A Study to Assess Whether PDE5 Inhibitors Increase the Chance of Triggering the Onset of Acute NAION: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Assess Whether PDE5 Inhibitors Increase the Chance of Triggering the Onset of Acute NAION (NCT00759174), a registered phase not stated study. The primary outcome is Number of Days Exposed to PDE5i During 1-Day Case Window and 29 1-Day Control Windows Among Participants Adjudicated as Definite NAION Cases.

2. Clinical Question

3. Trial Design

The study, NCT00759174, is titled “A Study to Assess Whether PDE5 Inhibitors Increase the Chance of Triggering the Onset of Acute NAION.” Its official title describes it as a case-crossover study of PDE5 inhibitor exposure as a potential trigger factor for acute NAION. The brief summary states that its objective was to examine whether use of sildenafil, vardenafil, or tadalafil triggers onset of acute NAION. The registry classifies it as non-interventional and records its status as completed.

A case-crossover approach uses a participant’s own exposure at different times as the basis for comparison. Here, the case window was the 1 day preceding the symptom-onset day, while the 29 control windows were the 29 days preceding the case window. Each participant contributed exposure information for 1 case window and 29 control windows. This structure differs from a randomized parallel-group trial: the registry does not describe assignment to a PDE5 inhibitor or to a control treatment. Allocation is not stated.

The exposure rule accounted for the medication used. A window was classified as exposed if sildenafil or vardenafil was used on that day or the previous day, or if tadalafil was used on that day or any of the previous 4 days. Thus, the measure is based on specified recent-use windows rather than a general record of whether a participant ever used a PDE5 inhibitor.

The registry reports enrollment of 673. The study start is recorded as 2008-11 and primary completion as 2012-10. It lists Optic Neuropathy, Ischemic as the condition and identifies “No intervention.” These details describe the study record, but enrollment alone does not establish how many participants contributed to the primary analysis. The analyzed population is described as definite NAION cases with PDE5 inhibitor exposure, but the eligibility sentence defining the number analyzed is cut off. The number analyzed is not reported in the registry record.

The registry lists 3 outcome measures and 3 analyses, but does not give enough information to characterize the secondary endpoints or their results. It also does not report the statistical model used for the primary analysis. The study’s observational, within-person structure is clear; the precise inferential procedure is not.

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:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set (FAS) included all participants who signed informed consent, were eligible for study, reported exposure to PDE5i within 60 days prior to participant-reported onset of NAION symptoms. N (number of participants analyzed) represents Definite NAION cases with PDE5i exposure on at least 1 day but not all 30 days prior to symptom onset.

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 Number of Days Exposed to PDE5i During 1-Day Case Window and 29 1-Day Control Windows Among Participants Adjudicated as Definite NAION Cases, measured over 30-day period prior to onset of NAION symptoms.

The primary endpoint counts days classified as exposed to a PDE5 inhibitor during the case window and control windows among participants adjudicated as definite NAION cases. The stated timeframe is the 30-day period before onset of NAION symptoms. The case window is the 1 day preceding the symptom-onset day, and the control windows are the 29 days before the case window. Each participant contributes exposure information for both the case and control periods.

The exposure definition reflects the drugs’ specified recent-use windows in the registry: sildenafil or vardenafil use on the window day or the previous day, and tadalafil use on the window day or any of the previous 4 days. The endpoint therefore measures whether recent use falls within these defined periods; it is not simply a count of prescriptions or a measure of treatment assignment.

In a case-crossover study, comparing a participant’s exposure near the event with exposure at other times can help focus the analysis on whether exposure was unusually timed around the event. As a general statistical rationale, using the same person as their own comparison can reduce the influence of characteristics that remain stable over the periods being compared. It does not, by itself, remove possible time-varying influences, establish that exposure caused the event, or ensure that the chosen control periods are suitable.

The registry record gives the endpoint definition but does not report the counts of exposed days, the number of analyzed participants, or a statistical comparison. Consequently, the endpoint’s observed value and its uncertainty are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The study uses a case-crossover design, a general observational approach for examining whether a transient exposure is associated with the timing of an event. Rather than assigning participants to separate treatment groups, this design compares exposure during an event-related period with exposure during other periods for the same person. The registry describes this framework for PDE5 inhibitor use around acute NAION symptom onset. It does not report the precise statistical method used to analyze the comparison.

