Suicide Prevention for Criminal Legal Involved Managed Care Subscribers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Suicide Prevention for Criminal Legal Involved Managed Care Subscribers (NCT05579600), a randomized na study. The primary outcome is Suicide Attempt.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group trial with no masking reported. The registry lists the study status as completed, enrollment as 45155, the start date as 2022-09-01, and primary completion as 2024-09-03. The sponsor was Butler Hospital. The study concerns suicide and self-harm among managed care subscribers following jail release.

Participants were assigned to Caring Contact Letters or Caring Contact Control. Letters were to be mailed to a randomly selected group over a 6-month period following jail release; the control arm was described as care as usual. The registry does not report how randomization was generated or concealed, or provide a fuller description of the care-as-usual condition. Masking was listed as none, but further details about who knew each assignment are not reported in the registry record.

The primary endpoint was the mean number of suicide attempts during 6-months, assessed using claims data and an established Mental Health Research Network algorithm described in the outcome definition. Four secondary outcomes have posted numerical summaries: all-cause injury and poisoning, outpatient behavioral health service linkage, CareSource Member Services Line use, and inpatient and emergency department mental health care visits is listed among the secondary outcomes. Numerical summaries for the latter listed outcome are not included in the posted secondary results. The registry lists 6 outcome measures but only 4 posted secondary summaries. Formal reported analyses are not listed. The protocol’s original sample-size assumptions and detailed analysis plan 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.

6. Primary Endpoint

The registry lists the primary outcome as Suicide Attempt, measured over 6-months.

The primary endpoint is the number of subsequent suicide attempts during 6-months. The registry defines a suicide attempt as engaging in potentially self-injurious behavior with at least some intent to die as a result of the act. It states that attempts would be identified from claims data using the Mental Health Research Network algorithm for extracting suicide-attempt information.

The endpoint is presented as a mean, with standard deviation describing variation in the number of attempts across participants. This approach summarizes the average recorded count in each group, while the standard deviation indicates how dispersed individual counts are around that average. A mean count gives a group-level description; it does not describe every participant’s experience or show how attempts were distributed among individuals.

Using a claims-based algorithm can provide a consistent way to identify events in administrative data, but the registry record does not report the algorithm’s operating characteristics in this trial, how incomplete claims were handled, or how events were confirmed. The endpoint therefore represents attempts identified under the stated claims-based definition, rather than every possible attempt regardless of whether it appeared in the data. The record’s primary outcome summaries are means and standard deviations, with a calculated unadjusted difference and confidence interval available for the comparison.

7. Results Posted in the Registry

• Caring Contact Letters: mean 0.02 versus 0.02 on control; difference -0.00 (95% CI -0.01 to 0.01), standardized effect d = -0.01; z-test p = 0.562.

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.

All-cause Injury and Poisoning

• 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:
  • Caring Contact Letters: mean 1.18 versus 1.17 on control; difference 0.01 (95% CI -0.06 to 0.08), standardized effect d = 0.00; z-test p = 0.772.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Caring Contact Letters and the control. The interval gives the range of effects that remain compatible with the data.

Outpatient Behavioral Health Service Linkage

• 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:
  • Caring Contact Letters: mean 0.32 versus 0.34 on control; difference -0.02 (95% CI -0.11 to 0.07), standardized effect d = -0.00; z-test p = 0.669.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Caring Contact Letters and the control. The interval gives the range of effects that remain compatible with the data.

CareSource Member Services (CMS) Line Use

• 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:
  • Caring Contact Letters: mean 0.01 versus 0.02 on control; difference -0.01 (95% CI -0.02 to 0.01), standardized effect d = -0.01; z-test p = 0.405.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Caring Contact Letters and the control. The interval gives the range of effects that remain compatible with the data.

The reported secondary comparisons do not show a consistent direction of separation across the different outcome domains. The mean differences vary in direction, while the confidence intervals for each posted comparison include zero. Considered together, the pattern does not indicate a clear, coherent shift across the measured injury, service-linkage, and member-services outcomes. The registry lists inpatient and emergency department mental health care visits as a secondary outcome, but numerical results for that listed measure are not among the posted secondary summaries.

