High Tech and High Touch (HT2): Transforming Patient Engagement Through Portal Technology at the Bedside: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for High Tech and High Touch (HT2): Transforming Patient Engagement Through Portal Technology at the Bedside (NCT02943109), a randomized na study. The primary outcome is Inpatient Portal Use Frequency.

2. Clinical Question

The study is registered as a randomized, single-masked factorial trial. The factorial structure is reflected in the combinations of technology version and training approach. The registry does not identify who was masked, describe the allocation procedure, or report a formal analysis of the separate technology and training effects or their interaction. The primary comparisons therefore address the posted combination-versus-reference contrasts, not every possible question about the factorial design.

3. Trial Design

The registry describes a randomized, single-masked factorial study of inpatient portal technology and training. It lists Ohio State University as the sponsor, a total enrollment of 3782, and a completed status. The recorded start is 2016-12 and the primary completion is 2020-02. The phase is listed as NA. The study was embedded in a larger mixed methods project examining patient experiences and outcomes as well as subsequent ambulatory portal use.

The four arm descriptions cross two versions of MyChart Bedside with two training approaches. High Tech, High Touch combined the full portal version with support from a technology navigator. Low Tech, High Touch combined the non-interactional version with navigator support. High Tech, Low Touch used the full version with online training only, and Low Tech, Low Touch used the non-interactional version with online training only. These descriptions identify the assigned approaches, but do not specify the amount of portal use participants were expected to complete.

The primary endpoint was recorded at hospital discharge and concerns sessions during the admission associated with enrollment. The registry does not report the randomization sequence, concealment procedures, masking details, adherence measures, or protocol deviations. It also does not report the original design assumptions or a sample-size justification. The posted results give arm-specific outcome summaries and unadjusted contrasts against High Tech, High Touch; they do not report a factorial model estimating main effects or interaction.

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: Of the 3,782 enrolled participants, 2,892 met the study inclusion criteria and were randomized to one of the four study arms. Of those excluded: * 734 had an enrollment admission \< 3 days; * 17 had no inpatient portal use on record; * 139 switched Tech assignment.

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 Inpatient Portal Use Frequency, measured over At hospital discharge..

The primary endpoint is the mean count of MyChart Bedside sessions for the admission associated with study enrollment, measured at hospital discharge. It represents frequency of inpatient portal use, not a measure of the quality of each session, the amount of time spent in the portal, or whether a patient found the portal useful. The outcome is a count, but the registry reports it as a mean with a standard deviation.

Session frequency is relevant to the study because the intervention combines access to different versions of an inpatient portal with different forms of training. The measure can describe whether use differed across those assigned combinations during the enrolled admission. It does not by itself establish why use differed, whether use changed patient experience, or whether it affected later outpatient portal use or other health outcomes.

The time point matters to interpretation: this endpoint concerns the hospitalization and is assessed at discharge. It should not be read as a measure of continued portal use after discharge. The record does not give a separate definition of what counted as a session beyond identifying the outcome as the count of MyChart Bedside sessions. It also does not state whether the reported mean was adjusted for other participant or admission characteristics.

7. Results Posted in the Registry

• Low Tech, High Touch: mean 21.00 versus 29.00 on control; difference -8.00 (95% CI -10.37 to -5.63), standardized effect d = -0.29; z-test p = <0.001.
• High Tech, Low Touch: mean 29.00 versus 29.00 on control; difference 0.00 (95% CI -2.91 to 2.91), standardized effect d = 0.00; z-test p = 1.000.
• Low Tech, Low Touch: mean 20.00 versus 29.00 on control; difference -9.00 (95% CI -12.52 to -5.48), standardized effect d = -0.30; 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.