Why use a within-person comparison?

As a general principle, comparing a person with themselves can control by design for characteristics that do not change across the periods being compared. Examples in general include stable personal attributes. This can be useful when the question concerns a short-term exposure and an event with a defined onset time. The approach does not control automatically for characteristics that change over time. Changes in health, behavior, or other exposures could coincide with medication use and with the period around symptom onset. The registry record does not describe whether such factors were measured, adjusted for, or considered in sensitivity analyses.

The design also depends on how the event time and control periods are defined. The registry places the case window on the day preceding the symptom-onset day. It defines 29 control windows as the 29 days preceding that case window. Each participant contributes exposure information for 1 case window and 29 control windows. Those definitions determine what counts as exposure near the event and what counts as exposure during the comparison period; they are not results about how frequently exposure occurred.

How is exposure classified?

The record uses drug-specific recent-use rules. Sildenafil or vardenafil use on the day of a window or the previous day makes that window exposed. For tadalafil, use on the day of a window or any of the previous 4 days makes it exposed. These rules mean that exposure classification depends on both the reported drug and its timing relative to each window. The record does not report the underlying medication-use histories or the number of windows classified as exposed.

The endpoint is expressed as a number of exposed days, with the unit stated as exposed days. This is a count outcome, not a mean difference or a time-to-event outcome as described in the record. Counts can have statistical features that affect analysis. As a general explanation, count data may be more variable than a simple model assumes, a feature often called over-dispersion. If present, over-dispersion can make standard errors from an unsuitable model too small and confidence intervals too narrow. The registry does not identify the model, report a dispersion estimate, or state whether over-dispersion was assessed, so it is not possible to say how the analysis handled this issue.

What does a confidence interval tell the reader?

In general, a confidence interval describes the uncertainty around an estimated quantity under a specified statistical model and its assumptions. A narrow interval indicates greater statistical precision than a wide interval, but neither width nor location can be assessed without an estimate and interval. No primary confidence interval is reported in the registry record. The same limitation applies to a p-value: without a reported test result, there is no basis for judging whether the primary comparison was statistically distinguishable from its null value. Statistical significance, even if reported, would not by itself establish a causal effect or its clinical importance.

The registry describes a Full Analysis Set that included participants who signed informed consent, were eligible, and reported PDE5 inhibitor exposure within 60 days before participant-reported symptom onset. It then describes the number analyzed as definite NAION cases with PDE5 inhibitor exposure on at least—after which the eligibility wording is incomplete. This leaves the precise analysis-population criterion unclear. The number analyzed is not reported in the registry record.

What analysis details are unavailable?

The record does not report the model specification, the test used, the handling of multiple windows within a participant, or any adjustment variables. It also does not report whether there were interim analyses, how missing exposure information was handled, or whether any sensitivity analyses were performed. These details matter because repeated windows from one person are related observations rather than independent measurements. As a general statistical principle, analyses should account for that dependence; treating related observations as independent can misstate uncertainty. The registry’s design description alone does not establish which method was used.

The record contains planning assumptions for an illustrative sample-size calculation: a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect defined as a stated fraction of the observed effect. These are described as planning illustrations based on the observed effect, not as the study’s own design. The numerical illustrations and observed effect are not reported in the registry record. A planning calculation is not observed power, and the equal-allocation assumption should not be mistaken for the allocation structure of this non-interventional case-crossover study.

In short, the available methodological description supports an account of the exposure windows and the intended self-comparison. It does not provide enough information to reproduce the primary analysis or verify its uncertainty estimates. The outcome values and the detailed statistical plan are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 3 outcome measures and 3 analyses. It gives a detailed description of the primary endpoint, but does not identify the other outcome measures or report their results. A count of outcomes and analyses does not establish which were prespecified as primary or secondary beyond the endpoint explicitly labeled primary, nor does it show which analyses were conducted or reported.

