Improving Accessibility and Personalization of CR for Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Improving Accessibility and Personalization of CR for Schizophrenia (NCT03576976), a randomized na study. The primary outcome is Treatment Satisfaction.

2. Clinical Question

The study addressed two practical adaptations of cognitive remediation: changing how sessions are delivered to increase accessibility, and personalizing treatment in relation to individual neurocognitive needs. The posted results concern satisfaction and neurocognition; they do not establish how participants or clinicians experienced the delivery formats more broadly. Qualitative stakeholder interviews are described in the study summary, but their findings are not reported in the registry record.

3. Trial Design

This was a randomized, single-masked, parallel-group study listed as completed. The registry lists the phase as not applicable. The recorded enrollment was 67. The study start date was 2018-07-16, and primary completion was recorded as 2020-12-31. The registry names the New York State Psychiatric Institute as sponsor.

The two groups differed in how cognitive remediation was delivered. Clinic-based cognitive remediation involved twice weekly group-based, clinician-led sessions. Hybrid cognitive remediation involved one weekly group-based, clinician-led session and independent cognitive practice. The study summary frames the comparison around accessibility: participants had reported that twice-weekly attendance could limit access. The registry also describes a separate adaptation aimed at personalizing cognitive remediation according to individual neurocognitive needs.

The primary outcome was treatment satisfaction through study completion, at 15 weeks. Change from baseline in neurocognition at baseline and 15 weeks was listed as a secondary outcome. The satisfaction summaries include 18 participants in the clinic-based group and 17 in the hybrid group. The neurocognition summaries include 17 and 15, respectively. The registry does not explain how these outcome-specific counts relate to the recorded enrollment or describe the analysis populations in further detail.

The registry record does not report who was masked, the method used to generate or conceal random assignments, the recruitment dates, or the trial's original sample-size assumptions. It also does not report whether the study used interim analyses or the detailed statistical analysis plan. The planned stakeholder interviews are described in the study summary, but their results 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: Enrolled participants were screened and assessed on outcome measures prior to randomization. Those who did not complete baseline assessments due to lost contact or no longer meeting inclusion criteria were not assigned to a treatment condition and were exited from the study.

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 Treatment Satisfaction, measured over Through study completion, 15 weeks.

The primary endpoint was treatment satisfaction, measured with a self-report Likert-type rating scale questionnaire reflecting specific components of the treatments used in the study. The outcome was the average of the scale items, with scores ranging from 1 to 6: 1 represented strong dissatisfaction and 6 represented strong satisfaction. The registry gives the timeframe as through study completion, at 15 weeks.

Satisfaction is a patient-reported outcome: it records participants' ratings of their experience of the treatment rather than a direct measure of cognitive performance. It is relevant to this comparison because the study examined a delivery adaptation intended to improve accessibility. Satisfaction can help describe how participants evaluated the two formats, but it does not by itself measure attendance, feasibility, acceptability among all stakeholders, or change in cognition or daily functioning. Those are distinct questions and should not be treated as interchangeable.

For a continuous outcome such as an average scale score, group means describe the central responses, while standard deviations describe variability among the recorded scores. The posted results present those summaries and a mean difference between groups. The registry record does not report additional details about how questionnaire items were handled, how scores were calculated when responses were incomplete, or whether the scale's measurement properties were evaluated in this study.

7. Results Posted in the Registry

• Hybrid Cognitive Remediation: mean 5.32 versus 5.25 on control; difference 0.07 (95% CI -0.58 to 0.72), standardized effect d = 0.07; z-test p = 0.832.

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.

Change From Baseline in Neurocognition

• 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:
  • Hybrid Cognitive Remediation: mean 0.97 versus -1.49 on control; difference 2.46 (95% CI -1.27 to 6.19), standardized effect d = 0.47; z-test p = 0.197.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Hybrid Cognitive Remediation and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcome adds a different perspective from treatment satisfaction: it concerns change from baseline in neurocognition rather than participants' ratings of treatment. The posted comparison does not provide conclusive evidence of a between-group difference, and its uncertainty permits values on either side of no difference. This broad pattern does not contradict the primary outcome; the two endpoints measure different aspects of the study and need not move together.

The secondary result should be interpreted as supportive or exploratory context rather than as a replacement for the primary question. The registry does not report a detailed analysis plan, a prespecified hierarchy among outcomes, or a multiplicity adjustment. When multiple outcomes are examined, unadjusted results carry less weight because the chance of a seemingly notable finding can increase across tests. This concern is not resolved merely because a result is secondary or because its direction appears plausible.

The study summary also describes an intended personalization adaptation and stakeholder interviews about implementation. The posted secondary endpoint is neurocognition, but the registry record does not provide results for qualitative interviews or outcomes that directly establish whether personalization improved delivery. Consequently, the secondary outcome cannot stand in for all the study's broader implementation aims.

