Implementing Health Plan-Level Care Management for Solo & Small Practices: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Implementing Health Plan-Level Care Management for Solo & Small Practices (NCT02041962), a randomized na study. The primary outcome is Health-related Quality of Life, as Measured by the Mental Health Component Score.

2. Clinical Question

3. Trial Design

The registry describes a randomized, single-masked, parallel-group study sponsored by the University of Michigan. It lists the phase as NA and the status as completed. The registry dates the study start to 2014-07 and primary completion to 2017-11. Planned enrollment is 280. The conditions listed are bipolar disorder and depression.

Participants were assigned to either Chronic Care Model for Mood Disorders or Educational Control. The intervention arm is described as Life Goals Collaborative Care. The control arm was to receive usual care from clinic providers and a self-guided workbook by mail. The study’s stated purpose was to determine whether a version of the chronic care model for people with mood disorders in small or solo practices could improve patient health.

The primary outcome is a continuous score on the SF-12 Mental Health Component, assessed at 12-months. The reported outcome summaries list 42 participants in Educational Control and 31 in Chronic Care Model for Mood Disorders. These are the group counts associated with the posted primary outcome, not an explanation of the flow of all enrolled participants. Reasons for the difference between planned enrollment and the reported outcome counts are not reported in the registry record.

The registry does not identify which party was masked, how random assignment was generated or concealed, or how clinics and participants were recruited. More detailed procedures for intervention delivery, follow-up, adherence, and outcome collection are not reported in the registry record. The protocol’s original sample-size rationale and detailed statistical analysis plan are also not reported in the registry record, so the planning illustration should not be read as a description of the trial’s original design.

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 total of 280 participants were consented and enrolled to participate. Of those 280 participants, only 238 completed the baseline survey and were randomized.

Overall Study

Reasons for not completing this period:

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 Health-related Quality of Life, as Measured by the Mental Health Component Score, measured over 12-months.

The primary endpoint is health-related quality of life measured by the Mental Health Component Score of the 12-item Short Form Survey, or SF-12, at 12-months. The registry identifies the outcome as a mean score on a scale and gives the scale range as 0-100. Higher values represent better outcomes. This endpoint focuses on the mental-health component of self-reported health-related quality of life rather than on a diagnosis count or a direct measure of service use.

A continuous score can retain differences across the scale instead of reducing each participant’s result to a yes-or-no category. Comparing group means therefore describes the average score in each arm and the mean difference between them. It does not show how every participant’s score changed or how many participants experienced a particular level of benefit. The posted summaries do not report baseline scores, individual changes, or the distribution of scores beyond the group means and standard deviations.

The registry states that the endpoint was assessed at 12-months, but does not report why that follow-up time was selected, whether the analysis used a particular adjustment, or whether a threshold for clinical importance was specified. Such details are not reported in the registry record. The score’s direction is clear from the description: higher values indicate better outcomes. Clinical meaning still depends on context, including the population, measurement properties, and a prespecified interpretation threshold, none of which is quantified in the posted result.

7. Results Posted in the Registry

• Chronic Care Model for Mood Disorders: mean 43.20 versus 38.20 on control; difference 5.00 (95% CI 0.36 to 9.64), standardized effect d = 0.50; z-test p = 0.035.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean difference describe?

A mean is the arithmetic average of a continuous outcome within a group. The registry reports a mean of 38.2 for Educational Control and 43.2 for Chronic Care Model for Mood Disorders. The comparison reports a mean difference of 5.0, calculated as the intervention mean relative to the control mean. A positive difference therefore indicates a higher reported average in the intervention arm on this scale, whose higher values represent better outcomes. The difference is expressed in score points, not as a percentage or as the proportion of participants who improved.

The mean difference condenses the groups’ results into a single contrast. It does not show whether the difference was similar across participants, whether a small subgroup drove the average, or how much scores changed from baseline. The registry’s arm summaries include standard deviations of 9.6 and 10.3. A standard deviation describes the spread of participant scores around a group mean. It is distinct from a standard error, which describes uncertainty in an estimated mean. The reported standard errors are 1.481 for control and 1.85 for the intervention.

How does the interval add information?

The comparison gives an interval from 0.355 to 9.645 around the estimated mean difference. A confidence interval is a method for expressing statistical uncertainty: under its assumptions, a procedure applied repeatedly would produce intervals that cover the underlying parameter at the stated rate. The confidence level is not reported in the registry record, so it should not be inferred from the interval alone. The interval’s width reflects uncertainty in the estimate; it leaves room for a range of possible mean differences, rather than presenting 5.0 as an exact population value.

