This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Assessing a Risk Model for G6PD Deficiency (NCT03337152), a randomized phase 4 study. The primary outcome is Change in Haemoglobin.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03337152 |
| Phase | Phase 4 |
| Condition | Malaria, Vivax, G6PD Deficiency |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 54 |
| Arms in the primary outcome | 2 |
| Lead sponsor | PATH |
| Results posted | Yes |
2. Clinical Question
Population
Volunteers eligible for primaquine treatment under national guidelines, with normal G6PD results by the fluorescent spot test, were studied. The registry describes male participants with wildtype G6PD and female participants who were either homozygous for wildtype G6PD or heterozygous for G6PD deficiency. The study was designed in the context of vivax malaria and G6PD deficiency, with particular interest in assessing haemolytic risk among females heterozygous for G6PD.
Intervention
One arm received primaquine for 14 days at 0.5 mg/Kg. Drug administration was directly observed. This was an active drug regimen rather than a placebo condition.
Comparator
The other arm received chloroquine for 3 days concomitantly with primaquine for 14 days at 0.5 mg/Kg. The registry describes direct observation of drug administration in this arm as well. The comparison is therefore between two regimens that both include primaquine.
Primary question
The primary outcome was change in haemoglobin from baseline through day 28 after enrollment, reported as a median with a full range. The study sought to develop and inform an individual risk model for haemolysis based on individual red blood cell G6PD levels. The posted primary outcome summaries describe the two arms separately; a formal estimate of the difference between regimens is not reported in the registry record.
3. Trial Design
This was a randomized, parallel-group, phase 4 clinical study sponsored by PATH. The registry record lists the study as open-label, with no masking. It was designed around two active treatment regimens: primaquine alone for 14 days, and chloroquine for 3 days given concomitantly with primaquine for 14 days. Both regimens used primaquine at 0.5 mg/Kg. The record describes drug administration as directly observed.
The planned participant groups included males hemizygous for wildtype G6PD, females homozygous for wildtype G6PD, and females heterozygous for G6PD deficiency. The arm descriptions give 12 participants in each of those categories for each arm. For the primary endpoint, the posted outcome population required available baseline data and at least one post-treatment timepoint. The recorded outcome count is 24 in each arm; the registry does not explain how these endpoint counts relate to the participant categories in the arm descriptions.
The study began on 2018-05-07, had a primary completion date of 2018-08-21, and is listed as terminated. The reason for termination is not reported in the registry record. The record gives enrollment as 54. Allocation concealment and the method used to generate the random sequence are not reported in the registry record. Nor is there a protocol-level explanation of the intended allocation ratio or the rationale for the enrollment target.
Follow-up was described as intensive during treatment and for 14 days after treatment, with haematologic measures, G6PD quantification, and drug-level assays. The primary outcome timeframe is stated as 28 days after enrollment. The registry lists five secondary outcome topics, including CYP2D6, drug levels, serious adverse events, reticulocyte count, and urobilinogen levels.
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: 6 participants were enrolled into the trial but had not been randomized at the time of the study halt. Those 6 participants were thus never assigned to a study group.
Overall Study
| Stage | 1A: Primaquine | 1B: Chloroquine + Primaquine |
|---|---|---|
| STARTED | 24 | 24 |
| COMPLETED | 13 | 13 |
| NOT COMPLETED | 11 | 11 |
Reasons for not completing this period:
| Reason | 1A: Primaquine | 1B: Chloroquine + Primaquine |
|---|---|---|
| Physician Decision | 1 | 0 |
| Lost to Follow-up | 1 | 3 |
| Adverse Event | 3 | 1 |
| Stopped drug dosing due to study halt | 6 | 7 |
Analysis population for the primary outcome, as defined in the registry: Among those with available data at baseline and at least one timepoint after study treatment began.
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.
