Identification of Fatty Liver With Advanced Fibrosis in Type 2 Diabetes Using Simple Fibrosis Scores and Electronic Reminder Messages: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Identification of Fatty Liver With Advanced Fibrosis in Type 2 Diabetes Using Simple Fibrosis Scores and Electronic Reminder Messages (NCT04241575), a randomized na study. The primary outcome is Action on Suspected Advanced Liver Fibrosis.

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel-group, triple-masked controlled trial. It lists the phase as not applicable and records enrollment of 1061. The sponsor is the Chinese University of Hong Kong. The study start is recorded as 2020-05-19, primary completion as 2022-10-31, and status as completed.

The intervention was a package: fibrosis scores were calculated and, for patients with increased scores, an electronic pop-up reminder was entered into the clinical management system. Control patients underwent the same assessments, but score results were not specifically shown to physicians and no reminder was used. This comparison therefore evaluates the combined approach against usual care as described in the registry; it does not separate the contribution of score calculation from the contribution of reminders.

The primary outcome population consisted of patients with abnormal fibrosis scores. For this outcome, the reported denominators are 165 in the intervention group and 131 in the control group. Other outcomes have their own recorded denominators, so these primary-outcome figures should not be treated as the group sizes for every analysis. The registry does not report how the masking was implemented or which roles were masked. It also does not report the original design assumptions for the trial’s sample size.

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: Patients with abnormal fibrosis scores

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 Action on Suspected Advanced Liver Fibrosis, measured over Within 1 year of the baseline visit.

The primary endpoint, Action on Suspected Advanced Liver Fibrosis, counts participants who were referred for specialist care or further liver assessment within 1 year of the baseline visit. Its stated population is patients with abnormal fibrosis scores, and its description frames the result as the proportion of patients with high scores who received one of those follow-up actions.

This endpoint addresses a step between risk identification and diagnostic evaluation. A fibrosis score can flag a patient for further attention, while referral or additional assessment records whether that signal was acted upon. The endpoint is therefore useful for evaluating a screening and reminder process in diabetes care: it focuses on whether physicians took a specified follow-up action after a high score was identified.

The outcome does not itself establish that advanced fibrosis was present, that an assessment was completed, or that a patient’s health changed. Referral and further assessment are combined in the primary definition, so the result does not distinguish between those actions. The registry does not report the specific criteria used to classify an abnormal or high fibrosis score, nor does it provide further detail about which assessments qualified.

7. Results Posted in the Registry

• Intervention Group: 33.3% (55/165) versus 3.1% (4/131) on control; risk difference 30.3 percentage points (95% CI 22.5 to 38.1), risk ratio 10.92 (4.06–29.34); unadjusted two-proportion 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.

Referral for Specialist Care

• 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:
  • Intervention Group: 10.7% (57/533) versus 1.1% (6/528) on control; risk difference 9.6 percentage points (95% CI 6.8 to 12.3), risk ratio 9.41 (4.09–21.64); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Intervention Group than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Inappropriate Referral for Specialist Care

• 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:
  • Intervention Group: 0.3% (1/368) versus 0.5% (2/397) on control; risk difference -0.2 percentage points (95% CI -1.1 to 0.6), risk ratio 0.54 (0.05–5.92); unadjusted two-proportion z-test p = 0.608.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intervention Group and the control. The interval gives the range of effects that remain compatible with the data.

Confirmed Diagnosis of Advanced Liver Fibrosis

• 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:
  • Intervention Group: 2.1% (11/533) versus 0.2% (1/528) on control; risk difference 1.9 percentage points (95% CI 0.6 to 3.1), risk ratio 10.90 (1.41–84.10); unadjusted two-proportion z-test p = 0.004.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Intervention Group than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes describe different points along a pathway from identifying a possible risk to taking action and confirming a diagnosis. Taken together, the posted pattern is consistent with more specialist referral and more confirmed diagnoses recorded in the intervention group, while inappropriate referrals were rare in both groups. This broad direction complements the primary outcome, which focuses on the combined action of referral or further assessment.