The registry lists one secondary outcome and no reported analyses. That absence of a detailed analysis record leaves uncertainty about how outcomes were selected, handled, and interpreted. The posted summaries are informative descriptions, but they do not support a comprehensive account of consistency across all planned outcomes or analyses.

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

• Change From Baseline in Neurocognition

9. Statistical Interpretation

10. Statistical Methods Explained

The posted outcomes are continuous measures summarized using means and standard deviations. A mean is the arithmetic average of recorded values within a group; it gives a concise description of the group's central score. A standard deviation describes how dispersed the scores are around that average. It is not the uncertainty in the mean itself, and it does not show the full distribution of participant scores. General statistical interpretation should therefore distinguish the spread of individual scores from the precision of a between-group estimate.

What does a mean difference represent?

The registry presents the primary comparison as a mean difference between hybrid and clinic-based cognitive remediation. Its sign depends on the order of subtraction: a positive value in this comparison means the hybrid group's mean is higher, while a negative value means the clinic-based group's mean is higher. For treatment satisfaction, the posted difference is 0.07. The estimate describes the observed separation in group averages; it does not show that the delivery format caused a difference, nor does it establish that the observed size is important to participants.

The secondary comparison is also expressed as a mean difference, for change from baseline in neurocognition. This kind of comparison summarizes the separation between the groups' average change scores. It does not show whether every participant changed in the same direction, and a group average may conceal variation among individuals. The registry record does not report the underlying participant-level scores or a detailed account of the change-score analysis.

How should the confidence interval be read?

A confidence interval expresses uncertainty around an estimated difference. In general, a wider interval indicates less precision, while a narrower interval indicates more precision, given the analysis and its assumptions. The intervals posted for both outcomes include zero, the value corresponding to no mean difference. They also extend to values on either side of zero. Thus, the posted results are compatible with more than one direction of group difference; they do not establish equivalence or rule out potentially meaningful differences.

A confidence interval is not a range containing a fixed proportion of individual participants' scores. Nor does it give the probability that the true effect lies within the displayed bounds. It is an interval estimate whose interpretation depends on the statistical procedure and assumptions. The registry record does not identify the procedure used to construct these intervals, so further method-specific claims cannot be made.

What does the p-value add?

The unadjusted p-value summarizes how unusual the observed comparison, or a more extreme one, would be under a statistical model in which the groups have no underlying mean difference. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of an effect. The primary p-value is 0.832, and the secondary p-value is 0.197. These values should be considered alongside the estimates and intervals rather than used as stand-alone conclusions.

The registry identifies the p-values as unadjusted. It does not report a multiplicity strategy or a detailed testing plan. When several outcomes or comparisons are examined, unadjusted p-values do not account for the possibility that chance findings can arise across the set of tests. This matters especially for secondary outcomes, which are generally less conclusive when they are not supported by a prespecified analysis plan and an account of multiplicity.

Why can over-dispersion matter?

General explanation: over-dispersion occurs when observations vary more than a statistical model assumes. If a model understates the variability, uncertainty estimates can be too small and tests can appear more decisive than warranted. The posted outcomes are continuous and include standard deviations, but the registry record does not report a model specification or any assessment of over-dispersion. The relevance of that issue cannot be determined for these analyses from the record.

The registry lists a mean and standard deviation for each group and provides mean differences, confidence intervals, and unadjusted p-values. It does not report the statistical test or model, whether analyses were adjusted for baseline or other characteristics, how missing outcome values were treated, or whether analyses followed an intention-to-treat principle. No sponsor-reported analysis is listed. These omissions limit the extent to which the posted summaries can be independently evaluated or reproduced.

Planning illustrations based on the observed effects are separate from the trial's observed findings. They describe sample sizes under stated assumptions for future planning, not the power achieved in this completed study. Their interpretation depends on the stated assumptions and on whether a future study's design and analysis match them. The registry record does not report the study's original sample-size justification, so the planning illustrations should not be read as evidence about the rationale for the completed study's enrollment.

11. Confidence Intervals and Precision

Precision concerns how narrowly the data locate the between-group difference. The treatment satisfaction estimate is 0.07, with a 95% confidence interval from -0.576 to 0.716. The interval includes zero and allows for differences favoring either group. In relation to the 1-to-6 satisfaction scale, the interval is broad enough that the observed estimate alone should not be treated as a precise description of the underlying group difference.

For change from baseline in neurocognition, the posted difference is 2.46 T score points, with a 95% confidence interval from -1.275 to 6.195. This interval also includes zero and permits effects in either direction. Its width reflects uncertainty in the comparison, not the range of individual participants' scores. The registry record does not report a prespecified threshold for a meaningful difference on either endpoint, making it difficult to judge which values within the intervals would matter in practice.