Population: Participants randomized to a treatment arm.
| Characteristic | Unit | 1A: Primaquine | 1B: Chloroquine + Primaquine | Total |
|---|---|---|---|---|
| Participants analysed | 24 | 24 | 48 | |
| Age, Continuous | years | 33.5 (19 to 63) | 38.5 (19 to 64) | 36 (19 to 64) |
| Sex: Female, Male: Female | Participants | 16 | 16 | 32 |
| Sex: Female, Male: Male | Participants | 8 | 8 | 16 |
| Race and Ethnicity Not Collected | Participants | — | — | 0 |
| Region of Enrollment: Thailand | participants | 24 | 24 | 48 |
| G6PD Status: Heterozygous Female | Participants | 5 | 4 | 9 |
| G6PD Status: Homozygous Female | Participants | 11 | 12 | 23 |
| G6PD Status: Hemizygous Male | Participants | 8 | 8 | 16 |
| Hemoglobin | g/dL | 13.2 (10.1 to 17.1) | 12.65 (9.6 to 15.7) | 12.9 (9.6 to 17.1) |
| G6PD Concentration | IU/g Hb | 7.66 (2.94 to 11.08) | 7.515 (3.55 to 12.66) | 7.58 (2.94 to 12.66) |
6. Primary Endpoint
The registry lists the primary outcome as Change in Haemoglobin, measured over 28 days after enrollment.
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | Full Range |
| Unit | g/dL |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is change in haemoglobin from baseline, measured in g/dL. The description refers to haemoglobin during the treatment course and at day 28. The outcome population includes participants with available baseline data and at least one timepoint after study treatment began. The registry reports the arm-level median change and full range, rather than a mean and standard deviation.
Haemoglobin is a measure of the blood’s oxygen-carrying protein. In this study, repeated haemoglobin measurements were relevant to monitoring haematologic changes during exposure to primaquine, in a study concerned with individual risk of haemolysis and G6PD levels. This is a general explanation of the clinical measurement, not a claim about the size or cause of any observed change.
A change score expresses how a participant’s measurement differs from their baseline measurement. The registry’s summary gives a median change for each arm, which describes the middle of the observed changes in the endpoint population. The full range communicates the lowest and highest recorded changes. A range is not a confidence interval: it describes the spread of observed values, not the precision of an estimated treatment contrast. The registry does not report a formal between-arm estimate or its uncertainty for this endpoint.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
For participants included in the primary outcome population, the median haemoglobin change was -1.1 g/dL in the primaquine arm, with a full range from -4.1 to 0.0. In the chloroquine plus primaquine arm, the median was -1.05 g/dL, with a full range from -4.5 to 0.0. These are separate arm-level summaries. The registry record does not report a formal statistical comparison of the arms, so the posted medians and ranges should not be treated as an estimate of the effect of adding chloroquine.
The ranges indicate that observed changes varied within each arm. They do not show how common any particular value was, how the changes were distributed between the minimum and maximum, or whether the difference between arm summaries is statistically distinguishable from chance variation. The values are reported as medians, not as adjusted estimates. No adjustment for participant characteristics or other variables is reported in the registry record.
The endpoint population was restricted to participants with a baseline value and at least one post-treatment measurement. The recorded count for this outcome is 24 in each arm. The registry does not report how many participants lacked endpoint data, why data were unavailable, or whether the observed summaries are sensitive to missing measurements. The enrollment entry and the endpoint counts describe different record elements; the registry does not provide enough detail to reconstruct the path from enrollment to endpoint analysis.
The study’s stated purpose was to develop and inform an individual haemolysis-risk model based on red blood cell G6PD levels. The posted primary result is an arm-level haemoglobin summary, not a reported model, model validation, or analysis of the relationship between an individual’s G6PD level and haemoglobin change. Results for that intended modeling objective are not reported in the registry record.
The safety entries list serious adverse events affecting 3 participants in the primaquine arm and 1 in the chloroquine plus primaquine arm. Other adverse events are recorded for 8 and 17 participants, respectively. No deaths are listed. These counts are descriptive and do not establish that one regimen caused more or fewer events. The record does not provide a formal safety comparison, event-level timing, or a full account of how events were collected and attributed.
Overall, the posted results give descriptive information about haemoglobin changes and adverse events in the recorded arms. They do not provide a precise comparative estimate, a confidence interval, or a reported hypothesis test for the primary outcome. The record also does not provide results that establish the performance of the proposed individual risk model.
9. Statistical Methods Explained
What does a median describe?