These outcomes answer related but distinct questions. A referral records a care decision, whereas a confirmed diagnosis records a later classification. Neither alone establishes the effect on longer-term health. Inappropriate referral provides a separate check on a potential consequence of increasing referral activity, but the small event counts limit how much can be concluded from that comparison. The registry does not supply additional clinical detail about the appropriateness criteria or diagnostic process.

The secondary results carry less inferential weight than a single prespecified primary comparison would because the registry lists several secondary endpoints and reports unadjusted comparisons. When multiple outcomes are tested, the chance of at least one apparently notable result can increase, even if there is no effect on any outcome. No multiplicity strategy is reported, so the secondary p-values should not be interpreted as if they were adjusted for the number of outcomes examined. Their value is mainly in describing whether the findings across related measures are coherent, while recognizing that they are not independent evidence of a definitive effect.

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

• Referral for Specialist Care
• Inappropriate Referral for Specialist Care
• Confirmed Diagnosis of Advanced Liver Fibrosis

9. Statistical Interpretation

10. Statistical Methods Explained

The reported outcomes are binary counts: participants either had a recorded event for a particular endpoint or did not. For the primary endpoint, the event is the combined action of specialist referral or further liver assessment within 1 year of the baseline visit. The basic comparison is between the observed event proportions in the intervention and control groups among the patients included in that endpoint.

What does the risk difference express?

The risk difference subtracts the control-group event proportion from the intervention-group event proportion. It is an absolute measure, expressed on the probability scale. A positive value indicates a higher observed event proportion in the intervention group; a negative value indicates a lower observed event proportion. For the primary endpoint, the posted risk difference is 0.303. This measure can be interpreted directly as a contrast in event probability, without relying on the size of the control-group probability as a multiplier.

How do relative measures differ?

A risk ratio compares the event proportions as a ratio, while an odds ratio compares the odds of an event. Ratios can describe how large one group’s event frequency is relative to the other group’s frequency, but their interpretation depends on the underlying event frequency. When events are sparse in one group, a small number of observations can have substantial influence on a ratio. For that reason, the primary comparison is best understood using the counts, the absolute risk difference, and the relative estimates together rather than relying on a single measure.

What does the p-value tell the reader?

The posted p-value is an unadjusted test result. In general, a p-value describes how unusual data at least as discrepant as those observed would be under a specified null model, assuming that model and the test’s conditions apply. It is not the probability that the null hypothesis is true, and it does not quantify the size or clinical importance of an effect. A small p-value does not correct for possible bias, incomplete follow-up, or selective outcome reporting.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the observed data under the method used to construct it. It displays precision as well as direction: a narrower interval represents less statistical uncertainty than a wider interval, all else equal. Its endpoints are not a guarantee that the true effect lies within that particular interval. The registry does not report the detailed analysis methods used to construct the posted intervals, so additional method-specific assumptions cannot be assessed.

Why distinguish endpoint populations?

The primary endpoint is explicitly restricted to patients with abnormal fibrosis scores. The denominators for the other outcomes differ, and the registry reports distinct denominator sets for those analyses. Results should therefore be read in relation to each endpoint’s own recorded population, rather than assuming that a single analysis population applies throughout. The registry does not explain all reasons for these denominator differences, such as eligibility rules, missing outcome information, or exclusions.

What does the intervention comparison establish?

Random allocation is intended to make comparison groups similar on average at the point of assignment, reducing the influence of confounding in expectation. The intervention combines fibrosis-score calculation with reminder messages, while control care includes the same assessments without specifically displayed scores or reminders. The observed comparison concerns this package, not the independent effect of either component. The registry record identifies the design as randomized and triple masked, but does not provide further details about allocation procedures, masking roles, adherence, or implementation fidelity.

11. Confidence Intervals and Precision

The primary risk-difference estimate is 0.303, with a 95% confidence interval from 0.225 to 0.381. The interval lies above zero, which is consistent with a higher recorded event proportion in the intervention group under the comparison used. Its width still matters: the estimate is not exact, and values across the interval represent a range of effects compatible with the data and the method’s assumptions.