When several outcomes or analyses are considered, the chance of at least one apparently notable finding can increase. As a general statistical principle, a plan may address this through a prespecified hierarchy, adjustment, or an explicitly exploratory interpretation. The registry record does not report a multiplicity strategy, so whether or how multiplicity was addressed is not reported in the registry record. No adjustment should be assumed.

The Full Analysis Set description includes consenting, eligible participants who reported PDE5 inhibitor exposure within 60 days before participant-reported symptom onset. For the primary endpoint, the record refers to definite NAION cases with exposure, but the final part of the number-analyzed criterion is incomplete. The number analyzed is not reported in the registry record, and the relationship between the full analysis set and the primary endpoint population cannot be fully reconstructed from the text.

Missing-data methods are not described. The record does not report how incomplete medication histories, uncertain symptom-onset dates, missing adjudication information, or unavailable exposure data were treated. It notes that an adjudication committee classified cases as definite, possible, or non-NAION, and could also classify information as insufficient or unable to adjudicate. It does not state how these categories were handled in analyses other than the primary endpoint’s reference to definite cases.

Interim analyses, analysis exclusions, and sensitivity analyses are not reported in the registry record. These omissions limit assessment of the completeness and robustness of the reported evidence. The registry’s completed status does not resolve those methodological details.

11. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 2% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: diarrhoea (0.1%), candidiasis (0.1%), influenza (0.1%), blood pressure decreased (0.1%), weight decreased (0.1%).

For the registry’s Potential NAION Cases group, 673 were listed as at risk for serious adverse events, with 2 affected. The same group lists 673 at risk for other adverse events, with 0 affected. Deaths are not reported in the registry record; a missing death value should not be read as a recorded count of zero.

The registry also lists serious adverse-event terms with displayed values of 0.1% each: Diarrhoea, Candidiasis, Influenza, Blood pressure decreased, Weight decreased, and Dizziness. These term listings are safety-record entries; they do not describe the primary endpoint’s case-window versus control-window exposure comparison.

These safety figures should be read as registry summaries, not as evidence that PDE5 inhibitor exposure caused or did not cause any listed event. The study is non-interventional, and the record does not provide a randomized safety comparison or enough detail to establish attribution. The group label “Potential NAION Cases” also differs from the primary endpoint population, which is defined around participants adjudicated as definite cases with PDE5 inhibitor exposure. The available safety summaries do not explain the relationship between those populations.

The record does not provide additional details about the timing, severity criteria, exposure status at the time of each event, or follow-up for the listed adverse events. Those details are not reported in the registry record, so the figures should not be used to infer comparative risk.

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

The study addresses a focused timing question: whether PDE5 inhibitor exposure was more common in the period immediately preceding acute NAION symptom onset than during control periods for the same participant. Its case-crossover structure is suited to examining short-term exposure timing, and the registry provides explicit rules for defining exposure windows for sildenafil, vardenafil, and tadalafil.

However, the registry record does not report the primary endpoint values or a statistical estimate. It therefore does not show the observed difference or association, its direction, or its precision. Without the number of analyzed definite cases and the corresponding exposure summaries, readers cannot assess how much information supports the comparison. The record’s enrollment figure of 673 is not a substitute for the primary analysis denominator.

Even with complete outcome results, a within-person observational comparison would not by itself prove that a medicine triggered an event. As a general principle, causal interpretation depends on the validity of the exposure and event timing, the suitability of control periods, and the handling of time-varying factors that could affect both exposure and event occurrence. The registry record does not provide enough information to assess all of these assumptions or the detailed analysis used.

The study also should not be read as a randomized comparison of PDE5 inhibitor use with non-use. The registry describes it as non-interventional, lists “No intervention,” and does not state an allocation scheme. Its comparison is between periods within participants, not between two assigned treatment groups.

The adverse-event summaries do not fill the gap in the primary results. They describe safety entries for a group labeled Potential NAION Cases, but they are not a measure of whether exposure was unusually timed around symptom onset. Nor do the listed counts establish that the exposure caused or prevented an adverse event.

The defensible conclusion from the posted registry information is therefore limited: the study’s objective, window definitions, and intended primary measurement are described, while the numerical primary result and key analysis details are not reported in the registry record. No conclusion about the size or precision of the association can be drawn from those descriptions alone.

