Evaluating Cancer Survivorship Care Models: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluating Cancer Survivorship Care Models (NCT02362750), a registered phase not stated study. The primary outcome is Percentage of Survivors Reporting "Yes Definitely" Across Items Within Each of 9 Factors Reflecting Survivorship Care Quality.

2. Clinical Question

3. Trial Design

Evaluating Cancer Survivorship Care Models is a study led by George Washington University. The registry record lists a start date of 2013-07, primary completion in 2016-10, a status of completed, and enrollment of 991. The study concerns survivorship-care delivery for people with non-metastatic breast, prostate, and colorectal cancer.

Three models are described. CSSC is characterized as consultative survivorship care after treatment has ended. LSSC is characterized as ongoing survivorship care through survivorship-specific visits. OESC is characterized as survivorship care integrated into standard post-treatment oncology follow-up. The record also distinguishes direct access for the two specialized clinic models from varied access for OESC. The descriptions of the clinic models are truncated in the registry record, so further operational detail is not reported in the registry record.

The record does not state how participants were allocated to the models. It explicitly cautions that the arms are not a like-for-like control comparison. Consequently, a reader should not interpret the groups as randomized arms or assume that differences between them isolate the effect of the care model. The allocation process, any matching or adjustment strategy, and the analysis plan are not reported in the registry record.

The study emphasizes patient-centered outcomes, including satisfaction and quality of life, and identifies patient-reported care quality as its primary endpoint. It lists four outcome measures and eight analyses, but does not report the outcome values or analysis results. The study description notes collaboration among cancer organizations, survivors, and health professionals; it does not provide further details of how that collaboration affected study conduct.

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.

Baseline prior to survivorship visit

Within 1 week post survivorship visit

Reasons for not completing this period:

3 months post survivorship visit

Reasons for not completing this period:

6 months post survivorship visit

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 Percentage of Survivors Reporting "Yes Definitely" Across Items Within Each of 9 Factors Reflecting Survivorship Care Quality, measured over Patients complete a survey within one week after their survivorship visit, and 6 months post visit..

The primary endpoint is the percentage of participants reporting “Yes Definitely” across items within each of nine factors reflecting survivorship-care quality. The registry describes the measure as the Survivorship Care Quality Index, comprising nine factors and 46 metrics grouped into domains. The listed topics include mental health and social support, information about survivorship care, patient involvement, clinician support and respect, communication, care coordination, transitions, access to care, practical support, and connection to a medical home.

Patient-reported care quality is relevant to a study focused on how follow-up care is delivered because it captures participants’ assessments of their experience, rather than only documenting that a visit took place. In general, reporting the percentage selecting a particular response makes the response threshold clear; it does not, by itself, describe the full distribution of answers or explain why participants chose that response.

The registry schedules surveys within one week after the survivorship visit and at 6 months post visit. These time points could describe early and later reports, but the registry record does not report results at either time point, how values from the two assessments were compared, or how missing responses were handled. It also does not report a threshold for interpreting a difference as important to patients.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help frame what this registry record can and cannot establish. The distinction between a descriptive outcome and a causal comparison is particularly important because the study involves three care models and the allocation method is not stated.

What does the endpoint summarize?

The endpoint is a percentage of participants reporting “Yes Definitely” across items within each care-quality factor. In general, a percentage translates responses into the share meeting a defined response category. This can make a result readily interpretable, but it also compresses information: it does not show the range of responses among those who did not select that category, or whether answers differed across the items that contribute to a factor.

The registry describes the Survivorship Care Quality Index as having nine factors and 46 metrics, with factors grouped into domains. It names content concerning support, information, communication, patient involvement, coordination, transitions, access, practical support, and a medical home. The endpoint description does not report the actual responses, factor-level percentages, or a combined score. The method used to combine item responses, if any, is not reported in the registry record.

Why does allocation matter?

In general, random allocation can help balance measured and unmeasured participant characteristics between groups. When allocation is not stated, group differences may reflect both the care model and differences in the people receiving it. The registry record explicitly says these arms are not a like-for-like control comparison. That warning limits causal interpretation: even if group percentages were available, an unadjusted difference would describe the groups as observed, not necessarily the effect of assigning a survivor to one model rather than another.

The registry record does not report the allocation process, whether the groups were selected from different settings, or whether analyses adjusted for participant characteristics. It also does not report a statistical model for the factor-level outcomes. No assumptions about these procedures should be made.