Outpatient Portal Adoption

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Low Tech, High Touch: 22.8% (66/290) versus 31.7% (218/688) on control; risk difference -8.9 percentage points (95% CI -14.9 to -3.0), risk ratio 0.72 (0.57–0.91); unadjusted two-proportion z-test p = 0.005.
  • High Tech, Low Touch: 24.8% (126/508) versus 31.7% (218/688) on control; risk difference -6.9 percentage points (95% CI -12.0 to -1.8), risk ratio 0.78 (0.65–0.94); unadjusted two-proportion z-test p = 0.009.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Low Tech, High Touch than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for High Tech, Low Touch than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Self-Efficacy

• 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:
  • Low Tech, High Touch: mean 4.30 versus 4.30 on control; difference 0.00 (95% CI -0.07 to 0.07), standardized effect d = 0.00; z-test p = 1.000.
  • High Tech, Low Touch: mean 4.30 versus 4.30 on control; difference 0.00 (95% CI -0.06 to 0.06), standardized effect d = 0.00; z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Low Tech, High Touch and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between High Tech, Low Touch and the control. The interval gives the range of effects that remain compatible with the data.

The secondary results cover different stages and concepts than the primary endpoint. Outpatient Portal Adoption concerns whether participants adopted an outpatient portal at 3 months post-discharge. Self-Efficacy is reported at baseline as a continuous score. The registry also lists outpatient portal frequency of use and additional self-efficacy outcomes, but corresponding quantitative results are not reported in the registry record. Thus, the secondary summaries do not provide a complete account of every secondary outcome named in the record.

Across the posted outpatient adoption contrasts, both compared combinations have lower observed adoption than the High Tech, High Touch reference, and the unadjusted intervals and p values indicate that these contrasts are not centered on no difference. This direction differs in outcome type from the primary result: adoption is a binary outpatient measure, while the primary endpoint is a count of inpatient sessions. The findings therefore describe related but distinct aspects of portal engagement, not a single continuous trajectory from inpatient use to later adoption.

The baseline self-efficacy summaries are similar across the reported arms. Since these measurements are labeled baseline, they are not evidence of a post-intervention change or an intervention effect on self-efficacy. No adjusted longitudinal analysis is reported. The primary session outcome and the secondary measures should not be combined into a single claim about patient experience or clinical outcomes.

Unadjusted secondary results carry less inferential weight because the registry lists multiple outcomes and comparisons, while a multiplicity procedure is not reported. When many outcomes or contrasts are examined, some apparently notable results can arise by chance. The absence of reported adjustment does not invalidate the descriptive summaries, but it makes confirmatory interpretation more cautious. In addition, the record does not provide analysis plans, a hierarchy among secondary outcomes, or complete analysis results for all listed secondary measures. The secondary pattern is best treated as descriptive context for the primary portal-use result rather than independent confirmation of a broad treatment effect.

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

• Outpatient Portal Adoption
• Outpatient Portal Frequency of Use
• Self-Efficacy
• Self-Efficacy
• Self-Efficacy

9. Statistical Interpretation

10. Statistical Methods Explained

The results report the primary endpoint as a mean and standard deviation within each arm. A mean gives the arithmetic average session count, while a standard deviation describes how spread out participant-level counts are around that average. A mean difference compares two arm means on the original session-count scale. Negative values in the posted contrasts indicate that the compared arm’s mean is lower than the reference mean; a value of zero indicates equal reported means.

What do the unadjusted contrasts represent?

Each primary contrast compares one intervention combination with High Tech, High Touch. The estimates are unadjusted comparisons based on the posted arm summaries. They do not account for participant characteristics, admission characteristics, or other covariates. The registry record does not report an adjusted analysis, a model for the factorial main effects, or an interaction analysis. Because the design is factorial, it is useful to distinguish these particular contrasts from a model that estimates the average effect of technology across training approaches, or the average effect of training across technology versions.

A confidence interval expresses uncertainty about the estimated difference, under the assumptions of the calculation. Intervals wholly on one side of zero indicate estimates that are less compatible with no mean difference under the interval procedure than intervals that cross zero. The interval does not represent the range of individual patients’ session counts, and it does not describe the spread of outcomes in an arm; the standard deviation addresses that separate feature. The confidence level and interval procedure are not reported in the registry record, so they should not be inferred from the displayed endpoints.

Why does variation matter?

