabilities are related but are not interchangeable; an odds ratio should not be described as a risk ratio or a direct percentage increase. With a binary outcome, the choice of scale affects the apparent size of the contrast.

How should the p-value be read?

The registry reports a Chi-squared method for its odds-ratio analysis and gives p = .018. The unadjusted recalculation gives p = 0.018. A p-value is a measure of how unusual the observed data, or data more extreme, would be under a specified null hypothesis and the assumptions of the test. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The registry record does not report whether additional statistical tests were planned or whether any adjustment for multiple testing was used.

What does a confidence interval add?

A confidence interval conveys uncertainty around an estimate on its particular scale. The registry’s odds-ratio interval is from 1.17 to 4.93 at the 95% level. The recalculated risk-difference interval is from 0.034 to 0.323; the recalculated odds-ratio interval is from 1.143 to 4.321; and the recalculated risk-ratio interval is from 1.044 to 1.661. Each interval belongs to its own measure and should not be compared as if the scales were the same. General statistical interpretation: a confidence interval procedure is designed to produce intervals that cover the target parameter at the stated rate over repeated use under its assumptions. It is not a guarantee that the particular interval contains the true parameter.

The registry’s reported statistical method is Chi-squared, and the posted results include an odds ratio, confidence limits, and p-value. Details of the exact test implementation, continuity corrections, stratification, covariate adjustment, and treatment of missing outcomes are not reported in the registry record. The unadjusted recalculations are based on the posted group summaries and are not substitutes for an analysis using participant-level data or the sponsor’s analytic procedures.

Why can a binary outcome have several valid contrasts?

Each measure emphasizes a different feature. The risk difference describes absolute separation between proportions. The risk ratio describes their relative proportion. The odds ratio compares odds, which are the ratio of the event probability to the probability of no event. The numerical values therefore need not match. Selecting a measure should follow the scientific question and the prespecified analysis plan, but the trial’s original statistical plan is not reported in the registry record. Presenting the scale explicitly helps prevent a reader from interpreting an odds ratio as an absolute change.

General statistical explanation: a Chi-squared test evaluates whether observed categorical counts depart from the counts expected under a null comparison. Its use relies on assumptions about the data and the test implementation. The registry labels the method but does not report its detailed assumptions or diagnostics. A reader can interpret the posted result as the registry’s categorical comparison, while recognizing that the available record does not provide enough detail to reconstruct every analytic choice.

The single primary endpoint combines qualifying engagement by either 12 or 24 weeks. It is not a comparison of the two follow-up times separately, nor does it count repeated visits. The design therefore answers whether the threshold was reached within the stated follow-up window. It does not characterize the timing of the first visit, the number of later visits, or the course of symptoms. Those distinctions follow from the registry’s endpoint definition and should guide interpretation of the statistical contrast.

10. Confidence Intervals and Precision

The posted arm proportions are 74.1% for Open Door and 56.3% for Services Referral. The confidence interval for the recalculated risk difference runs from 0.034 to 0.323. This interval gives a range of estimates on the absolute difference scale, rather than an interval for either arm’s percentage alone. Its endpoints indicate that the size of the contrast remains uncertain even though the interval is above zero.

For the odds ratio, the registry reports an estimate of 2.4 with a 95% confidence interval from 1.17 to 4.93. The recalculated odds-ratio interval is from 1.143 to 4.321, while the recalculated risk-ratio interval is from 1.044 to 1.661. These intervals are not directly interchangeable because their estimates use different scales. The registry’s confidence limits and the unadjusted recalculated limits also arise from distinct reported calculations, so they should be kept associated with their respective estimates.

Precision is not the same as validity or importance. A narrower interval would represent less statistical uncertainty under its method, while a wider interval would represent more. The interval does not account for unreported design features, missing-data handling, measurement error, or limits to generalizability. The registry does not report confidence intervals for the arm-specific percentages themselves, nor does it report adjusted intervals. Consequently, the available precision statements concern the listed contrasts, not every possible interpretation of engagement.

The p-value of .018 in the registry’s analysis and the recalculated p-value of 0.018 are consistent with the respective posted tests, but a p-value alone does not show how precise the estimated effect is. The confidence intervals add information about the range of effect sizes compatible with the data and method. They should be read with attention to their scale and the analysis from which each was obtained.

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 sample sizes or total events. They are based on the observed effect and use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The planning scenarios consider detecting 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The trial’s original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the number of participants per arm in the planning illustration for a design intended to have an 80% chance of detecting the specified effect under its assumptions. The 90% power column gives the corresponding per-arm planning number for a design intended to have a 90% chance. For example, planning for 100% of the observed effect gives 111 participants per arm for 80% power and 149 participants per arm for 90% power. Planning for 75% of the observed effect gives 204 and 273 participants per arm, respectively; planning for 50% gives 471 and 630 participants per arm, respectively.

