- Email: [email protected]
Labor epidural analgesia and the risk of postpartum depression: A meta-analysis of observational studies
Journal of Clinical Anesthesia xxx (xxxx) xxxx
Contents lists available at ScienceDirect
Journal of Clinical Anesthesia journal homepage: www.elsevier.com/locate/jclinane
Original contribution
Labor epidural analgesia and the risk of postpartum depression: A metaanalysis of observational studies ⁎
Joanna Kountanis (MD)a, , Christie Vahabzadeh (MD)a, Samuel Bauer (MD)b, Maria Muzik (MD)c, Ruth Cassidy (MS)a, Casey Aman (BS)a, Mark MacEachern (MLIS)d, Melissa Bauer (DO)a a
Department of Anesthesiology, University of Michigan Health System, 1500 East Medical Center Drive, Ann Arbor, MI 48109, United States of America Department of Obstetrics and Gynecology, Oakland University William Beaumont School of Medicine, 586 Pioneer Drive, Rochester, MI 48309, United States of America c Department of Psychiatry, University of Michigan Health System, 1500 East Medical Center Drive, Ann Arbor, MI 48109, United States of America d University of Michigan Taubman Health Sciences Library, 1135 Catherine Street, Ann Arbor, MI 48109, United States of America b
A R T I C LE I N FO
A B S T R A C T
Keywords: Analgesia Epidural Depression Postpartum Labor pain Obstetrics Risk factors
Study objective: This study aims to systematically review the literature to evaluate the association between labor epidural analgesia (LEA) and postpartum depression (PPD). Design: Meta-analysis. Setting: Obstetric patients delivering vaginally with or without LEA in a hospital. Interventions: This study aimed to investigate the effects of providing LEA on developing PPD. Measurements: Pooled odds ratios (OR) and 95% confidence intervals (CI) were calculated using the random effects model. Results: A total of 356 full text articles were reviewed. Eleven articles studying 85,928 patients met inclusion criteria. The pooled unadjusted OR 1.03 and 95% CI (0.77, 1.37) suggest that LEA is not associated with a decreased risk of developing PPD. Conclusions: Labor epidural analgesia was not shown to confer protection against developing PPD according to this meta-analysis. Future studies are needed to explore whether other aspects of LEA, beyond its presence or absence, influence the onset of PPD.
1. Introduction Developing a psychological sequela from childbirth is one of the most frequent complications seen in women postpartum [1]. As many as one in five mothers experience postpartum depression, with the incidence of postpartum depression (PPD) ranging from 11% to as high as 19.2% within the first three months [2,3]. The impact of suffering from a mental illness or subsyndromal symptoms postpartum has consequences for mother and child, and is a financial burden to the health care system. In a cost analysis study, women with PPD incur up to 90% higher health services expenditures than non-depressed women for themselves, as well as their infant, after delivery [4]. Children born to mothers with PPD also suffer negative health consequences; mothers suffering from PPD may have impaired maternal-infant bonding or have children with lower cognitive and behavioral development up to 5 years old [5,6]. It has been suggested that even if the mother
subsequently gets treatment, there could be long lasting mental health consequences for her child [7]. Postpartum depression also has direct consequences to the health and safety of the mother. Depression has been associated with an increased risk of suicide. Approximately 4% of depressed individuals die from suicide [8]. A recent review of total maternal deaths in Colorado from 2004 to 2012 demonstrated that the leading cause of death was due to self-harm, comprising 30% of all maternal deaths and with the majority occurring postpartum [9]. The Mothers and Babies: Reducing Risk through Audits and Confidential Enquiries across the United Kingdom 2014–2016 report cites 31% of maternal deaths between six weeks and one year after pregnancy were from mental-health related causes (suicide, drug/alcohol, other psychiatric causes). Maternal suicide is the leading cause of direct deaths in this time period [10]. Therefore, the sequelae from PPD include adverse effects to maternal and neonatal health, in addition to contributing to mortality.
⁎
Corresponding author at: 1500 East Medical Center Drive SPC 5278, University Hospital South, L3626 Ann Arbor, MI 48109-5278, United States of America. E-mail addresses: [email protected] (J. Kountanis), [email protected] (C. Vahabzadeh), [email protected] (S. Bauer), [email protected] (M. Muzik), [email protected] (R. Cassidy), [email protected] (C. Aman), [email protected] (M. MacEachern), [email protected] (M. Bauer). https://doi.org/10.1016/j.jclinane.2019.109658 Received 24 June 2019; Received in revised form 9 September 2019; Accepted 16 November 2019 0952-8180/ © 2019 Elsevier Inc. All rights reserved.
Please cite this article as: Joanna Kountanis, et al., Journal of Clinical Anesthesia, https://doi.org/10.1016/j.jclinane.2019.109658
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
Google Translate to assess whether the subject matter met inclusion criteria. The quality of individual selected studies was assessed in relation to its ability to investigate the association between LEA and PPD with unadjusted data using the Newcastle Ottawa Quality Assessment scale. Therefore, well-designed studies may be assessed as “poor” quality if the study was not primarily designed for the outcome of interest in this meta-analysis. In addition, studies appropriately controlling for multiple confounding factors in their adjusted analysis may also be rated poorly due to this meta-analysis's use of their unadjusted data. Each study was graded based on selection of an exposed and non-exposed cohort that were truly representative of an appropriate study population, that cohorts were drawn from the same community, ascertainment of exposure, demonstration that outcome of interest was not present at start of study, comparability of the exposed with the unexposed cohort, assessment of outcome, follow up long enough for outcome to occur, and lost to follow up rate. The Newcastle Ottawa Quality Assessment rating for each study was then converted to the Agency for Healthcare Research and Quality standard (good, fair, and poor). To estimate the effect size, most studies reported either unadjusted odds ratios and 95% confidence intervals or appropriate counts for which to calculate odds ratio and confidence interval estimates. We directly contacted authors to obtain additional information if not available in the original publication. To obtain a pooled measure of association between LEA and incidence of PPD, we calculated a pooled unadjusted odds ratio using a random effects logistic model. Unadjusted odds ratios were collected for each study by review of the manuscript or by using raw data obtained by direct email correspondence with the corresponding author. A random effects model was utilized rather than a fixed effects model to account for heterogeneity between studies due to the fact that included studies had been conducted under variable conditions with variable patient populations. The fixed effects model is chosen when studies have been conducted under similar conditions and have a common effect size, which was not appropriate for this meta-analysis. To assess pooled effect size and statistical significance at alpha = 0.05, associated 95% confidence intervals were calculated. We performed a test for heterogeneity using the Q-statistic, and calculated the I2 statistic and associated confidence interval. In cases where the reported odds ratio confidence interval is not symmetric around the odds ratio, the upper limit was recalculated based on the distance between the odds ratio and the lower confidence limit. All statistical analyses were performed using Comprehensive Meta-Analysis Version 3.3.070 (Biostat, Inc. Englewood, NJ, USA) and plots were made using R 3.1.1 (The R Foundation, Vienna, Austria).