The primary outcome uses the median change in haemoglobin. In general, the median is the middle value when observations are ordered. It is a measure of location that is less affected by extreme observations than the arithmetic mean. That can be useful when a distribution is skewed or contains unusually large or small values, although the registry does not state why the median was selected for this endpoint.
The reported median changes are -1.1 g/dL for primaquine and -1.05 g/dL for chloroquine plus primaquine. These values describe the centers of the observed arm-specific distributions. They do not describe every participant’s change, and they cannot by themselves show whether a treatment contrast is statistically credible. A formal contrast would need an estimate of the difference and a method for quantifying its uncertainty. Neither is reported in the registry record.
How should the full range be read?
For the primaquine arm, the full range is -4.1 to 0.0; for the chloroquine plus primaquine arm, it is -4.5 to 0.0. In general, a full range gives the observed minimum and maximum. It is sensitive to the values at the extremes and does not show the distribution between them. Two groups can have similar medians and overlapping ranges while having different distributions, or different medians without that difference being precise. The range is not a confidence interval and should not be read as one.
A confidence interval is a general statistical tool for expressing uncertainty around an estimated quantity. For a treatment comparison, the quantity would commonly be a contrast between arms. The registry record does not report a confidence interval for the primary endpoint contrast, so the precision of such a contrast cannot be assessed from the posted median and range alone.
What is the role of the randomized design?
In general, random allocation is intended to make comparison groups similar, on average, with respect to measured and unmeasured factors at the point of allocation. That supports interpretation of a between-arm contrast, provided allocation and follow-up are implemented and analyzed appropriately. Randomization does not make the groups identical in a particular study, and it does not replace reporting the actual comparison and its uncertainty.
The registry identifies the allocation as randomized, the design as parallel, and masking as none. The random-sequence generation method, concealment procedures, and details of the statistical analysis plan are not reported in the registry record. No formal primary-outcome comparison is posted. Thus, the design label alone is not enough to infer the method used to analyze the observed endpoint or the strength of evidence for a treatment difference.
What can the endpoint counts tell the reader?
The primary endpoint has 24 recorded participants in each arm. The endpoint population is defined as participants with baseline data and at least one post-treatment timepoint. In general, restricting an analysis to available measurements can reduce the information used and can introduce bias if having a measurement is related to the outcome or to treatment. The registry does not describe the extent or reasons for missing measurements, or any method used to address them. It is therefore not possible to assess the direction or size of any resulting bias from the record.
The number listed for an outcome is not, by itself, a complete account of the analysis population. Information about exclusions, timing of measurements, repeated observations, and handling of incomplete records would be needed to reconstruct the analysis. Those details are not reported in the registry record.
What does the sample-size planning illustration mean?
The registry record describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The illustration is framed around detecting a stated fraction of the observed effect. In general, such calculations explore how a future study’s size might relate to a chosen effect and assumptions about statistical testing. They are planning illustrations, not evidence about the probability that the present study detected an effect.
The numerical results of that illustration are not reported in the registry record. The record also does not report the trial’s own sample-size justification or establish that the planning assumptions were part of the original design. The calculation should not be described as observed power. No time-to-event primary outcome is reported; the general note about counting total events for time-to-event calculations does not change the interpretation of the haemoglobin endpoint.
How should the adverse-event summaries be interpreted?
The registry reports participant counts for serious and other adverse events by arm, and lists hemolysis as the top serious adverse-event term at 12.5%. In general, event counts and participant counts answer different questions: a participant can experience more than one event, while a participant count records whether an event category affected a person. The record’s adverse-event entries are presented as counts of affected participants, but it does not give event-level detail, timing, attribution, or a statistical comparison.
Because the record does not describe the event definitions or ascertainment procedures in detail, the counts should be read as registry safety summaries rather than as a complete causal assessment. The listed counts do not establish whether differences between arms are meaningful or attributable to the regimens.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists five secondary outcome topics: significance of CYP2D6, association of drug levels, serious adverse events, significance of reticulocyte count, and significance of urobilinogen levels. It also reports seven outcome measures overall. The remaining outcome measure titles and results are not reported in the registry record. The record does not state which outcomes were confirmatory, how the primary endpoint was prioritized relative to secondary outcomes, or whether a multiplicity adjustment was planned.
