Adapting Critical Time Intervention to Support Inpatient Medical Care Transition: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Adapting Critical Time Intervention to Support Inpatient Medical Care Transition (NCT03637296), a non-randomized na study. The primary outcome is 7 Days Hospital Inpatient All-cause Readmission Rates.

2. Clinical Question

3. Trial Design

The registry record describes a parallel-group pilot study comparing Critical Time Intervention with treatment as usual among adults with schizophrenia admitted for inpatient care of medical complications. The listed allocation is nonrandomized, and masking is listed as none. The study phase is listed as NA. These design details matter: without random allocation, the groups may differ at the outset, and without masking, awareness of treatment assignment may affect care or outcome ascertainment. The posted comparison is not a like-for-like control comparison.

The brief summary describes a plan to randomize 80 eligible inpatients, while also giving planned arm counts of N=20 for TAU and N=40 for CTI and TAU. Those planning statements do not agree with one another. Separately, the registry lists enrollment as 15 and allocation as nonrandomized. The record does not explain the difference between its planned description and the registered design or enrollment. The safety listings give at-risk counts of 10 for CTI and 5 for TAU; those figures are safety denominators and should not be taken as additional evidence of the allocation process.

CTI is described as intensive care management during the inpatient stay and after discharge, with the brief summary specifying a 3-month follow-up period. TAU is described as routine care management during and following discharge. The primary outcome is all-cause hospital readmission within 7 days; the brief summary also identifies readmissions at 30 days and follow-up with outpatient medical and mental health services. The record lists the study status as terminated, a start date of 2019-05-01, and a primary completion date of 2020-12-31. The reason for termination is 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 7 Days Hospital Inpatient All-cause Readmission Rates, measured over Within 7 days following discharge.

The primary endpoint is described as the number of participants readmitted to hospital for any cause within 7 days following discharge. It concerns whether a participant experienced at least one qualifying readmission during the stated period, rather than the number of admissions per participant. The registry lists the unit as participants and the measure as a count of participants. Although the description calls it a readmission rate, the record does not report a rate estimate or a numerical result.

Readmission is used as a short-term marker of what happens after inpatient discharge. In general, a measure focused on the period following discharge can capture early returns to hospital, while an all-cause definition does not restrict the event to a particular diagnosis. Such an endpoint may be useful for describing care transitions, but it does not by itself explain why a person was readmitted or whether a readmission was preventable. Nor does a difference between nonrandomized groups necessarily represent an intervention effect.

The brief summary also refers to readmissions at 30 days and to outpatient medical and mental health follow-up at 7 and 30 days. These are additional outcomes, not substitutes for the stated primary endpoint. The primary outcome result is not reported in the registry record, so the endpoint's observed frequency and any comparison between CTI and TAU cannot be described from the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a count endpoint mean? The primary measure is a count of participants with an event: participants readmitted within 7 days following discharge. In general, a participant-level binary outcome records whether the event occurred, rather than counting repeated instances of the event. A group count can describe how many participants met the endpoint, but interpreting it comparatively also requires knowing the relevant group denominator. Here, the registry record states the endpoint definition but does not report its outcome counts or denominators. The safety denominators should not be substituted for primary-outcome denominators.

What does a comparative estimate tell a reader? In general, an effect estimate expresses the observed contrast between groups for an outcome. Its meaning depends on the outcome definition, the study design, and whether the estimate is adjusted for relevant differences. A confidence interval describes the range of estimates compatible with the data and the method's assumptions; it conveys precision as well as direction and magnitude. A narrow interval indicates greater precision than a wide interval, but neither interval width nor a point estimate establishes causation by itself. No primary effect estimate or confidence interval is reported in the registry record.

Why does nonrandom allocation matter? In general, random allocation helps make groups comparable, on average, with respect to both measured and unmeasured characteristics. Nonrandom allocation does not ensure that comparability. For example, reasons for receiving intensive care management could also be related to the likelihood of follow-up or readmission. These are general design concerns, not claims about what happened to participants in this study. The record identifies the allocation as nonrandomized and says the arms are not a like-for-like control comparison. It reports no statistical analyses, so it is not possible to assess whether any adjustment was attempted.

How should planning calculations be read? The planning assumptions describe illustrations based on fractions of an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. In general, a sample-size calculation estimates how much information a planned study might need under specified assumptions. It does not show that the study had a particular probability of detecting an effect once conducted. Sample sizes calculated from an effect estimate are planning illustrations, not observed power. Their relevance depends on the accuracy of the effect assumption and the appropriateness of the calculation method. No planning values or observed primary effect are reported in the registry information used for this interpretation.