A prior history of mental illness is one of the strongest predictors of postpartum psychological pathology [11,12]. The development of PPD is multi-factorial and encompasses risks such as low levels of social support, breast feeding problems, and neonatal intensive care admission [13]. However, the current body of literature is inconsistent in identifying all risk factors. Despite studies reporting associations between labor pain and PPD, American College of Obstetrics and Gynecology (ACOG) and the United States Preventive Services Task Force do not consider labor pain as a risk factor for developing PPD [13,14]. Boudou et al. demonstrated a positive association between the intensity of childbirth pain and severity of postpartum blues. The authors speculated possible explanations as an evoked stress response from intense pain as well as possible maternal feelings of failure from not overcoming the pain [15]. Ding et al. also speculated that severe labor pain may induce a strong stress stimulus, resulting in neuropsychological consequences postpartum [16]. This theory supported his finding that the presence of labor epidural analgesia (LEA) was protective against the development of postpartum depression. However other studies contradict these findings [17,18]. It would seem prudent to identify whether a low risk procedure such as LEA could be protective since other known risk factors for developing PPD, such as poor social support or a history of depression, are not modifiable. This meta-analysis aims to evaluate the association of labor epidural analgesia and the development of post-partum depression. 2. Methods This study was registered on the international prospective register of systematic reviews, PROSPERO (CRD42017079565). The meta-analysis was conducted and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. Three Ovid MEDLINE databases (MEDLINE, In-Process & Other Non-Indexed Citations, and Epub Ahead of Print), Embase.com, CINAHL via EBSCOhost, Cochrane Central Register of Controlled Trials via Wiley, and clinicaltrials.gov were searched on August 23, 2017 to identify observational studies evaluating the presence of LEA and subsequent development of PPD. Studies were included even if their primary outcome was not investigating the association of LEA and PPD. The search was conducted again on December 6, 2017 and July 30, 2019 to ensure recent publications were identified. The search was conducted using a combination of controlled terms (MeSH or EMTREE) and title or title and abstract text words. The searches included sentinel articles that were identified by the first author in an initial literature review. A citation and bibliography search of all citations associated with the included studies was conducted in Clarivate Web of Sciences and Google Scholar. Duplicate citations were identified and discarded in Endnote X6 (Clarivate Analytics) and hand review of citations. The complete searches can be found in Appendix A. Two reviewers independently reviewed relevant publications that were found in the search for inclusion articles (JK and CV). Bibliographies of selected articles and articles citing the selected articles were screened for inclusion. Once relevant articles were identified, two authors extracted data and discrepancies were resolved by discussion. Data was extracted from studies providing presence or absence of LEA and Edinburgh postnatal depression scale scores (EPDS), Beck Depression Inventory, Diagnostic and Statistical Manual of Mental Disorders (DSM) diagnosis of postpartum depression, or other validated methods of identifying PPD. Inclusion criteria were observational studies reporting the presence or absence of LEA, participants > 18 years of age, and subsequent validated methods of identifying PPD. Exclusion criteria were studies not specifying the type of assessment for postpartum depression, unspecified timing of the assessment, unclear separation between those receiving and not receiving epidural analgesia, case reports, case series, editorials, and correspondences. The titles and text of foreign language articles were placed into
3. Results A total of 356 full text articles were reviewed after the search was run on August 23, 2017, December 6, 2017, and July 30, 2019. Fig. 1 shows the selection process. Google Translate could translate all studies written in a foreign language and none were found to meet inclusion criteria. Eleven articles with 85,928 patients met inclusion criteria [16–26]. (Table 1) All studies identified probable postpartum depression within the first year postpartum utilizing the EPDS screening tool, expect for Wu et al. which used health care services for depression as their criteria. The diagnosis of postpartum depression varied between studies with some using EPDS scores ≥10, > 12, ≥12, > 6, DSM IV criteria, or utilization of health care services for depression. (Table 1) Nine studies were designed as prospective observational, Suhitharan et al. designed a retrospective case control, and Wu et al. designed a population based matched cohort. The cohort without LEA varied in reporting of other forms of analgesia provided, including not reported in the manuscript, nitrous oxide, spinal, pudendal block, paracervical block, doula support, narcotics, transcutaneous electrical nerve stimulation, pethidine, and no other form of analgesia provided. 2
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
Fig. 1. Flow chart of search strategy.
several strengths. These include appropriate eligibility criteria, clear reporting of the presence of LEA, and using a verified screening tool to detect PPD. Subject inclusion criteria was liberal and characteristics between groups similar among all studies, as well as all studies appropriately measured exposure and outcome. Another strength of this study was to reduce the effect of publication bias by including peer reviewed Internet based journal articles that were found on supplemental searches outside PubMed in the analysis. Furthermore, all studies published in a non-English language were also considered due to the ability to translate the text. However, heterogeneity was a concern given each study measured a variety of confounding risk factors when testing the relationship between LEA and PDD. For example, Ding et al. controlled for risk factors such as satisfaction with medical service and attendance of a childbirth class, whereas Orbach-Zinger et al. controlled for marital status. In addition, some studies controlled for confounding between cohorts that had PPD and those who did not have PPD, and others controlled for confounding between cohorts that had LEA and those that did not. Due to the complexity of PPD and its development, it is near impossible for each study to adequately control for all known prognostic factors. There are many other suspected risk factors that were left unmeasured by many studies; high levels of non-pregnancy related stresses, poor social support, history of physical or sexual abuse, unintended pregnancy,
Analysis of Johnstone et al., Orbach-Zinger et al., and Tobin et al. was accomplished by obtaining unpublished data from the corresponding authors to calculate unadjusted odd ratios. The pooled unadjusted OR 1.03 and associated 95% CI (0.77, 1.37) suggest that there is insufficient evidence to conclude that LEA is significantly associated with PPD. (Fig. 2) The Q-statistic of 15.994 with 10 degrees of freedom supports the assumption of heterogeneity at a significance level of 0.10 (p-value = 0.0998). The estimate and 95% uncertainty interval for I2 is 37.48 (0–69.25). The quality assessment assigning “good” for Ding et al., OrbachZinger et al., Riazanova et al., Wu et al., and Zhang et al. compared to the other studies assigned as “poor” was primarily distinguished by cohort comparability when using unadjusted OR and adequacy of follow up (Table 2).
4. Discussion This meta-analysis aimed to review and quantify the strength of the association between presence of LEA and onset of PPD among women. According to the analysis, LEA was not shown to confer protection against developing PPD. This negative finding concurs with the majority of studies in this analysis. The individual studies included in this meta-analysis also have 3
Population based longitudinal cohort Prospective cohort Prospective cohort
Eckerdal
4
Prospective nonrandomized observational trial Retrospective case control Secondary analysis of a prospective cohort Population-based matched cohort
Prospective cohort
Riazanova
Suhitharan Tobin
Zhang
2–4 weeks
Up to one year
EPDS
Seeking medical care for depression
EPDS EPDS
EPDS
6 weeks
4–8 weeks 6–8 weeks
EPDS
EPDS
EPDS EPDS
EPDS
EPDS
Screening tool
6 weeks
6 weeks or 6 months
6–8 weeks 8 weeks
3 days and 6 weeks 6 weeks
PP timing
NR = Not reported. a Corrected the upper limit to make the confidence interval symmetric. b Additional data obtained via email correspondence from author.
Wu
Prospective cohort
Orbach-Zinger
Nahirney
Secondary analysis of a prospective cohort
Prospective cohort
Ding
Gaillard Johnstone
Study type
Author
Table 1 Summary of observational studies included in the review.
206
≥10
565
80,606
NR
>6
479 65
DSM IV ≥10
210
264 490
≥12 > 12
≥10
1503
1326
214
≥10 at 6 weeks ≥12
≥10
Total N
Diagnostic criteria
2.38 (1.38,4.10)
1.00 (0.86,1.16)
0.47 (0.27,0.8) 4.35 (0.54,35.04)a,b
0.79 (0.23,2.68)
1.45 (0.87,2.42)b
0.86 (0.69,1.07)a
1.1 (0.4,3.03)a 1.36 (0.61,3.00)b
1.52 (1.12,2.08)
0.31 (0.16,0.61)
Unadjusted OR (95% CI)
General anesthesia, local anesthetic, narcotics, nitrous oxide, nonpharmacological, spinal anesthesia, pudendal block Doula, transcutaneous electrical nerve stimulation
Nitrous oxide or intramuscular pethidine NR
No analgesia
NR
NR
Spinal anesthesia, nitrous oxide, pudendal block, paracervical block NR NR
No other form of analgesia
Non-epidural analgesia group
Yes, excluded women with a history of depression
Yes, excluded women with a history of depression
No No
No Excluded only women with a history of extreme mental illness; psychotic illness, drug or alcohol dependence history, history of severe personal dysfunction as identified by history of self-harm. Included women with a history of depression and anxiety Yes, excluded women with EPDS ≥13 at time of delivery which would indicate strong probability of preexisting depression Yes, excluded women if they were taking antidepressants during pregnancy or had a history of depression Yes, excluded women with psychiatric disorders in the decompensation stage
No
Yes, excluded women with a history of psychiatric illness
Exclusion for mental illness
J. Kountanis, et al.
Journal of Clinical Anesthesia xxx (xxxx) xxxx
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
Fig. 2. Labor epidural analgesia and the risk of postpartum depression.