When several outcomes or analyses are considered, the chance of at least one apparently notable result can increase. In general, multiplicity procedures can control that risk, but the appropriate approach depends on the analysis plan and the role of each outcome. The statistical plan, any multiplicity strategy, and detailed results for the secondary outcomes are not reported in the registry record. The record lists zero analyses, so no analysis-level results are available to assess across the outcome set.
The primary endpoint population requires baseline data and at least one measurement after treatment began. The registry does not report the number or characteristics of participants who did not meet this requirement, reasons for unavailable measurements, or any imputation or sensitivity analyses. It also does not report whether all randomized participants were analyzed in their assigned arms or whether other analysis populations were used. Without those details, the potential effect of missing data and selection into the endpoint population cannot be evaluated.
Information about protocol deviations, treatment adherence, and any interim analyses is not reported in the registry record. The absence of those details does not establish that such procedures did or did not occur; it limits what can be learned from the posted record.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| 1A: Primaquine | 3 | 24 | 12.5% |
| 1B: Chloroquine + Primaquine | 1 | 24 | 4.2% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| 1A: Primaquine | 0 of 24 |
| 1B: Chloroquine + Primaquine | 0 of 24 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| 1A: Primaquine | 8 of 24 |
| 1B: Chloroquine + Primaquine | 17 of 24 |
Most frequent serious adverse event terms
| Event term | Organ system | 1A: Primaquine | 1B: Chloroquine + Primaquine |
|---|---|---|---|
| Hemolysis | Blood and lymphatic system disorders | 3 of 24 | 1 of 24 |
Most frequent non-serious adverse event terms
| Event term | Organ system | 1A: Primaquine | 1B: Chloroquine + Primaquine |
|---|---|---|---|
| Dizziness | General disorders | 1 of 24 | 11 of 24 |
| Blurred vision | Eye disorders | 1 of 24 | 6 of 24 |
| Decreased appetite | Gastrointestinal disorders | 2 of 24 | 5 of 24 |
| Difficulty sleeping | General disorders | 2 of 24 | 5 of 24 |
| Headache | General disorders | 4 of 24 | 3 of 24 |
| Fatigue | General disorders | 2 of 24 | 4 of 24 |
| Common cold | Respiratory, thoracic and mediastinal disorders | 2 of 24 | 2 of 24 |
| Anemia | Blood and lymphatic system disorders | 3 of 24 | 0 of 24 |
| Diarrhea | Gastrointestinal disorders | 0 of 24 | 3 of 24 |
| Wound | Injury, poisoning and procedural complications | 1 of 24 | 1 of 24 |
Serious events with the highest arm-level rate: hemolysis (12.5%).
The registry records serious adverse events for 3 participants in the primaquine arm and 1 participant in the chloroquine plus primaquine arm. Other adverse events are recorded for 8 and 17 participants, respectively. Deaths are recorded as 0 in each arm. Hemolysis is listed as the top serious adverse-event term at 12.5%.
These entries are descriptive safety counts. They are not a comparative risk estimate, and the registry does not report a statistical test or uncertainty interval for a difference between arms. The serious-event counts do not show the event timing, severity detail, relationship to study treatment, or whether a participant had multiple events. The denominator associated with the listed hemolysis percentage is not reported in the registry record.
Other adverse events and serious adverse events are separate categories in the registry summary. The counts should not be combined, and the record does not clarify the extent of any overlap between participants counted in the two categories. They should therefore be read as reported categories rather than as a complete, mutually exclusive accounting of participants or events.
12. What the Result Does, and Does Not, Mean
The results describe recorded haemoglobin changes and adverse events in two randomized treatment arms. The median change was -1.1 g/dL with primaquine and -1.05 g/dL with chloroquine plus primaquine; the full ranges were -4.1 to 0.0 and -4.5 to 0.0, respectively. These summaries show the center and observed span of the endpoint values in each arm, not a formal estimate of the effect of one regimen relative to the other.
The registry does not report a comparative analysis of the primary outcome, a confidence interval, or a p-value. It is not possible from the posted summaries alone to determine the precision of a between-arm difference or whether a statistical analysis found evidence of a difference. The fact that the medians are numerically close does not establish equivalence, and the ranges do not demonstrate equivalence either.