The pattern reflects a general principle: smaller effects are harder to distinguish from random variation, so detecting them at the same target power generally requires more participants. These are planning illustrations calculated from the observed effect under stated assumptions. They are not observed power and do not describe the probability that the completed trial’s result is correct. The registry record does not report whether these assumptions were used when the trial was originally designed.

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 = 0.562; treatment proportion p₁ = 0.741; average p̄ = 0.652.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6738 = 1.3206.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6619 = 0.5571.
4. n = (1.3206 + 0.5571)² / (0.1782)² = 111.0, rounded up to 111 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 <- 0.5625                          # control proportion
p1 <- 0.740740740740741                          # 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(605358)
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 = 111 (80% power), 149 (90% power)
 75% of observed effect: n per arm = 204 (80% power), 273 (90% power)
 50% of observed effect: n per arm = 471 (80% power), 630 (90% power)
power.prop.test: 111 
Simulated power at n = 111 per arm: 0.806 (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. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as a Chi-squared comparison with an odds ratio of 2.4, a 95% confidence interval from 1.17 to 4.93, and p = .018. The unadjusted recalculation gives an odds ratio of 2.222, with a confidence interval from 1.143 to 4.321, and p = 0.018. It also gives a risk difference of 0.178 and a risk ratio of 1.317. The registry record does not flag a discrepancy.

The sponsor’s odds ratio and the recalculated odds ratio are close but not identical. These are distinct reported results, and the sponsor’s analysis is the registry’s own reported analysis. The available record does not explain the difference between the estimates or confidence limits, so no cause should be inferred. The unadjusted recalculation is a simple comparison from the posted summary statistics, not a reconstruction of the sponsor’s analytic workflow or a replacement for participant-level analysis.

The p-values are reported as .018 for the sponsor’s analysis and 0.018 for the recalculation. Agreement in the displayed p-values does not make the estimates or their methods identical. Interpretation should retain the measure, interval, and source together: the sponsor’s odds ratio describes odds, while the recalculated risk difference describes an absolute contrast in event proportions.

14. Multiplicity, Missing Data and Analysis Populations

The registry records one primary outcome measure and one analysis, and it does not post secondary outcome results. The endpoint is defined as a single binary engagement outcome, met by a qualifying visit at either 12 or 24 weeks. The registry describes the two follow-up opportunities as ways to fulfill that one endpoint, not as separate primary outcomes. This structure is relevant when considering multiplicity: the outcome definition itself combines the follow-up times, but the record does not provide the full analysis plan.

Whether the trial prespecified additional tests, subgroup analyses, or alternate outcomes is not reported in the registry record. A multiplicity adjustment, or a statement that none was planned, is also not reported in the registry record. The single analysis and single primary outcome recorded in the results do not establish that no other analyses were conducted. Accordingly, the posted p-value should be understood as the reported test result, without assuming a particular family of planned comparisons.

The registry does not report how missing endpoint data were handled, whether participants with incomplete follow-up were included in the analysis, or whether any imputation was used. The analysis population is not reported in the registry record; it is not possible to establish from the posted information whether the analysis followed an intention-to-treat principle or another population definition. The arm denominators are recorded, but that information alone does not resolve how missing observations or exclusions were treated.

Further details about protocol deviations, outcome ascertainment procedures, and analysis-set exclusions are not reported in the registry record. These omissions limit assessment of potential bias and of how closely the result reflects the randomized groups as assigned. They also make it difficult to determine whether the published comparison accounts for all randomized participants or only those with available outcome data. The results can be described as posted, while these aspects of the analysis remain uncertain.

15. Safety Results

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

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: hospitalization (62.5%), death (4.9%), active suicidal ideation (2.5%).

The serious adverse event figures list 14 affected participants among 81 at risk in the Open Door arm and 17 among 80 at risk in the Services Referral arm. Other adverse events list 0 affected among 80 at risk for Open Door and 0 among 81 at risk for Services Referral. The at-risk denominators are recorded differently across the serious and other adverse event summaries, so each count should be read with its own stated denominator.

The registry lists hospitalization, death, and active suicidal ideation among the top serious-event terms, with 62.5%, 4.9%, and 2.5%, respectively. These are event-term percentages in the record; their denominators and the way terms were grouped are not reported in the registry record. The deaths field for the arm summaries is null, which means a death count is not reported there, rather than establishing that there were no deaths.

These safety figures are descriptive registry entries. They do not establish that an event was caused by an intervention, nor do they provide enough detail to compare event timing, severity, or relatedness. The registry record does not report further context for adjudication or attribution.

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

The posted results show a higher engagement proportion in the Open Door arm than in the Services Referral arm: 74.1% compared with 56.3%. The unadjusted risk-difference estimate is 0.178, and the registry’s reported odds-ratio analysis has p = .018. These results describe the trial’s defined outcome: at least one visit with a mental health provider able to offer depression treatment by either 12 or 24 weeks.