of delivery. Subjects that were previously not diagnosed by eight weeks postpartum may have converted to a positive EPDS screening score at a later date. Studies should include long term follow up of subjects in order to more accurately assess the effect of an intervention on the development of postpartum depression. Although in this meta-analysis the presence of LEA was not shown to be significant in reducing the risk of PPD, individual studies did report an association and future studies should focus on other aspects involved in LEA. For example, recent literature has investigated the relationship between the quality of LEA provided and PPD [29]. Lim et al. found that LEA may be protective if a woman experiences significant pain relief. This invites further investigation into whether it is LEA or any method of adequate pain control that is protective. Conversely, LEA may increase the risk of PPD if her birth plan did not include labor epidural analgesia [22]. Many women have a desired individualized ideal birth plan which may or may not include LEA, or may have expectations of when and how LEA is given. This hypothesis is supported by the findings of Orbach-Zinger et al., which demonstrated that delivering LEA was protective against PPD when women had intentions to receive LEA. However, there was a negative additive interaction when the women did not originally plan on receiving LEA and received LEA. These findings suggest that unmet expectations, such as inadequate pain control from LEA or failing to have a planned natural childbirth, may contribute to mood disturbances postpartum. Overall, there remains a lack of literature examining associated factors of receiving labor epidural analgesia and its impact on a woman's overall birth satisfaction and development of a postpartum mood disorder. Investigation could be aimed at quality of analgesia, number of epidural top-up boluses, type of medication administered, ease of placement of LEA, or the anesthetist's communication during placement of LEA. Furthermore, studies varied in the exclusion or adjustment for women with a pre-existing history of mental illness. Exploration of the
negative attitudes towards pregnancy, breastfeeding difficulty/shorter duration/cessation, and infant outcomes that may have modified the link between LEA and PPD. This common study limitation leads to confounding and potential bias, as expected in investigating this complex disease. This also necessitated the inclusion of only unadjusted odds ratios in this meta-analysis as variation exists between studies in the adjustment of differing variables. Another limitation is that 5 studies were rated “good”, and 6 studies were assessed as “poor” based on AHRQ criteria. This is largely due to lack of comparability between cohorts, and lack of description of subjects lost to follow up. While many of the studies were well designed for their primary outcome or appropriately accounted for confounding in their adjusted analysis, they were not rated as highly due to not being designed to detect a difference between LEA and no LEA cohorts as well as our use of their unadjusted data. Furthermore, a significant source of bias was that all but two studies, Nahirney et al. and Orbach-Zinger et al., failed to account for possible antepartum depression. Nahirney et al. excluded women who had an EPDS score ≥13 at delivery. While Orbach-Zinger et al. excluded women taking antidepressants during their pregnancy, they did not assess if women were depressed prior to delivery. In order to appropriately assess the effect of LEA on PPD, it is important to consider if the patient is already suffering from depression at the time of delivery. Further limitations of this study are the variation in the diagnosis of depression. Initial research validating EPDS as a screening tool reported a value ≥10 as indicative of probable minor depression, and ≥13 as indicative of probable major depression [27,28]. Therefore, conceivably the findings of a meta-analysis could be altered if all studies used the same value for diagnosis, and there was a distinction between possible minor and major depression. In addition, all studies but two assessed for postpartum depression within the first two months after delivery. By definition, postpartum depression may occur within a year 5
6
* * * * * * * * * * *
Representative-ness of the exposed cohort: a) Truly representative (one star); b) Somewhat representative (one star); c) Selected group; d) No description of the derivation of the cohort
Selection
* * * * * * * * * * *
Selection of the nonexposed Cohort: a) Drawn from the same community as the exposed cohort (one star); b) Drawn from a different source; c) No description of the derivation of the nonexposed cohort
* * * * * * * * * * *
Ascertainment of exposure: a) Secure record (e.g., medical record) (one star); b) Structured interview (one star); c) Written self-report; d) No description; e) Other
b b b b * b b b b b b
Demonstration that outcome of interest was not present at start of study: a) Yes (one star); b) no
* * * * * * * * * * *
Assessment of Outcome: independent blind assessment (one star); b) record linkage (one star); c) selfreport; d) no description; e) other
Comparability of cohorts (LEA vs No LEA) on the basis of design when using unadjusted OR: a) The study cohorts have comparable history of mental illness (one star); b) Study cohorts have comparable known PPD risk factors (one star); c) Cohorts are not comparable on the basis of the design d) no statement
** c d d c * * d d * **
Outcome
Comparability
*Thresholds for converting the Newcastle-Ottawa scales to Agency for Healthcare Research and Quality (AHRQ) standards (good, fair, and poor): Good quality: 3 or 4 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain. Fair quality: 2 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain. Poor quality: 0 or 1 star in selection domain OR 0 stars in comparability domain OR 0 or 1 stars in outcome/exposure domain.
Ding Eckerdal Gaillard Johnstone Nahirney Orbach-Zinger Riazanova Suhitharan Tobin Wu Zhang
Author
Table 2 Newcastle-Ottawa scale quality assessment of observational studies.
* * * * * * * * * * b
Was follow up long enough for outcome to occur: a) Yes (one star); b) no
* * * * c * * * d * *
Adequacy of follow-up cohorts: a) complete follow up - all subjects accounted for (one star); b) Subjects lost to follow up unlikely to introduce bias- number lost less than or equal to 20% or description of those lost suggested no different from those followed. (one star); c) follow up rate < 80% or no description of those lost; d) no statement
8 6 6 6 6 7 7 6 5 7 7
Total quality score
Good Poor Poor Poor Poor Good Good Poor Poor Good Good
AHRQ standard*
J. Kountanis, et al.
Journal of Clinical Anesthesia xxx (xxxx) xxxx
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
o Embase: L610738777, L52929687, L53126246, L360073482, L38338865 o CINAHL Accession Number: 104949429,
impact of LEA on these high-risk populations is warranted and may yield different findings. According to this meta-analysis, the presence of labor epidural analgesia was not associated with a decreased risk of developing postpartum depression. However, the findings are limited by the quality of studies and there is a need for larger well-controlled investigations. Future studies are needed to explore whether the other aspects involved in receiving labor epidural analgesia, beyond its presence or absence, have an influence on developing postpartum depression.
Ovid MEDLINE; Ovid MEDLINE in-process & other non-indexed citations; Ovid MEDLINE Epub Ahead of Print (101 results on August 23, 2017; 110 results on December 6, 2017; 132 results on July 30, 2019). (exp depression, postpartum/ or ((maternal or postnatal or postnatal or postpartum or post-partum or prenatal or pre-natal or puerper*) adj3 depress*).tw. or (*depression/and (exp *pregnancy/or exp. *postpartum period/))) and (exp analgesia, epidural/or (analges* or epidural* or extradural* or peridural* or (pain* adj3 (manag* or relie* or treat*))).tw) Embase.com (166 results on August 23, 2017; 169 results on December 6, 2017; 215 results on July 30, 2019). (‘postnatal depression’/exp. OR (((maternal OR postnatal OR ‘postnatal’ OR postpartum OR ‘post-partum’ OR prenatal OR ‘pre-natal’ OR puerper*) NEAR/3 depress*):ti,ab)) AND (‘epidural anesthesia’/exp OR ‘epidural analgesia’/exp OR (analges* OR epidural* OR extradural* OR peridural* OR (pain* NEAR/3 (manag* OR relie* OR treat*))): ab,ti) Cochrane Central Register of Controlled Trials (27 results on August 23, 2017; 30 results on December 6, 2017; 62 results on July 30, 2019).
Abbreviations EPDS LEA PPD AHRQ
Edinburgh Postnatal Depression Scale Labor epidural analgesia Postpartum depression Agency for Healthcare Research and Quality
Declarations Ethics approval – N/A. Consent for publication – N/A. Availability of data and materials All data generated or analyzed during this study are included in this published article.