How should uncertainty be assessed?

In general, a confidence interval describes a range of estimates compatible with the data and the method used to calculate it. Its width conveys precision: a wider interval indicates more uncertainty about the estimate. A p-value, when reported, addresses how compatible the data are with a specified null hypothesis under the analysis assumptions; it does not measure the size or importance of an effect. Neither confidence intervals nor p-values for the primary endpoint are reported in the registry record, so statistical precision and evidence for group differences cannot be evaluated.

Percentages are also affected by the number of participants whose responses contribute to them. The overall enrollment is reported as 991, but the registry record does not give the number contributing to each factor, time point, or care model. Enrollment alone therefore cannot be treated as the denominator for every endpoint result.

What does the planning illustration represent?

The registry’s planning assumptions describe an illustration based on an observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. In general, such calculations estimate the number of participants per arm that would be needed under those assumptions for a targeted power level; they are not observed power and do not show what the completed study achieved.

The record does not report a sample-size justification for the original study design. Nor does the planning description establish that the study used equal allocation, a two-arm design, or those assumptions in its actual analyses. It is a planning illustration, not a substitute for the protocol or an account of the design. The numerical planning outputs and the effect fractions are not reported in the registry record.

What cannot be inferred from the posted record?

The record lists four outcome measures and eight analyses, but it does not provide the primary endpoint results, the analysis methods, or a statistical analysis plan. It does not report whether analyses were adjusted, how repeated surveys were handled, or how any departures from assumptions were assessed. These omissions prevent evaluation of the full statistical evidence. General methods can clarify what such information would mean, but they cannot fill in results or choices that are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists four outcome measures and eight analyses, while the primary endpoint concerns multiple care-quality factors and two scheduled survey times. When several outcomes, factors, time points, or group contrasts are examined, the chance of finding at least one apparently notable result can increase. In general, multiplicity procedures can address this issue by controlling a defined error rate or by identifying a primary comparison in advance. The registry record does not report whether any such procedure was planned or used, which outcomes or contrasts were designated primary for analysis, or how the eight analyses relate to the listed outcome measures.

Missing responses can also affect a patient-reported endpoint. In general, the percentage calculated among respondents may differ from the percentage that would be observed if everyone completed the survey, particularly if response completion is related to care experience. The record gives the survey schedule but does not report completion rates at either time point, the amount or pattern of missing data, or the method used to handle missing responses. A reader cannot determine from the enrollment figure alone how many participants contributed to each outcome.

The analysis population is not described in the registry record. It does not report whether analyses included all enrolled participants, only survey respondents, or another defined group. Rules for exclusions, withdrawals, and incomplete surveys are also not reported. As a result, the population represented by any endpoint estimate cannot be established from this record. These are limits on interpretation, not evidence that any particular approach was or was not used.

11. 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 three groups: CSSC, LSSC, and OESC. For each group it reports the number at risk for serious and other adverse events, the number affected, and deaths. The affected counts and deaths are recorded as 0 for all three groups. The serious-event terms list is empty.

These entries are a narrow account of the fields reported in the registry. A zero affected count means no affected participants are recorded in that field; it does not establish that no symptoms, concerns, or events occurred in any broader sense. The record does not provide event descriptions, event severity beyond the serious-versus-other distinction, timing, relatedness to care, or detail about how events were collected. It also does not report a safety-analysis plan or explain the observation period used for these entries.

Because the allocation method is not stated and the groups are not described as a like-for-like control comparison, the entries should not be treated as a comparative estimate of safety between models. They are registry counts, not evidence of equivalence. No conclusion about the relative safety of the care models follows from these counts alone.

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

The study addresses a relevant service-delivery question: how survivors experience different ways of organizing care after cancer treatment. It identifies three models and a patient-reported quality-of-care endpoint, with surveys scheduled within one week after a survivorship visit and at 6 months. The registry reports enrollment of 991 and a completed study status.

What the record does not show is the primary endpoint result. It gives no factor-level percentages, group contrasts, confidence intervals, or p-values. Therefore, it cannot establish whether participants in one model reported care quality differently from participants in another model, how large any difference was, or how precise an estimate might be. No conclusion about the direction of the outcome is supported by the record.

