HIV Screening in the Emergency Department Setting: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for HIV Screening in the Emergency Department Setting (NCT00667186), a randomized na study. The primary outcome is Percentage of Tested Participants Newly Diagnosed as HIV Infected.

2. Clinical Question

The comparison concerns two ways of deciding whom to approach for voluntary counseling and testing in the emergency department. It does not compare testing with no testing: both groups were offered testing, but the approach to selecting people differed. The primary endpoint is framed as the percentage of tested participants newly diagnosed as HIV infected, rather than the percentage of everyone enrolled or approached.

The registry identifies the study as randomized and parallel. It records 9572 participants enrolled, while the primary endpoint population is restricted to participants who consented to testing. The arm-level endpoint records list 1915.0 for routine screening and 1454.0 for targeted screening as the analyzed group sizes. Those figures describe the endpoint population, not everyone enrolled. The record does not report the proportion approached who consented in each arm as an outcome value.

3. Trial Design

The registry describes a randomized, parallel-group study with no masking. It was conducted in the emergency department setting and compared targeted screening with routine screening. In the targeted group, participants were approached for voluntary HIV counseling and testing based on risk for HIV. In the routine group, participants were approached regardless of established risk according to age criteria. The specific risk assessment and age criteria are not reported in the registry record.

The record lists enrollment as 9572 and status as completed. The recorded study start is 2008-01, and primary completion is listed as 2010-12. Phase is recorded as NA. The sponsor is the National Institute of Allergy and Infectious Diseases (NIAID). The study concerns HIV infections.

Random assignment is intended, in general, to create groups that are comparable on average at the start of a study, so differences in outcomes can be interpreted in light of the assigned approaches. That purpose does not guarantee that every characteristic will be balanced in a particular trial. The registry record does not report details of the randomization procedure, allocation concealment, or any checks of baseline comparability.

The absence of masking means the record does not describe participants or study staff as unaware of assigned screening approach. In general, awareness of assignment can influence how an intervention is delivered or how outcomes are recorded. For an outcome based on newly diagnosed infection, the registry does not report whether outcome assessment was masked or whether testing procedures were standardized across groups.

The record describes a 3 years timeframe for the primary endpoint. It does not report site-level information, the number of emergency departments, or operational details about how screening was implemented. Although the study description names resource costs as an additional objective, the registry record does not report cost outcomes.

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: Patients were ineligible if age was \<18 or \>64, they were known to be positive, or they had been previously approached for HIV testing in the ED that day. Post-exposure testing (e.g. occupational, assault, etc.) was not provided and persons for whom such testing was indicated were excluded.

Overall Study

Analysis population for the primary outcome, as defined in the registry: participants analyzed is restricted to the participants that consented to testing

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 Tested Participants Newly Diagnosed as HIV Infected, measured over 3 years.

The primary endpoint is titled “Percentage of Tested Participants Newly Diagnosed as HIV Infected.” Its stated unit is the percentage of tested participants who are positive, and the description specifies that the analyzed population is restricted to those who consented to testing. This denominator matters: the endpoint is not a rate among all enrolled participants, all people approached, or all emergency-department visitors.

The registry lists endpoint values of 6.0 for routine screening and 3.0 for targeted screening. Given the stated unit, these are arm-level percentage values. The record does not provide a usable two-arm comparison estimate or a statistical analysis of the difference. No confidence interval or p-value is reported in the registry record, so the precision of the contrast cannot be evaluated from the posted outcome information.

In general, this endpoint is used to describe the yield of screening among those who actually accept testing: it indicates how often a newly diagnosed infection is identified within that tested group. It is relevant to comparing screening strategies because the strategies differ in whom they approach. However, a yield among consenting testers alone does not describe the full pathway from being eligible or approached through accepting a test and receiving a diagnosis. The separate secondary outcome is percentage consenting to testing, but its result is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the study as randomized, parallel, and unmasked. The primary outcome is a percentage among participants who consented to testing. Although this identifies the basic comparison and outcome population, the registry record does not report an analysis model or a statistical test for the primary endpoint. It records no posted analyses. The comparison therefore cannot be reconstructed as a sponsor-reported inferential result from the available registry information.

What does an arm-level percentage tell us?

In general, an arm-level percentage describes the fraction of people in a defined group who had the outcome. Its meaning depends on the denominator. Here, the registry defines the denominator as participants who consented to testing, rather than all enrolled participants or all people approached in the emergency department. The outcome values are 6.0 for routine screening and 3.0 for targeted screening. The record does not provide a two-arm estimate of the contrast, and the endpoint is not reduced to one comparison in the posted analysis.

To assess a contrast statistically, an analysis ordinarily needs to account for the numbers of people with and without the outcome in each group, as well as the study design and chosen model. The registry lists group sizes for the endpoint, but does not report a usable comparison result or a model. No independent reanalysis is presented. There is no reported standard error, confidence interval, or p-value with which to judge uncertainty or compatibility with a null contrast.