The calculation description distinguishes between participant counts per arm for outcomes other than time-to-event outcomes, and total event counts for time-to-event calculations using the Schoenfeld method. The registered primary outcome is a participant count rather than a stated time-to-event measure. In general, event-driven sample-size methods and participant-based binary-outcome methods answer different planning questions; they should not be conflated. The registry record does not provide a statistical analysis plan, the inputs used for a specific calculation, or an explanation of whether a particular planning method was intended for the primary endpoint.

What can the safety counts establish? In general, adverse-event counts describe events recorded under the study's definitions and observation procedures. A zero count means no event was recorded in the stated category and listed group; it does not prove that the underlying risk is zero. The size of the group observed, the duration of observation, definitions, and completeness of reporting all affect interpretation. The record lists serious events, other events, and deaths, but does not give further definitions or a detailed safety-analysis method.

For a binary outcome, over-dispersion is not usually described in the same way as for a count model, but general model choice still matters: an analysis must match how the outcome is recorded and account appropriately for the design. If repeated events or follow-up times were analyzed, methods would need to reflect those features. The registry record does not report such analyses, so no model, covariate adjustment, or assumptions about repeated events can be attributed to this study.

Finally, statistical significance and clinical or practical importance are different considerations. In general, a small p-value does not quantify the size or importance of an effect, and a non-small p-value does not establish equivalence or absence of an effect. Precision, design limitations, and the actual endpoint estimates matter. Since the primary result and analysis are not reported, neither significance nor effect size can be interpreted for this study.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Critical Time Intervention", "Treatment as Usual")
n      <- c(10, 5)
events <- c(3, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Critical Time Intervention  10  3  30.0  6.7  65.2  10.8  60.3  10.9  61.0
ARM  Treatment as Usual  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 6 outcome measures and identifies several follow-up outcomes in outpatient medical and mental health care, including measures at 7 and 30 days. The brief summary additionally describes CTI assessments at 6 and 12 week, covering satisfaction with CTI services, psychiatric symptoms, community function, and involvement in medical care decisions. The registry record does not report how these outcomes were prioritized in the analysis, whether multiple comparisons were planned, or whether any multiplicity adjustment was used. Without that information, readers cannot determine how the potential for chance findings across several outcomes was handled.

The record lists 0 analyses. It does not report an analysis plan, statistical model, covariate adjustment, or a plan for handling missing outcomes. Missing-data amounts, reasons for missingness, and methods for addressing missing values are not reported in the registry record. The record also does not specify an analysis population, such as whether outcomes were analyzed according to original assignment, treatment received, or another rule. Those omissions prevent assessment of how participants with incomplete follow-up were handled.

These reporting gaps are especially relevant because allocation is nonrandomized and enrollment is listed as 15. In general, missing outcome data and selective emphasis among multiple outcomes can affect interpretation, while nonrandom allocation limits causal conclusions. The posted record does not support a judgment about the direction or size of any such effects. The absence of reported analyses should not be read as evidence that an analysis was or was not performed outside the registry record.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry lists serious adverse events for the Critical Time Intervention group as 0 among 10 participants at risk, and for Treatment as Usual as 0 among 5 participants at risk. It lists other adverse events as 0 among 10 at risk in the CTI group and 0 among 5 at risk in the TAU group. Deaths are listed as 0 in each group. These figures describe the safety entries recorded in the registry; they are not primary readmission results.

In general, zero recorded events means that none were reported in that category for the stated group during the study's safety reporting, not that the true possibility of an event is zero. Interpretation depends on event definitions, observation time, completeness of collection, and the number of participants observed. The registry record does not provide further detail about these procedures, the duration of safety follow-up, or individual event types. It also lists no serious adverse event terms.

The at-risk counts are not described as the primary-outcome denominators, so they should not be used to infer readmission counts or rates. The record does not report a formal comparative safety analysis. With no reported event counts beyond the zeros, and with the study identified as nonrandomized, these entries should be read as limited descriptive safety information rather than evidence of a comparative safety effect.

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

The registry record describes a pilot study intended to assess feasibility and preliminary effectiveness of adapting Critical Time Intervention for adults with schizophrenia transitioning from inpatient medical care. The intervention is described as intensive care management during the hospital stay and following discharge, while treatment as usual is described as routine care management. The primary question concerns all-cause hospital readmission within 7 days following discharge.

What the record does not provide is the primary endpoint result. It gives no readmission count by group, comparative estimate, confidence interval, or analysis result. Therefore, the posted information cannot establish how often readmission occurred, whether the groups differed, how large any observed difference was, or how precisely it was estimated. The listed zero adverse-event counts do not answer those questions.

