Auriculotherapy During Medication Abortion: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Auriculotherapy During Medication Abortion (NCT03900728), a randomized na study. The primary outcome is Pain VAS Score.

2. Clinical Question

3. Trial Design

The study is registered as a randomized, quadruple-masked, parallel, three-arm trial. It evaluates two forms of auriculotherapy—needles and gold beads—each added to usual care, against placebo adhesive disks plus usual care. The record lists Columbia University as sponsor and the trial status as completed. The recorded study start was 2019-04-01, and primary completion was 2020-03-30.

Participants received their assigned treatment immediately after the initial abortion medication, mifepristone. Misoprostol was taken 1-3 days later. Pain and anxiety were to be assessed by text message once daily for four days, with an in-person visit at one-week follow-up. For the primary outcome, the highest of the four post-misoprostol pain scores was selected for analysis. The registry describes the devices as remaining in place until removal at the follow-up visit.

The registry enrollment is 138. The posted primary-outcome summaries list 40 participants for the needle group, 57 for the gold-bead group, and 39 for the placebo group. These are the outcome-specific group counts; the record does not explain how they relate to the enrollment figure or whether any participants were excluded from the outcome summaries.

The record identifies anxiety VAS score as a secondary outcome but does not post its group results. It also does not report a statistical analysis plan, the intended primary contrast, or the method used to account for repeated assessments or other design features. The randomized design can support a causal comparison when its allocation and analysis are implemented appropriately, but the registry’s descriptive results alone do not establish such a 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.

Before assignment: Out of 138 enrolled subjects, 2 subjects were not randomized. 1 subject did not have a study auriculotherapist available during the visit and 1 subject declined randomization. Only data information for the randomized subjects were analyzed (n = 136).

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 Pain VAS Score, measured over Auriculotherapy is applied on the day of mifepristone. Participants take misoprostol 1-3 days later. Pain scores are recorded following misoprostol at 8pm x 4 days (selecting the highest of the 4 for analysis)..

The primary endpoint is maximum self-assessed pain after misoprostol, measured using a 0-100 mm visual analog scale. A score of 100 signifies maximum pain. Participants reported pain at 8pm for four days following misoprostol, and the highest of those four values was selected. The posted summaries are medians with interquartile ranges.

Using a maximum focuses on the most intense reported pain during the specified reporting period, rather than the average score or the pattern of pain over time. This can be useful when the research question concerns peak experience. It also means that the endpoint does not describe when the peak occurred, how long pain persisted, or whether scores rose or fell across the reporting days. Because it selects the highest of repeated self-reports, it is a summary of both pain intensity and the chosen observation window.

The registry reports a median of 75.0 with an interquartile range of 45.0 to 90.0 for needles, a median of 60.0 with an interquartile range of 20.0 to 85.0 for gold beads, and a median of 55.0 with an interquartile range of 30.0 to 85.0 for placebo. These are descriptive summaries, not estimates of adjusted treatment effects. No confidence intervals or formal tests for this endpoint are reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is reported with medians and interquartile ranges rather than means and standard deviations. A median describes the midpoint of ordered observations and is less affected by unusually high or low values than a mean. The interquartile range gives the middle half of the observed values, bounded by the lower and upper quartiles. These summaries can be informative when distributions are skewed or when the endpoint is understood through its central rank and spread. They do not, on their own, estimate the uncertainty of a difference between treatment groups.

What does randomization contribute?

Random allocation is intended to make treatment groups comparable, on average, with respect to both measured and unmeasured characteristics at the time of assignment. This is the basis for interpreting a properly conducted between-group analysis as an estimate of a treatment effect. Randomization does not guarantee exact balance in a particular trial, nor does it remove the need to report how outcomes were analyzed. The registry identifies the trial as randomized but does not report the allocation method, allocation concealment procedures, or a statistical plan for the primary comparison.

The design is parallel: participants are assigned to a treatment group and the groups are compared. It has three arms, representing needles, gold beads, and placebo disks, all alongside usual care. A trial with multiple arms can support several possible contrasts, but those contrasts should be specified and interpreted carefully. The registry does not state which contrast was primary, whether all pairwise comparisons were planned, or how results across multiple comparisons would be handled.

What does masking mean for interpretation?

The record describes masking as quadruple. Masking is intended to limit the influence of knowing treatment assignment on participants’ reports, care, or outcome assessment. The placebo disks were designed to resemble those used with active treatments, which is relevant to the plausibility of masking. The registry does not specify which four roles were masked or report whether masking was assessed. Because pain is self-reported, participant awareness could matter to the recorded outcome; the design’s masking description is useful context, but it does not establish that masking was successful.

