GER Poses a Potential Risk for Late Complications of BPD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for GER Poses a Potential Risk for Late Complications of BPD (NCT03014453), a registered phase not stated study. The primary outcome is the Late Complications of BPD Infants.

2. Clinical Question

The study is described as a prospective observation of infants with BPD. Its stated motivation is that respiratory symptoms, slower weight gain, and other complications may persist after treatment during infancy. The listed outcome description includes respiratory symptoms and support, vomiting during feeding, hypoxic ischemic injury, retinopathy of prematurity, rehospitalization, and sudden death. The record does not report a separate effect estimate for any of these components or for an overall complication outcome.

The planned follow-up is described in more than one way: the summary mentions monthly follow-up for 18 months after discharge, while the primary outcome description says questionnaires were used at 3, 6, 9, and 12 months corrected for premature age. The exact relationship between those descriptions is not reported in the registry record. This matters when interpreting what period the recorded outcome represents.

3. Trial Design

The registry identifies this as a prospective study of GER and late complications of BPD. Participants were observed rather than assigned at random to GER exposure groups. The allocation field is not stated, and the record explicitly cautions that the listed arms are not a like-for-like control comparison. Consequently, any association between GER and later complications would need to be interpreted as an observational association, not as an effect established by randomized treatment assignment.

The enrollment is 187. The study began on 2017-01-09, and primary completion is dated 2018-06-30; the registry status is COMPLETED. Shengjing Hospital is listed as the sponsor. The summary describes follow-up after discharge, with late complications assessed through follow-up questionnaires. It refers to monthly follow-up for 18 months, while the outcome description specifies questionnaire assessments at 3, 6, 9, and 12 months corrected for premature age. How these follow-up schedules relate to one another is not reported in the registry record.

The study summary says logistic regression was used to analyze the prevalence and risk factors of late complications. The registry does not include a primary analysis record, and it reports zero analyses. It therefore does not give the model specification, the number included in any model, covariates, handling of repeated observations, or an estimate of the association between GER and complications. The study’s stated objective and the results available in the record are distinct: the objective describes the question, but the record does not provide results that resolve it.

The adverse-event listing describes one group titled “Extremely Premature Infants With Bronchopulmonary Dysplasia.” It reports 116 at risk in each of two separately listed adverse-event columns, with 64 and 69 affected, respectively. The record does not explain how those columns relate to the study’s GER exposure or to the primary endpoint. They should not be treated as a randomized comparison.

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 the Late Complications of BPD Infants, measured over 18 months.

The primary outcome is titled “the Late Complications of BPD Infants,” and its unit is participants. It is described as a count of participants with an event. The listed complications include respiratory symptoms and respiratory support, respiratory medication use, cough without a cold at least once per week, rehospitalization due to respiratory diseases, vomiting when feeding, hypoxic ischemic injury, retinopathy of prematurity, rehospitalization, and sudden death.

Combining several complications into an outcome can provide a broad description of late morbidity, particularly when the study question concerns a range of possible consequences rather than one diagnosis. But a count of participants with an event does not, by itself, show which component occurred, how often it occurred, or whether multiple components occurred in the same participant. The registry does not report component-specific primary outcome counts or explain how overlapping events were handled.

The outcome description says questionnaires were used at 3, 6, 9, and 12 months corrected for premature age. The listed timeframe is 18 months, and the summary describes monthly follow-up for 18 months after discharge. The exact endpoint window and how information from the different follow-up descriptions was combined are not reported in the registry record. No primary outcome event count or comparison between GER groups is reported.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The study summary names logistic regression as the method used to analyze late complications and potential risk factors. The registry does not report the model, its coefficients, or its results. The explanation below describes the method generally; it does not imply that any unreported analysis was performed in a particular way.

What does logistic regression estimate?

In general, logistic regression is used when an outcome is binary, such as whether a participant did or did not experience an event. It models the relationship between an exposure and the probability of that outcome. A model can include several explanatory variables so that the association for one variable is adjusted for the others included in the model. The result is commonly expressed as an odds ratio, with a confidence interval and a statistical test. None of these model results is reported in the registry record for the primary question.

An adjusted estimate is conditional on the variables actually included and on the model assumptions. Adjustment does not guarantee that groups are comparable, because unmeasured factors cannot be controlled through variables that were not recorded or included. In an observational study, the exposure groups are not formed by random assignment; differences between them may therefore have more than one possible explanation. The registry does not identify the adjustment variables or describe how confounding was handled.

Why does outcome definition matter?

The primary outcome is described as participants with late complications and lists several distinct complications. General statistical practice requires clarity about whether a participant is counted once if any listed event occurs, whether each complication is analyzed separately, and how repeated events are treated. The record does not report those details. When outcomes are combined, the overall result can conceal different patterns across components; a broad composite may be driven by some components more than others.