The endpoint is a count, and the standard deviations are large relative to the arm means. This indicates substantial dispersion in use. Count outcomes can be skewed, with many participants having relatively few sessions and some having more; the registry does not report the participant-level distribution, so the actual shape cannot be established from the summaries alone. General statistical point: when a count outcome has greater variability than a simple model assumes, standard errors and uncertainty estimates can be inaccurate unless the analysis accounts for that dispersion. The posted record does not identify a count model or explain how distributional assumptions were assessed.

The reported comparisons are expressed as mean differences, which retain the outcome’s original unit. They are straightforward to interpret as differences in average sessions, but can conceal distributional features such as clustering at zero or a long upper tail. The record does not provide medians, quantiles, proportions with no sessions, or participant-level data. Consequently, the mean and standard deviation are the available description of distribution, and they do not allow a full assessment of whether a different summary or model would be more suitable.

How should the p values be read?

The p values in the primary comparisons are labeled unadjusted. A p value is a measure of how compatible the observed data are with a specified null model, given the analysis assumptions; it is not the probability that the null hypothesis is true, nor a measure of the size or practical importance of a difference. The registry record does not state a significance threshold for the trial’s own analysis or report a multiplicity procedure. The planning illustration separately specifies a two-sided significance level of 0.05, but that is a planning assumption, not evidence of the analysis rule used for the posted primary results.

The study has a factorial allocation, yet the listed contrasts use the high-tech, high-touch combination as a common reference. These comparisons are useful for describing each combination relative to that reference, but they do not isolate the contribution of each factor. For example, a contrast that changes both technology and training cannot attribute a difference to only one of those changes. A factorial analysis could address distinct questions about factor effects and their interaction, but no such results are reported in the registry record.

Random allocation supports comparisons between assigned groups, but it does not eliminate uncertainty, missing outcomes, or departures from the intended analysis. The registry record does not state whether the posted summaries include all randomized participants, only those with available endpoint data, or another analysis population. It also does not report missing-data handling, adherence, or sensitivity analyses. These omissions limit assessment of how the results might depend on exclusions or incomplete outcome measurement.

11. Confidence Intervals and Precision

The intervals around the primary mean differences describe uncertainty in the estimated contrasts, not the range of session counts among individual participants. For Low Tech, High Touch, the estimated difference is -8.0 sessions, with an interval from -10.37 to -5.629. For High Tech, Low Touch, the estimate is 0.0, with an interval from -2.911 to 2.911. For Low Tech, Low Touch, the estimate is -9.0, with an interval from -12.52 to -5.476. The registry record does not specify the confidence level or method used to construct these intervals.

The intervals for the low-tech comparisons are entirely below zero, while the interval for High Tech, Low Touch includes zero. In statistical terms, the posted estimates for the low-tech comparisons are more directionally constrained by their intervals; the interval for High Tech, Low Touch remains compatible with both a lower and a higher mean relative to the reference. This is not proof that the groups are identical in the latter comparison. The interval’s width indicates that uncertainty remains around the point estimate.

The standard deviations—31.0, 18.0, 36.0, and 26.0 sessions across the four arm summaries—show that participant-level use varied considerably. The precision of a mean contrast depends not only on the estimated difference but also on variability and the number of observations contributing to each summary. The registry reports outcome counts by arm, but does not explain whether those counts differ from the randomized population because of missing outcomes or other exclusions. Therefore, precision should be considered alongside the unreported analysis population and missing-data procedures.

Intervals and p values address statistical uncertainty under their assumptions. They do not determine whether a difference is important to patients, whether a particular portal approach should be selected, or whether the result applies to settings beyond the study. The posted intervals also do not account for any multiplicity adjustment, which is not reported.

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 planning figures are participants per arm, not total events. They are illustrations calculated from the observed effect rather than a reconstruction of the trial’s original design. The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning scenario sets the target effect to a stated fraction of the observed effect. The scenarios include planning for 100%, 75%, and 50% of the observed effect.

The 80% power column gives the estimated number of participants per arm for a future two-arm comparison to have an 80% chance of detecting the assumed effect under the planning assumptions, if that effect is real and the assumptions hold. The 90% power column gives the corresponding per-arm requirement for a 90% chance. These are design calculations, not measures of what the completed trial achieved. This is not observed power.

