Osimertinib Study in Indian Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Osimertinib Study in Indian Patients (NCT03853551), a na phase 4 study. The primary outcome is Safety of Osimertinib.

2. Clinical Question

The registry lists enrollment of 60 and reports that the study is completed. It does not report a separate analysis population or explain whether every enrolled participant contributed to each outcome. The primary outcome is expressed as a participant count, rather than as a comparison between randomized groups. Accordingly, the central statistical question is how to describe the event counts and their scope, while recognizing that the record does not establish whether observed events were caused by treatment.

3. Trial Design

This was a prospective, multicenter, phase-IV study of osimertinib in Indian adult patients with metastatic EGFR T790M mutation-positive NSCLC. The record describes a single arm and identifies safety as the study focus. Enrollment is listed as 60. The study started on 2019-04-18, had a primary completion date of 2020-04-15, and is recorded as completed.

Because there was no comparator and no randomized allocation, the design supports a descriptive account of events observed among participants in the study. It cannot separate the treatment experience from the background risk of events in a comparable untreated or differently treated group. Nor can an observed count by itself establish that treatment caused an event. The record lists masking as none; it does not report further details about how adverse events were collected, assessed, or attributed.

The primary safety outcome is described as the number of patients with treatment-emergent adverse events, with participants as the unit and a timeframe of 5 months. The listed outcome description does not specify an operational definition of “treatment-emergent,” severity grading rules, or how repeated events in one participant were handled. The registry record also does not report a protocol rationale for the enrollment target, a sample-size justification, or whether any interim analyses were planned.

Although the registry marks the study completed, completion does not by itself establish that every planned analysis was reported. The record lists one outcome measure and zero analyses. No comparative statistical analysis is reported, which is consistent with the single-group structure. The design should therefore be read as a safety study with descriptive event reporting, rather than as a trial estimating a treatment contrast.

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

Reasons for not completing this period:

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 Safety of Osimertinib, measured over 5 months.

The primary endpoint is titled “Safety of Osimertinib” and is described as the number of patients with treatment-emergent adverse events. Its unit is participants, and its timeframe is 5 months. In practical terms, this endpoint counts participants who experienced an event meeting the registry’s treatment-emergent classification during that stated period. It is a participant count, not a count of every event episode; the record does not explain how multiple events in one participant were represented in this outcome.

Safety outcomes are used to characterize what adverse events were observed during treatment and to identify event types that warrant attention in a clinical study. A participant-based measure can convey how many people experienced an event, while event-level information may describe the kinds of events reported. The registry does not provide the detailed endpoint definition, severity grading criteria, attribution rules, or analysis-population definition needed to interpret those distinctions fully.

The endpoint is not a comparative measure. With no control group, the count does not show whether events occurred more or less often than they would have under another treatment or without treatment. The record’s listed timeframe also matters: the outcome is bounded by 5 months, and it should not be read as a statement about events outside that period. The registry record does not report additional follow-up details for this endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes a single-group, prospective study and lists allocation as NA and masking as none. These design features determine what statistical conclusions can be drawn. There is no randomized treatment contrast to estimate, and no comparator group against which to assess event counts. The appropriate interpretation is descriptive: the results report events recorded in the study arm, not a causal effect of treatment.

What does a participant count measure?

The primary outcome uses participants as its unit. In general, a participant-based adverse-event measure records whether a person experienced an event meeting the outcome definition during the observation period. It differs from an event count, which may count multiple episodes in one person. The registry describes the outcome as the number of patients with treatment-emergent adverse events over 5 months, but does not state how repeated episodes were handled or provide a detailed definition of treatment-emergent.

The adverse-event record separately lists participants at risk and participants affected for serious events and for another event category. Those figures describe the posted counts in their respective categories. The record does not explain the category definitions, whether categories overlap, or whether the same participant can appear in both. Counts from these categories should not be combined, and no total across them can be inferred from the record.

Why does the absence of a comparator matter?

In general, a comparative trial estimates how outcomes differ between study groups, subject to its design and analysis assumptions. This trial has one study arm, so a difference between treatments cannot be estimated. The study can document the observed safety experience, but cannot determine from its own design whether the event pattern differs from what would have occurred with another treatment or without treatment. A single-arm count also cannot by itself distinguish treatment effects from the illness or other factors.

The record’s completed status and the presence of posted event counts do not change this design limitation. A statistical method cannot create a control group that the study did not include. The record lists zero analyses, so it does not report an inferential model or an adjusted estimate for the primary outcome.

How should uncertainty be considered?

In general, a confidence interval communicates uncertainty around an estimate under a specified model and sampling framework. It can show that an observed estimate is compatible with a range of underlying values, rather than being exact. No confidence interval is reported for the primary outcome in the registry record. The observed count is therefore available, but its statistical precision is not quantified there.