The interval’s lower end is above zero. In the model underlying this unadjusted comparison, zero represents no mean difference between arms. The positive interval is consistent with a higher intervention-arm average under that model. However, the registry does not report the assumptions, confidence level, or any prespecified threshold for interpreting the size of the difference. The interval is not a range containing every participant’s result and is not a guarantee that the true effect falls inside it.

What does the p-value say?

The unadjusted p-value is 0.035. A p-value evaluates the compatibility of the observed result with a null model that posits no mean difference, conditional on the model and its assumptions. A smaller p-value indicates that results at least as discordant with that null model would be less common if the model were correct. It does not measure effect size, clinical importance, or the probability that the intervention is effective. Nor does it establish that the analysis was the one planned before the results were observed; the trial’s original statistical plan is not reported in the registry record.

The planning illustration lists a two-sided significance level of 0.05, but that is an assumption for the planning calculations. It is not evidence that the trial’s own primary analysis used this threshold. The registry does not report the trial’s testing procedure, any adjustment for multiple outcomes or analyses, or whether the posted p-value was prespecified. The p-value should therefore be read as an unadjusted result attached to the available summary-statistic comparison.

What is the standardized effect?

The comparison also reports a pooled standard deviation of 9.902 and a standardized effect, d, of 0.505. A standardized mean difference expresses the contrast in units of a pooled standard deviation, rather than in SF-12 score points. This can help describe the size of a contrast on a scale that does not depend on the score’s original units. It does not change the clinical interpretation automatically: the meaning of a standardized difference depends on the outcome, population, and context. The registry does not report a threshold for what would be clinically important on either the original or standardized scale.

The standardized effect and the raw mean difference are transformations of the same group-level information, not independent confirmations. The raw estimate remains directly interpretable as a score-point contrast. The standardized estimate can be useful for planning illustrations, but it should not be treated as a separate outcome or as a substitute for a clinically grounded interpretation.

What can be concluded from the posted analysis?

The registry lists zero analysis records. Thus, the comparison figures are not accompanied by a posted analysis description that would establish model adjustments, handling of baseline values, or an analysis population. The registry does not report whether the comparison follows an intention-to-treat principle, whether participants were analyzed according to assignment, or whether only participants with observed follow-up scores were included. These absences do not prove that no such procedures were used; they mean the relevant details are not reported in the registry record.

The posted outcome counts, means, standard deviations, interval, and p-value can be described as an unadjusted summary-statistic comparison. The interval and p-value quantify uncertainty conditional on their statistical assumptions, not uncertainty about unreported design or missing-data decisions. Those design and reporting limitations remain important when judging how representative the estimate is of the randomized groups and how broadly it may apply.

10. Confidence Intervals and Precision

The estimated mean difference is 5.0 score points, with an interval from 0.355 to 9.645. The interval communicates that the point estimate is uncertain: values across this range are compatible with the reported procedure and data, subject to the procedure’s assumptions. The estimate is not precise enough to identify a single value as the population difference. The registry does not state the confidence level, so the interval should not be assigned a confidence percentage that is absent from the record.

The interval spans values that are positive but vary in magnitude. Its lower end is close to zero relative to the point estimate, while its upper end is substantially higher than the point estimate. This breadth makes the estimated size less certain than the direction indicated by the interval. The posted p-value of 0.035 is related to the same comparison and should not be read as an independent measure of precision. A p-value alone does not show the range of effect sizes compatible with the data.

The outcome summaries include 31 intervention participants and 42 control participants. The group standard deviations are 10.3 and 9.6. The registry does not explain the smaller outcome counts relative to planned enrollment, report the amount or pattern of missing outcome data, or describe adjustments that might affect precision. The available interval reflects the calculation attached to the posted summary, not the additional uncertainty that might arise from unreported design choices, missingness, or differences between the observed participants and all randomized participants.

Precision is also distinct from clinical importance. The interval concerns the mean score difference; the registry does not report a threshold by which to judge whether any value in the interval represents an important change for patients. The data therefore permit a description of the estimated average contrast and its statistical uncertainty, but not a definitive judgment about practical importance.

11. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total participants and not total events. They are illustrations calculated from the observed effect, rather than a report of the trial’s original design calculation. The assumptions listed for this illustration are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect equal to the stated fraction of the observed effect. The trial’s original sample-size assumptions and justification are not reported in the registry record.

For planning for 100% of the observed effect, the illustration gives 62 participants per arm for 80% power and 83 participants per arm for 90% power. For planning for 75% of the observed effect, it gives 110 participants per arm for 80% power and 147 participants per arm for 90% power. For planning for 50% of the observed effect, it gives 247 participants per arm for 80% power and 330 participants per arm for 90% power.

