This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for REACH Implicit Bias Training Project (NCT03415308), a registered phase not stated study. The primary outcome is Feasibility of Training Measured by Participation.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03415308 |
| Phase | Phase not stated |
| Condition | Communication, Bias, Racial, Relation, Doctor Patient |
| Allocation | Not Stated |
| Design model | Not Stated |
| Masking | Not Stated |
| Enrollment | 114 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Duke University |
| Results posted | Yes |
2. Clinical Question
Population
The pilot concerns healthcare providers eligible to participate in implicit bias training. The broader project also included patient focus groups to identify relevant perceptions and outcomes, and cognitive interviews with provider and health-system stakeholders to inform refinement of the training.
Intervention
The intervention activity was pilot testing of a refined implicit bias training program for providers. The project also included focus groups and cognitive interviews as development activities; these are not described as treatment arms in a like-for-like comparison.
Comparator
A control condition or a like-for-like comparator is not reported in the registry record. Allocation is not stated, and the listed activities have different purposes, so they should not be interpreted as comparable intervention and control groups.
Primary question
The primary feasibility question was whether at least 50% of eligible providers would agree to participate. The registry describes this as a judgment of feasibility, with a timeframe of 4 hours (post-training), but states that primary outcome data were not collected.
3. Trial Design
REACH Implicit Bias Training Project is a development and pilot-testing project, rather than a registry record of a randomized treatment comparison. Its stated purpose was to prepare for a definitive trial of whether provider training could affect racial and ethnic disparities in patient-centered care. The registry record lists enrollment as 114 and status as completed. The recorded study dates are 2018-04-18 for study start and 2022-06-01 for primary completion.
The project activities included patient focus groups, stakeholder cognitive interviews, and pilot testing of implicit bias training. The focus groups were intended to identify patient preferences for constructs and related outcomes reflecting implicit bias in clinical encounters. The stakeholder activity is described as semi-structured interviews, intended to inform refinement of existing training. Pilot testing involved the refined intervention emerging from that development work. The registry does not describe these activities as arms in a randomized comparison.
The primary outcome was feasibility measured by provider participation. The registry says investigators would judge the trial feasible if at least 50% of eligible providers agreed to participate. Listed secondary outcomes concern variability in trial outcomes by provider-related characteristics and change in IAT outcomes from baseline to immediately post-intervention. The registry record does not report primary outcome data, and it does not give results for those secondary outcomes.
Because the project combines formative activities and a pilot, the enrollment figure should not be read as the number assigned to a single intervention-versus-control comparison. The registry lists activity-specific at-risk counts in its adverse-event record, but those counts do not establish randomized allocation or comparability. Allocation is not stated, and the activities are not a like-for-like control comparison.
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
| Stage | Patient Focus Groups | Stakeholders Cognitive Interviews | Pilot Testing of Implicit Bias Training |
|---|---|---|---|
| STARTED | 50 | 27 | 37 |
| COMPLETED | 50 | 27 | 29 |
| NOT COMPLETED | 0 | 0 | 8 |
Reasons for not completing this period:
| Reason | Patient Focus Groups | Stakeholders Cognitive Interviews | Pilot Testing of Implicit Bias Training |
|---|---|---|---|
| Lost to Follow-up | 0 | 0 | 1 |
| Scheduling Conflicts | 0 | 0 | 7 |
Analysis population for the primary outcome, as defined in the registry: Data not collected.
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.
| Characteristic | Unit | Patient Focus Groups | Stakeholders Cognitive Interviews | Pilot Testing of Implicit Bias Training | Total |
|---|---|---|---|---|---|
| Participants analysed | 50 | 27 | 37 | 114 | |
| Age, Continuous | years | 42.4 (SD 15.9) | — | — | 42.4 (SD 15.9) |
| Sex: Female, Male: Female | Participants | 40 | 16 | 29 | 85 |
| Sex: Female, Male: Male | Participants | 10 | 11 | 8 | 29 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 11 | 1 | 1 | 13 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 39 | 11 | 36 | 86 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 15 | 0 | 15 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 0 | 0 | 10 | 10 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 1 | 0 | 1 |
| Race (NIH/OMB): Black or African American | Participants | 20 | 8 | 2 | 30 |
| Race (NIH/OMB): White | Participants | 26 | 18 | 24 | 68 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 1 | 1 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 4 | 0 | 0 | 4 |
6. Primary Endpoint
The registry lists the primary outcome as Feasibility of Training Measured by Participation, measured over 4 hours (post-training).
| Element | Registry entry |
|---|---|
| Result type | N/A |
| Measure of spread | not applicable |
| Unit | not stated |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is feasibility of training, measured by participation. Specifically, the registry describes a decision rule: investigators would judge the trial feasible if at least 50% of eligible providers agreed to participate. This is an implementation measure, not a measure of patient health, clinical effectiveness, or change in provider bias.
