Methodologic Concerns Regarding a Study Concluding That Epidural Labor Analgesia Is Associated with a Decreased Risk of Postpartum Depression
Bibliographic record
Abstract
To the Editor We advance the following concerns with the observational study of epidural analgesia and postpartum depression (PPD) by Ding et al.1: Inadequately Addressed Bias Misclassification bias: Although defined as such, the study by Ding et al.1 does not meet the definition of a prospective cohort, because their sample was not depression-free at the time of exposure. The authors do not estimate the magnitude and direction of this misclassification bias on their findings. Given that the prevalence of depression is 20% among Chinese women, and that depressed women are less likely to advocate for pain control, the study results could be moderately biased toward a protective effect. Selection bias: Obese subjects were excluded without clear justification. Obesity is not a contraindication for epidural analgesia and is actually associated with an increased incidence of epidural and a greater risk of PPD.2–8 The study reports a negative association between epidural analgesia and PPD. Given the proportion of obese women who were excluded (35% of the final sample size), there may be substantial bias in the direction of a significant protective association between epidural analgesia and PPD. Second, the authors present a greater loss to follow-up among the epidural group than the nonepidural group (24 vs 13). The reasons cited for this differential loss include known risk factors for PPD, namely, “too tired after delivery” and “bad mood because infant in the neonatal intensive care unit.” Evidence shows that mothers of infants admitted to neonatal intensive care unit have greater-than-average rates of depression and anxiety.9–14 It is well established that patients with depression are more likely to be lost to follow-up. In Chinese society, the family makes medical decisions rather than the individual. Thus, the family’s refusal to continue with the study could reflect their concern for the mother’s mental health. We estimate that a greater proportion of depressed mothers were lost from the exposed group than the unexposed group, which further biases the study results, in the direction of a protective effect. Uncontrolled Confounding Appropriate confounding control is essential in observational studies.15 By definition, a confounder has a demonstrated association with both exposure and outcome.16,17 The forward elimination method used by the investigators identifies variables associated with the outcome only. This may lead to discarding important confounders and including variables that cannot confound. A backward elimination process is a more rigorous method because all potential confounders are included in the initial model and then eliminated on the basis of significant effects.18 It is possible that the author’s method has led to uncontrolled confounding, which may further affect the results.15,19 Ding et al.1 present their study as an initial step in assessing the association between epidural and PPD, calling for further research using larger samples. We would add to this insightful recommendation that a depression-free sample be recruited at delivery, that obese women not be excluded, and that a complete set of potential confounders be included in an analysis using appropriate methods.15,19 Given the potential impact of this study on clinical practice, its conclusions should be interpreted with caution. Kathleen H. Chaput, PhD Department of Paediatrics Alberta Children’s Hospital Research Institute for Child and Maternal Health Cumming School of Medicine University of Calgary Calgary, Alberta, Canada [email protected] Angela Vinturache, MD, PhD Department of Paediatrics Cumming School of Medicine University of Calgary Calgary, Alberta, Canada
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.211 | 0.611 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.005 | 0.003 |
| Bibliometrics | 0.006 | 0.006 |
| Science and technology studies | 0.002 | 0.005 |
| Scholarly communication | 0.005 | 0.003 |
| Open science | 0.007 | 0.002 |
| Research integrity | 0.012 | 0.010 |
| Insufficient payload (model declined to judge) | 0.007 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".