Why is uncertainty important?

In general, a confidence interval communicates the precision of an estimated effect: a narrower interval indicates greater precision under the method’s assumptions, while a wider interval indicates greater uncertainty. An interval can also show whether a range of effects remains compatible with the data. No interval is reported in the registry record. Without one, the arm-level values cannot show how precisely the underlying screening yields have been estimated or whether the apparent difference could reflect sampling variation.

A p-value, in general, assesses how unusual data at least as incompatible with a specified null hypothesis would be if that null hypothesis and the analysis assumptions were true. It is not a measure of the size or importance of an effect. No p-value or null hypothesis is reported for this outcome, so statistical significance cannot be evaluated from the registry results.

What does the planning illustration represent?

In general, a sample-size calculation is a planning tool: it estimates how many participants would be needed under stated assumptions to detect a chosen effect with a chosen probability. The record describes a planning illustration based on a fraction of the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. It is explicitly a planning illustration, not the trial’s design justification. The registry record does not report the trial’s sample-size justification, and no planning-illustration results are stated in the registry record.

A planning calculation cannot establish the probability that this completed study detected an effect, and it is not observed power. Its interpretation depends on the assumed effect and statistical model. In general, a planning exercise based on an observed effect is especially dependent on that effect estimate; if the estimate is uncertain, the resulting planning quantities inherit that uncertainty. The registry record does not report the details needed to evaluate such assumptions beyond the general calculation description.

The outcome is a count-related endpoint expressed as a percentage, but the record does not report a comparison model that could account for features such as over-dispersion. In general, over-dispersion occurs when the variability in count data is greater than a model assumes; ignoring it can make uncertainty estimates too small. This is a general methodological consideration, not a claim that over-dispersion was present in this trial. No dispersion estimate or model assessment is reported.

The record gives no indication of an interim analysis, data-monitoring rule, or prespecified statistical analysis plan. Those details are not reported in the registry record. Without them, readers cannot determine whether the posted outcome analysis followed a detailed prespecified method or whether any adaptations were made. The most defensible statistical reading is descriptive: the record gives arm-level outcome values and analyzed group sizes, while leaving the inferential comparison and its precision unreported.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, percentage of tested participants newly diagnosed as HIV infected, and one secondary outcome, percentage consenting to testing. It records two outcome measures overall. The secondary outcome’s values and analysis are not reported in the registry record. The study description also names comparison of resource costs as an objective, but a cost outcome is not reported among the posted results.

When a study evaluates multiple outcomes or multiple analyses, the chance of finding an apparently notable result can increase if no adjustment or prioritization is used. This is a general statistical concern. The record does not report a multiplicity strategy, a hierarchy among outcomes, or adjustments for multiple testing. It also does not report a statistical analysis plan that would clarify which comparisons were prespecified.

The primary endpoint population is explicitly restricted to participants who consented to testing. Arm-level analyzed group sizes are listed, but the registry record does not describe how missing outcome data were handled, whether any participants were excluded after consenting, or whether imputation was used. The number approached and the number consenting by group are not available as outcome results, although consent percentage is listed as a secondary outcome.

The registry reports no posted analyses. It does not report an intention-to-treat analysis, a per-protocol analysis, or another analysis population beyond the endpoint description that restricts the primary outcome to consenters. The method used to account for participants who did not consent is not reported in the registry record. These omissions limit interpretation of how the arm-level outcome values relate to the full randomized groups and to the screening process as a whole.

11. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry’s serious-adverse-event entries list 4692 at risk and 0 affected for routine screening, and 4880 at risk and 0 affected for targeted screening. The entries for other adverse events likewise list 4692 at risk and 0 affected for routine screening, and 4880 at risk and 0 affected for targeted screening. These are the figures recorded in the registry fields; the record does not supply event terms for serious adverse events.

These entries should be read as registry-reported counts, not as a complete clinical safety assessment. A zero affected count in a posted category means that no affected participants were recorded in that field; it does not establish that no events of any kind occurred, that every event was collected, or that the two approaches have equivalent safety. The record does not report a detailed adverse-event collection method, attribution to screening, or event-specific follow-up.

Deaths are not reported in the registry record. The absence of listed serious-event terms also leaves the nature and severity of any events uncharacterized. The main study question concerns screening yield, and the posted safety entries do not supply evidence about the broader burden or consequences of screening.

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

The registry record describes a randomized comparison of two approaches to offering voluntary HIV counseling and testing in emergency departments. The primary endpoint concerns the percentage of consenting, tested participants newly diagnosed as HIV infected. The posted values are 6.0 for routine screening and 3.0 for targeted screening. They are descriptive arm-level results, not a reported estimate of the difference between approaches.