The relative estimates have broader intervals: the risk ratio is 10.92, with a 95% confidence interval from 4.061 to 29.34, and the odds ratio is 15.88, with a 95% confidence interval from 5.574 to 45.21. The small number of control-group events makes relative estimates sensitive to sparse data. The event counts, 55 of 165 and 4 of 131, help explain why the relative comparisons should not be read as precise multipliers.

For inappropriate referral, the risk-difference interval ranges from -0.011 to 0.006 and includes zero; the unadjusted p-value is 0.608. For confirmed diagnosis, the risk-difference interval ranges from 0.006 to 0.031; the unadjusted p-value is 0.004. These are secondary comparisons and are not adjusted for multiplicity in the posted results. Confidence intervals describe sampling uncertainty under their analysis methods, not uncertainty from unreported design, measurement, or missing-data issues.

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 participants or total events. They are illustrations based on the observed effect, rather than a reconstruction of the trial’s original design. The calculation assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row considers an effect to detect expressed as a stated fraction of the observed effect.

The 80% power column represents the number of participants per arm used in a planning calculation intended to give an 80% chance of detecting the specified effect under the calculation assumptions, if that effect were the true effect and the assumptions held. The 90% power column expresses the corresponding planning target with a higher chance of detection. These are design calculations, not measures of how much power the completed trial had. This is not observed power.

Smaller target effects require more participants per arm because a smaller difference is harder to distinguish from sampling variation. The figures are therefore sensitive to the effect chosen for planning. Since they are tied to fractions of the observed effect, they should not be treated as independent confirmation that the observed effect is stable or likely to recur. The original sample-size justification and design assumptions are not reported in the registry record. The planning figures do not account for any additional design features unless those are included in the stated assumptions.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.031; treatment proportion p₁ = 0.333; average p̄ = 0.182.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5456 = 1.0693.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5018 = 0.4223.
4. n = (1.0693 + 0.4223)² / (0.3028)² = 24.3, rounded up to 25 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
p0 <- 0.0305343511450382                          # control proportion
p1 <- 0.333333333333333                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(4241575)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 25 (80% power), 32 (90% power)
 75% of observed effect: n per arm = 37 (80% power), 49 (90% power)
 50% of observed effect: n per arm = 64 (80% power), 85 (90% power)
power.prop.test: 24.3 
Simulated power at n = 25 per arm: 0.884 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 three secondary outcomes in addition to the primary outcome and reports unadjusted comparisons. It does not report a multiplicity adjustment, a hierarchy for testing outcomes, or a rule for interpreting secondary p-values. When several outcomes are examined, the chance of at least one low p-value under a collection of null effects can increase. Without a stated approach to multiplicity, isolated secondary findings should be interpreted cautiously and in context rather than as confirmatory results.

The registry record does not report the statistical analysis plan, including whether the primary outcome was analyzed according to an intention-to-treat population or another analysis population. It also does not report how missing outcome data were handled, whether participants with missing records were excluded, or whether sensitivity analyses were conducted. The different outcome denominators indicate that the analyzed populations are not identical across all outcomes, but the reasons are not reported in the registry record.

Details of allocation concealment, the roles covered by triple masking, protocol deviations, adherence to the intervention, and handling of potential clustering or site effects are not reported in the registry record. The record reports zero analyses in its analysis listing, while posting outcome comparisons; no further inferential models or adjusted estimates are described. These omissions do not establish that such procedures were absent, but they limit assessment of how the estimates were produced and how robust they are to alternative analytic choices.

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 adverse-event record lists 0 serious adverse events among 533 participants at risk in the intervention group and 0 among 528 at risk in the control group. The record also lists 9 deaths in the intervention group and 3 in the control group. These figures are counts as recorded; they do not establish that a death was caused by the intervention or that it was related to a serious adverse event.

The registry does not report the definitions, observation period, attribution process, or further details for these safety outcomes. Zero recorded serious adverse events should not be read as proof that no adverse events of any kind occurred, nor can the death counts be interpreted as comparative risks without further context. The posted safety information is limited to the event counts and the recorded groups at risk.

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

The primary result supports a narrow interpretation: among patients with abnormal fibrosis scores included in the endpoint comparison, more participants in the intervention group had a recorded referral or further liver assessment within 1 year of the baseline visit. The estimated risk difference is 0.303, and its confidence interval is above zero. This is a comparison of documented follow-up action under the trial’s intervention package and control approach.

