RTLS for Healthcare Process Validation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for RTLS for Healthcare Process Validation (NCT04116957), a randomized na study. The primary outcome is In-Room Activity Time Difference Between RTLS System and Manual Observation.

2. Clinical Question

3. Trial Design

The completed study, NCT04116957, is titled “RTLS for Healthcare Process Validation.” It is registered as a randomized, single-masked, parallel study, with phase listed as NA. The sponsor is the University of Missouri-Columbia. The registry record gives a start date of 2020-02-17 and a primary completion date of 2021-04-16.

The study aimed to develop and assess a real-time location system for understanding patient-care processes. The registry describes deployment of the system in an ICU at the University Hospital, University of Missouri Health Care in Columbia, Missouri. The intended validation compared location-system performance with manual observation of nursing activity. The summary also describes an intended correlation of nursing activity metrics with patient outcomes measured by SOFA score.

The arm descriptions distinguish nurse location data collected with the original NFER system from data collected with an upgraded NFER system. The intervention list identifies the devices as “RTLS current tag” and “RTLS tag revised.” The posted primary outcome summaries list 9 participants for the Tracking System arm and 10 for the New Tracking System arm. Those are the outcome-specific arm counts; the registry does not explain whether every enrolled participant contributed to this outcome or how the counts relate to the enrolled population.

The record marks the study as completed. Completion status alone does not establish that the stated validation aims were met. The registry record does not describe the masking procedure, the allocation process, the unit of randomization, or whether the two arms were balanced on relevant characteristics. It also does not report implementation details such as how the manual observations were scheduled or how observer judgments were standardized. These omissions limit what can be concluded about the design beyond its registered randomized, single-masked, parallel classification.

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

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 In-Room Activity Time Difference Between RTLS System and Manual Observation, measured over Up to 12 hours.

The primary endpoint is titled “In-Room Activity Time Difference Between RTLS System and Manual Observation.” It is a continuous measure reported as a mean with standard deviation, in minutes, over a timeframe of up to 12 hours. The registry description says in-room activity times collected through RTLS are statistically compared with in-room activity time collected during observer time studies.

Manual observation serves as the reference method for assessing whether the tracking system captures activity time accurately. Comparing the methods is relevant because the project’s stated goal is to produce location-based data that can describe, analyze, or validate healthcare processes. The registry describes an anticipated maximum error of 5% relative to manual observation and says the time difference should be less than 15 minutes during the day shift, up to 12 hours. These are stated expectations, not evidence that the posted results meet those criteria.

The registry proposes a paired t-test for determining whether the means of the RTLS and manual-observation measurements differ. A difference in means addresses average discrepancy, but does not by itself describe how close individual paired measurements are, or whether discrepancy varies across activity levels or observation periods. The record does not report a numerical estimate of the paired difference, its confidence interval, or a test result. It also does not explain whether the posted arm-specific summaries are the paired RTLS-versus-manual measurements described in the endpoint definition.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The endpoint is continuous: it records time in minutes, and the registry identifies the reported summary as a mean with standard deviation. Continuous outcomes can take values across a scale rather than being limited to event categories. A mean describes the average recorded value, while a standard deviation describes how dispersed observations are around that average. Neither summary alone shows the shape of the data or the degree of uncertainty in an estimated treatment or system contrast.

Why does pairing matter?

The endpoint description says the RTLS and manual-observation activity times are compared using a paired t-test. Pairing is appropriate when each measurement from one method has a meaningful counterpart from the other method, such as measurements of the same activity period evaluated by both methods. The analysis then focuses on within-pair differences, rather than treating every measurement from each method as unrelated. This can account for shared features of a matched observation, but it requires a clearly defined matching process.

The registry record does not explain how RTLS measurements were matched to manual observations, how many paired observations were available, or how the analysis handled incomplete or unmatched observations. It also does not post a paired difference estimate or test result. The arm summaries are presented by tracking-system group, while the endpoint description concerns comparison with manual observation. Without more detail, the relationship between those summaries and the specified paired analysis is not clear.

What does the paired t-test assess?

In general, a paired t-test evaluates whether the average within-pair difference is compatible with a null mean difference, under the assumptions of the test. A test result does not by itself establish that the methods agree closely enough for a practical purpose. The size and uncertainty of the estimated difference matter, as does whether the acceptable level of discrepancy was defined before the data were examined. The registry supplies a target of less than 15 minutes during the day shift, up to 12 hours, and refers to a maximum error of 5% relative to manual observation. It does not report a confidence interval or a test against either target.

