Sun Safe Workplaces: A Campaign on Sun Protection Policies for Outdoor Workers Phase I: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Sun Safe Workplaces: A Campaign on Sun Protection Policies for Outdoor Workers Phase I (NCT02824289), a randomized na study. The primary outcome is Presence of a Sun Protection Policy for Outdoor Workers Score at 2-Year Follow-up.

2. Clinical Question

3. Trial Design

The registry describes a randomized, group-randomized pretest-posttest controlled trial with parallel arms and no masking. The organizations were the stated participants, and policy outcomes were compared between organizations assigned to the Sun Safe Workplaces Program and those assigned to attention control. The record lists an enrollment of 825. The posted primary outcome has arm counts of 50 for the campaign and 48 for attention control; the record does not explain how these endpoint counts relate to the overall enrollment figure.

The intervention was a workplace campaign delivered through visits with senior managers and in-person training for outdoor workers over two years. The comparison arm received two mailings with educational materials and presentations at state professional meetings. The design therefore compared two different approaches to promoting occupational sun protection rather than comparing the campaign with an explicitly described absence of activity.

The primary outcome was assessed at two-year follow-up by coding written workplace policy documents. The registry also lists a secondary measure of the extent of sun-protection policy from baseline to two-year follow-up. The trial is marked completed. Its recorded start date is 2009-07-01, and its primary completion date is 2015-04-30. The phase is listed as not applicable, and the sponsor is Klein Buendel, Inc.

The registry record does not report the details of the randomization procedure, the specific unit or method used to allocate organizations, any stratification or blocking, or the original sample-size justification. It also does not report whether outcome coders were blinded to assignment, how organizations were recruited, or what procedures were used to account for clustering in the analysis. Those omissions limit what can be concluded about the planned analysis and implementation of the 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.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Policies were compared between organizations assigned to intervention and control arms. Organizations were public sector employers including city and county governments and special taxing districts providing fire and sanitation services. These organizations were the participants. A single policy score was obtained for each organization based on assessment of the organization's written policy documents. Thus, there were no human participants, just organizational units in the primary analysis.

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 Presence of a Sun Protection Policy for Outdoor Workers Score at 2-Year Follow-up, measured over 2 year follow-up.

The primary endpoint is the Presence of a Sun Protection Policy for Outdoor Workers Score at two-year follow-up. Project staff coded written workplace policy documents for the presence of content categories in three domains: environmental controls, administrative procedures, and personal protection practices. The total score was calculated by summing the number of policy components in the documents, with a stated range of 0 to 15. Higher values indicate more comprehensive written sun-protection policy.

The description also defines the presence of any policy component as a value of 1 and no content components as a value of 0. The reported endpoint, however, is a score based on the sum of components, and the posted arm summaries are means on that score. The endpoint therefore captures the extent of policy content, rather than simply whether an organization had any written policy at all.

A document-based measure is relevant to a campaign designed to promote formal workplace policies: it offers a way to assess whether organizational rules include specific protective components. It is distinct from implementation. A written policy score cannot by itself establish that a policy was put into practice, that workers followed it, or that workers experienced any particular health outcome. Those outcomes are not what the primary score measures.

7. Results Posted in the Registry

• Sun Safe Workplaces Program: mean 2.68 versus 0.59 on control; difference 2.09 (95% CI 2.02 to 2.16), standardized effect d = 11.12; z-test p = <0.001.

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.

Extent of Sun Protection Policy for Outdoor Workers at Posttest

• 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:
  • Sun Safe Workplaces Program: mean 4.26 versus 1.35 on control; difference 2.91 (95% CI 2.77 to 3.05), standardized effect d = 8.48; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Sun Safe Workplaces Program than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcome concerns the extent of sun-protection policy from baseline to two-year follow-up. Its arm summaries and unadjusted comparison follow the same direction as the primary outcome: the intervention arm’s policy score is higher than the attention-control arm’s. Taken together, the two posted policy outcomes show a consistent pattern in the direction of more documented policy content in the campaign arm.