The missing two-arm comparison is central to interpretation. No confidence interval, p-value, or analysis model is reported, and the registry indicates that a comparison was not estimable from the posted endpoint data. The values therefore do not establish the magnitude or precision of any contrast. They also do not establish statistical superiority, equivalence, or absence of an effect.

The denominator limits what can be concluded. The primary outcome includes only participants who consented to testing. It does not show the percentage of all approached people who were tested, and the secondary outcome on consent is listed without results. Because the strategies differ in whom they approach, understanding the whole screening process would require results describing approach, acceptance, and diagnosis together. The registry record does not provide that complete pathway.

The study description says that identifying more HIV-infected participants and comparing resource costs were objectives. The posted primary endpoint is a percentage among testers, rather than a total number of diagnoses, and cost results are not reported in the registry record. The available information therefore does not answer every objective described for the trial.

In general, a trial’s randomized design can support a causal comparison when the assigned groups, outcome definition, follow-up, and analysis are appropriately reported and analyzed. Here, randomization is recorded, but details of allocation, outcome analysis, missing-data handling, and comparison estimates are absent. The result should be read as a limited registry summary of arm-level screening yield, not as a definitive assessment of the relative effectiveness, costs, or overall consequences of either strategy.

13. Why This Trial Matters Statistically

This trial poses a statistical question that is broader than whether a test detects infection. It compares two approaches to selecting people for an offer of testing: risk-based targeting and routine approach according to age criteria. Evaluating such strategies requires attention to the entire pathway from approach to consent to diagnosis. The primary endpoint captures only the final yield among consenting testers.

That distinction illustrates why denominator choice matters. A higher positivity percentage among tested participants would not necessarily mean that more infections were identified overall, because the number and characteristics of people approached and tested also matter. The registry lists consent percentage as a secondary outcome but does not report its results, and it does not report total diagnoses by strategy. The posted primary values cannot independently describe the full reach of screening.

The trial also illustrates the difference between descriptive and inferential results. Arm-level values can tell readers what the registry records for each group, but a comparison estimate and its uncertainty are needed to judge how compatible the data are with a range of differences. No such estimate, confidence interval, or p-value is reported. The result is thus informative about the form of the endpoint and its posted values, while remaining limited as evidence about the comparative effect.

Finally, the record names resource use and costs as an objective but does not report cost results. This matters statistically because a service strategy may be assessed on several dimensions, and an outcome focused on diagnostic yield does not substitute for a cost comparison. The trial is a useful example of why a registry’s design description, outcome definition, analyzed population, and posted analysis results must be read together—and why an absent comparison should not be filled in by inference from arm values alone.

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.

Several trial-specific details that would help interpret the results are not reported in the registry record. The risk criteria used to determine targeted approaches and the age criteria used for routine screening are not stated. Without those definitions, it is difficult to judge how the two strategies were operationalized or whether they would be implemented similarly in another emergency department.

The record does not report the number approached, the number eligible under the respective screening criteria, or the consent outcome values by arm. The primary endpoint is restricted to consenting participants, so the posted values do not capture the complete screening pathway. The trial enrolled 9572 participants, while the endpoint group sizes are smaller and differ between arms; the registry does not explain the flow from enrollment to endpoint population in detail.

The registry does not report a statistical comparison, uncertainty interval, analysis model, missing-data method, or detailed analysis population. No baseline characteristics or randomization procedures are described in the posted information. The absence of masking is recorded, but the record does not explain how testing and outcome ascertainment were standardized. These gaps prevent a detailed assessment of potential selection, measurement, and analysis effects.

Generalizability is also difficult to assess because the registry record does not identify sites, describe local emergency-department procedures, or report participant characteristics. Cost outcomes are not reported even though cost comparison is named as an objective. The registry provides no serious-event terms, and deaths are not reported. These limitations are specific to the information available in the record; they do not establish that the unreported procedures or outcomes were absent from the conduct of the study.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: HIV Screening in the Emergency Department Setting — NCT00667186.
• Registry record last updated 2013-02-04; status last verified 2013-01.
• Results first posted 2013-01-24.
• 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

The completed study is recorded as a randomized, unmasked, parallel comparison of targeted and routine HIV screening in the emergency department setting. Enrollment is listed as 9572. The study start is 2008-01, and primary completion is 2010-12.

The primary endpoint is the percentage of tested participants newly diagnosed as HIV infected, among those who consented to testing. The registry lists values of 6.0 for routine screening and 3.0 for targeted screening, with endpoint analyzed group sizes of 1915.0 and 1454.0, respectively. It does not report an estimable two-arm comparison, confidence interval, p-value, or analysis model.

Percentage consenting to testing is listed as a secondary outcome, but its result is not reported in the registry record. The study description also names resource costs as an objective; cost results are not reported. Serious-adverse-event entries list 0 affected in each group, while deaths and event terms are not reported.