The stated timeframe is 4 hours (post-training). The registry record marks the primary outcome population as “Data not collected.” It therefore does not report how many eligible providers were invited, how many agreed, or whether the feasibility criterion was met. No participation estimate or uncertainty interval is reported in the registry record.
Participation is useful in a pilot because it addresses whether providers are willing to take part in the training activity. General statistical context: feasibility measures can help determine whether a later evaluation is practicable, but they do not by themselves establish that an intervention changes patient care or health outcomes. The registry’s stated aim was to lay groundwork for a definitive trial; the participation endpoint should be interpreted within that developmental purpose.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry describes a completed project, but completion status is not a result for the primary endpoint. For the participation measure, the record explicitly says “Data not collected.” As a result, the posted information does not show whether provider participation reached the stated feasibility threshold of at least 50% of eligible providers. It also does not report an estimate of participation or its precision.
The project’s stated rationale concerns implicit bias training and disparities in patient-centered care. That rationale is not itself evidence of an observed treatment effect. The registry identifies pilot testing and development activities, but it does not report comparative results showing whether training changed provider behavior, patient experience, or disparities. The listed secondary outcomes—variability by provider-related characteristics and change in IAT outcomes from baseline to immediately post-intervention—have no results reported in the registry record.
Enrollment is recorded as 114, but this figure should not be treated as a denominator for the primary participation endpoint or as the size of a single comparison group. The registry does not report how many providers were eligible, invited, or agreed to participate. It also does not state allocation, and the named activities are not like-for-like control and intervention groups. Thus, a comparative treatment effect cannot be inferred from the activity labels or the enrollment figure.
The adverse-event record lists zero serious events, zero other adverse events, and zero deaths for each named activity group. These are the posted figures for the registry’s adverse-event fields; they do not establish that every possible harm was measured or that the intervention has no risk. The at-risk counts are activity-specific, and no comparative safety analysis is reported.
The registry lists 0 analyses. This is a description of the registry record, not proof that no analyses were conducted outside it. No statistical analysis plan, model, confidence interval, or inferential result is reported. The most defensible reading is therefore descriptive: the record documents a development and pilot project and its intended feasibility criterion, while its primary outcome data and outcome analyses are not reported.
9. Statistical Methods Explained
The registry presents a mixed set of development and pilot activities, not a single randomized comparison. That distinction matters when interpreting the statistics: a treatment-effect estimate ordinarily depends on a defined comparison between groups, while the activities have different roles. Allocation is not stated, and the record cautions that the activities are not a like-for-like control comparison.
What does the primary measure capture?
The primary measure is participation as a feasibility indicator. Its stated decision rule is that investigators would judge the trial feasible if at least 50% of eligible providers agreed to participate. To estimate that proportion, a record would need a clearly defined eligible-provider denominator and the number who agreed. Neither count is reported for this endpoint, and the registry states that the primary outcome data were not collected. Therefore no observed participation proportion can be interpreted from the record.
General statistical context: a proportion describes how often an event occurs within a defined group. Its interpretation depends on the event definition and denominator. For participation, the distinction between eligible, invited, and consenting providers can be important; the registry record does not report those counts. No denominator should be inferred from the overall enrollment figure because the record includes distinct project activities and does not specify how enrollment relates to eligible providers.
How should uncertainty be understood?
General statistical context: a confidence interval expresses the uncertainty around an estimate under a statistical model and its assumptions. A narrow interval indicates greater precision than a wide interval, but it does not resolve limitations in the way participants were selected or an outcome was measured. The registry record reports no primary endpoint estimate or confidence interval, so the precision of the participation result cannot be assessed. There is likewise no reported estimate or interval for change in IAT outcomes.
A significance test, when used, evaluates compatibility between data and a specified null hypothesis under a model. A p-value does not quantify the probability that an intervention works, and statistical significance alone does not establish practical importance. No p-value or inferential result is reported for the outcomes in this registry record. The record also does not report the statistical plan or the hypotheses used for any analyses.
What do the planning assumptions mean?
