Secondary Logo

Journal Logo

Methodologic Concerns Regarding a Study Concluding That Epidural Labor Analgesia Is Associated with a Decreased Risk of Postpartum Depression

Chaput, Kathleen H. PhD; Vinturache, Angela MD, PhD

doi: 10.1213/ANE.0000000000000878
Letters to the Editor: Letter to the Editor
Free

Department of Paediatrics, Alberta Children’s Hospital Research Institute for Child and Maternal Health, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada, khchaput@ucalgary.ca

Department of Paediatrics, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada

Back to Top | Article Outline

To the Editor

We advance the following concerns with the observational study of epidural analgesia and postpartum depression (PPD) by Ding et al.1:

Back to Top | Article Outline

Inadequately Addressed Bias

  1. 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.
  2. 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.

Back to Top | Article Outline

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

khchaput@ucalgary.ca

Angela Vinturache, MD, PhD

Department of Paediatrics

Cumming School of Medicine

University of Calgary

Calgary, Alberta, Canada

Back to Top | Article Outline

REFERENCES

1. Ding T, Wang DX, Qu Y, Chen Q, Zhu SN. Epidural labor analgesia is associated with a decreased risk of postpartum depression: a prospective cohort study. Anesth Analg. 2014;119:383–92
2. Schwartz AJ. Learning the essentials of epidural anesthesia. Nursing. 2006;36:44–9
3. Vincent RD Jr, Chestnut DH. Epidural analgesia during labor. Am Fam Physician. 1998;58:1785–92
4. Hader FC. Epidural analgesia in the critically ill. Nurs Crit Care. 2007;2:20–30
5. Bernard CMBarash PG, Cullen BF, Stoelting RK, Cahalan MK, Stock MC. Epidural and spinal anesthesia. In: Clinical Anesthesia. 20096th ed Philadelphia, PA Lippincott Williams & Wilkins:950
6. Anim-Somuah M, Smyth RMD, Jones L. Epidural versus non-epidural or no analgesia in labour. Cochrane Database Syst Rev. 2011:CD000331
7. Molyneaux E, Poston L, Ashurst-Williams S, Howard LM. Obesity and mental disorders during pregnancy and postpartum: a systematic review and meta-analysis. Obstet Gynecol. 2014;123:857–67
8. Weiss JL, Malone FD, Emig D, Ball RH, Nyberg DA, Comstock CH, Saade G, Eddleman K, Carter SM, Craigo SD, Carr SR, D’Alton MEFASTER Research Consortium. . Obesity, obstetric complications and cesarean delivery rate—a population-based screening study. Am J Obstet Gynecol. 2004;190:1091–7
9. Wigert H, Johansson R, Berg M, Hellström AL. Mothers’ experiences of having their newborn child in a neonatal intensive care unit. Scand J Caring Sci. 2006;20:35–41
10. Chourasia N, Surianarayanan P, Adhisivam B, Vishnu Bhat B. NICU admissions and maternal stress levels. Indian J Pediatr. 2013;80:380–4
11. Discenza D. Taking care of the NICU mom. Neonatal Netw. 2009;28:351–2
12. Hynan MT, Mounts KO, Vanderbilt DL. Screening parents of high-risk infants for emotional distress: rationale and recommendations. J Perinatol. 2013;33:748–53
13. Lefkowitz DS, Baxt C, Evans JR. Prevalence and correlates of posttraumatic stress and postpartum depression in parents of infants in the neonatal intensive care unit (NICU). J Clin Psychol Med Settings. 2010;17:230–7
14. Alkozei A, McMahon E, Lahav A. Stress levels and depressive symptoms in NICU mothers in the early postpartum period. J Matern Fetal Neonatal Med. 2014;27:1738–43
15. Christenfeld NJ, Sloan RP, Carroll D, Greenland S. Risk factors, confounding, and the illusion of statistical control. Psychosom Med. 2004;66:868–75
16. Last J A Dictionary of Epidemiology. 20014th ed. New York, NY Oxford University Press:37–8
17. Rothman K, Greenland SRothman K, Greenland S, Lash T. Introduction to stratified analysis. In: Modern Epidemiology. 20083rd ed Philadelphia, PA Lippincott Williams & Wilkins:259–60
18. Sun GW, Shook TL, Kay GL. Inappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis. J Clin Epidemiol. 1996;49:907–16
19. Moons KG, Altman DG, Reitsma JB, Ioannidis JP, Macaskill P, Steyerberg EW, Vickers AJ, Ransohoff DF, Collins GS. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med. 2015;162:W1–73
© 2015 International Anesthesia Research Society