These secondary measures address aspects of healthcare use and injury that are related to, but distinct from, the primary count of suicide attempts. They can add context about whether service engagement or other recorded events changed alongside the primary outcome. However, each measure has its own definition, follow-up period, data source, and possible sources of variation. In particular, outpatient behavioral health linkage is reported over 18 months, whereas the primary endpoint is reported over 6-months. The results should not be treated as interchangeable evidence about one underlying outcome.

The secondary p-values are unadjusted, and the registry record does not report a multiplicity strategy. When several outcomes are examined, the chance of seeing an apparently notable result by chance can increase. Without a prespecified hierarchy or adjustment, isolated secondary findings generally carry less weight than a clearly defined primary comparison. Here, the uncertainty across the posted comparisons reinforces the need to regard them as exploratory or descriptive rather than as independent confirmation of the primary result.

The secondary summaries do not resolve the primary endpoint’s uncertainty. They neither establish a broader effect across all measured outcomes nor provide a substitute for the missing numerical result for the listed inpatient and emergency department outcome. The record also does not report formal secondary analyses, adjusted estimates, or how missing observations were handled. Overall, the secondary results provide limited contextual evidence and should be interpreted alongside those reporting gaps.

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

• All-cause Injury and Poisoning
• Outpatient Behavioral Health Service Linkage
• CareSource Member Services (CMS) Line Use
• Inpatient and Emergency Department Mental Health Care Visits

9. Statistical Interpretation

10. Statistical Methods Explained

The posted primary and secondary outcomes are continuous measures summarized as means and standard deviations. For the attempt endpoint, the mean represents the average number of recorded attempts per participant within an arm during the stated follow-up period. The standard deviation describes variation in individual counts around the arm mean. These are distinct quantities: a standard deviation is not a confidence interval for the mean, and it does not directly indicate how precisely the group difference has been estimated.

What does the mean difference represent?

The comparison is expressed as Caring Contact Letters minus Caring Contact Control. A negative estimate therefore means that the posted mean was lower in the letters arm; a positive estimate means it was higher. For the primary endpoint, the calculated unadjusted difference was -0.003. The comparison is derived from posted arm summaries and is not identified in the registry record as a sponsor-reported adjusted analysis. A mean difference is on the original outcome scale, so its practical interpretation depends on the outcome definition and the population being summarized.

What does the confidence interval add?

A confidence interval describes the uncertainty around an estimated difference under the assumptions of the method used to construct it. A narrow interval indicates greater precision than a wide interval, while an interval that includes zero leaves a range of possible differences that includes no average difference. It is not a range containing a fixed proportion of individual participants’ outcomes, nor does it assign a probability to the true value after the data have been observed. The registry does not report the exact inferential procedure used for the trial’s own primary analysis; the displayed intervals accompany unadjusted comparisons calculated from summary statistics.

Why are standard deviations and means not the whole distribution?

A mean and standard deviation compress the outcome data into two summary values. They cannot show whether the counts are symmetric, concentrated at zero, or driven by a small number of participants with larger counts. Count outcomes can be skewed or have more variability than a simple model assumes. Over-dispersion means that the observed variation exceeds what a model’s assumed distribution would predict. In general, that can make standard errors and tests based on a poorly fitting model unreliable. The registry record does not report whether over-dispersion was assessed or whether a count-specific model was used, so the summaries alone cannot establish the model’s suitability.

What do the p-values say?

The p-values are unadjusted values associated with the calculated comparisons. A p-value describes how unusual data at least as inconsistent with a specified no-difference model would be, if that model and the analysis assumptions were correct. It is not the probability that the no-difference model is true, and a larger p-value does not establish that the groups are equivalent. Interpretation should also consider the effect estimate, confidence interval, study design, and analysis plan. The registry record does not report a formal analysis plan or a trial-specific decision threshold for the p-values.

What is the role of standardized effects?