The secondary measure is closely related to the primary endpoint. Both concern written workplace sun-protection policy, and both are based on policy scores rather than worker behavior or health outcomes. Their agreement is therefore useful as a coherent pattern across related measures, but it is not equivalent to replication in an independent population or confirmation by a different type of endpoint. The secondary result should be interpreted as supporting context for the primary policy-document finding, not as evidence about outcomes it did not measure.

The secondary comparison is unadjusted. When multiple outcomes are tested, unadjusted p-values do not account for the possibility of chance findings across the set of tests. The registry record does not report a multiplicity plan, a hierarchy among outcomes beyond identifying the primary endpoint, or a correction method. This is one reason the secondary evidence carries less weight than the primary result, especially when the outcomes overlap conceptually.

The record lists two outcome measures and one posted secondary outcome, but it does not provide a reported-analysis entry for the secondary measure. The registry summaries still give arm values and a comparison, yet they do not establish whether the comparison was prespecified, what model was used, or how baseline information was incorporated. Those missing details limit the strength of claims based on the secondary result. The overall pattern remains descriptive of the posted policy outcomes and should not be expanded to implementation, worker sun-protection practices, or health effects.

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

• Extent of Sun Protection Policy for Outdoor Workers at Posttest

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a continuous score, and the posted summaries use least-squares means with standard deviations. A mean represents the average score across organizations in an arm. A standard deviation describes variation in scores around that arm’s mean; it does not describe uncertainty in the mean itself. The posted arm summaries are 2.68 with a standard deviation of 0.24 for the Sun Safe Workplaces Program and 0.59 with a standard deviation of 0.11 for attention control.

What does the mean difference represent?

The posted comparison is a mean difference: the campaign-arm mean minus the attention-control mean. Its value is 2.09. In plain terms, the campaign arm’s average score is higher by that amount on the policy-component scale. A mean difference retains the endpoint’s original units, which makes it interpretable against the score definition. It does not indicate that every organization had a higher score or that the difference was the same for every organization.

Least-squares means are adjusted means in many statistical settings, but the registry record does not identify a model or covariates for this reported primary summary. It is therefore not possible to describe which adjustments, if any, were used to obtain the posted means. Nor does the record state how the group-randomized design was reflected in the calculation. The comparison available for statistical interpretation is the posted summary-statistic contrast and its accompanying interval and unadjusted p-value.

How should the confidence interval be understood?

A confidence interval expresses uncertainty around an estimated quantity under the assumptions of the calculation. For the primary mean difference, the posted interval extends from 2.017 to 2.163. In general, a narrower interval indicates greater statistical precision than a wider interval for the same kind of estimate. The interval does not account automatically for every possible source of uncertainty; its meaning depends on the statistical model and design assumptions used to construct it. The registry record does not describe those details.

General statistical explanation: a confidence interval is not a range that contains a known proportion of individual organization scores. It concerns uncertainty in the estimated group contrast. It also does not establish that the campaign caused the difference unless the design, analysis, and assumptions support that interpretation. Random allocation is part of the trial design, but the record’s limited reporting about allocation and analysis means that several relevant assumptions cannot be evaluated from the registry record alone.

What does the p-value add?

The unadjusted p-value for the primary comparison is <0.001. A p-value describes how incompatible the observed data are with a specified null model, assuming the model and its assumptions hold. It is not the probability that the null model is true, and it does not measure the size or importance of the difference. A small p-value can occur alongside uncertainty about the model, outcome measurement, or generalizability. The registry record does not report an adjusted p-value or a multiplicity strategy.

General statistical explanation: the designation unadjusted means the reported comparison is not described as having been modified to account for multiple statistical tests. When several outcomes or comparisons are examined, the chance of at least one apparently notable result can increase. The registry lists a primary and a secondary policy outcome, but it does not report how the analyses were prioritized or whether any correction or gatekeeping procedure was planned.

Why does the group-randomized design matter?

Organizations, not individual workers, are described as the participants for the policy comparison. In a group-randomized trial, observations within the same organization may be more alike than observations from different organizations. General statistical explanation: when outcomes are correlated within groups, treating every observation as independent can understate uncertainty. Appropriate analysis can account for grouping through design-aware methods, but the registry record does not identify the model or variance method used for the posted results. The endpoint counts of 50 and 48 are organization-level counts as described in the trial population, while the overall enrollment is 825; the record does not explain the relationship between these figures.