Confidence intervals are useful because they show both the direction and the range of estimates compatible with the data under the analysis used. They do not prove that there is no difference when they include zero. Nor do they demonstrate equivalence: an equivalence conclusion requires an appropriate design and a prespecified margin, neither of which is reported in the registry record. The intervals should be considered together with the group summaries, outcome-specific sample sizes, and incomplete information about the analysis methods.

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 illustrations are planning scenarios derived from the observed effect, not a report of the trial's original design assumptions. The record describes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario uses a stated fraction of the observed effect. The displayed figures are participants per arm, not a total across arms or a count of events. The time-to-event convention of total events does not apply to these continuous outcomes.

The 80% power column gives the planned number of participants per arm for a future study designed to have an 80% chance of detecting the specified effect under the planning assumptions, if that effect is present and the assumptions hold. The 90% power column gives the corresponding per-arm planning number for a 90% chance. These are design targets, not guarantees that a future study will detect an effect.

Rows that plan for 100%, 75%, or 50% of the observed effect illustrate how the assumed effect size changes the planning requirement. Smaller effects are harder to distinguish from random variation and therefore require more participants under the same assumptions. The observed effect is estimated with uncertainty, so using it as the basis for planning can yield very different projections depending on the fraction selected.

These values are not observed power. They do not describe the probability that the completed study detected an effect, and they are not evidence that the original trial was designed around the observed results. The original sample-size justification and its assumptions are not reported in the registry record.

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 = 0.070; pooled standard deviation σ = 0.975.
2. n = 2 × (1.96 + 0.842)² × 0.975² / 0.070² = 3,044.6, rounded up to 3,045 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 <- 0.0700000000000003                      # observed difference in means
sigma <- 0.974861295495182                     # 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(3576976)
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 = 3,045 (80% power), 4,076 (90% power)
 75% of observed effect: n per arm = 5,413 (80% power), 7,246 (90% power)
 50% of observed effect: n per arm = 12,179 (80% power), 16,304 (90% power)
power.t.test: 3045.5 
Simulated power at n = 3,045 per arm: 0.797 (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 record does not report a multiplicity strategy, such as a prespecified ordering of outcome tests or an adjustment to account for multiple comparisons. The posted p-values are identified as unadjusted. When researchers test multiple outcomes, unadjusted p-values do not control the overall chance of a false-positive result across the full set. This is one reason secondary findings generally warrant more caution, particularly when the analysis plan and the full set of planned tests are not described.

Missing-data methods are not reported in the registry record. It does not say how incomplete questionnaires or unavailable follow-up measurements were handled, whether any values were imputed, or what assumptions were used for missing observations. The outcome summaries include different numbers of participants across measures and groups. The record does not explain the reasons for those differences, so they should not be attributed to a particular cause.

The registry does not describe the analysis population or state whether the primary analysis followed intention-to-treat principles, included only participants with observed outcomes, or used another population definition. It also does not report whether the comparisons were adjusted for baseline values or other variables. Those details affect how an estimated difference should be interpreted and how it can be compared with other studies.

No reported analyses are listed in the registry record, and the statistical model or test used for the posted intervals and p-values is not specified. The registry does provide summary results, but it does not provide enough methodological detail to reconstruct the inferential analysis. The original sample-size justification, interim analysis plan, and detailed statistical analysis plan are not reported in the registry record.

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.

The registry lists zero serious adverse events in the clinic-based group among 27 participants at risk, and zero in the hybrid group among 28 participants at risk. It also lists zero other adverse events among 27 participants at risk in the clinic-based group and zero among 28 at risk in the hybrid group. Deaths are listed as zero in both groups.

These figures should be read as the adverse-event counts and at-risk denominators reported in the registry, not as proof that no adverse events of any kind occurred beyond what was recorded. The registry does not list serious adverse event terms, and it does not provide further descriptions of event ascertainment, severity, relatedness, or timing. The safety denominators differ from the outcome-specific participant counts; the record does not explain the difference. The figures are descriptive and do not, by themselves, establish a comparative safety profile.

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

The posted results describe satisfaction ratings and a neurocognition comparison for participants represented in the registry summaries. For treatment satisfaction, the observed group means are close, but the confidence interval includes zero and allows differences favoring either format. For neurocognition, the interval likewise includes zero. The estimates therefore do not establish that one delivery approach produced better results on these measures, and they do not establish that the approaches are equivalent.