The finding is specific to initial engagement during that follow-up window. It does not show how many visits participants made, whether they continued treatment, whether the care was received as intended, or whether depression symptoms changed. The registry does not post secondary outcome results. It also does not provide enough information to determine the effect of missing observations or to assess all potential sources of bias.

An odds ratio is not a direct measure of the difference in event proportions. The recalculated risk difference and risk ratio offer other ways to express the same general comparison, each with its own scale and interpretation. The confidence intervals show uncertainty in the estimated contrast, and the registry’s reported interval is distinct from the unadjusted recalculated interval. A p-value should not be treated as a measure of effect size, importance, or the probability that a finding is true.

The trial’s recorded population and service context also bound interpretation. The registry identifies major depression and describes community-dwelling adults, with its brief summary referring to depressed community elderly. Eligibility details and information needed to judge applicability to other populations or service settings are not reported in the registry record. The result therefore describes the posted comparison in this trial, not every mental health service or population.

These statistics do not establish that Open Door caused every observed difference or that a particular treatment choice is appropriate. They provide a record-based comparison of engagement under two assigned service approaches. Broader claims about clinical outcomes, long-term use, or implementation require evidence beyond the endpoint and details recorded.

17. Why This Trial Matters Statistically

This trial illustrates why defining a binary clinical-service endpoint carefully matters. Engagement was not defined as symptom improvement or completion of a course of care. It meant a visit with a provider able to offer treatment, and a visit could count by either 12 or 24 weeks. Combining these follow-up opportunities into one endpoint gives a clear threshold for initial contact, while leaving the timing and continuity of care unanswered.

It also demonstrates how the same binary result can be expressed with different effect measures. The risk difference describes an absolute contrast, the risk ratio compares proportions, and the odds ratio compares odds. Their values are not interchangeable. Naming the scale and pairing an estimate with its interval makes the comparison more interpretable than relying on a p-value alone.

The registry reports both its own Chi-squared odds-ratio analysis and summary statistics from which unadjusted comparisons can be recalculated. Keeping those results distinct is important: the posted estimates are similar in direction, but the record does not explain why the odds-ratio estimates and intervals differ. Recalculation can help readers understand the data on several scales, but it cannot recreate undocumented analysis choices.

Finally, the record highlights the value of transparent reporting. Details about the original sample-size plan, missing-data handling, analysis populations, multiplicity, and the role of masking are not reported in the registry record. Those gaps do not erase the posted results, but they limit how fully readers can assess precision, bias, and generalizability. Statistical interpretation depends on both the estimates and the design information available to support them.

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

Several trial-specific details needed for a fuller assessment are not reported in the registry record. The masking entry says single, but does not say who was masked. The allocation sequence and concealment procedures are not described. Eligibility criteria beyond the broad description of community-dwelling adults with depression are absent, and the brief summary’s reference to older adults does not supply age limits or other criteria.

The registry does not report the original sample-size justification or the assumptions used in the trial’s design. The planning illustrations based on the observed effect cannot establish what enrollment was targeted in advance. The record also does not specify outcome-data completeness, missing-data handling, the analysis population, or whether any participants were excluded from the primary analysis.

The control-arm description is incomplete, and the record does not provide additional detail on how the referral service was delivered. It does not report intervention fidelity, counselor training, the timing or number of counseling contacts, or adherence to the assigned approach. Such information could help readers assess whether the arms differed only as intended, but its absence does not establish that any particular implementation problem occurred.

The endpoint measures initial service engagement within the stated follow-up window, not sustained treatment, symptoms, or other clinical outcomes. No secondary results are posted. The safety entries also lack reported context about event relatedness and timing, and the deaths field is null. These reporting limits restrict interpretation of outcomes beyond the stated primary endpoint and the descriptive adverse-event counts.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Increasing Use of Mental Health Services — NCT00605358.
• Registry record last updated 2017-03-03; status last verified 2017-01.
• Results first posted 2017-03-03.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

22. Record Summary

NCT00605358 is a completed, randomized, single-masked, parallel-group trial titled “Increasing Use of Mental Health Services.” It evaluated Open Door, a brief psychosocial intervention, against Services Referral for community-dwelling adults with depression. The registry lists an enrollment of 161 and records 81 participants in the Open Door arm and 80 in the control arm.

The primary outcome was engagement with a mental health provider able to offer depression treatment, defined as at least one visit by either 12 or 24 weeks. The posted engagement proportions were 74.1% for Open Door and 56.3% for Services Referral. The registry reports a Chi-squared odds ratio of 2.4, with a 95% confidence interval from 1.17 to 4.93 and p = .018. Further details about missing data, analysis populations, and the original design assumptions are not reported in the registry record.