The comparison summaries include a standardized effect, which expresses a mean difference relative to a pooled standard deviation. Standardization can help put effects from measures on different scales into a comparable unit, but it does not replace the original-scale difference or its confidence interval. The standardized values in the posted comparisons are close to zero. Because the observed differences are also used in the planning illustrations, those illustrations depend on these estimates and should not be interpreted as proof that the same effect would recur in another study.

The trial is randomized and parallel-group, which supports comparing outcomes between assigned arms, but the record does not provide details needed to evaluate the implementation of randomization, adherence, analysis populations, missingness, or model adjustment. No formal reported analyses are listed. Accordingly, the statistical summaries should be understood as unadjusted comparisons derived from the recorded arm-level means and standard deviations, not as a complete account of the trial’s planned or sponsor-conducted analysis.

11. Confidence Intervals and Precision

The primary mean difference is -0.003, with a 95% confidence interval from -0.012 to 0.006. The interval includes zero, so the estimated direction is not precise enough to exclude either a negative or positive difference on the mean-count scale. Its width indicates that the data summaries remain compatible with a range of possible group differences, rather than pinpointing a single value.

Precision is different from whether a result is important. The interval concerns uncertainty in the average difference, not the range of outcomes for individual participants. Nor does it establish that the groups are equivalent: an equivalence conclusion would require an appropriate margin and analysis, neither of which is reported in the registry record. The unadjusted p-value of 0.562 is consistent with the interval’s inclusion of zero, but the interval is more informative about both direction and the range of differences compatible with the estimate.

The arm standard deviations for the primary outcome are 0.387 and 0.584, while the means are 0.020 and 0.023. Large variation relative to the means can occur with count data and makes it especially important not to infer the full outcome distribution from the averages alone. The record does not provide an individual-level distribution or a model-based account that would clarify the sources of variation.

The same broad precision issue applies to the posted secondary comparisons: each reported confidence interval includes zero. They therefore do not sharply distinguish a modest difference in either direction from no average difference. These intervals are unadjusted, and the record does not report whether their construction accounts for multiple outcomes. The confidence intervals should be read as uncertainty around each comparison, not as assurance about unreported outcomes or the reliability of every possible 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 display gives planning illustrations, not the trial’s observed power. The figures are participants per arm for both the 80% power and 90% power columns; they are not total events. The assumptions listed for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario uses the stated fraction of the observed effect as the effect to detect. The trial’s own original design assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation would require for an 80% chance of detecting the specified effect under the listed assumptions, if that effect and assumptions were correct. The 90% column gives the corresponding planning number for a higher chance of detection. Power is a property of a design under assumed conditions, not a measure of whether a particular result is true.

Rows for 100% of the observed effect use the full observed estimate as the target for planning; rows for 75% and 50% use smaller target effects. The smaller the effect to be detected, the more participants are needed to distinguish it from random variation under the same planning framework. The figures can become very large when the observed effect is small, as in these comparisons.

These calculations are illustrative rather than a reconstruction of the completed trial’s design. They rely on the observed effects, which are themselves uncertain, and use assumptions that may not match the original protocol. They should not be used to infer that the trial was adequately or inadequately powered without the original design rationale and analysis plan. They are not observed power.

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 = -0.003; pooled standard deviation σ = 0.496.
2. n = 2 × (1.96 + 0.842)² × 0.496² / -0.003² = 529,830.6, rounded up to 529,831 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 <- -0.0027                      # observed difference in means
sigma <- 0.496036285164959                     # 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(5579600)
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 = 529,831 (80% power), 709,293 (90% power)
 75% of observed effect: n per arm = 941,922 (80% power), 1,260,966 (90% power)
 50% of observed effect: n per arm = 2,119,323 (80% power), 2,837,172 (90% power)
power.t.test: 529831.6 
Simulated power at n = 529,831 per arm: 0.806 (20000 replicates)

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

14. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and multiple secondary outcomes, with 6 outcome measures recorded and 4 secondary summaries posted. Examining several outcomes creates a multiplicity issue: even when each analysis is conducted correctly, the chance of at least one apparently unusual result can rise as the number of analyses increases. A prespecified hierarchy, family-wise adjustment, or other strategy can help define how results should be interpreted. The registry record does not report a multiplicity procedure, so the displayed secondary p-values should be treated as unadjusted rather than as family-wise controlled evidence.