The planning entries are tied to each comparison’s observed effect, so their values should not be treated as a universal sample-size recommendation. Smaller assumed effects require more participants to distinguish them from random variation under otherwise similar assumptions. Equal allocation is part of the illustration; the observed arm sizes in this factorial trial are not equal. The registry record does not report the original sample-size assumptions, the trial’s intended power, or a justification for the enrollment target.

The planning curve should be read as the same relationship expressed across effect assumptions: as the target effect becomes smaller relative to the observed effect, the required participants per arm increase. The calculation is a simplified planning aid. It does not incorporate any unreported clustering, attrition, missingness, factorial modeling, or multiplicity considerations, and should not be read as evidence that the original study was designed using these calculations.

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 = -8.000; pooled standard deviation σ = 28.041.
2. n = 2 × (1.96 + 0.842)² × 28.041² / -8.000² = 192.9, rounded up to 193 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 <- -8                      # observed difference in means
sigma <- 28.040874509809                     # 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(2943109)
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 = 193 (80% power), 259 (90% power)
 75% of observed effect: n per arm = 343 (80% power), 459 (90% power)
 50% of observed effect: n per arm = 772 (80% power), 1,033 (90% power)
power.t.test: 193.8 
Simulated power at n = 193 per arm: 0.800 (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 lists 15 outcome measures and identifies multiple secondary outcome topics, including outpatient portal adoption, outpatient portal frequency of use, and self-efficacy. It reports no analysis count and no multiplicity adjustment. The primary comparisons are labeled unadjusted, and the secondary comparisons are also unadjusted. The record does not identify a hierarchy of hypotheses, a prespecified family of comparisons, or whether any results were designated exploratory.

Multiplicity matters because repeated testing increases the opportunity for chance findings. A correction or prespecified testing strategy can help control that risk, but the registry record does not report whether one was used outside the posted summaries. As a result, the p values should not be treated as if they account for all outcomes and contrasts examined. This caution is particularly relevant to secondary outcomes, which are numerous and not fully represented by posted quantitative summaries.

Missing-data procedures are not reported in the registry record. The record does not say how participants without a recorded session count or follow-up measure were handled, whether missingness differed by arm, or whether imputation or sensitivity analyses were conducted. Arm-specific outcome counts are available, but the record does not explain their relationship to the randomized population. It is therefore not possible to determine from the registry summaries whether the analysis followed an intention-to-treat approach or another population definition.

The registry also does not report the protocol-defined analysis population, adherence criteria, exclusions after randomization, or handling of protocol deviations. These details are important for judging whether comparisons preserve the benefits of random assignment and how representative the analyzed participants are of those assigned. Without them, interpretation must remain focused on the posted arm summaries and unadjusted comparisons, with uncertainty about the analysis choices that produced them.

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

The results describe portal session frequency during the enrolled admission. The High Tech, High Touch arm had a reported mean of 29.0 sessions. The mean was lower in the two low-tech combinations, while the High Tech, Low Touch mean matched the reference mean. The unadjusted contrasts and intervals quantify the differences between these combinations and the reference. They do not establish that the technology version alone caused the observed pattern, because the registry does not report a factorial analysis separating technology and training effects.

The study concerns engagement with a portal, not a direct measure of health status, quality of care, or clinical benefit. More sessions do not necessarily mean a better patient experience, and fewer sessions do not necessarily mean a worse one. The record does not report a validated threshold for meaningful session frequency or connect the primary endpoint to a patient-important outcome. It also does not establish whether differences in inpatient sessions led to later outpatient use.

The posted outpatient adoption results provide a separate observation at 3 months post-discharge, but do not establish that inpatient session frequency caused adoption. The baseline self-efficacy results do not show change after assignment. Further, several listed secondary outcomes lack quantitative results in the registry record. The record therefore supports a limited description of the posted outcomes, not a comprehensive account of all study aims.

Randomization is a strength for comparing assigned groups, but the registry summary does not include enough detail to assess masking, missing data, adherence, or the analysis population. The unadjusted comparisons also need to be read in light of multiple outcomes and the absence of a reported multiplicity strategy. These limitations do not erase the reported differences; they define the boundaries of what can be concluded from them.