1. [mh “depression, postpartum”] or ((maternal or postnatal or “postnatal” or postpartum or “post-partum” or prenatal or “pre-natal” or puerper*) near/3 depress*):ab,ti or ([mh depression] and ([mh pregnancy] or [mh “postpartum period”])). 2. [mh “analgesia, epidural”] or (analgesi* or epidural* or extradural* or peridural* or (pain* near/3 (manag* or relief* or treat*))): ab,ti. 3. #1 AND #2.
Authors' contributions JK was a major contributor in reviewing and selecting articles, and writing the manuscript. CV was a major contributor in reviewing and selecting articles, and writing the manuscript. SB assisted with manuscript authorship. MM assisted with manuscript authorship. RC performed all statistical analyses and contributed to writing the manuscript. CA assisted with article collection, organization, and creation of tables and images. MM performed literature searches and contributed to writing the manuscript. MB was a major contributor in manuscript design and writing the manuscript. All authors read and approved the final manuscript.
CINAHL (61 results on August 25, 2017; 62 results on December 6, 2017; 77 results on July 30, 2019). S1 TI ((maternal OR postnatal OR “post-natal” OR postpartum OR “post-partum” OR prenatal OR “pre-natal” OR puerper*) N3 depress*) OR AB ((maternal OR postnatal OR “post-natal” OR postpartum OR “post-partum” OR prenatal OR “pre-natal” OR puerper*) N3 depress*). S2 MH “depression, postpartum” OR MH “edinburgh postnatal depression scale” OR (MH depression AND (MH pregnancy OR MH “postpartum period”)). S3 S1 OR S2. S4 TI (analges* OR epidural* OR extradural* OR peridural* OR (pain* N3 (manag* OR relie* OR treat*))) OR AB (analges* OR epidural* OR extradural* OR peridural* OR (pain* N3 (manag* OR relie* OR treat*))). S5 MH “analgesia, epidural”. S6 S4 OR S5. S7 S3 AND S6.
Disclosures This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Declaration of competing interest The authors declare that they have no competing interests. Acknowledgements The authors are grateful to Dr. Philip Boyce, Dr. Sharon OrbachZinger, Dr. Catherine Tobin, and Dr. Bethany Wolf for providing additional data from their original work.
References
Appendix A. Search strategies
[1] Weissman M, Olfson M. Depression in women: implications for health care research. Science 1995;269:799–801. [2] Gavin NI, Gaynes BN, Lohr KN, Meltzer-Brody S, Gartlehner G, Swinson T. Perinatal depression: a systematic review of prevalence and incidence. Obstet Gynecol 2005;106:1071–83. [3] Koy JY, Rockill KM, Tong VT, Morrow B, Farr SL. Trends in postpartum depressive symtpoms – 27 states, 2004, 2008, and 2012. MMWR Morb Mortal Wkly Rep 2017;66:153–8. [4] Dagher RK, McGovern PM, Dowd BE, Gjerdingen DK. Postpartum depression and health services expenditures among employed women. J Occup Environ Med 2012;54:210–5. [5] Grace SL, Evindar A, Stewart DE. The effect of postpartum depression on child cognitive development and behavior: a review and critical analysis of the literature. Arch Womens Ment Health 2003;6:263–74.
Reproducible search strategies
• Searches run on August 23, 2017, December 6, 2017, and July 30, 2019
Epidural, labor pain, postpartum depression
• Sentinel articles:
o Ovid: 24370337, 24797120, 20831438, 14995921 7
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
[19] Eckerdal P, Kollia N, Karlsson L, Skoog-Svanberg A, Wikstrom A-K, Hogberg U, et al. Epidural analgesia during childbirth and postpartum depressive symptoms: a population-based longitudinal cohort study. Anesth Analg: Epub ahead of print; 2019 Jun 24. [20] Gaillard A, Le Strat Y, Mandelbrot L, Keita H, Dubertret C. Predictors of postpartum depression: prospective study of 264 women followed during pregnancy and postpartum. Psychiatry Res 2014;215:341–6. [21] Johnstone SJ, Boyce PM, Hickey AR, Morris-Yates AD, Harris MG. Obstetric risk factors for postnatal depression in urban and rural community samples. Aust N Z J Psychiatry 2001;35:69–74. [22] Orbach-Zinger S, Landau R, Harousch AB, et al. The relationship between women’s intention to request a labor epidural analgesia, actually delivering with labor epidural analgesia, and postpartum depression at 6 weeks: a prospective observational study. Anesth Analg 2018;126:1590–7. [23] Riazanova OV, Alexandrovich YS, Ioscovich AM. The relationship between labor pain management, cortisol level and risk of postpartum depression development: a prospective nonrandomized observational monocentric trial. Rom J of Anaesth Intensive Care. 2018;25:123. [24] Suhitharan T, Pham TPT, Chen H, et al. Investigating analgesic and psychological factors associated with risk of postpartum depression development: a case–control study. Neuropsychiatric Disease and Treatment 2016;12:1333–9. [25] Wu YM, McArthur E, Dixon S, Dirk JS, Welk BK. Association between intrapartum epidural use and maternal postpartum depression presenting for medical care: a population-based. Matched Cohort Study. Int J Obstet Anesth 2018;35:10–6. [26] Zhang Y, Johnstone L, Ma D, Wang F, Zheng X, Xu X. An exploratory study of the effect of labor pain management on postpartum depression among Chinese women. Ginekol Pol. 2018;89:627–36. [27] Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh postnatal depression scale. Br J Psychiatry 1987;150:782–6. [28] Matthey S, Henshaw C, Elliott S, Barnett B. Variability in use of cut-off scores and formats on the Edinburgh postnatal depression scale: implications for clinical and research practice. Arch Womens Ment Health 2006;9:309–15. [29] Lim G, Farrell LM, Facco FL, Gold MS, Wasan AD. Labor analgesia as a predictor for reduced postpartum depression scores: a retrospective observational study. Anesth Analg 2018;126:1598–605.
[6] Muzik M, Bocknek EL, Broderick A, et al. Mother-infant bonding impairment across the first six months postpartum: the primacy of psychopathology in women with childhood abuse and neglect histories. Arch Womens Ment Health 2013 February;16(1):29–38. [7] Marcus SM. Depression during pregnancy: rates, risks and consequences–motherisk update 2008. Can J Clin Pharmacol 2009;16:e15–22. [8] Hawton K, van Heeringen K. Suicide. Lancet 2009;373:1372–81. [9] Metz TD, Rovner P, Hoffman MC, Allshouse AA, Beckwith KM, Binswanger IA. Maternal deaths from suicide and overdose in Colorado, 2004–2012. Obstet Gynecol 2016;128:1233–40. [10] Knight M, Nair M, Tuffnell D, Shakespeare J, Kenyon S, Kurinczuk JJ, editors. Saving Lives, improving mothers’ care-lessons learned to inform maternity care from the UK and Ireland confidential enquiries into maternal deaths and morbidity 2014–16. Oxford: National Perinatal Epidemiology Unit, University of Oxford; 2018. [11] O’Hara MW, Swain AM. Rates and risk of postpartum depression—a meta-analysis. Int Rev Psychiatry 1996;8:37–54. [12] Robertson E, Grace S, Wallington T, Stewart DE. Antenatal risk factors for postpartum depression: a synthesis of recent literature. Gen Hosp Psychiatry 2004;26:289–95. [13] Screening for perinatal depression. ACOG committee opinion no. 757. American College of Obstetricians and Gynecologists. Obstet Gynecol 2018;132:e208–12. [14] USPTF Interventions to Prevent Perinatal Depression US Preventive Services Task Force Recommendation Statement JAMA. 2019;321(6):580-587. [15] Boudou M, Teissèdre F, Walburg V, Chabrol H. Association between the intensity of childbirth pain and the intensity of postpartum blues. L Encéphale 2007;33:805–10. [16] 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. [17] Tobin CD, Wilson SH, Hebbar L, Roberts LL, Wolf BJ, Guille C. Labor epidural analgesia and postpartum depression. Arch Depress Anxiety 2016;2:44–6. [18] Nahirney M, Metcalfe A, Chaput KH. Administration of epidural labor analgesia is not associated with a decreased risk of postpartum depression in an urban Canadian population of mothers: a secondary analysis of prospective cohort data. Local Reg Anesth 2017;10:99–104.