The posted sample-size planning illustrations are described as planning calculations based on the observed effect, not as the trial’s own design. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. For time-to-event illustrations, the reported quantity is total events; other illustrations are expressed as participants per arm.
These assumptions describe hypothetical planning calculations. They do not establish that the pilot used equal allocation, that it had a two-arm randomized design, or that the illustrations were part of the protocol. The registry does not report a protocol-based sample-size justification. Planning sample sizes are not observed power. In general, power is a design property calculated under assumptions about an effect and variability; it is not a retrospective measure of whether an observed study result is reliable.
General statistical context: a normal approximation can simplify calculations when its conditions are suitable. If outcomes are more variable than the assumed model allows, uncertainty can be understated; over-dispersion is one reason a model may fit count data poorly. This general point does not imply that a particular model, approximation, or dispersion adjustment was used for this project. The analysis methods are not reported in the registry record.
Why does the project structure matter?
Focus groups and semi-structured interviews are development activities intended to inform constructs, outcomes, and intervention refinement. Their role differs from that of an outcome comparison designed to estimate an intervention effect. The pilot-testing activity is also described as groundwork for proposing a definitive trial. Without reported outcome data, group definitions, allocation details, or an analysis plan, there is no basis for reconstructing a comparative effect from the registry information.
The registry lists variability by provider-related characteristics and change in IAT outcomes as secondary outcomes, but it gives no results for them. A baseline-to-post-intervention change would require measurements at the relevant times and a defined analytic approach. The record does not report the observed measurements, the model, or whether analyses were adjusted for other factors. These omissions limit assessment of both the direction and precision of any change.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists a primary feasibility outcome and secondary outcomes concerning provider-related variability and change in IAT outcomes. It also reports 4 outcome measures and 0 analyses in the registry record. The relationship among the outcome measures, any planned hypotheses, and the listed secondary outcomes is not described in enough detail to determine how multiplicity was handled.
Multiplicity arises when several outcomes or hypotheses are examined: the chance of at least one apparently notable result can increase when many tests are performed. General statistical context: researchers may address this through a prespecified hierarchy, adjusted testing thresholds, or by treating analyses as exploratory. The registry does not report whether any such approach was planned or used. No multiplicity adjustment should be assumed.
Missing-data handling is also not reported in the registry record. In particular, the primary outcome population is marked “Data not collected,” rather than having a reported participation value with an explanation of missing observations. The record does not describe how incomplete IAT measurements, if any, were handled. It does not report imputation, exclusions, or sensitivity analyses.
The analysis populations are not defined in the registry record. It is therefore not possible to determine which participants, providers, or activity groups would be included in any outcome analysis, or how eligibility and participation were operationalized. General statistical context: analysis-population definitions affect the interpretation of an estimate, especially when inclusion depends on attendance, post-intervention completion, or other events after enrollment. No such analytic choices should be inferred for this project.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Patient Focus Groups | 0 | 50 | 0.0% |
| Stakeholders Cognitive Interviews | 0 | 27 | 0.0% |
| Pilot Testing of Implicit Bias Training | 0 | 37 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Patient Focus Groups | 0 of 50 |
| Stakeholders Cognitive Interviews | 0 of 27 |
| Pilot Testing of Implicit Bias Training | 0 of 37 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Patient Focus Groups | 0 of 50 |
| Stakeholders Cognitive Interviews | 0 of 27 |
| Pilot Testing of Implicit Bias Training | 0 of 37 |
The registry’s adverse-event record reports activity-specific counts. For Patient Focus Groups, 50 were listed at risk for serious and other adverse events; the record lists 0 affected for serious events, 0 affected for other events, and 0 deaths. For Stakeholders Cognitive Interviews, 27 were listed at risk in each adverse-event category; the corresponding affected counts and deaths are 0. For Pilot Testing of Implicit Bias Training, 37 were listed at risk in each category; the corresponding affected counts and deaths are also 0.
These figures describe what is recorded for the specified groups and fields. They are not a comparison of randomized treatment arms: the activities are not like-for-like control groups, and allocation is not stated. The registry does not provide event descriptions or a broader safety analysis. Zero recorded events should not be interpreted as proof that no harms occurred or that the training is risk-free; it means the posted adverse-event figures show zero affected in those fields.
12. What the Result Does, and Does Not, Mean
The record documents a project aimed at developing and pilot-testing implicit bias training for healthcare providers. Its development activities included patient focus groups and stakeholder interviews intended to inform relevant outcomes and refine training. The stated primary feasibility criterion was provider agreement to participate, with a threshold of at least 50% of eligible providers.