The findings should not be interpreted as a recommendation for a particular portal configuration. They are evidence about recorded use within the trial’s reported arms and setting. Applicability to other hospitals, patient groups, or portal systems is not established by the registry summaries alone.

16. Why This Trial Matters Statistically

This trial is statistically interesting because it studies a factorial combination of portal technology and training rather than a single intervention contrasted with usual care. The four assigned combinations make it possible in principle to ask whether the portal version, the training approach, or their combination relates to use. The posted results, however, use contrasts against one reference combination and do not report separate factorial effects. This distinction illustrates why a design’s structure and the analysis actually reported must be considered together.

The outcome also highlights a common challenge in digital health research: engagement is often measured by counts of use. A count can be an informative behavioral outcome, but its mean may not describe the whole distribution, especially when participant use varies widely. The arm standard deviations are substantial, and the record does not provide distributional summaries or the model used to analyze the outcome. Readers therefore need to distinguish the reported average contrast from the diversity of individual use patterns.

The trial further links an inpatient measure with secondary outcomes concerning outpatient portal adoption and self-efficacy. These outcomes reflect different constructs and time points, and their analyses should not be collapsed into a general claim about engagement. The incomplete secondary reporting and absence of a reported multiplicity plan also show why outcome registration and transparent reporting matter for evaluating a broad study.

Finally, the record illustrates the limits of interpreting registry summaries alone. Randomized allocation is reported, but important details about masking, analysis populations, missing data, and factorial modeling are not. A careful statistical reading can identify the observed contrasts while keeping causal, clinical, and generalizability claims within the boundaries of what the record documents.

17. 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 needed to assess the results are not reported in the registry record. These include the randomization procedure, allocation concealment, who was masked, the intended analysis population, missing-data handling, adherence to assigned training, and how deviations from the assigned approach were handled. The record also does not state whether the posted primary outcome summaries include all randomized participants with available data or a selected subset.

The reported primary contrasts compare each available combination with High Tech, High Touch. No factorial main-effect or interaction estimates are reported, so the results do not isolate technology from training. Nor does the registry specify whether the session count was modeled as a count distribution, analyzed using another method, or adjusted for any covariates. The large standard deviations relative to the means make the unavailable distributional details relevant to interpretation.

There is no reported analysis plan, original sample-size justification, or trial-specific significance threshold. The planning illustrations use stated assumptions, but those are not the original design assumptions. The record lists multiple outcomes, gives no multiplicity procedure, and does not include quantitative summaries for all listed secondary outcomes. These omissions make it difficult to assess selective reporting or the role of unreported analyses.

The primary measure is limited to portal sessions during the enrolled admission. The registry does not report a threshold for meaningful use, session duration, or a direct link between session counts and patient experience or clinical outcomes. Finally, the summaries alone do not establish how the findings generalize beyond the study setting or to other portal platforms.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: High Tech and High Touch (HT2): Transforming Patient Engagement Through Portal Technology at the Bedside — NCT02943109.
• Registry record last updated 2024-07-22; status last verified 2024-02.
• Results first posted 2024-07-22.
• 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.

21. Record Summary

NCT02943109 is a completed randomized, single-masked factorial study sponsored by Ohio State University. It enrolled 3782 and examined combinations of MyChart Bedside portal version and training approach. The primary endpoint was the mean number of inpatient portal sessions during the admission associated with study enrollment, assessed at hospital discharge.

The posted primary contrasts use High Tech, High Touch as the reference. The low-tech combinations had lower reported mean session counts, while the High Tech, Low Touch mean matched the reference. Confidence intervals and unadjusted p values are reported for these comparisons. Secondary summaries include outpatient portal adoption at 3 months post-discharge and baseline self-efficacy; other listed secondary results are not reported in the registry record.

The registry does not report the original design assumptions, analysis population, missing-data procedures, multiplicity adjustment, or a factorial analysis separating technology and training effects. The results therefore support a description of the posted arm summaries and unadjusted contrasts, with these limitations in mind.