8
Contents lists available at ScienceDirect
Journal of Clinical Anesthesia journal homepage: www.elsevier.com/locate/jclinane
Original contribution
Labor epidural analgesia and the risk of postpartum depression: A metaanalysis of observational studies ⁎
Joanna Kountanis (MD)a, , Christie Vahabzadeh (MD)a, Samuel Bauer (MD)b, Maria Muzik (MD)c, Ruth Cassidy (MS)a, Casey Aman (BS)a, Mark MacEachern (MLIS)d, Melissa Bauer (DO)a a
Department of Anesthesiology, University of Michigan Health System, 1500 East Medical Center Drive, Ann Arbor, MI 48109, United States of America Department of Obstetrics and Gynecology, Oakland University William Beaumont School of Medicine, 586 Pioneer Drive, Rochester, MI 48309, United States of America c Department of Psychiatry, University of Michigan Health System, 1500 East Medical Center Drive, Ann Arbor, MI 48109, United States of America d University of Michigan Taubman Health Sciences Library, 1135 Catherine Street, Ann Arbor, MI 48109, United States of America b
A R T I C LE I N FO
A B S T R A C T
Keywords: Analgesia Epidural Depression Postpartum Labor pain Obstetrics Risk factors
Study objective: This study aims to systematically review the literature to evaluate the association between labor epidural analgesia (LEA) and postpartum depression (PPD). Design: Meta-analysis. Setting: Obstetric patients delivering vaginally with or without LEA in a hospital. Interventions: This study aimed to investigate the effects of providing LEA on developing PPD. Measurements: Pooled odds ratios (OR) and 95% confidence intervals (CI) were calculated using the random effects model. Results: A total of 356 full text articles were reviewed. Eleven articles studying 85,928 patients met inclusion criteria. The pooled unadjusted OR 1.03 and 95% CI (0.77, 1.37) suggest that LEA is not associated with a decreased risk of developing PPD. Conclusions: Labor epidural analgesia was not shown to confer protection against developing PPD according to this meta-analysis. Future studies are needed to explore whether other aspects of LEA, beyond its presence or absence, influence the onset of PPD.
1. Introduction Developing a psychological sequela from childbirth is one of the most frequent complications seen in women postpartum [1]. As many as one in five mothers experience postpartum depression, with the incidence of postpartum depression (PPD) ranging from 11% to as high as 19.2% within the first three months [2,3]. The impact of suffering from a mental illness or subsyndromal symptoms postpartum has consequences for mother and child, and is a financial burden to the health care system. In a cost analysis study, women with PPD incur up to 90% higher health services expenditures than non-depressed women for themselves, as well as their infant, after delivery [4]. Children born to mothers with PPD also suffer negative health consequences; mothers suffering from PPD may have impaired maternal-infant bonding or have children with lower cognitive and behavioral development up to 5 years old [5,6]. It has been suggested that even if the mother
subsequently gets treatment, there could be long lasting mental health consequences for her child [7]. Postpartum depression also has direct consequences to the health and safety of the mother. Depression has been associated with an increased risk of suicide. Approximately 4% of depressed individuals die from suicide [8]. A recent review of total maternal deaths in Colorado from 2004 to 2012 demonstrated that the leading cause of death was due to self-harm, comprising 30% of all maternal deaths and with the majority occurring postpartum [9]. The Mothers and Babies: Reducing Risk through Audits and Confidential Enquiries across the United Kingdom 2014–2016 report cites 31% of maternal deaths between six weeks and one year after pregnancy were from mental-health related causes (suicide, drug/alcohol, other psychiatric causes). Maternal suicide is the leading cause of direct deaths in this time period [10]. Therefore, the sequelae from PPD include adverse effects to maternal and neonatal health, in addition to contributing to mortality.
⁎
Corresponding author at: 1500 East Medical Center Drive SPC 5278, University Hospital South, L3626 Ann Arbor, MI 48109-5278, United States of America. E-mail addresses: [email protected] (J. Kountanis), [email protected] (C. Vahabzadeh), [email protected] (S. Bauer), [email protected] (M. Muzik), [email protected] (R. Cassidy), [email protected] (C. Aman), [email protected] (M. MacEachern), [email protected] (M. Bauer). https://doi.org/10.1016/j.jclinane.2019.109658 Received 24 June 2019; Received in revised form 9 September 2019; Accepted 16 November 2019 0952-8180/ © 2019 Elsevier Inc. All rights reserved.
Please cite this article as: Joanna Kountanis, et al., Journal of Clinical Anesthesia, https://doi.org/10.1016/j.jclinane.2019.109658
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
Google Translate to assess whether the subject matter met inclusion criteria. The quality of individual selected studies was assessed in relation to its ability to investigate the association between LEA and PPD with unadjusted data using the Newcastle Ottawa Quality Assessment scale. Therefore, well-designed studies may be assessed as “poor” quality if the study was not primarily designed for the outcome of interest in this meta-analysis. In addition, studies appropriately controlling for multiple confounding factors in their adjusted analysis may also be rated poorly due to this meta-analysis's use of their unadjusted data. Each study was graded based on selection of an exposed and non-exposed cohort that were truly representative of an appropriate study population, that cohorts were drawn from the same community, ascertainment of exposure, demonstration that outcome of interest was not present at start of study, comparability of the exposed with the unexposed cohort, assessment of outcome, follow up long enough for outcome to occur, and lost to follow up rate. The Newcastle Ottawa Quality Assessment rating for each study was then converted to the Agency for Healthcare Research and Quality standard (good, fair, and poor). To estimate the effect size, most studies reported either unadjusted odds ratios and 95% confidence intervals or appropriate counts for which to calculate odds ratio and confidence interval estimates. We directly contacted authors to obtain additional information if not available in the original publication. To obtain a pooled measure of association between LEA and incidence of PPD, we calculated a pooled unadjusted odds ratio using a random effects logistic model. Unadjusted odds ratios were collected for each study by review of the manuscript or by using raw data obtained by direct email correspondence with the corresponding author. A random effects model was utilized rather than a fixed effects model to account for heterogeneity between studies due to the fact that included studies had been conducted under variable conditions with variable patient populations. The fixed effects model is chosen when studies have been conducted under similar conditions and have a common effect size, which was not appropriate for this meta-analysis. To assess pooled effect size and statistical significance at alpha = 0.05, associated 95% confidence intervals were calculated. We performed a test for heterogeneity using the Q-statistic, and calculated the I2 statistic and associated confidence interval. In cases where the reported odds ratio confidence interval is not symmetric around the odds ratio, the upper limit was recalculated based on the distance between the odds ratio and the lower confidence limit. All statistical analyses were performed using Comprehensive Meta-Analysis Version 3.3.070 (Biostat, Inc. Englewood, NJ, USA) and plots were made using R 3.1.1 (The R Foundation, Vienna, Austria).