The primary outcome data were not collected, according to the registry. Consequently, the record does not establish whether the stated feasibility criterion was reached. It also does not report results for the listed secondary outcomes. Completion status and enrollment do not substitute for those results: neither tells readers how many eligible providers agreed, whether IAT outcomes changed, or how precisely any outcome was estimated.
The record does not establish whether implicit bias training reduces racial or ethnic disparities in patient-centered care. That goal appears in the project rationale, while the project is described as groundwork for proposing a definitive trial. The activities are not described as a randomized, like-for-like comparison, and allocation is not stated. There is no reported comparative estimate from which to infer an effect.
The posted adverse-event figures list zero affected for serious and other events in each named activity group, and zero deaths. Those entries are descriptive registry data, not proof that all harms were captured or that risk is absent. Overall, the record supports an account of the project’s purpose, activities, and intended feasibility rule, but not a conclusion about intervention effectiveness or the achievement of the primary feasibility endpoint.
13. Why This Trial Matters Statistically
Development projects can address a measurement problem before testing an intervention’s impact: researchers need outcomes that reflect the experiences and priorities relevant to the clinical encounter. This project’s patient focus groups were intended to identify preferences for constructs and related outcomes reflecting implicit bias. Stakeholder interviews were intended to inform refinement of existing training. Those activities help explain why the registry includes formative work alongside pilot testing rather than only an intervention comparison.
The primary endpoint also illustrates a distinction between feasibility and effectiveness. Provider participation can indicate whether a training activity is acceptable enough to study further, but it is not a direct measure of patient-centered care or disparities. The stated threshold gives the project a practical feasibility criterion; however, because primary outcome data were not collected, the registry cannot show whether that criterion was met.
Statistically, the project matters as a reminder that the purpose of a pilot and the evidence it can support depend on its design and reported outcomes. A completed record with an enrollment figure is not automatically a comparative effectiveness result. Here, allocation is not stated, the activities are not like-for-like comparison groups, and the record reports no outcome analyses. These features limit what can be inferred, while leaving the developmental intent clear.
The record’s stated aim was to lay groundwork for a definitive trial. A later evaluation would need clearly specified comparison conditions and outcome definitions to support a causal interpretation. That is a general design consideration, not a claim that a particular protocol or future study was undertaken. The statistical value of this record is therefore in distinguishing the project’s feasibility question and development activities from evidence about clinical impact.
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 principal limitation for interpreting the primary question is that the registry marks its data as not collected. The record therefore provides a feasibility rule without an observed participation result. It does not report the eligible-provider denominator, the number invited, or the number who agreed, so the participation threshold cannot be evaluated from the posted information.
The record does not state allocation and cautions that the listed activities are not a like-for-like control comparison. The activity descriptions cover patient focus groups, stakeholder cognitive interviews, and pilot testing of training. They cannot be treated as randomized arms or used to infer an intervention effect. The enrollment figure of 114 does not identify the number included in any particular outcome analysis.
Although secondary outcomes are listed, results are not reported for them. The registry does not report measurement estimates, confidence intervals, statistical tests, or a statistical analysis plan. It also does not specify how analysis populations were defined, how missing outcome data would be handled, or whether any multiplicity procedure was planned. These absences make it impossible to assess the precision or robustness of outcome findings from the registry record.
The adverse-event record gives zero affected in its serious and other event fields for the named activities, but it does not provide event narratives or details about ascertainment. Those figures should not be taken as evidence that all possible harms were captured. Finally, the registry’s planning assumptions do not establish the study’s actual design or protocol-based sample-size rationale. A sample-size justification for the trial is not reported in the registry record.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: REACH Implicit Bias Training Project — NCT03415308.
- Registry record last updated 2024-03-18; status last verified 2024-03.
- Results first posted 2024-03-18.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
REACH Implicit Bias Training Project (NCT03415308) is listed as completed, with Duke University as sponsor and enrollment of 114. The registry records a study start of 2018-04-18 and primary completion of 2022-06-01.
The project combined patient focus groups, stakeholder cognitive interviews, and pilot testing of implicit bias training for providers. Its primary outcome was feasibility measured by participation, with a stated criterion of at least 50% of eligible providers agreeing to participate. The primary outcome population is marked “Data not collected.”
Secondary outcomes concern variability by provider-related characteristics and change in IAT outcomes from baseline to immediately post-intervention. Results and analyses are not reported in the registry record. Allocation is not stated, and the activities are not a like-for-like control comparison.