The primary endpoint description says staff coded policy documents for components in three domains. General measurement explanation: a coded document score depends on the coding rules and their consistent application. The registry describes the domains and score construction but does not report coder training, reliability checks, adjudication procedures, or blinding. The score is therefore interpretable as the registry-defined count of documented components, while uncertainty about the coding process remains.

The posted secondary endpoint is also continuous and has arm means and an unadjusted comparison. Its measure concerns extent of policy from baseline to two-year follow-up. The registry record does not report a detailed longitudinal model, an adjusted analysis, or a reported analysis entry. Consequently, the summary-statistic contrast should not be mistaken for a fully described repeated-measures analysis. Across both outcomes, the numerical summaries convey the observed direction and scale of the differences, while the record leaves important details of statistical specification unavailable.

11. Confidence Intervals and Precision

For the primary mean difference, the posted estimate is 2.09, with a confidence interval from 2.017 to 2.163. The interval is narrow relative to the scale of the estimate and indicates high precision under the calculation used to produce it. The primary arm means are 2.68 and 0.59, with standard deviations of 0.24 and 0.11. These standard deviations describe variation among scores, whereas the confidence interval describes uncertainty around the estimated difference.

Precision should be interpreted in light of the design and analysis. Organizations were randomized as groups, but the registry record does not report the statistical model or how within-organization dependence was handled. The record also does not explain the derivation of the primary standard deviations, and the posted calculation flags the primary dispersion as approximate. Without a full analysis description, readers cannot verify whether the interval reflects all relevant features of the group-randomized design.

The secondary comparison also has a posted interval, from 2.775 to 3.045 around a mean difference of 2.91. Its existence does not remove concerns about multiplicity or substitute for a reported analysis plan. More generally, a precise interval can coexist with limitations in outcome measurement, missing-data reporting, and external validity. The interval quantifies uncertainty within its statistical calculation, not all uncertainty about the trial’s interpretation.

The primary interval does not imply that the effect is identical across organizations or that every workplace would show a similar policy change. The registry does not provide organization-level scores, a distribution of changes, or subgroup estimates. It also does not report sensitivity analyses. The precision of the posted average contrast should therefore be kept separate from the breadth of conclusions that can be drawn.

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 display is a planning illustration based on the observed effect, not a record of the trial’s original design calculation. The listed 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 defined as a stated fraction of the observed effect. For these continuous outcomes, the displayed sample sizes are participants per arm, not total events. The total-event convention applies to time-to-event displays, which are not the outcomes considered in this trial.

The 80% power column gives the estimated number of participants per arm for a design with an 80% chance of detecting the specified effect under the planning assumptions, if that effect and the assumptions were correct. The 90% power column uses a higher target chance of detection and will generally require at least as many participants per arm. Smaller effects need more participants to distinguish them from random variation with the same target power.

The display considers planning for 100%, 75%, and 50% of the observed effect. The posted primary planning figures are 1 participant per arm in each listed power-and-effect scenario. For the secondary outcome, the displayed figures are also 1 per arm except for the 90% power scenario planning for 50% of the observed effect, which gives 2 per arm. These very small outputs follow from basing a planning calculation on the unusually large observed contrasts relative to the posted dispersion; they should not be read as realistic evidence that such small trials would reliably answer the question. The registry record does not report the trial’s own original sample-size justification or planning assumptions.

The curve should be read as illustrating how the required per-arm sample size changes as the effect to detect changes under the stated planning assumptions. It is not a description of the number enrolled or analyzed. This is not observed power: it is a hypothetical planning calculation derived from the observed effect, and the observed effect can be an unstable basis for planning a future study.