The design description also limits what a group comparison could mean. Allocation is nonrandomized, and the record cautions that the arms are not like-for-like. In general, differences in outcomes could reflect differences between participants or care circumstances rather than CTI itself. An observed contrast, if reported, would need to be interpreted in light of that limitation and of the analysis methods. No analyses are listed in the registry record.

The brief summary's description of randomized plans and the recorded nonrandomized allocation do not match; its planned total and arm targets also differ as stated. Enrollment is listed as 15, and status is listed as terminated. The record does not explain the discrepancies or the reason for termination. These details constrain what can be inferred about feasibility, recruitment, and the study's conduct. They do not reveal whether CTI is beneficial or harmful, and they do not support medical advice. The appropriate conclusion is limited to what is documented: the study was registered as a nonrandomized pilot, and its primary readmission result is not reported in the registry record.

14. Why This Trial Matters Statistically

This study raises a statistical issue common to care-transition research: how to evaluate an intervention when treatment groups are not randomly assigned. The question is clinically recognizable, and the endpoint is defined around readmission after discharge. But an outcome comparison is not automatically an intervention effect. With nonrandom allocation, baseline differences and care-selection processes can be associated with both treatment and outcome. That makes design information central to interpretation rather than a technical footnote.

The record also illustrates why registration details need to be read together. The brief summary describes randomization and a planned enrollment of 80, yet gives arm targets of N=20 and N=40; the registered allocation is nonrandomized and enrollment is listed as 15. The registry does not reconcile these statements. A reader therefore cannot assume that the trial followed the randomization plan in the summary, or treat planned figures as actual enrollment.

Outcome reporting is equally important. A clearly named endpoint does not convey a result unless counts, denominators, and an analysis are reported. Here, the primary readmission result is absent, and the registry lists 0 analyses. As a result, there is no reported effect estimate or precision to interpret. The listed safety counts are descriptive and do not fill that gap.

The study's pilot framing is relevant: feasibility and preliminary effectiveness are distinct aims, and a pilot can inform future research without providing a definitive causal estimate. However, the registry record does not report feasibility outcomes or explain the terminated status. Statistical planning illustrations, when available, should be understood as conditional planning calculations, not evidence of observed power or a result. The record thus makes the limitations of incomplete outcome reporting and nonrandomized comparisons especially visible.

15. 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 registry identifies allocation as nonrandomized and explicitly warns that the arms are not a like-for-like control comparison. This limits causal interpretation of any between-group contrast. The record does not report baseline characteristics, reasons for treatment assignment, or adjustment methods, so comparability cannot be assessed from the posted information.

The brief summary describes randomizing 80 eligible inpatients but specifies arm targets of N=20 and N=40. In contrast, the registered allocation is nonrandomized and enrollment is listed as 15. The record does not reconcile these descriptions. The status is terminated, but the reason is not reported. Therefore, the circumstances of recruitment, implementation, and early termination cannot be evaluated from the registry record.

The primary endpoint result is not reported, and the record lists 0 analyses. No group-specific readmission counts, denominators, effect estimates, confidence intervals, or statistical tests are available in the registry information used. The record also does not report missing-data handling, analysis populations, a statistical analysis plan, or any adjustment for multiple outcomes. These omissions prevent assessment of outcome precision and analysis choices.

The registry describes additional medical and mental health follow-up outcomes and further CTI assessments, but does not report their results. Safety entries list zero serious adverse events, zero other adverse events, and zero deaths in the stated groups; definitions and observation details are not reported. Taken together, these limitations mean the record supports description of the intended question and registered design, but not a conclusion about comparative outcomes or intervention effectiveness.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Adapting Critical Time Intervention to Support Inpatient Medical Care Transition — NCT03637296.
• Registry record last updated 2021-09-28; status last verified 2021-09.
• Results first posted 2021-09-28.
• 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.

19. Record Summary

The registry record concerns NCT03637296, titled “Adapting Critical Time Intervention to Support Inpatient Medical Care Transition,” sponsored by New York State Psychiatric Institute. It describes a parallel pilot study of CTI and treatment as usual for adults with schizophrenia admitted for inpatient treatment of medical complications. The listed allocation is nonrandomized, masking is none, phase is NA, and status is terminated. Enrollment is listed as 15.

The stated primary endpoint is the count of participants readmitted to hospital for any cause within 7 days following discharge. The brief summary also discusses 30-day readmissions and outpatient medical and mental health follow-up. The primary readmission result and analyses are not reported in the registry record. The listed safety entries show zero serious adverse events, zero other adverse events, and zero deaths in each group's stated at-risk population.

The brief summary's randomized plan is inconsistent with the registered nonrandomized allocation and enrollment details. The registry does not explain these differences or the terminated status. The record therefore describes the research question and design features, but does not provide results that establish a comparative effect.