The record lists no formal reported analyses. It does not describe whether the primary comparison was adjusted for baseline characteristics, which statistical model was used in the trial’s own analysis, or whether any subgroup or sensitivity analyses were prespecified. The unadjusted comparisons calculated from the posted summary statistics do not answer those questions. The original statistical analysis plan is not reported in the registry record.

Missing-data handling is also not reported. The registry does not state how participants without complete claims coverage, incomplete follow-up, or unavailable outcome data were treated. It does not identify an analysis population, such as an intention-to-treat population, or report whether participants were analyzed according to assigned group. Without this information, readers cannot determine whether the summaries include all randomized participants or only those with outcome data available.

Random assignment is a design feature intended to support fair group comparisons, but interpreting the results also depends on maintaining follow-up and analyzing participants in a way that respects assignment. The registry record does not report allocation concealment, adherence to assigned intervention, contamination between groups, or the handling of departures from assignment. These omissions do not prove that such issues occurred; they limit assessment of how they were addressed.

The outcome definition uses claims data and an established algorithm, but the record does not report trial-specific validation, event adjudication, or missing-claims procedures. These are relevant because events not captured in available claims may not appear in the endpoint summaries. In combination, the lack of reported multiplicity methods, missing-data procedures, analysis populations, and formal analyses means that the posted numbers offer a useful descriptive comparison but not a complete statistical account of the trial.

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: medically-treated suicide attempt (5.3%), inpatient hospitalization (potentially suicide-related) (5.3%).

The registry reports serious adverse event counts among 22380 participants at risk in Caring Contact Letters and 22775 participants at risk in Caring Contact Control. Serious adverse events affected 1177 participants in the letters group and 1071 in the control group. Other adverse events affected 0 participants in each group, and deaths were recorded as 0 in each group.

The registry also lists Medically-treated Suicide Attempt and Inpatient hospitalization (potentially suicide-related) among the top serious adverse event terms, each at 5.3%. The record does not clarify in this summary whether these terms overlap with each other or how they relate to the all-serious-event counts. These are counts and terms as recorded; they do not by themselves establish whether an event was caused by the intervention or differed statistically between arms.

The at-risk group sizes for safety are not identical to the arm sizes reported for the primary outcome. The registry record does not explain the difference or provide detailed event definitions, exposure time, adjudication procedures, or a comparative safety analysis. Therefore, the figures describe the posted safety reporting but do not support a causal conclusion about adverse events.

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

The posted results describe a randomized comparison of Caring Contact letters with care as usual among managed care subscribers released from jail. For the primary endpoint, the group means were close, and the unadjusted mean difference had a confidence interval that included zero. The record therefore does not precisely determine the direction or size of the average difference in recorded suicide-attempt counts.

This is not the same as demonstrating that the groups were equivalent, or that there was no possible difference. Such claims require a design and analysis aimed at testing equivalence or non-inferiority, with prespecified margins; the registry record does not report those features. Likewise, a non-small p-value is not proof that an intervention has no effect. The posted estimates and their uncertainty are the appropriate basis for describing what these results do and do not establish.

The findings are specific to the enrolled population, the intervention as described, the care-as-usual comparator, the follow-up periods, and outcomes identified from the data sources stated in the registry. They do not show whether individual participants experienced benefit or harm, explain mechanisms, or establish what would happen in other settings or populations. The primary measure is a group mean count, and the summaries do not show the individual distribution of attempts.

Secondary summaries add some information about service-use and injury measures, but their unadjusted nature and incomplete reporting limit the strength of conclusions drawn from them. One listed secondary outcome lacks a numerical summary among the posted results. The record also omits details about analysis populations, missing data, formal analysis methods, and multiplicity handling. Those gaps matter when interpreting both the apparent direction of estimates and how broadly they can be generalized.