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 = 2.090; pooled standard deviation σ = 0.188.
2. n = 2 × (1.96 + 0.842)² × 0.188² / 2.090² = 0.1, rounded up to 1 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 <- 2.09                      # observed difference in means
sigma <- 0.18794669013668                     # 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(2824289)
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 = 1 (80% power), 1 (90% power)
 75% of observed effect: n per arm = 1 (80% power), 1 (90% power)
 50% of observed effect: n per arm = 1 (80% power), 1 (90% power)
power.t.test: 1.6 
Simulated power at n = 1 per arm: NA (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 identifies a primary policy-score outcome and a secondary policy-extent outcome. It does not report a multiplicity strategy, such as an adjustment to significance thresholds, a testing hierarchy, or a plan to control the chance of false-positive findings across outcomes. The posted comparisons are unadjusted. General statistical explanation: when several outcomes are tested, the probability of at least one result appearing statistically notable by chance can be greater than for a single prespecified test. This does not invalidate the estimates, but it is relevant when interpreting the secondary result and the full set of findings.

The registry record does not report missing-data amounts, reasons for missingness, or procedures for handling missing outcome scores. Although endpoint arm counts are posted, the record does not explain whether all organizations assigned to each arm contributed a policy document, whether any observations were excluded, or how missing documents were treated. It also does not report whether analyses used intention-to-treat, per-protocol, or another analysis population. These omissions prevent assessment of possible bias from missing or excluded data.

The overall enrollment is 825, while the primary endpoint arm counts are 50 and 48. The organizations are described as the participants for the policy comparison, but the registry record does not explain how the overall enrollment relates to the organization counts or the endpoint analysis. It would be inappropriate to infer the number of workers assessed from these summaries.

For a group-randomized trial, the analysis should reflect that organizations, rather than individual workers, were assigned. The registry record does not report the analysis model, the handling of correlation within organizations, or whether any covariates were included. It also does not describe the coding quality checks for policy documents. These are not proof of a flaw; they are unreported methodological details that constrain independent evaluation of the posted comparisons.

The record lists no reported analysis entries. That absence means the analysis specification is not documented in the registry’s analysis listings; it does not by itself establish that no analyses were conducted. The posted summaries remain interpretable as reported outcome statistics, while the missing information limits assessment of prespecification, missing-data sensitivity, and analytic robustness.

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 reports serious adverse events for 413 people at risk in the Sun Safe Workplaces Program group and 412 in the attention-control group. The number affected is recorded as 0 in each group. Other adverse events are also recorded as 0 affected among 413 at risk in the campaign group and 412 at risk in the attention-control group. Deaths are recorded as 0 in both groups.

These figures describe what is entered in the registry record for the listed groups and at-risk counts. They do not provide event definitions, details about how adverse events were collected, or a narrative account of monitoring. The registry lists no serious adverse-event terms. Zero recorded affected does not establish that no events of any kind occurred outside the record or that every possible event was captured. These safety summaries should be read as registry reporting, not as a complete independent assessment of safety.

The adverse-event groups have counts that differ from the endpoint arm counts, and the registry does not explain the relationship between those populations. The safety figures therefore should not be combined with the policy-score summaries or used to infer the number of organizations contributing to the primary endpoint.

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

The posted primary result is a comparison of written workplace policy scores at two-year follow-up. The campaign arm has the higher reported mean, and the posted unadjusted interval and p-value indicate a clear statistical separation under the calculation used. This supports the descriptive conclusion that organizations assigned to the Sun Safe Workplaces Program had more policy components documented on average than organizations assigned to attention control in the posted endpoint summaries.

The result does not establish that outdoor workers used protective practices more often, that workplace policies were implemented consistently, or that any health outcome changed. The primary score was derived from written policy documents. The registry’s broader study summary mentions implementation and worker practices as secondary interests, but those outcomes are not represented by the primary policy score. Nor does this result show that every organization improved or that any particular policy component was adopted.

Random allocation supports a causal comparison in principle, but a strong causal interpretation also depends on how the design was carried out and how the analysis handled group assignment, missing documents, and other relevant factors. The registry record does not report the allocation procedure, the statistical model, missing-data methods, analysis populations, or multiplicity plan. Those gaps make it difficult to evaluate the full strength of the causal inference from the registry entry alone.

The attention-control arm received educational mailings and presentations, so the contrast is between the described campaign approaches. It is not a comparison with no workplace sun-protection activity. The results also apply most directly to the participating public-sector organizations described in the registry. The record does not provide enough information to determine how well the findings generalize to other employers or settings.