The outcome description refers to questionnaires at 3, 6, 9, and 12 months corrected for premature age. The study summary also describes monthly follow-up for 18 months after discharge, and the outcome timeframe is 18 months. These descriptions do not by themselves specify the exact analytic window, how partial follow-up was treated, or how observations from separate visits contributed to the participant-level count. Such details are not reported in the registry record.

How should a confidence interval be read?

In general, a confidence interval communicates the precision of an estimated association by showing a range of values compatible with the data and the statistical method. A narrower interval indicates greater precision than a wider interval, while an interval spanning values on both sides of a null association may leave the direction or presence of an association uncertain. A p-value addresses compatibility with a specified null model, but it does not measure the size or clinical importance of an association. The registry does not report a confidence interval or p-value for GER and the primary outcome.

These summaries also depend on the analysis being appropriate to the data. For example, repeated measurements can require methods that account for within-participant dependence rather than treating every observation as independent. This is a general consideration, not a claim about the method used in this study. The registry does not state whether the analysis used participant-level outcomes, repeated observations, or another approach.

What do the planning assumptions mean?

The registry describes a planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It says the effect to detect is a stated fraction of the observed effect. These are assumptions for an illustrative planning calculation, not evidence about the design choices made for this cohort. The record does not report the study’s own sample-size justification, and the planning assumptions should not be interpreted as observed power.

For a general planning exercise, power is the chance that a specified statistical procedure will reject a null hypothesis when a particular effect is true under the assumed design. A planning calculation depends on its assumed effect and analysis conditions; it does not establish the probability that a reported result is true. The registry does not give a numerical planning result in the record, nor does it provide a primary effect estimate from which readers could assess the study’s observed precision.

Why does over-dispersion matter in general?

Over-dispersion occurs when observed variability is greater than a statistical model expects. In settings where it applies, ignoring it can make standard errors too small and uncertainty appear narrower than warranted. This is a general concept and does not imply that the study used a count model or encountered over-dispersion. The registry describes a participant count outcome and logistic regression in its summary, but does not report an over-dispersion assessment or model diagnostics.

10. Multiplicity, Missing Data and Analysis Populations

The registry reports one outcome measure and zero analysis records. It does not report how many statistical tests were conducted in the study, whether analyses were prespecified, or whether any adjustment for multiple comparisons was used. When many outcomes, subgroups, or models are tested, the chance of at least one apparently unusual result can increase; this is a general reason that multiplicity methods and transparent reporting can matter. It is not possible to determine from this record whether that issue arose in the study.

Missing-data methods are not reported in the registry record. In particular, the record does not give the number of participants with complete primary outcome follow-up, the extent or pattern of missing questionnaire data, or how missing outcomes were handled. Follow-up is described both as monthly for 18 months after discharge and as questionnaire assessments at 3, 6, 9, and 12 months corrected for premature age. The record does not explain whether every participant was expected to contribute data at each time or how incomplete follow-up affected the analysis.

The analysis population is also not described. Although enrollment is 187, the registry does not state how many participants were included in the primary analysis, whether all enrolled participants were analyzed, or whether exclusions were made. These omissions prevent assessment of the relationship between enrollment and the reported outcome figures.

In general, missing outcomes can bias an estimate when the likelihood of missingness is related to participants’ health or exposure. Statistical methods can address some missing-data settings under assumptions, but those assumptions require explanation and cannot be evaluated from the enrollment count alone. The registry does not report a missing-data strategy or sensitivity analysis. As a result, the potential influence of missing follow-up on the primary question is not reported in the registry record.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: re-hospitalization (26.7%), retinopathy of prematurity (25.9%), pneumonia (8.6%), dyspnea (8.6%), cyanosis (8.6%).

The adverse-event listing identifies a group titled “Extremely Premature Infants With Bronchopulmonary Dysplasia.” It reports 116 at risk in each of two separately listed columns, with 64 and 69 affected. The record does not explain the meaning of the two columns or connect them to GER exposure, so they should not be interpreted as a direct comparison of GER groups. The listing also reports one death; its timing, cause, and relationship to the study outcome are not stated.

The serious-event terms and percentages listed are Re-hospitalization at 26.7%, retinopathy of prematurity at 25.9%, pneumonia at 8.6%, dyspnea at 8.6%, cyanosis at 8.6%, and intermittent apnea at 4.3%. The registry does not report the denominator or clarify whether each percentage refers to participants or events. These terms describe events reported in the record; they do not establish that GER caused them or that they belong to the primary endpoint.

Serious adverse-event summaries should be read separately from an efficacy or risk-factor analysis unless the registry explicitly links them. Here, the primary comparison and the relationship between adverse-event columns and study exposure are not reported in the registry record. The figures therefore do not answer whether GER is associated with late complications.