What does the maximum-score endpoint capture?

Participants were asked to report pain at 8pm for four days after misoprostol, and the largest value was used. Selecting the maximum emphasizes peak reported pain during that window. It discards information about the sequence of scores, including whether the peak was isolated or part of a sustained pattern. It also does not capture symptoms outside the reporting times. As a general statistical point, a maximum drawn from repeated observations can be affected by the number and completeness of observations, as well as by measurement variability. The registry does not describe how incomplete daily reports were handled or how the maximum was derived when a participant had fewer than four reports.

The endpoint is on a 0-100 mm visual analog scale, with 100 representing maximum pain. A scale score is a measurement, not a direct biological quantity. Interpretation depends on consistent administration and on what differences in the scale mean to participants. The registry provides no threshold for a clinically important difference, and no such threshold should be inferred from the observed medians.

How should uncertainty and testing be assessed?

A confidence interval, in general, gives a range of values compatible with the data under a specified statistical procedure and communicates uncertainty in an estimate. The registry does not report confidence intervals for the group summaries or any treatment contrast. A p-value, in general, assesses how compatible data are with a null hypothesis under a specified model; it does not measure the probability that a treatment works. No p-value or formal test is posted for the primary endpoint. Consequently, readers cannot use the registry results to assess inferential precision or statistical evidence for a difference.

The sample-size assumptions in the registry describe a planning illustration based on a stated fraction of an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such calculations are planning exercises, not observed power and not evidence that the trial was adequately powered. Any planning illustration based on an observed effect is also dependent on the effect chosen as its input. The registry record does not report a prospective sample-size justification or enough information to evaluate the original design assumptions. Statistical methods and sample-size calculations can support interpretation, but they cannot substitute for reported estimates, uncertainty, and transparent outcome analyses.

Why do distribution shape and spread matter?

General statistical reasoning favors methods appropriate to the outcome’s distribution and study design. If data are skewed, an approach that assumes normally distributed observations may not describe the data well. If repeated observations are analyzed, their within-participant dependence matters because measurements from the same person are not generally independent. Over-dispersion, in general, means that observed variability is greater than a model expects; ignoring it can make uncertainty estimates too small. The registry does not report which methods, if any, were used to address distribution shape, repeated observations, or over-dispersion. The posted medians and interquartile ranges are descriptive and do not reveal the model used for any unreported analysis.

10. Multiplicity, Missing Data and Analysis Populations

The study includes a primary pain outcome and a secondary anxiety outcome, and the design has three treatment arms. This creates potential multiplicity: several outcomes or treatment contrasts may be considered, increasing the chance of finding an apparently notable result when many tests are performed. In general, a prespecified analysis plan can identify the principal contrast and define how additional comparisons will be interpreted or adjusted. The registry record does not report a multiplicity strategy, the planned contrast hierarchy, or formal analyses for the posted pain summaries. Anxiety is listed as a secondary outcome, but its results are not reported.

Missing data can affect both descriptive summaries and treatment comparisons. The registry describes daily text-message assessments and an in-person follow-up, but it does not report completion rates, missingness by group, reasons for missing reports, or procedures for handling incomplete pain records. Since the endpoint uses the highest value over four days, the handling of a participant with fewer than four daily scores could influence the resulting summary. No imputation or sensitivity-analysis approach is reported in the registry record.

The record does not define the analysis population—for example, whether the primary outcome was analyzed according to randomized assignment or only among participants with recorded outcomes. It also does not report exclusions after randomization or participant flow. These omissions limit assessment of whether the posted group summaries retain the protection of randomization and how representative they are of the enrolled participants.

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 5% reporting threshold.

The registry records zero serious adverse events among 40 participants at risk in the needle group, among 57 at risk in the gold-bead group, and among 39 at risk in the placebo group. It also records zero other adverse events in each of those groups and zero deaths in each group. These entries describe the event counts as reported in the registry.

A zero count in a registry entry does not establish that an event is impossible or that the treatment is free of risk. Safety interpretation depends on what events were sought, the period of observation, definitions, ascertainment, and completeness of reporting. The record does not provide details about safety surveillance or event definitions beyond distinguishing serious and other adverse events. It also does not list serious-event terms. The displayed counts should therefore be read as the reported safety results, not as a guarantee that no adverse events occurred or could occur.

The group-specific at-risk counts are the relevant denominators for these entries. They are not necessarily the denominators for every outcome, and the registry does not explain whether safety follow-up was complete for everyone at risk. The absence of reported event counts cannot determine comparative safety, particularly without inferential estimates and fuller reporting of the ascertainment process.