In plain terms, the 80% power column represents the number of participants per arm in the illustration for a design with an 80% chance of detecting the specified effect under its assumptions. The 90% power column uses a higher target chance of detection, which requires more participants in each corresponding scenario. Smaller effects are harder to distinguish from random variation, so the illustration calls for more participants per arm as the assumed effect gets smaller.

These are planning illustrations, not observed power. They do not show the probability that the completed trial’s result is true, nor do they establish whether the original trial was adequately sized. Their values depend on the stated assumptions and the observed effect used as the reference. The registry does not report the protocol’s own target effect, allocation details beyond randomized assignment, or original sample-size calculation.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 5.000; pooled standard deviation σ = 9.902.
2. n = 2 × (1.96 + 0.842)² × 9.902² / 5.000² = 61.6, rounded up to 62 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 <- 5                      # observed difference in means
sigma <- 9.90181374764027                     # 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(2041962)
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 = 62 (80% power), 83 (90% power)
 75% of observed effect: n per arm = 110 (80% power), 147 (90% power)
 50% of observed effect: n per arm = 247 (80% power), 330 (90% power)
power.t.test: 62.5 
Simulated power at n = 62 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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures and zero posted analysis records, but it does not provide enough detail to determine how multiple outcomes or analyses were handled. A multiplicity plan can specify which outcome is primary, how testing across outcomes is controlled, and which comparisons are confirmatory. The registry identifies the SF-12 Mental Health Component Score as the primary endpoint, but a broader testing strategy, including any adjustment or hierarchy among outcomes, is not reported in the registry record.

The posted comparison is unadjusted. The registry does not report whether the analysis accounted for baseline outcome values, clinic-level clustering, prognostic factors, or other covariates. It also does not state whether there were interim analyses or any changes to the statistical plan. These details should not be inferred from the randomized, single-masked, parallel design description.

Missing-data methods and analysis populations are also not reported in the registry record. The primary-outcome summaries include 42 control participants and 31 intervention participants, while planned enrollment is 280. The record does not explain the participant flow from enrollment to these summaries, the number with missing 12-month scores, reasons for missingness, or whether any imputation or sensitivity analyses were used. It does not specify whether the result represents all randomized participants, participants with observed outcomes, or another analysis population.

Missingness can affect both the estimate and its uncertainty if participants without outcome data differ from those with observed scores. A comparison limited to observed outcomes may not preserve the balance created by random assignment if availability of a score is related to prognosis or treatment experience. The posted summaries alone cannot establish whether that concern applies in this study. Without reported methods, the reader cannot assess how robust the result is to alternative assumptions about missing outcomes.

The absence of these details is a reporting limitation, not proof that the trial investigators did not consider multiplicity, missing data, or analysis populations. Those procedures are simply not reported in the registry record. The unadjusted comparison is therefore best interpreted as a description of the posted group summaries rather than a fully documented account of the trial’s inferential analysis.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The registry reports serious adverse events for 123 participants at risk in Educational Control and 115 at risk in Chronic Care Model for Mood Disorders. Serious adverse events affected 0 participants in each arm. The registry also reports other adverse events for the same at-risk counts, with 0 affected participants in each arm. These figures are participant counts in the registry’s adverse-event summaries; they do not describe the primary-outcome analysis groups, which have different reported counts.

The registry lists 1 death in Educational Control and 1 in Chronic Care Model for Mood Disorders. It does not report whether either death was judged related to the intervention, the timing or cause, or the circumstances of these deaths. No top serious adverse-event terms are listed. The record does not provide enough detail to infer that no adverse events of any kind occurred, to assess event definitions or ascertainment, or to compare event risk reliably.

Zero affected participants in the posted adverse-event categories means that no affected participants are recorded in those categories for the stated at-risk groups. It is not proof that the intervention has no safety risks. The group sizes and the registry’s limited event detail constrain what can be concluded. These statistics should be read as registry reporting, not as individualized safety guidance.

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

The posted result shows a higher mean 12-month SF-12 Mental Health Component Score in the Chronic Care Model for Mood Disorders arm than in Educational Control: 43.2 compared with 38.2. The reported mean difference is 5.0 score points, with an interval from 0.355 to 9.645 and an unadjusted p-value of 0.035. Since higher scores represent better outcomes, the direction of the group-average difference favors the intervention arm on this measure. This is a description of the posted data, not a conclusion that the intervention works or does not work.

The result does not say that every participant in the intervention arm had a higher score, that individual participants improved, or that the average difference was caused by the intervention. The posted summaries do not include baseline scores or individual changes. Nor do they establish whether the difference is important to patients: a clinical-importance threshold is not reported in the registry record.