It does not show that every patient with a high score had advanced fibrosis, that all referrals led to completed specialist assessments, or that diagnoses resulted in improved health. The primary outcome combines referral and further assessment, so their separate contributions cannot be determined from this endpoint. The secondary diagnosis result offers additional information about confirmed diagnoses, but it is based on a separate outcome with its own denominator and an unadjusted comparison.

The intervention combines score calculation and electronic reminders. The trial result cannot isolate whether the scores, the reminders, their combination, or other elements of implementation account for the observed pattern. The record also does not establish how well the same approach would work in settings with different patient populations, electronic systems, staffing, or referral pathways.

Statistical significance is not equivalent to clinical importance, and a confidence interval does not remove uncertainty about selection, measurement, missing data, or analysis choices. The strongest conclusion is about the recorded care-process outcome in this randomized trial. Longer-term effects and broader applicability are not reported in the registry record.

17. Why This Trial Matters Statistically

This trial is statistically informative because it separates a screening signal from the clinical action that may follow it. The primary endpoint measures whether patients with abnormal scores received referral or further assessment, making the outcome relevant to the implementation pathway rather than merely to score calculation. The randomized comparison provides a framework for comparing that process with control care as described in the registry.

The result also illustrates why absolute and relative measures should be considered together. A risk difference communicates the contrast in event probability, while risk ratios and odds ratios describe relative contrasts. With few events in the control group, relative estimates can be large yet imprecise; the counts and confidence intervals are essential context rather than optional details.

The secondary measures add a view of what happened after the initial care-process outcome, including referral and recorded diagnosis. Their unadjusted status illustrates a common interpretive issue: multiple related outcomes can help describe a pathway, but they do not all carry the same evidential weight as the primary comparison. The registry’s limited reporting of analysis populations, missing-data handling, and multiplicity further shows why a statistically striking result is not a substitute for a transparent analysis plan.

Overall, the record helps clarify what a randomized trial can say about a clinical workflow and what remains unresolved about downstream outcomes and implementation beyond the trial.

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 primary endpoint is a process measure, and its definition combines specialist referral with further liver assessment. The posted result does not separate those actions or establish that an assessment was completed. The registry does not report detailed criteria for abnormal scores, referral appropriateness, or confirmation of advanced fibrosis, which limits interpretation of how the outcome classifications were made.

The intervention is a combined strategy of score calculation and electronic reminders. Its components cannot be evaluated separately from this comparison. The registry does not report the degree to which clinicians saw or acted on reminders, or whether implementation varied across clinical settings. Generalizability to other healthcare systems and patient groups cannot be assessed from the registry record alone.

Several outcomes have sparse event counts, particularly for inappropriate referral and confirmed diagnosis. Their estimates can therefore be unstable, and secondary comparisons are unadjusted. The registry does not report missing-data methods, analysis-population definitions, adjusted models, sensitivity analyses, or the statistical analysis plan. It also does not provide enough detail to assess allocation concealment or the practical implementation of masking. These limitations concern the information available in the record and do not establish that the trial did not address them in other materials.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Identification of Fatty Liver With Advanced Fibrosis in Type 2 Diabetes Using Simple Fibrosis Scores and Electronic Reminder Messages — NCT04241575.
• Registry record last updated 2024-08-09; status last verified 2024-08.
• Results first posted 2024-08-09.
• 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

NCT04241575 is a completed randomized, parallel-group, triple-masked trial sponsored by the Chinese University of Hong Kong. It evaluated simple fibrosis scores paired with electronic reminder messages against assessments without displayed fibrosis scores or reminder messages in patients with type 2 diabetes.

The primary endpoint was referral for specialist care or further liver assessment among patients with abnormal fibrosis scores within 1 year of the baseline visit. The posted comparison favored the intervention group for this recorded care-process outcome, with a risk difference of 0.303 and a 95% confidence interval from 0.225 to 0.381. Secondary outcomes address referral, inappropriate referral, and confirmed diagnosis. The registry does not report the original sample-size assumptions, detailed statistical analysis plan, multiplicity approach, or missing-data handling.