A prior history of mental illness is one of the strongest predictors of postpartum psychological pathology [11,12]. The development of PPD is multi-factorial and encompasses risks such as low levels of social support, breast feeding problems, and neonatal intensive care admission [13]. However, the current body of literature is inconsistent in identifying all risk factors. Despite studies reporting associations between labor pain and PPD, American College of Obstetrics and Gynecology (ACOG) and the United States Preventive Services Task Force do not consider labor pain as a risk factor for developing PPD [13,14]. Boudou et al. demonstrated a positive association between the intensity of childbirth pain and severity of postpartum blues. The authors speculated possible explanations as an evoked stress response from intense pain as well as possible maternal feelings of failure from not overcoming the pain [15]. Ding et al. also speculated that severe labor pain may induce a strong stress stimulus, resulting in neuropsychological consequences postpartum [16]. This theory supported his finding that the presence of labor epidural analgesia (LEA) was protective against the development of postpartum depression. However other studies contradict these findings [17,18]. It would seem prudent to identify whether a low risk procedure such as LEA could be protective since other known risk factors for developing PPD, such as poor social support or a history of depression, are not modifiable. This meta-analysis aims to evaluate the association of labor epidural analgesia and the development of post-partum depression. 2. Methods This study was registered on the international prospective register of systematic reviews, PROSPERO (CRD42017079565). The meta-analysis was conducted and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. Three Ovid MEDLINE databases (MEDLINE, In-Process & Other Non-Indexed Citations, and Epub Ahead of Print), Embase.com, CINAHL via EBSCOhost, Cochrane Central Register of Controlled Trials via Wiley, and clinicaltrials.gov were searched on August 23, 2017 to identify observational studies evaluating the presence of LEA and subsequent development of PPD. Studies were included even if their primary outcome was not investigating the association of LEA and PPD. The search was conducted again on December 6, 2017 and July 30, 2019 to ensure recent publications were identified. The search was conducted using a combination of controlled terms (MeSH or EMTREE) and title or title and abstract text words. The searches included sentinel articles that were identified by the first author in an initial literature review. A citation and bibliography search of all citations associated with the included studies was conducted in Clarivate Web of Sciences and Google Scholar. Duplicate citations were identified and discarded in Endnote X6 (Clarivate Analytics) and hand review of citations. The complete searches can be found in Appendix A. Two reviewers independently reviewed relevant publications that were found in the search for inclusion articles (JK and CV). Bibliographies of selected articles and articles citing the selected articles were screened for inclusion. Once relevant articles were identified, two authors extracted data and discrepancies were resolved by discussion. Data was extracted from studies providing presence or absence of LEA and Edinburgh postnatal depression scale scores (EPDS), Beck Depression Inventory, Diagnostic and Statistical Manual of Mental Disorders (DSM) diagnosis of postpartum depression, or other validated methods of identifying PPD. Inclusion criteria were observational studies reporting the presence or absence of LEA, participants > 18 years of age, and subsequent validated methods of identifying PPD. Exclusion criteria were studies not specifying the type of assessment for postpartum depression, unspecified timing of the assessment, unclear separation between those receiving and not receiving epidural analgesia, case reports, case series, editorials, and correspondences. The titles and text of foreign language articles were placed into
3. Results A total of 356 full text articles were reviewed after the search was run on August 23, 2017, December 6, 2017, and July 30, 2019. Fig. 1 shows the selection process. Google Translate could translate all studies written in a foreign language and none were found to meet inclusion criteria. Eleven articles with 85,928 patients met inclusion criteria [16–26]. (Table 1) All studies identified probable postpartum depression within the first year postpartum utilizing the EPDS screening tool, expect for Wu et al. which used health care services for depression as their criteria. The diagnosis of postpartum depression varied between studies with some using EPDS scores ≥10, > 12, ≥12, > 6, DSM IV criteria, or utilization of health care services for depression. (Table 1) Nine studies were designed as prospective observational, Suhitharan et al. designed a retrospective case control, and Wu et al. designed a population based matched cohort. The cohort without LEA varied in reporting of other forms of analgesia provided, including not reported in the manuscript, nitrous oxide, spinal, pudendal block, paracervical block, doula support, narcotics, transcutaneous electrical nerve stimulation, pethidine, and no other form of analgesia provided. 2
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
Fig. 1. Flow chart of search strategy.
several strengths. These include appropriate eligibility criteria, clear reporting of the presence of LEA, and using a verified screening tool to detect PPD. Subject inclusion criteria was liberal and characteristics between groups similar among all studies, as well as all studies appropriately measured exposure and outcome. Another strength of this study was to reduce the effect of publication bias by including peer reviewed Internet based journal articles that were found on supplemental searches outside PubMed in the analysis. Furthermore, all studies published in a non-English language were also considered due to the ability to translate the text. However, heterogeneity was a concern given each study measured a variety of confounding risk factors when testing the relationship between LEA and PDD. For example, Ding et al. controlled for risk factors such as satisfaction with medical service and attendance of a childbirth class, whereas Orbach-Zinger et al. controlled for marital status. In addition, some studies controlled for confounding between cohorts that had PPD and those who did not have PPD, and others controlled for confounding between cohorts that had LEA and those that did not. Due to the complexity of PPD and its development, it is near impossible for each study to adequately control for all known prognostic factors. There are many other suspected risk factors that were left unmeasured by many studies; high levels of non-pregnancy related stresses, poor social support, history of physical or sexual abuse, unintended pregnancy,
Analysis of Johnstone et al., Orbach-Zinger et al., and Tobin et al. was accomplished by obtaining unpublished data from the corresponding authors to calculate unadjusted odd ratios. The pooled unadjusted OR 1.03 and associated 95% CI (0.77, 1.37) suggest that there is insufficient evidence to conclude that LEA is significantly associated with PPD. (Fig. 2) The Q-statistic of 15.994 with 10 degrees of freedom supports the assumption of heterogeneity at a significance level of 0.10 (p-value = 0.0998). The estimate and 95% uncertainty interval for I2 is 37.48 (0–69.25). The quality assessment assigning “good” for Ding et al., OrbachZinger et al., Riazanova et al., Wu et al., and Zhang et al. compared to the other studies assigned as “poor” was primarily distinguished by cohort comparability when using unadjusted OR and adequacy of follow up (Table 2).
4. Discussion This meta-analysis aimed to review and quantify the strength of the association between presence of LEA and onset of PPD among women. According to the analysis, LEA was not shown to confer protection against developing PPD. This negative finding concurs with the majority of studies in this analysis. The individual studies included in this meta-analysis also have 3
Population based longitudinal cohort Prospective cohort Prospective cohort
Eckerdal
4
Prospective nonrandomized observational trial Retrospective case control Secondary analysis of a prospective cohort Population-based matched cohort
Prospective cohort
Riazanova
Suhitharan Tobin
Zhang
2–4 weeks
Up to one year
EPDS
Seeking medical care for depression
EPDS EPDS
EPDS
6 weeks
4–8 weeks 6–8 weeks
EPDS
EPDS
EPDS EPDS
EPDS
EPDS
Screening tool
6 weeks
6 weeks or 6 months
6–8 weeks 8 weeks
3 days and 6 weeks 6 weeks
PP timing
NR = Not reported. a Corrected the upper limit to make the confidence interval symmetric. b Additional data obtained via email correspondence from author.
Wu
Prospective cohort
Orbach-Zinger
Nahirney
Secondary analysis of a prospective cohort
Prospective cohort
Ding
Gaillard Johnstone
Study type
Author
Table 1 Summary of observational studies included in the review.
206
≥10
565
80,606
NR
>6
479 65
DSM IV ≥10
210
264 490
≥12 > 12
≥10
1503
1326
214
≥10 at 6 weeks ≥12
≥10
Total N
Diagnostic criteria
2.38 (1.38,4.10)
1.00 (0.86,1.16)
0.47 (0.27,0.8) 4.35 (0.54,35.04)a,b
0.79 (0.23,2.68)
1.45 (0.87,2.42)b
0.86 (0.69,1.07)a
1.1 (0.4,3.03)a 1.36 (0.61,3.00)b
1.52 (1.12,2.08)
0.31 (0.16,0.61)
Unadjusted OR (95% CI)
General anesthesia, local anesthetic, narcotics, nitrous oxide, nonpharmacological, spinal anesthesia, pudendal block Doula, transcutaneous electrical nerve stimulation
Nitrous oxide or intramuscular pethidine NR
No analgesia
NR
NR
Spinal anesthesia, nitrous oxide, pudendal block, paracervical block NR NR
No other form of analgesia
Non-epidural analgesia group
Yes, excluded women with a history of depression
Yes, excluded women with a history of depression
No No
No Excluded only women with a history of extreme mental illness; psychotic illness, drug or alcohol dependence history, history of severe personal dysfunction as identified by history of self-harm. Included women with a history of depression and anxiety Yes, excluded women with EPDS ≥13 at time of delivery which would indicate strong probability of preexisting depression Yes, excluded women if they were taking antidepressants during pregnancy or had a history of depression Yes, excluded women with psychiatric disorders in the decompensation stage
No
Yes, excluded women with a history of psychiatric illness
Exclusion for mental illness
J. Kountanis, et al.
Journal of Clinical Anesthesia xxx (xxxx) xxxx
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
Fig. 2. Labor epidural analgesia and the risk of postpartum depression.