The study was intended to develop and inform an individual risk model based on red blood cell G6PD levels. The registry record does not report a completed risk model, its predictive performance, or a validation analysis. It also does not report detailed results for the listed secondary outcome topics. The haemoglobin summaries therefore should not be interpreted as a direct test of the model’s ability to predict individual haemolytic risk.
The safety counts provide a record of reported adverse events in the two arms, including serious events and deaths. They do not by themselves show that either regimen caused a particular event or that one regimen is safer than the other. The study status is listed as terminated, but the reason is not reported in the registry record. Taken together, the posted data support a descriptive account of this study’s recorded outcomes, while leaving important questions about comparative effects, model performance, missing data, and safety attribution unanswered.
13. Why This Trial Matters Statistically
This trial is statistically interesting because its stated aim goes beyond a simple comparison of average treatment outcomes. It sought to inform an individual risk model for haemolysis using red blood cell G6PD levels, in a population that included females heterozygous for G6PD deficiency as well as other G6PD groups. That aim raises questions about individual-level prediction, subgroup representation, and the relationship between a biological measurement and a clinical outcome.
The posted primary result, however, is an arm-level median and range for change in haemoglobin. Those summaries can describe the observed outcome distributions at a basic level, but they do not show how individual G6PD values relate to haemoglobin changes. Nor do they establish how well a model would classify or predict risk. A model-development objective generally requires a clearly specified model and assessment of its predictive performance; those results are not reported in the registry record.
The record also illustrates why endpoint population definitions matter. The primary outcome includes participants with baseline data and at least one post-treatment measurement, while the treatment descriptions give planned participant categories. Without a fuller account of missing measurements and analysis populations, readers cannot determine how closely the endpoint summaries represent all randomized participants.
Finally, randomized treatment assignment and descriptive outcomes are not substitutes for a reported treatment contrast. A direct estimate with an uncertainty measure would be needed to assess the comparative haemoglobin outcome. The record’s limited statistical reporting makes it important to distinguish a clinical study’s stated objective, its posted descriptive summaries, and the conclusions those summaries can support.
14. Limitations
- The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
- Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
The registry record provides limited detail on the conduct and analysis of this study. Although allocation is identified as randomized, the sequence-generation method and allocation concealment are not reported. Masking is listed as none. The record does not report the reason for termination, and it does not explain the relationship between the enrollment entry, the planned participant categories, and the primary endpoint counts.
The primary result is reported using medians and full ranges, without a formal between-arm estimate or uncertainty interval. No adjustment strategy, model for repeated measurements, or detailed statistical analysis plan is reported in the registry record. As a result, the comparative interpretation of the primary outcome is limited to descriptive arm-level summaries.
The endpoint population includes only participants with baseline data and at least one post-treatment timepoint. The record does not describe missingness, exclusions, protocol deviations, or methods used to address incomplete follow-up. The possible effect of these factors on the endpoint summaries cannot be assessed.
The study’s individual risk-model objective is not accompanied by a reported model specification, model performance results, or validation results. Detailed results for the secondary outcomes are also not reported. Safety information is summarized by participant counts, without event-level timing, treatment attribution, or a formal comparative analysis. These limits constrain conclusions about comparative effects, individual prediction, and safety differences.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Assessing a Risk Model for G6PD Deficiency — NCT03337152.
- Registry record last updated 2021-11-12; status last verified 2019-05.
- Results first posted 2021-11-12.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT03337152, titled “Assessing a Risk Model for G6PD Deficiency,” is a phase 4 randomized, open-label, parallel-group study sponsored by PATH. It compared 14 days of primaquine with 3 days of chloroquine given concomitantly with 14 days of primaquine. The registry describes planned groups of males hemizygous for wildtype G6PD, females homozygous for wildtype G6PD, and females heterozygous for G6PD deficiency.
The primary outcome was change in haemoglobin from baseline through day 28 after enrollment, reported as a median and full range. The posted endpoint count is 24 in each arm. The study began on 2018-05-07, had a primary completion date of 2018-08-21, and is listed as terminated. Its termination reason is not reported in the registry record. Arm-level primary outcome summaries and adverse-event counts are available, but a formal comparative analysis and results for the intended individual risk model are not reported.