The study addresses an implementation question with potential relevance to delivery in outpatient settings: whether cognitive remediation can be offered in a hybrid format rather than through twice weekly clinic-based sessions. Its design compares those two formats, but the posted numerical outcomes do not directly answer every question about accessibility, feasibility, acceptability, or personalization. The registry describes stakeholder interviews, but their results are not reported in the registry record. It also does not provide posted outcomes establishing effects on daily functioning or functional recovery.

A non-significant result should not be translated into proof of no effect. The confidence intervals show that the estimates remain uncertain, and the registry does not give a threshold for determining what size of difference would be important. Satisfaction scores reflect participants' ratings on the stated questionnaire; they are not a direct measure of cognition or of the practical burden of attending sessions.

The findings should also be understood in light of the information not reported in the registry record: detailed analysis methods, missing-data handling, analysis populations, and multiplicity procedures. These limitations do not erase the posted summaries, but they restrict how firmly the results can be interpreted or generalized. The registry record is best read as a report of observed outcomes with uncertainty, rather than a definitive answer about the effectiveness or broader implementation value of either format.

17. Why This Trial Matters Statistically

This trial is statistically informative because it places an implementation adaptation beside a more established clinic-based delivery format and reports outcomes that address different questions. Treatment satisfaction reflects participants' ratings, while the secondary neurocognition measure concerns change in performance. Keeping these constructs distinct helps prevent a favorable or unfavorable rating from being mistaken for evidence of cognitive change, or vice versa.

The comparison also illustrates why estimates and confidence intervals matter more than a binary reading of a p-value. The posted estimates are accompanied by intervals that include zero and allow for differences in both directions. That pattern signals uncertainty, not equivalence. The record does not provide a threshold for meaningful differences, so interpretation requires attention to the scale and the practical context rather than statistical significance alone.

Planning illustrations based on the observed effect highlight how future-study requirements depend on the effect a study aims to detect. Smaller target effects require more participants under the stated assumptions. Because these illustrations are not the original trial's design justification, they should not be mistaken for evidence that the completed study was planned around a particular effect or achieved a particular level of power.

Finally, the record highlights the importance of reporting analysis populations, missing-data methods, multiplicity plans, and model details. Without those elements, readers have less ability to assess robustness or reproduce the analysis. Transparent reporting of these details is especially useful in adaptation research, where implementation outcomes and clinical outcomes may require separate interpretation.

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.

The outcome-specific participant counts are limited: the satisfaction summaries include 18 participants in the clinic-based group and 17 in the hybrid group, while the neurocognition summaries include 17 and 15. These counts and the recorded enrollment of 67 are not reconciled in the registry record. Reasons for differences among enrollment, safety denominators, and outcome summaries are not reported, so the amount or pattern of missing outcome data cannot be determined.

The registry record does not report participant characteristics beyond the conditions listed, or enough information to assess how representative the participants were of other outpatient settings. It describes a New York State implementation context, which is important to the study question, but the extent to which the findings would generalize to different services or delivery systems is not established by the posted summaries.

Important methods are also not detailed: the record does not identify the person or people masked, explain randomization procedures, specify the analysis model, or describe missing-data handling and analysis populations. It does not state whether the outcome comparisons were adjusted for baseline or other factors. These gaps limit evaluation of possible bias and reproducibility.

The primary satisfaction outcome is a self-report average on a Likert-type scale, and the registry does not report further scale-validation details or a threshold for a meaningful difference. Qualitative interview findings and direct results on broader implementation goals are not reported. These limitations mean the posted outcome summaries give a narrow view of the trial's aims and should not be treated as a complete evaluation of accessibility, acceptability, personalization, or functional recovery.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Improving Accessibility and Personalization of CR for Schizophrenia — NCT03576976.
• Registry record last updated 2022-07-13; status last verified 2022-06.
• Results first posted 2022-02-02.
• 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

NCT03576976, titled “Improving Accessibility and Personalization of CR for Schizophrenia,” is recorded as a completed, randomized, single-masked, parallel-group study sponsored by the New York State Psychiatric Institute. The listed conditions are schizophrenia and schizoaffective disorder, and enrollment is recorded as 67.

The study compared clinic-based cognitive remediation, consisting of twice weekly group-based, clinician-led sessions, with hybrid cognitive remediation, consisting of one weekly group-based, clinician-led session plus independent cognitive practice. The primary outcome was treatment satisfaction through study completion, at 15 weeks. Change from baseline in neurocognition was listed as a secondary outcome.

The posted satisfaction comparison showed a mean difference of 0.07, with a 95% confidence interval from -0.576 to 0.716. The posted secondary comparison showed a mean difference of 2.46, with a 95% confidence interval from -1.275 to 6.195. Both intervals include zero. Detailed analysis methods, missing-data procedures, and multiplicity plans are not reported in the registry record.