More generally, a confidence interval would not answer every question. It would not establish causation in this single-arm design, resolve an unclear outcome definition, or account automatically for participants with missing follow-up. The registry record does not report the missing-data approach, if any, or an analysis-population definition. Those omissions limit assessment of how representative the posted counts are of the intended population.

What is the role of event percentages?

The serious-event listing gives percentages for several event terms. These are registry-reported values, and they should be read as descriptive entries. The record does not include term-specific counts, detailed denominators, or rules for handling participants with multiple event terms. As a result, the entries do not support a reconstruction of event-level totals or a comparison between groups.

In general, percentages can make frequencies easier to interpret when their denominator and counting method are clear. Those details are important because a participant can experience multiple event types, and event terms may be grouped or coded in different ways. When the record does not state the counting rules, the listed percentages should not be treated as a complete or independently verifiable description of event frequency.

Why can over-dispersion matter in other analyses?

In general, event counts can vary more than a simple statistical model expects, a feature often called over-dispersion. If over-dispersion is present, a model that assumes less variability may understate uncertainty and produce intervals or tests that are too confident. That issue is relevant when choosing a model for count outcomes, but the registry record does not report a count model, event-level analysis, or evidence that over-dispersion was assessed. It would not be appropriate to claim that it occurred in this study.

The posted primary outcome is a participant count rather than a reported model-based rate or treatment contrast. There is therefore no basis in the registry record to infer that a particular model, correction, or adjustment was used. A statistical model can address specified features of data, but it cannot compensate for unreported definitions or absent comparison data.

What can planning calculations say?

Sample-size calculations are planning illustrations, not summaries of what a trial observed. In general, such calculations depend on assumptions about the effect to detect, the error rate, the allocation, and the statistical model. The registry record does not report the trial’s own sample-size justification or design assumptions. Any planning illustration based on the observed effect should not be interpreted as observed power or as evidence that the trial was adequately sized. No power conclusion about this study can be drawn from the registry record.

For this study, the key methodological distinction is between describing the observed safety experience and estimating a comparative treatment effect. The record supports the former through event counts and listed terms. It does not report the analysis details needed for inferential claims, and the single-arm design does not support a randomized comparison.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Single Arm")
n      <- c(60)
events <- c(60)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Single Arm  60  60  100.0  94.0  100.0  94.0  100.0  94.1  100.0

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and zero analyses. It does not report a multiplicity strategy, such as how the interpretation of several outcomes or repeated statistical tests would be controlled. The serious-event listing contains several event terms, but the record does not state whether these were prespecified outcomes, selected for reporting under a particular rule, or assessed through formal hypothesis tests. No multiplicity-adjusted inference is reported.

In general, multiplicity arises when many outcomes, groups, time points, or tests are considered. Testing many hypotheses can increase the chance of observing an apparently notable result by chance. Adjustments or a prespecified hierarchy can address some of that issue, but the registry record does not describe such a plan. The event-term percentages should therefore be treated as descriptive entries, not as the results of a set of confirmatory comparisons.

Missing-data methods and analysis populations are also not reported in the registry record. Although enrollment is listed as 60 and the event record lists 60 at risk for each category, the record does not explain whether all enrolled participants completed the relevant follow-up, how incomplete records were handled, or whether any participants were excluded from an analysis. It does not define an intention-to-treat, safety, or other analysis population.

In general, missing safety follow-up can affect descriptive event frequencies if participants with incomplete observation differ from those with complete observation. The direction and size of any such effect cannot be inferred without information about missingness and the analysis approach. No imputation or sensitivity analysis is reported. These gaps limit assessment of the completeness and robustness of the posted safety counts.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: dyspnoea (3.3%), cardio-respiratory arrest (1.7%), cataract (1.7%), deep vein thrombosis (1.7%), mouth ulceration (1.7%).

The registry’s adverse-event record lists 7 participants affected among 60 at risk for serious events, and 51 affected among 60 at risk in its other-event category. It also lists 3 deaths. The record does not state whether deaths were considered related to treatment, whether they were included in the serious-event count, or what circumstances surrounded them. The counts should not be interpreted as evidence of causation.

The listed serious-event terms are Dyspnoea (3.3%), Cardio-respiratory arrest (1.7%), Cataract (1.7%), Deep Vein Thrombosis (1.7%), Mouth Ulceration (1.7%), and Diarrhoea (1.7%). These are term-level entries as reported in the registry. It does not provide event counts for each term or clarify whether a participant could contribute to multiple terms. The percentages should not be added or treated as mutually exclusive categories.

In general, serious adverse events and other adverse events are distinct safety classifications, but the record does not provide the criteria used for these categories. It also does not report severity grades, duration, resolution, treatment attribution, or completeness of follow-up. The figures characterize what is posted, not a complete clinical adjudication of the events.