13. Why This Trial Matters Statistically

This study is statistically notable because its question concerns the timing of a potentially transient exposure relative to an event, rather than the effect of an assigned intervention over a treatment period. A case-crossover design makes that timing explicit by comparing exposure in an event-related window with exposure during control windows for the same person. This can be a useful design choice when event onset can be placed in time and exposure changes over time.

The design also makes measurement decisions especially important. The registry defines one case window and 29 control windows, and applies different recent-use rules to sildenafil or vardenafil than to tadalafil. Those rules determine which medication-use days contribute to the exposure count. A clear definition improves interpretability, but it does not show whether the classification is complete or accurate in the underlying records.

Repeated windows from the same person are statistically dependent. As a general principle, the analysis needs to reflect that structure when estimating uncertainty. The registry does not report the model or inferential method, so readers cannot assess whether the dependence was handled appropriately or reproduce the analysis.

The endpoint is a count of exposed days rather than a reported effect estimate. Without the counts and their analysis, the design description cannot answer the substantive question. The registry’s enrollment figure does not reveal the primary analysis size, which is restricted to definite NAION cases with exposure and is incompletely defined in the record.

Finally, the record illustrates why design information and results must be distinguished. It gives a detailed research question and exposure-window framework, but does not report the primary numerical findings or their uncertainty. That distinction is essential when evaluating observational evidence: a coherent design can clarify what was intended to be measured, while missing results prevent assessment of what the study observed.

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 most direct limitation of the registry record is that it describes the primary endpoint without reporting its values. No exposed-day counts, effect estimate, confidence interval, or p-value are reported. As a result, the central comparison cannot be quantitatively evaluated, and its precision cannot be judged.

The analysis population is also not fully specified. The Full Analysis Set description refers to eligible participants with consent and reported PDE5 inhibitor exposure within 60 days before symptom onset. The subsequent criterion for the number analyzed ends after “at least,” leaving the exposure requirement incomplete. The number analyzed is not reported in the registry record, so the primary analysis denominator cannot be established from the description.

The endpoint relies on adjudication of NAION status and participant-reported symptom onset. The record explains that an adjudication committee classified participants as definite, possible, non-NAION, insufficient information, or unable to adjudicate. It does not report the number in each category or how non-definite and unclassifiable cases were handled outside the stated primary endpoint definition. It also does not describe the process for establishing the onset date or validating medication-use timing.

The exposure definitions allow recent-use windows that differ by drug. The record does not report the source or completeness of medication-use information, or how uncertain or missing exposure dates were treated. Nor does it report whether other time-varying factors were measured or included in analysis. These issues can affect the interpretation of an observational timing comparison.

Other key details are not reported in the registry record, including the statistical model, treatment of repeated windows, multiplicity handling, missing-data methods, interim analyses, and sensitivity analyses. The non-interventional design also means there is no randomized assigned-treatment comparison. These limitations restrict both reproducibility and causal interpretation.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Study to Assess Whether PDE5 Inhibitors Increase the Chance of Triggering the Onset of Acute NAION — NCT00759174.
• Registry record last updated 2021-02-01; status last verified 2021-01.
• Results first posted 2013-10-17.
• 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

NCT00759174 is a completed, non-interventional case-crossover study examining PDE5 inhibitor exposure as a potential trigger factor for acute NAION. The registry lists enrollment of 673, a study start of 2008-11, and primary completion of 2012-10. It identifies Optic Neuropathy, Ischemic as the condition and lists “No intervention.”

The primary endpoint counts exposed days in a case window and control windows among participants adjudicated as definite NAION cases. The case window is the 1 day before symptom onset, with 29 control windows before it. Drug-specific exposure rules are described for sildenafil, vardenafil, and tadalafil.

The registry lists 3 outcome measures and 3 analyses, but does not report the primary endpoint values, the number analyzed, or the statistical estimate and uncertainty. The analysis-population wording is incomplete, and methods for missing data, multiplicity, and the primary statistical model are not reported in the registry record. Safety summaries list 2 affected for serious adverse events and 0 for other adverse events in the Potential NAION Cases group; deaths are not reported.