Even if comparative percentages were available, the allocation method would remain important. The record does not state how participants entered each model and cautions that the arms are not a like-for-like control comparison. In general, an observed group difference in a non-randomized comparison can be associated with differences in participant mix or care setting, as well as with the model itself. The record does not report enough design or analysis detail to separate these explanations.

The endpoint focuses on the “Yes Definitely” response across items within care-quality factors. This is one way to summarize patient reports, not a complete description of every experience captured by a survey. The record does not report the distribution of other responses or how missing surveys were handled. Interpretation should therefore remain tied to the intended measure and to the information actually recorded.

The zero adverse-event counts are likewise limited to the registry entries and do not provide a complete account of safety. Taken together, the record documents the study question and its intended outcomes, but not the results needed to judge comparative differences. Missing results must not be interpreted as equal outcomes, evidence of benefit, or evidence of no benefit.

13. Why This Trial Matters Statistically

Statistically, this study illustrates why the research question, design, outcome definition, and result reporting must be read together. The question concerns care models rather than a drug or procedure, and its primary outcome is a patient-reported assessment of care quality. That makes the meaning of the response threshold and the timing of surveys central to understanding what an estimate would represent.

The endpoint is structured around factors and items, rather than a single simple event. Such measures can capture several dimensions of care, but they also raise questions about how responses are combined, which factors are emphasized, and how results across factors are interpreted. The registry record describes the measure’s scope but does not report the analysis details or results needed to answer those questions.

The comparison also calls for careful reasoning about study design. Three groups are named, allocation is not stated, and the registry warns that the arms are not a like-for-like control comparison. These registry entries show why a difference between groups, if reported, would need to be interpreted in light of how participants came to receive each model. Statistical adjustment can sometimes address measured differences, but the record does not report whether adjustment was used, and adjustment cannot be assumed to resolve every source of bias.

Finally, the completed status and enrollment do not substitute for outcome reporting. A study can be completed without the registry record presenting estimates, uncertainty, missing-data handling, or multiplicity procedures. The record is informative about the intended evaluation, while leaving the central comparative statistical conclusions unreported. This distinction helps prevent both overstatement and unwarranted conclusions from absent results.

14. 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 main limitation for statistical interpretation is that the registry record does not report primary outcome values or comparative results. Without percentages, estimates of differences, or measures of uncertainty, the study’s findings on care quality cannot be evaluated. The completed status and enrollment count do not reveal what the analyses found.

The record does not state the allocation method, and it explicitly cautions that the groups are not a like-for-like control comparison. This limits causal interpretation. Further details about eligibility, recruitment, group comparability, and any adjustment for participant characteristics are not reported in the registry record.

The endpoint is described as “Yes Definitely” responses across items within each of nine factors, with surveys scheduled within one week after the visit and at 6 months. However, the record does not report the response distributions, the number contributing at each time point, the approach to combining item responses, or the handling of missing data. It also does not report whether the same participants contributed at both assessments.

The record lists multiple outcome measures and analyses but does not report a statistical analysis plan or how multiplicity was addressed. It does not describe the analysis population, model, adjustment strategy, or handling of repeated assessments. The descriptions of the clinic models are truncated, so some operational differences are not reported in the registry record.

Finally, the adverse-event entries contain zero affected counts, but event details and safety-monitoring methods are not reported. These limits restrict what can be concluded from the registry record; they do not establish that unreported methods or events were absent.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Evaluating Cancer Survivorship Care Models — NCT02362750.
• Registry record last updated 2026-03-25; status last verified 2026-03.
• Results first posted 2026-03-25.
• 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.

18. Record Summary

Evaluating Cancer Survivorship Care Models is a completed study led by George Washington University, with enrollment of 991. The registry lists a start date of 2013-07 and primary completion in 2016-10. The population of interest was non-metastatic breast, prostate, and colorectal cancer survivors receiving one of three survivorship-care models: CSSC, LSSC, or OESC.

The primary endpoint concerns the percentage reporting “Yes Definitely” across items within each of nine survivorship-care quality factors, measured using the Survivorship Care Quality Index. Surveys were scheduled within one week after the survivorship visit and at 6 months. The registry record does not report endpoint results or comparative estimates.

Allocation is not stated, and the registry cautions that the arms are not a like-for-like control comparison. Serious and other adverse-event fields report zero affected participants and zero deaths in each group. These counts do not provide a complete comparative safety assessment. The analysis plan, missing-data methods, and multiplicity procedures are not reported in the registry record.