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

The results provide a descriptive account of adverse events recorded in a completed, single-arm study of osimertinib. They include participant counts for serious events and another event category, deaths, and a listing of serious-event terms with percentages. The primary outcome concerns participants with treatment-emergent adverse events over 5 months. These details can inform a reader about what the registry reports as observed during the study.

They do not establish that osimertinib caused any particular event. In general, adverse events occurring during treatment may have several possible causes, including the underlying disease and other care. Without a comparator group, the study cannot estimate how the event experience would differ under another treatment or no treatment. No randomized treatment contrast is available from this design.

The result also does not establish that a listed event was or was not related to treatment. The registry record does not report attribution rules or the circumstances of the deaths. Nor does it provide detailed information on event severity, duration, resolution, or the handling of recurrent events. The listed percentages are descriptive and lack term-specific counts and an explanation of whether categories overlap.

Precision is not quantified for the primary outcome: the registry record does not report a confidence interval or other uncertainty estimate. The record also does not define the analysis population or explain how incomplete follow-up was handled. Thus, the posted counts should not be read as exact estimates of an underlying event risk in all patients with this condition.

In general, a safety study can help describe observed events and identify issues for further evaluation, while comparative conclusions require suitable comparison data and clearly specified analyses. The registry record supports the descriptive reading, but does not report enough detail to support broader claims about comparative safety, causation, or the frequency of events beyond the study’s stated observation period.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of the difference between observing safety events and estimating comparative effects. Its prospective, multicenter, single-arm phase-IV design records an experience with osimertinib in Indian adult patients with metastatic EGFR T790M mutation-positive NSCLC. The registry reports participant counts and serious-event terms, but does not report a comparator or a formal analysis of treatment differences.

For clinical researchers, the key question is not only which events appear in a registry, but also how they were defined, counted, attributed, and followed. The record gives a primary outcome description and timeframe, yet leaves the operational definition and analysis-population details unreported. Those aspects affect how faithfully a count represents the safety experience and how it can be compared with other evidence.

The record also illustrates why percentages and participant counts need their denominators and counting rules. A participant may experience more than one event type, and the registry does not explain whether its listed serious-event terms are mutually exclusive. Without term-specific counts or a stated coding and analysis approach, the list is best read as descriptive rather than as a complete event distribution.

Finally, an absence of comparative analysis is not evidence of no difference. It reflects the design and the analysis information reported in the registry. A single-arm safety study can contribute observations about events recorded during treatment, but its counts do not establish comparative risk or causation. Understanding this distinction helps prevent descriptive safety reporting from being interpreted as a controlled treatment-effect estimate.

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.

Several limitations specific to this record constrain interpretation. The study has a single arm and no comparator, so it cannot provide a randomized comparison of safety outcomes. Allocation is listed as NA and masking as none. The record does not describe procedures for event adjudication, treatment attribution, severity grading, or outcome coding.

The primary outcome is described as the number of patients with treatment-emergent adverse events over 5 months, but its detailed definition is not reported in the registry record. It is also not clear how repeated events in one participant were handled. The record lists 60 as enrollment and as the at-risk count for each adverse-event category, but does not define the analysis population or state whether follow-up was complete for all participants.

The registry reports 7 participants affected in the serious-event category, 51 in the other-event category, and 3 deaths. It does not explain whether these categories overlap or how deaths relate to the event classifications. The listed serious-event percentages lack term-specific counts and information about whether participants could contribute to several terms. No uncertainty estimates are reported for the primary outcome.

The record lists one outcome measure and zero analyses. It does not report a statistical analysis plan, multiplicity approach, missing-data method, interim analysis plan, or sample-size justification. These omissions prevent a fuller assessment of the planned and completed analyses. The trial is recorded as completed, but that status does not resolve the unreported methodological details or support conclusions beyond the posted descriptive results.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Osimertinib Study in Indian Patients — NCT03853551.
• Registry record last updated 2021-06-03; status last verified 2021-05.
• Results first posted 2021-06-03.
• 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

NCT03853551 is a completed prospective, multicenter phase-IV, single-arm study of osimertinib in Indian adult patients with metastatic EGFR T790M mutation-positive NSCLC. Enrollment is listed as 60. The study started on 2019-04-18, with primary completion on 2020-04-15.

The primary outcome is the number of patients with treatment-emergent adverse events over 5 months. The adverse-event record lists 7 affected among 60 at risk for serious events, 51 affected among 60 at risk in another event category, and 3 deaths. It also lists serious-event terms with percentages. No comparator is present, and the registry lists zero analyses.

The record supports descriptive reporting of the posted safety experience. It does not report a comparative treatment-effect estimate, uncertainty interval, detailed outcome definitions, analysis-population criteria, or missing-data methods. These limitations should guide interpretation of the counts and event-term entries.