of delivery. Subjects that were previously not diagnosed by eight weeks postpartum may have converted to a positive EPDS screening score at a later date. Studies should include long term follow up of subjects in order to more accurately assess the effect of an intervention on the development of postpartum depression. Although in this meta-analysis the presence of LEA was not shown to be significant in reducing the risk of PPD, individual studies did report an association and future studies should focus on other aspects involved in LEA. For example, recent literature has investigated the relationship between the quality of LEA provided and PPD [29]. Lim et al. found that LEA may be protective if a woman experiences significant pain relief. This invites further investigation into whether it is LEA or any method of adequate pain control that is protective. Conversely, LEA may increase the risk of PPD if her birth plan did not include labor epidural analgesia [22]. Many women have a desired individualized ideal birth plan which may or may not include LEA, or may have expectations of when and how LEA is given. This hypothesis is supported by the findings of Orbach-Zinger et al., which demonstrated that delivering LEA was protective against PPD when women had intentions to receive LEA. However, there was a negative additive interaction when the women did not originally plan on receiving LEA and received LEA. These findings suggest that unmet expectations, such as inadequate pain control from LEA or failing to have a planned natural childbirth, may contribute to mood disturbances postpartum. Overall, there remains a lack of literature examining associated factors of receiving labor epidural analgesia and its impact on a woman's overall birth satisfaction and development of a postpartum mood disorder. Investigation could be aimed at quality of analgesia, number of epidural top-up boluses, type of medication administered, ease of placement of LEA, or the anesthetist's communication during placement of LEA. Furthermore, studies varied in the exclusion or adjustment for women with a pre-existing history of mental illness. Exploration of the
negative attitudes towards pregnancy, breastfeeding difficulty/shorter duration/cessation, and infant outcomes that may have modified the link between LEA and PPD. This common study limitation leads to confounding and potential bias, as expected in investigating this complex disease. This also necessitated the inclusion of only unadjusted odds ratios in this meta-analysis as variation exists between studies in the adjustment of differing variables. Another limitation is that 5 studies were rated “good”, and 6 studies were assessed as “poor” based on AHRQ criteria. This is largely due to lack of comparability between cohorts, and lack of description of subjects lost to follow up. While many of the studies were well designed for their primary outcome or appropriately accounted for confounding in their adjusted analysis, they were not rated as highly due to not being designed to detect a difference between LEA and no LEA cohorts as well as our use of their unadjusted data. Furthermore, a significant source of bias was that all but two studies, Nahirney et al. and Orbach-Zinger et al., failed to account for possible antepartum depression. Nahirney et al. excluded women who had an EPDS score ≥13 at delivery. While Orbach-Zinger et al. excluded women taking antidepressants during their pregnancy, they did not assess if women were depressed prior to delivery. In order to appropriately assess the effect of LEA on PPD, it is important to consider if the patient is already suffering from depression at the time of delivery. Further limitations of this study are the variation in the diagnosis of depression. Initial research validating EPDS as a screening tool reported a value ≥10 as indicative of probable minor depression, and ≥13 as indicative of probable major depression [27,28]. Therefore, conceivably the findings of a meta-analysis could be altered if all studies used the same value for diagnosis, and there was a distinction between possible minor and major depression. In addition, all studies but two assessed for postpartum depression within the first two months after delivery. By definition, postpartum depression may occur within a year 5
6
* * * * * * * * * * *
Representative-ness of the exposed cohort: a) Truly representative (one star); b) Somewhat representative (one star); c) Selected group; d) No description of the derivation of the cohort
Selection
* * * * * * * * * * *
Selection of the nonexposed Cohort: a) Drawn from the same community as the exposed cohort (one star); b) Drawn from a different source; c) No description of the derivation of the nonexposed cohort
* * * * * * * * * * *
Ascertainment of exposure: a) Secure record (e.g., medical record) (one star); b) Structured interview (one star); c) Written self-report; d) No description; e) Other
b b b b * b b b b b b
Demonstration that outcome of interest was not present at start of study: a) Yes (one star); b) no
* * * * * * * * * * *
Assessment of Outcome: independent blind assessment (one star); b) record linkage (one star); c) selfreport; d) no description; e) other
Comparability of cohorts (LEA vs No LEA) on the basis of design when using unadjusted OR: a) The study cohorts have comparable history of mental illness (one star); b) Study cohorts have comparable known PPD risk factors (one star); c) Cohorts are not comparable on the basis of the design d) no statement
** c d d c * * d d * **
Outcome
Comparability
*Thresholds for converting the Newcastle-Ottawa scales to Agency for Healthcare Research and Quality (AHRQ) standards (good, fair, and poor): Good quality: 3 or 4 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain. Fair quality: 2 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain. Poor quality: 0 or 1 star in selection domain OR 0 stars in comparability domain OR 0 or 1 stars in outcome/exposure domain.
Ding Eckerdal Gaillard Johnstone Nahirney Orbach-Zinger Riazanova Suhitharan Tobin Wu Zhang
Author
Table 2 Newcastle-Ottawa scale quality assessment of observational studies.
* * * * * * * * * * b
Was follow up long enough for outcome to occur: a) Yes (one star); b) no
* * * * c * * * d * *
Adequacy of follow-up cohorts: a) complete follow up - all subjects accounted for (one star); b) Subjects lost to follow up unlikely to introduce bias- number lost less than or equal to 20% or description of those lost suggested no different from those followed. (one star); c) follow up rate < 80% or no description of those lost; d) no statement
8 6 6 6 6 7 7 6 5 7 7
Total quality score
Good Poor Poor Poor Poor Good Good Poor Poor Good Good
AHRQ standard*
J. Kountanis, et al.
Journal of Clinical Anesthesia xxx (xxxx) xxxx
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
o Embase: L610738777, L52929687, L53126246, L360073482, L38338865 o CINAHL Accession Number: 104949429,
impact of LEA on these high-risk populations is warranted and may yield different findings. According to this meta-analysis, the presence of labor epidural analgesia was not associated with a decreased risk of developing postpartum depression. However, the findings are limited by the quality of studies and there is a need for larger well-controlled investigations. Future studies are needed to explore whether the other aspects involved in receiving labor epidural analgesia, beyond its presence or absence, have an influence on developing postpartum depression.
Ovid MEDLINE; Ovid MEDLINE in-process & other non-indexed citations; Ovid MEDLINE Epub Ahead of Print (101 results on August 23, 2017; 110 results on December 6, 2017; 132 results on July 30, 2019). (exp depression, postpartum/ or ((maternal or postnatal or postnatal or postpartum or post-partum or prenatal or pre-natal or puerper*) adj3 depress*).tw. or (*depression/and (exp *pregnancy/or exp. *postpartum period/))) and (exp analgesia, epidural/or (analges* or epidural* or extradural* or peridural* or (pain* adj3 (manag* or relie* or treat*))).tw) Embase.com (166 results on August 23, 2017; 169 results on December 6, 2017; 215 results on July 30, 2019). (‘postnatal depression’/exp. OR (((maternal OR postnatal OR ‘postnatal’ OR postpartum OR ‘post-partum’ OR prenatal OR ‘pre-natal’ OR puerper*) NEAR/3 depress*):ti,ab)) AND (‘epidural anesthesia’/exp OR ‘epidural analgesia’/exp OR (analges* OR epidural* OR extradural* OR peridural* OR (pain* NEAR/3 (manag* OR relie* OR treat*))): ab,ti) Cochrane Central Register of Controlled Trials (27 results on August 23, 2017; 30 results on December 6, 2017; 62 results on July 30, 2019).
Abbreviations EPDS LEA PPD AHRQ
Edinburgh Postnatal Depression Scale Labor epidural analgesia Postpartum depression Agency for Healthcare Research and Quality
Declarations Ethics approval – N/A. Consent for publication – N/A. Availability of data and materials All data generated or analyzed during this study are included in this published article.
1. [mh “depression, postpartum”] or ((maternal or postnatal or “postnatal” or postpartum or “post-partum” or prenatal or “pre-natal” or puerper*) near/3 depress*):ab,ti or ([mh depression] and ([mh pregnancy] or [mh “postpartum period”])). 2. [mh “analgesia, epidural”] or (analgesi* or epidural* or extradural* or peridural* or (pain* near/3 (manag* or relief* or treat*))): ab,ti. 3. #1 AND #2.