In short, the posted data show group-level estimates with substantial uncertainty and do not settle whether the intervention altered the primary outcome. They should not be interpreted as medical advice or as a definitive judgment about treatment effects.

17. Why This Trial Matters Statistically

This trial raises statistical questions that are common in pragmatic research using administrative data: how to compare randomized groups on a count outcome, how to communicate uncertainty when estimates are close to zero, and how to interpret several related outcomes without overstating isolated findings.

The primary endpoint is a count summarized as a mean and standard deviation. Such summaries are easy to communicate, but count data can have many zero values and a long upper tail. A mean difference can therefore conceal very different distributions among participants. The posted standard deviations, together with the small means, make it especially important to distinguish the average from the experience of any particular participant. The registry record does not report enough model detail to assess whether the trial’s own analysis accommodated the outcome distribution.

The confidence interval adds essential information beyond a single estimate or p-value. Its inclusion of zero means the posted summary comparison does not precisely resolve whether the average difference is positive or negative. The multiple secondary outcomes add context, but their unadjusted comparisons and incomplete reporting make them less conclusive than a prespecified, fully reported analysis would be.

Finally, the planning illustrations demonstrate how sample-size requirements depend on the effect chosen for detection. Because smaller target effects require more participants under the listed assumptions, apparently large enrollment alone does not guarantee precise detection of a very small mean difference. These illustrations are not the trial’s original justification; that design rationale is not reported. The case is a reminder to separate observed estimates, uncertainty, planning assumptions, and the protocol’s intended analysis when reading clinical-trial results.

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 trial-specific limitations affect interpretation. The population is described broadly as managed care subscribers released from jail, but eligibility criteria and participant characteristics are not reported in the registry record. This makes it difficult to assess how representative the enrolled group may be of other people released from jail or other managed care populations.

The registry describes letters mailed over a 6-month period, but does not report the content, number, delivery success, or participant engagement with the letters. The control is described as care as usual without further detail. Masking is listed as none, and the record does not explain who was aware of group assignment or whether outcome ascertainment could have differed by group. The primary endpoint relies on claims and an algorithm; trial-specific validation and the handling of unavailable claims are not described.

The posted primary summaries are group means and standard deviations. They do not display the distribution of counts or the number of participants with any recorded attempt. Formal reported analyses are not listed, and information about adjustment, analysis populations, missing data, adherence, and multiplicity is not reported. These gaps prevent full evaluation of whether the posted comparisons match the planned analysis.

Secondary reporting is incomplete: 4 secondary outcomes have numerical summaries, while the registry lists 6 outcome measures. The outcomes also have different timeframes, including 6-months and 18 months, so they do not all represent the same period of observation. Safety group denominators differ from the primary outcome group sizes, without an explanation in the record. These limitations do not establish bias or invalidate the findings, but they leave important aspects of study conduct and analysis uncertain.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Suicide Prevention for Criminal Legal Involved Managed Care Subscribers — NCT05579600.
• Registry record last updated 2026-10-21; status last verified 2025-10.
• Results first posted 2025-11-12.
• 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

NCT05579600 is a completed, randomized, parallel-group trial sponsored by Butler Hospital. It enrolled 45155 participants and compared Caring Contact letters mailed after jail release with care as usual. Masking was listed as none. The registry lists 2022-09-01 as the start date and 2024-09-03 as the primary completion date.

The primary endpoint was the mean number of suicide attempts during 6-months, identified from claims using the Mental Health Research Network algorithm. Posted arm means were 0.020 for Caring Contact Letters and 0.023 for Caring Contact Control. The calculated unadjusted mean difference was -0.003, with a 95% confidence interval from -0.012 to 0.006 and an unadjusted p-value of 0.562.

Four secondary outcomes have numerical summaries. No formal reported analyses are listed. The trial’s original sample-size assumptions, detailed analysis plan, multiplicity procedures, analysis populations, and missing-data methods are not reported in the registry record.