General interpretation of confidence intervals is useful even though none is posted for this endpoint. A confidence interval represents a range of values compatible with the data and the statistical method, under its assumptions. A narrow interval indicates greater precision than a wide interval; an interval can also help readers judge whether estimates are compatible with a meaningful threshold. Without an interval for the paired discrepancy, the precision of the comparison cannot be assessed from the registry record.

Why consider agreement as well as average difference?

In general, a mean difference can be close to zero even when some individual measurements differ materially in opposite directions. Agreement assessment therefore often considers the spread and pattern of paired differences, not only their average. A plot of paired differences against the magnitude of the measurements, for example, can help identify whether discrepancies change across the measurement range. Such analyses require appropriate paired data and clear definitions. The registry record does not report these agreement summaries, so no conclusion about the distribution of individual errors can be drawn from the means and standard deviations alone.

How should variability be read?

The posted standard deviations are 21.19 minutes for the Tracking System arm and 12.46 minutes for the New Tracking System arm. They describe observed dispersion in the listed endpoint values, rather than precision of the arm means or of an arm contrast. Precision depends on the amount of information and the variability, and is normally conveyed with standard errors or confidence intervals. Those uncertainty measures are not reported for the primary endpoint. The arm-specific counts are 9 and 10, respectively, and the record does not report how missing observations or repeated measurements were handled.

General methods also depend on the outcome distribution and study structure. The t-test uses assumptions about the distribution of differences and independence between the paired units. If there were repeated observations from the same nurse, patient, or observation period, those measurements could be correlated; the registry record does not state whether repeated measures were collected or how clustering was addressed. No alternative model or diagnostic assessment is reported. These missing details prevent a full evaluation of whether the proposed analysis matched the data structure.

The record describes a planning illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. Such illustrations are planning aids, not observed power and not evidence that the trial was designed prospectively to detect a particular effect. A prospective sample-size justification is not reported in the registry record. The illustration’s numerical estimates are not reported in the registry record.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Tracking System", "New Tracking System")
n   <- c(9, 10)
m   <- c(34.61, 6.63)
sp  <- c(21.19, 12.46)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Tracking System  9  34.610  21.190  18.322  50.898
CARM  New Tracking System  10  6.630  12.460  -2.283  15.543

11. Multiplicity, Missing Data and Analysis Populations

The registry lists three outcome measures and reports no secondary outcome results. It lists no analyses. The primary endpoint description identifies a paired t-test, but the posted record does not include its result or any other inferential analysis. It is therefore not possible to determine from the registry whether multiple outcomes, comparisons, or analyses were considered in the study’s statistical plan, or whether any adjustment for multiple testing was planned or used. A multiplicity strategy is not reported in the registry record.

Missing-data handling is also not reported in the registry record. The outcome-specific arm counts are 9 for the Tracking System and 10 for the New Tracking System, but the record does not explain whether these counts represent all randomized participants, all participants with an observed endpoint, or a subset with usable measurements. It gives no account of missing time observations, incomplete paired measurements, exclusions, or reasons for nonavailability. Without that information, the completeness of the endpoint data and the possibility of bias from missing observations cannot be evaluated.

The registry record does not define an analysis population, such as an intention-to-treat population or a per-protocol population, and does not describe how participants were included in the reported summaries. Randomization is listed, but the record does not explain whether analyses followed randomized assignment or how any deviations were handled. The available descriptive results should therefore be read as posted arm summaries, without assuming a particular analysis population or missing-data method.

12. 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.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: fatigue caused by wearing rtls tag (0.0%).

The registry reports no serious adverse events and no other adverse events in either arm: the Tracking System group lists 0 among 9 at risk for each category, and the New Tracking System group lists 0 among 10 at risk for each category. Deaths are also recorded as 0 in each group. “Fatigue caused by wearing RTLS tag” is listed among the top serious-event terms at 0.0%.

These figures describe events recorded in the registry, not proof that no unrecorded or unmeasured events occurred. The record does not explain how adverse events were elicited, what observation period applied, or how the listed term relates to the group-level serious-event counts. The event summaries should be interpreted as registry reporting for this study, rather than as a complete assessment of device safety or tolerability.

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

The record’s main contribution is a small, randomized comparison of nursing activity tracking systems in an ICU, with a stated focus on validating location-derived activity time against manual observation. It posts arm-specific means and standard deviations for the primary continuous outcome, along with the outcome-specific participant counts. The posted values are 34.61 minutes with a standard deviation of 21.19 in the Tracking System arm and 6.63 minutes with a standard deviation of 12.46 in the New Tracking System arm.