Authors' contributions JK was a major contributor in reviewing and selecting articles, and writing the manuscript. CV was a major contributor in reviewing and selecting articles, and writing the manuscript. SB assisted with manuscript authorship. MM assisted with manuscript authorship. RC performed all statistical analyses and contributed to writing the manuscript. CA assisted with article collection, organization, and creation of tables and images. MM performed literature searches and contributed to writing the manuscript. MB was a major contributor in manuscript design and writing the manuscript. All authors read and approved the final manuscript.
CINAHL (61 results on August 25, 2017; 62 results on December 6, 2017; 77 results on July 30, 2019). S1 TI ((maternal OR postnatal OR “post-natal” OR postpartum OR “post-partum” OR prenatal OR “pre-natal” OR puerper*) N3 depress*) OR AB ((maternal OR postnatal OR “post-natal” OR postpartum OR “post-partum” OR prenatal OR “pre-natal” OR puerper*) N3 depress*). S2 MH “depression, postpartum” OR MH “edinburgh postnatal depression scale” OR (MH depression AND (MH pregnancy OR MH “postpartum period”)). S3 S1 OR S2. S4 TI (analges* OR epidural* OR extradural* OR peridural* OR (pain* N3 (manag* OR relie* OR treat*))) OR AB (analges* OR epidural* OR extradural* OR peridural* OR (pain* N3 (manag* OR relie* OR treat*))). S5 MH “analgesia, epidural”. S6 S4 OR S5. S7 S3 AND S6.
Disclosures This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Declaration of competing interest The authors declare that they have no competing interests. Acknowledgements The authors are grateful to Dr. Philip Boyce, Dr. Sharon OrbachZinger, Dr. Catherine Tobin, and Dr. Bethany Wolf for providing additional data from their original work.
References
Appendix A. Search strategies
[1] Weissman M, Olfson M. Depression in women: implications for health care research. Science 1995;269:799–801. [2] Gavin NI, Gaynes BN, Lohr KN, Meltzer-Brody S, Gartlehner G, Swinson T. Perinatal depression: a systematic review of prevalence and incidence. Obstet Gynecol 2005;106:1071–83. [3] Koy JY, Rockill KM, Tong VT, Morrow B, Farr SL. Trends in postpartum depressive symtpoms – 27 states, 2004, 2008, and 2012. MMWR Morb Mortal Wkly Rep 2017;66:153–8. [4] Dagher RK, McGovern PM, Dowd BE, Gjerdingen DK. Postpartum depression and health services expenditures among employed women. J Occup Environ Med 2012;54:210–5. [5] Grace SL, Evindar A, Stewart DE. The effect of postpartum depression on child cognitive development and behavior: a review and critical analysis of the literature. Arch Womens Ment Health 2003;6:263–74.
Reproducible search strategies
• Searches run on August 23, 2017, December 6, 2017, and July 30, 2019
Epidural, labor pain, postpartum depression
• Sentinel articles:
o Ovid: 24370337, 24797120, 20831438, 14995921 7
Journal of Clinical Anesthesia xxx (xxxx) xxxx
J. Kountanis, et al.
[19] Eckerdal P, Kollia N, Karlsson L, Skoog-Svanberg A, Wikstrom A-K, Hogberg U, et al. Epidural analgesia during childbirth and postpartum depressive symptoms: a population-based longitudinal cohort study. Anesth Analg: Epub ahead of print; 2019 Jun 24. [20] Gaillard A, Le Strat Y, Mandelbrot L, Keita H, Dubertret C. Predictors of postpartum depression: prospective study of 264 women followed during pregnancy and postpartum. Psychiatry Res 2014;215:341–6. [21] Johnstone SJ, Boyce PM, Hickey AR, Morris-Yates AD, Harris MG. Obstetric risk factors for postnatal depression in urban and rural community samples. Aust N Z J Psychiatry 2001;35:69–74. [22] Orbach-Zinger S, Landau R, Harousch AB, et al. The relationship between women’s intention to request a labor epidural analgesia, actually delivering with labor epidural analgesia, and postpartum depression at 6 weeks: a prospective observational study. Anesth Analg 2018;126:1590–7. [23] Riazanova OV, Alexandrovich YS, Ioscovich AM. The relationship between labor pain management, cortisol level and risk of postpartum depression development: a prospective nonrandomized observational monocentric trial. Rom J of Anaesth Intensive Care. 2018;25:123. [24] Suhitharan T, Pham TPT, Chen H, et al. Investigating analgesic and psychological factors associated with risk of postpartum depression development: a case–control study. Neuropsychiatric Disease and Treatment 2016;12:1333–9. [25] Wu YM, McArthur E, Dixon S, Dirk JS, Welk BK. Association between intrapartum epidural use and maternal postpartum depression presenting for medical care: a population-based. Matched Cohort Study. Int J Obstet Anesth 2018;35:10–6. [26] Zhang Y, Johnstone L, Ma D, Wang F, Zheng X, Xu X. An exploratory study of the effect of labor pain management on postpartum depression among Chinese women. Ginekol Pol. 2018;89:627–36. [27] Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh postnatal depression scale. Br J Psychiatry 1987;150:782–6. [28] Matthey S, Henshaw C, Elliott S, Barnett B. Variability in use of cut-off scores and formats on the Edinburgh postnatal depression scale: implications for clinical and research practice. Arch Womens Ment Health 2006;9:309–15. [29] Lim G, Farrell LM, Facco FL, Gold MS, Wasan AD. Labor analgesia as a predictor for reduced postpartum depression scores: a retrospective observational study. Anesth Analg 2018;126:1598–605.
[6] Muzik M, Bocknek EL, Broderick A, et al. Mother-infant bonding impairment across the first six months postpartum: the primacy of psychopathology in women with childhood abuse and neglect histories. Arch Womens Ment Health 2013 February;16(1):29–38. [7] Marcus SM. Depression during pregnancy: rates, risks and consequences–motherisk update 2008. Can J Clin Pharmacol 2009;16:e15–22. [8] Hawton K, van Heeringen K. Suicide. Lancet 2009;373:1372–81. [9] Metz TD, Rovner P, Hoffman MC, Allshouse AA, Beckwith KM, Binswanger IA. Maternal deaths from suicide and overdose in Colorado, 2004–2012. Obstet Gynecol 2016;128:1233–40. [10] Knight M, Nair M, Tuffnell D, Shakespeare J, Kenyon S, Kurinczuk JJ, editors. Saving Lives, improving mothers’ care-lessons learned to inform maternity care from the UK and Ireland confidential enquiries into maternal deaths and morbidity 2014–16. Oxford: National Perinatal Epidemiology Unit, University of Oxford; 2018. [11] O’Hara MW, Swain AM. Rates and risk of postpartum depression—a meta-analysis. Int Rev Psychiatry 1996;8:37–54. [12] Robertson E, Grace S, Wallington T, Stewart DE. Antenatal risk factors for postpartum depression: a synthesis of recent literature. Gen Hosp Psychiatry 2004;26:289–95. [13] Screening for perinatal depression. ACOG committee opinion no. 757. American College of Obstetricians and Gynecologists. Obstet Gynecol 2018;132:e208–12. [14] USPTF Interventions to Prevent Perinatal Depression US Preventive Services Task Force Recommendation Statement JAMA. 2019;321(6):580-587. [15] Boudou M, Teissèdre F, Walburg V, Chabrol H. Association between the intensity of childbirth pain and the intensity of postpartum blues. L Encéphale 2007;33:805–10. [16] 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. [17] Tobin CD, Wilson SH, Hebbar L, Roberts LL, Wolf BJ, Guille C. Labor epidural analgesia and postpartum depression. Arch Depress Anxiety 2016;2:44–6. [18] Nahirney M, Metcalfe A, Chaput KH. Administration of epidural labor analgesia is not associated with a decreased risk of postpartum depression in an urban Canadian population of mothers: a secondary analysis of prospective cohort data. Local Reg Anesth 2017;10:99–104.
8