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

The study asks whether GER may be a risk factor for late complications among infants with BPD, and the registry summary describes a prospective cohort and logistic regression as the analytic approach. The record also provides an outcome description and adverse-event listings. These details explain the intended question and some reported event information, but they are not an estimate of the association between GER and complications.

No primary outcome comparison, logistic-regression coefficient, confidence interval, or p-value is reported in the registry record. There is therefore no reported estimate of how much more or less often complications occurred according to GER status, and no reported measure of the precision of such an estimate. The title’s wording that GER poses a “potential risk” should be read as the study’s framing, not as a finding established by the results recorded.

Even a reported association from an observational cohort would not by itself establish causality. In general, exposure groups can differ in ways that affect outcomes, and statistical adjustment can address only measured variables included in an appropriate model. Here, the model details and adjustment variables are not reported. The absence of those details, together with the absence of the primary analysis result, prevents assessment of the strength or robustness of the proposed relationship.

The outcome itself includes a range of complications, and the record does not provide component-specific primary results. It also gives different descriptions of the follow-up schedule: monthly follow-up for 18 months after discharge in the study summary, and questionnaires at 3, 6, 9, and 12 months corrected for premature age in the outcome description. The exact endpoint construction and treatment of incomplete follow-up are not reported.

Accordingly, the registry record supports describing the study’s question, design, and listed event information. It does not support a conclusion about whether GER is associated with the primary outcome, nor about the direction, size, or clinical meaning of any such association.

13. Why This Trial Matters Statistically

This study is statistically interesting because it frames GER as a possible risk factor in a prospective cohort of infants with BPD, while the central question depends on a clearly defined exposure, outcome, and follow-up period. The registry identifies the broad endpoint and names logistic regression in its summary. Those elements make the intended risk-factor analysis understandable even though the primary result is not reported.

The record also illustrates why a study’s objective and its posted results should be distinguished. A stated hypothesis, a named statistical method, and an enrollment count do not show the size or precision of an association. For interpretation, readers need the outcome counts by exposure group, an analysis estimate, uncertainty around that estimate, and enough information about adjustment and missing data to understand what the estimate represents. Those details are not reported in the registry record.

Endpoint definition is especially important when the outcome combines respiratory symptoms, rehospitalization, feeding-related vomiting, other diagnoses, and death. General statistical reasoning indicates that a composite can offer a broad summary but may obscure differences among its components. The record’s varying descriptions of follow-up also make it important to know how outcome timing was handled. Without those details, even a future comparison would be harder to interpret consistently.

The adverse-event listing supplies additional descriptive information, but its two columns are not explained in relation to GER status or the primary outcome. Treating them as a direct exposure comparison would go beyond what the registry reports. The record is therefore useful for identifying the research question and reporting gaps, while it does not provide the analysis needed to judge the proposed association.

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 main limitation for interpretation is that the registry does not report a primary analysis result. It gives no outcome comparison by GER status, effect estimate, confidence interval, or p-value. The stated use of logistic regression cannot be evaluated without the model specification and results.

The exposure definition and comparison groups are not described in the primary outcome record. The allocation field is not stated, and the registry cautions that the listed arms are not a like-for-like control comparison. The relationship between the two adverse-event columns and GER exposure is also not explained.

The endpoint description lists multiple complications, but does not report component-specific counts or explain how overlapping events were handled. Follow-up is described as monthly for 18 months after discharge in the summary, while the outcome description specifies questionnaires at 3, 6, 9, and 12 months corrected for premature age. The exact analytic window and endpoint construction are not reported in the registry record.

Although enrollment is 187, the record does not state the number with primary outcome data or the number included in any analysis. Missing-data handling, the analysis population, model covariates, and sensitivity analyses are not reported. These gaps limit assessment of selection, confounding, precision, and robustness. The sample-size assumptions describe a planning illustration, not the study’s own design justification or observed power. Overall, the available registry information is insufficient to assess the magnitude or credibility of the proposed GER association.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: GER Poses a Potential Risk for Late Complications of BPD — NCT03014453.
• Registry record last updated 2020-03-26; status last verified 2020-03.
• Results first posted 2020-03-26.
• 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.

18. Record Summary

NCT03014453 is a completed prospective study sponsored by Shengjing Hospital, titled “GER Poses a Potential Risk for Late Complications of BPD.” It reports enrollment of 187 and lists Gastroesophageal Reflux and Bronchopulmonary Dysplasia as conditions. The study began on 2017-01-09, with primary completion on 2018-06-30.

The stated question is whether GER is related to late complications among infants with BPD. The outcome description lists several complications and describes questionnaire assessments at 3, 6, 9, and 12 months corrected for premature age; the summary also describes monthly follow-up for 18 months after discharge. Logistic regression is named in the summary, but the registry reports zero analyses and no primary-effect estimate.

Adverse-event terms and affected counts are listed, but their relation to GER exposure is not explained. The registry therefore records the study’s purpose and some descriptive information without reporting the analysis needed to answer its primary question.