Those summaries do not establish how accurately either system reproduced manual observations. The endpoint title and description refer to a time difference, but the record does not specify whether the values are signed or absolute differences, nor does it provide a paired comparison, confidence interval, or test result. The registered expectation of less than 15 minutes during the day shift, up to 12 hours, and maximum error of 5% relative to manual observation cannot be judged from the posted summaries alone.

The results also do not show whether nursing activity metrics correlate with patient outcomes measured by SOFA score. Although that relationship is described as an aim, no secondary outcome results are posted. Nor do the summaries address whether workflow measurement improves care, changes nursing practice, or affects patient outcomes. Those conclusions go beyond the information in the registry record.

More generally, a completed study and a numerical result do not guarantee that a measurement tool has been validated for all settings or uses. The record concerns an ICU at one named hospital, and does not report further sites or settings. It also does not give enough detail to assess the consistency of observer measurement, the representativeness of the measured activities, or the performance of the system across different circumstances. The posted results are best read as descriptive evidence from this registered study, with important questions about comparison, precision, and generalizability unresolved.

14. Why This Trial Matters Statistically

This study raises a practical measurement question with direct statistical implications: whether location data can represent observed nursing activity. Validation depends not just on a difference between group means, but on how well paired measurements correspond across the range of activity times and on how much uncertainty surrounds that assessment. A registry entry that distinguishes the intended reference method from the tracking system helps readers identify the measurement problem being studied.

The record also illustrates why outcome definitions and analysis descriptions need to align. The primary endpoint is framed as a comparison between RTLS and manual observation, and the description names a paired t-test. The posted results, however, are arm-specific summaries for the two tracking-system groups, without a reported paired comparison. Clear reporting of the matching unit, the analysis population, and the estimated discrepancy would make it easier to connect the design to the result.

Another statistical issue is the difference between average accuracy and agreement for individual observations. A mean can describe overall tendency, while variability and the pattern of paired discrepancies determine whether an average is representative of performance in particular observations. The record’s stated target—less than 15 minutes during the day shift, up to 12 hours, and a maximum error of 5% relative to manual observation—makes precision and agreement particularly relevant, but the necessary inferential results are not posted.

Finally, the intended link between activity metrics and SOFA score concerns a different question from measurement validation. Separating the performance of the measurement system from associations with patient outcomes is important for interpretation. With no secondary results or analyses posted, the record leaves those questions open rather than resolving them.

15. 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 record gives limited detail about how activity observations were collected. It does not report the observation schedule, observer training, blinding of observers to tracking-system data, or procedures for resolving disagreements. These details matter because manual observation is the reference method, yet it may itself be subject to measurement variation. The record also does not describe how RTLS data were cleaned, assigned to in-room activity, or matched with observation periods.

The setting is an ICU at the University Hospital, University of Missouri Health Care in Columbia, Missouri. The record does not report whether other units or sites contributed data. Consequently, the posted results cannot establish performance in different clinical environments, workflows, staffing conditions, or system configurations. Participant characteristics and the exact unit of randomization are also not described, limiting assessment of representativeness and the possibility of dependence among observations.

The registry gives no direct estimate of the paired discrepancy, confidence interval, or inferential test result, despite describing a paired t-test. The meaning of the arm-specific endpoint values is not fully specified, and missing-data procedures and analysis populations are not reported. These gaps limit both interpretation of the posted means and evaluation of uncertainty. The record also does not report whether repeated measurements were made or whether their correlation was handled.

Finally, the stated association with SOFA score is not accompanied by posted secondary results. Thus, the registry record does not permit evaluation of that aim or of any relationship between tracking-system performance and patient outcomes. These limitations reflect information not reported in the registry record; they do not establish that the corresponding procedures were or were not carried out.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: RTLS for Healthcare Process Validation — NCT04116957.
• Registry record last updated 2024-08-19; status last verified 2024-07.
• Results first posted 2024-08-19.
• 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.

19. Record Summary

NCT04116957 is a completed randomized, single-masked, parallel study sponsored by the University of Missouri-Columbia. It evaluated an original and upgraded NFER real-time location tracking system for nursing activity in an ICU at the University Hospital, University of Missouri Health Care in Columbia, Missouri. The registry records an enrollment of 19 and lists a start date of 2020-02-17 and primary completion date of 2021-04-16.

The primary endpoint concerns in-room activity time recorded by RTLS compared with manual observation, reported as a mean and standard deviation in minutes over up to 12 hours. The arm summaries are 34.61 with standard deviation 21.19 for Tracking System and 6.63 with standard deviation 12.46 for New Tracking System. The registry describes a paired t-test but posts no analyses or inferential results. It lists three outcome measures and no secondary outcome results. Adverse-event summaries record zero serious events, other events, and deaths in each arm.