Statistical uncertainty remains. The interval covers a range of positive mean differences, with a lower value close to zero and an upper value considerably larger than the point estimate. The p-value describes the compatibility of the result with a no-difference model under the unadjusted calculation; it does not represent the probability that the intervention is effective. The registry does not report the trial’s original statistical plan or a confidence level for the interval.

Interpretation is further limited by the outcome counts of 42 and 31, which are below planned enrollment of 280. The reasons for this difference, missing-outcome methods, and analysis populations are not reported in the registry record. The available comparison may therefore not represent every randomized participant, and the posted information does not permit evaluation of how missing data affected the estimate. The registry’s randomization designation is relevant design information, but details of assignment procedures and masking are not reported.

Finally, these findings concern the stated outcome, population, intervention, comparator, and follow-up. They should not be generalized automatically to other care settings, populations, interventions, or outcomes. The record supports a carefully qualified account of a group-average score difference; it does not supply all the information needed for a definitive causal or clinical interpretation.

16. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the distinction between a point estimate and the uncertainty around it. The reported mean difference of 5.0 is easy to state, but the interval from 0.355 to 9.645 shows that the estimate does not pinpoint a single value. Looking at both the difference and its interval gives a more complete account than relying on the p-value alone.

The continuous SF-12 score also makes clear why outcome scale matters. The mean difference is expressed in score points, while the standardized effect of 0.505 expresses the contrast relative to pooled variability. Neither scale independently tells the reader what participants would regard as an important difference. A clinically meaningful threshold is not reported in the registry record, so statistical evidence and clinical interpretation must remain distinct.

The planning illustration highlights the dependence of sample-size requirements on the target effect and power. Its participant-per-arm figures rise as the effect assumed for planning becomes smaller, and the higher power target also requires more participants. These values are illustrations based on the observed effect and stated assumptions, not evidence about the trial’s original design or its achieved power.

Finally, the registry’s limited reporting makes transparent interpretation important. The primary outcome counts are smaller than planned enrollment, while missing-data handling, analysis populations, multiplicity procedures, and detailed analysis methods are not reported in the registry record. Recognizing those limits helps prevent an unadjusted result from being read as a complete account of the trial. The statistics are most useful when the effect estimate, its uncertainty, the reporting gaps, and the scope of the measured outcome are considered together.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several trial-specific limitations follow from the registry information. Planned enrollment is 280, but the primary-outcome summaries report 42 participants in Educational Control and 31 in Chronic Care Model for Mood Disorders. The registry does not explain the difference between planned enrollment and these outcome counts. It does not report participant flow, follow-up completion, or reasons for missing outcomes, so the extent and possible consequences of missing data cannot be assessed.

The comparison is unadjusted and no analysis records are posted. Baseline scores, baseline balance, covariate adjustments, clinic-level analysis, and sensitivity analyses are not reported in the registry record. The analysis population and the method used for missing data are also not reported. These omissions limit assessment of whether the summary represents all randomized participants and whether alternative analytic choices would materially affect the estimate.

The registry identifies randomization and single masking but does not report the randomization method, allocation concealment, or who was masked. It gives limited detail about the intervention beyond the Life Goals Collaborative Care name, and does not specify how the intervention was delivered or measured. The control description mentions usual clinic care and a mailed self-guided workbook, but does not characterize usual care across sites.

The reported outcome is a single mental-health quality-of-life score at 12-months. Baseline outcome data and a threshold for clinical importance are not reported. The registry also does not report the confidence level for the interval or the trial’s prespecified testing and sample-size plans. Consequently, the result can be described, but its clinical importance, robustness to missing data, and relationship to the original analysis plan cannot be fully evaluated from the registry record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Implementing Health Plan-Level Care Management for Solo & Small Practices — NCT02041962.
• Registry record last updated 2019-04-23; status last verified 2019-04.
• Results first posted 2019-04-16.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

The University of Michigan sponsored this completed randomized, single-masked, parallel-group study of care management for people with bipolar disorder or depression in solo and small practices. The registry lists planned enrollment of 280, a study start of 2014-07, and primary completion of 2017-11. Chronic Care Model for Mood Disorders was identified as Life Goals Collaborative Care; Educational Control involved usual clinic care and a mailed self-guided workbook.

The primary outcome was the SF-12 Mental Health Component Score at 12-months. The posted means were 43.2 among 31 intervention participants and 38.2 among 42 control participants. The unadjusted mean difference was 5.0, with an interval from 0.355 to 9.645 and p-value of 0.035. The registry lists 0 serious adverse events affecting participants in either arm and 1 death in each arm.

Detailed trial analysis methods, missing-data handling, participant flow, and the original sample-size rationale are not reported in the registry record. Interpretation is therefore limited to the posted summaries and their stated scope.