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

The registry record documents a randomized comparison of auriculotherapy with needles, auriculotherapy with gold beads, and placebo adhesive disks, each alongside usual care. For maximum post-misoprostol pain, it reports group medians and interquartile ranges. The posted medians are 75.0 for needles, 60.0 for gold beads, and 55.0 for placebo. The corresponding interquartile ranges are 45.0 to 90.0, 20.0 to 85.0, and 30.0 to 85.0. These values describe the recorded outcome distributions; they are not a substitute for a treatment-effect estimate with uncertainty.

The record does not report a formal statistical comparison, confidence interval, or p-value for pain. It therefore does not show whether the differences among the descriptive summaries are statistically distinguishable, how precise a treatment contrast would be, or whether any difference meets a threshold of clinical importance. It does not support a claim that any of the three approaches is superior, equivalent, or ineffective.

The outcome is specifically the highest pain score reported at 8pm during the four days after misoprostol. It is not a measure of average pain over that period, total pain burden, duration of symptoms, or pain at other times. Results for anxiety, although the outcome is listed as secondary, are not posted. The registry’s zero event counts are limited to the safety information it records and do not establish absence of risk.

These distinctions matter because a descriptive result can be useful for understanding what was observed while remaining insufficient for a causal or clinical conclusion. Interpretation would be strengthened by a clearly stated primary contrast, the analysis method, uncertainty estimates, participant flow, missing-data handling, and the secondary outcome results. Those details are not reported in the registry record.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of why a randomized design and descriptive summaries are not the same as a reported treatment-effect analysis. Randomization establishes a design intended to support group comparisons, while the registry’s posted pain results consist of arm-level medians and interquartile ranges. Without a defined contrast and an estimate of uncertainty, readers cannot determine how much evidence the observed summaries provide about differences between interventions.

The three-arm structure also illustrates the importance of specifying which comparisons answer the primary question. Needle and bead procedures are each compared in the study against a placebo procedure, but the registry does not state the intended contrast hierarchy or how multiple comparisons were handled. Naming a primary endpoint does not, by itself, clarify which statistical test or estimand was used.

Endpoint construction is another important feature. Selecting the highest of repeated daily pain scores targets peak reported pain during a defined window, not the overall trajectory. That choice may be appropriate for a particular research question, but it changes what the result means and makes transparent reporting of missing daily responses especially relevant. The registry describes assessment timing but does not explain how incomplete reports were treated.

Finally, the record shows why outcome completeness and precision matter. It identifies anxiety as a secondary outcome without posting results, and it presents no confidence intervals or formal tests for the primary outcome. These gaps do not invalidate the descriptive summaries, but they constrain the conclusions readers can draw from a completed randomized trial.

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’s primary outcome is a self-reported maximum score, selected from scheduled evening reports after misoprostol. It does not represent pain at every point during the process, and the maximum alone cannot show the duration or trajectory of pain. The record does not state how many daily reports each participant completed or how incomplete reporting affected the endpoint.

The posted results are descriptive. The registry does not report a treatment-effect estimate, confidence interval, p-value, statistical model, or planned primary contrast. It also does not identify a multiplicity approach for the three-arm design or explain whether any adjustment was made for baseline characteristics. The listed medians and interquartile ranges cannot answer those questions.

Although anxiety VAS score is listed as a secondary outcome, its results are absent. The record also does not report detailed participant characteristics, participant flow, analysis-population definitions, outcome-specific missingness, or sensitivity analyses. The enrollment figure and the group counts attached to the pain outcome are both recorded, but the registry does not reconcile them.

Quadruple masking is stated, and the placebo disks resembled the disks used with active procedures. The record does not specify which roles were masked or whether masking remained effective. Safety reporting gives zero serious and other adverse events and zero deaths by group, but does not give detailed ascertainment procedures or event definitions. These reporting limitations restrict assessment of precision, bias, and generalizability from the registry record alone.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Auriculotherapy During Medication Abortion — NCT03900728.
• Registry record last updated 2021-07-06; status last verified 2021-06.
• Results first posted 2021-04-15.
• 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

This completed randomized, quadruple-masked, parallel trial compared auriculotherapy with Pyonex needles or gold beads, each added to usual care, against placebo adhesive disks plus usual care. Its primary outcome was maximum self-reported pain on a 0-100 mm visual analog scale, using the highest of four daily scores after misoprostol. The registry reports medians and interquartile ranges for each group, but no formal treatment comparisons or uncertainty estimates. Anxiety VAS score is listed as a secondary outcome without posted results. The registry records zero serious adverse events, zero other adverse events, and zero deaths in each group. Details of the statistical analysis, missing-data handling, participant flow, and analysis populations are not reported in the registry record.