The findings should therefore be read narrowly: the posted summaries show higher average written-policy scores in the campaign arm, with a precise unadjusted contrast under the displayed calculation. They do not by themselves establish implementation, worker behavior, health benefit, or universal effectiveness, and they are not a basis for individual medical advice.

17. Why This Trial Matters Statistically

This trial is statistically informative because its intervention targets organizational policy while its primary outcome is measured from written documents. That pairing makes the estimand concrete: the comparison concerns the documented content of workplace policies, not a loosely defined impression of workplace sun protection. The score’s domains—environmental controls, administrative procedures, and personal protection practices—also show how a broad policy concept was operationalized for comparison.

The group-randomized design highlights a central issue in clinical and public-health statistics: the unit assigned to an intervention may differ from the people or records used to describe outcomes. Here, organizations are described as participants, and the total enrollment is 825, while endpoint counts are much smaller. Correctly understanding the unit of analysis and the relationship between organizational and individual counts is essential for interpreting precision and avoiding false assumptions about independent observations.

The result also illustrates why a narrow confidence interval does not settle every statistical question. The posted primary interval is precise under its calculation, yet the registry record does not identify the model or explain how clustering was handled. Transparent reporting of those methods would help readers judge whether the interval reflects the randomized design. The same applies to missing documents, analysis populations, and multiplicity: these details influence how much confidence to place in a summary without changing the fact that the posted arm means differ.

Finally, the distinction between policy adoption and policy implementation is important. The primary endpoint measures written content, which is a meaningful organizational outcome but not a direct measure of worker practice. Reading the estimate in the units and scope actually measured prevents a policy-document finding from being extended to behavior or health outcomes that the endpoint does not establish.

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 registry entry leaves several trial-specific limitations difficult to assess. It describes a group-randomized design but does not report the allocation procedure, the statistical model, or how correlation among organizations or observations within organizations was handled. Because organizations were participants in the policy comparison, these details are important to interpreting the precision of the posted estimate. The record also does not explain the relationship between overall enrollment of 825 and the endpoint arm counts of 50 and 48.

The primary outcome relies on staff coding written policy documents. The registry provides the component domains and scoring approach, but does not report coder blinding, inter-rater reliability, training, adjudication, or handling of unavailable documents. It therefore is not possible to assess how reproducibly the documents were scored from the registry record. A document score also cannot establish whether policy components were put into practice or followed by outdoor workers.

Missing-data methods and analysis populations are not reported. The registry does not state whether all randomized organizations were included in the endpoint summaries, how missing policy documents were addressed, or whether the analysis followed assignment regardless of participation. A multiplicity plan is also not reported, limiting interpretation of the secondary comparison alongside the primary result. The registry lists no reported analysis entries, so details such as covariate adjustment and sensitivity analyses are unavailable.

The posted summary provides no organization-level distribution or subgroup results, and it does not describe the participating organizations in enough detail to assess representativeness. Generalizability beyond the public-sector employers described is therefore uncertain. These limitations concern what can be evaluated from the registry record; they do not establish that the study lacked procedures that were not reported.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Sun Safe Workplaces: A Campaign on Sun Protection Policies for Outdoor Workers Phase I — NCT02824289.
• Registry record last updated 2026-08-18; status last verified 2026-07.
• Results first posted 2026-08-18.
• 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.

22. Record Summary

NCT02824289 is a completed, parallel, randomized trial of the Sun Safe Workplaces Program compared with attention control among local government organizations. The registry describes it as a group-randomized pretest-posttest controlled trial with no masking. Enrollment is 825, and the primary endpoint is a written-policy score at two-year follow-up. The posted primary summaries show a higher mean policy score in the campaign arm, with a mean difference of 2.09 and an unadjusted p-value of <0.001.

A secondary policy-extent outcome follows the same overall direction. The safety record lists 0 serious adverse events, 0 other adverse events, and 0 deaths in both groups. The registry does not report the original sample-size justification, detailed statistical model, missing-data handling, analysis populations, or multiplicity strategy. The results concern documented workplace policy content and do not, on their own, establish implementation, worker practices, or health outcomes.