Factors associated with persistent pain after childbirth: a narrative review

Summary

A systematic literature search was performed to identify studies that reported risk factors for persistent pain after childbirth. Many studies have sought to identify risk factors for post-delivery pain in different populations, using different methodologies and different outcome variables. Studies of several different but interrelated post-partum pain syndromes have been conducted. Factors strongly and specifically associated with persistent incisional scar pain after Caesarean delivery include a coexisting persistent pain problem in another part of the body and severe acute postoperative pain. For persistent vaginal and perineal pain, operative vaginal delivery and the magnitude of perineal trauma have been consistently linked. History of pregnancy-related and pre-pregnancy back pain and heavier body weight are robust risk factors for persistent back pain after pregnancy. Unfortunately, limitations, particularly small samples and lack of a priori sample size calculation designed to detect specific effect sizes for risk of persistent pain outcomes, preclude definitive conclusions about many other predictors and the strength of outcome associations. In future studies, assessments of specific phenotypes using a rigorous analysis with appropriate predetermined sample sizes and validated instruments are needed to allow elucidation of stronger and reliable associations. Interventional studies targeting the most robustly associated, modifiable risk factors, such as acute post-partum pain, may lead to solutions for the prevention and treatment of these common problems that impact a large population.

Editor's key points.

• •The authors examined the evidence base for the risk factors relating to the important problem of persistent post-partum pain. They conducted a systematic literature search to identify studies reporting risk factors for several persistent post-partum pain syndromes.
• •Persistent Caesarean scar pain appears to be associated with coexisting persistent pain problems and severe, acute postoperative pain.
• •Persistent vaginal and perineal pain appears to be associated with operative vaginal delivery and the magnitude of perineal trauma.
• •A history of pregnancy-related and pre-pregnancy back pain and heavier body weight are associated with persistent post-partum back pain.

Persistent pain in the large population of women after childbirth has attracted increasing attention from governments, medical societies, and the public over the past decade.^1,2 One hundred and forty million births occur yearly worldwide.^3,4 The reported incidence of persistent pain after childbirth varies depending on the method of delivery; study population; and, perhaps most importantly, study design. A meta-analysis published in 2016 that included prospective and retrospective observational studies and RCTs concluded that 11% of women undergoing Caesarean delivery have persistent pain at 12 months.⁵ Thus far, some have attempted to use antepartum psychophysical testing to identify women who are at risk for persistent post-partum pain, and one study has identified catastrophising assessed by the Pain-Related Self-Statements Scale-Catastrophizing Subscale as a predictor for persistent perineal pain after vaginal delivery.⁶

Because of the size, youth, vulnerability, and therefore the societal importance of the population at risk, we undertook a narrative review of the underlying literature. Our objective was to provide a critical review of the current literature regarding predictive factors for persistent pain after childbirth, with the aim of improving the identification of individuals at risk. The elucidation of modifiable targets will point to interventions to be tested in treatment studies. The identification of non-modifiable targets will enhance screening effectiveness.

Methods

MEDLINE database searches were conducted to identify studies focused on the identification of risk factors for persistent pain after childbirth using the following search string for matches in titles or abstracts: ‘pain [ti/ab] AND (childbirth [ti/ab] OR “Caesarean delivery” [ti/ab] OR “vaginal delivery” [ti/ab] OR delivery [ti/ab] OR postpartum [ti/ab])’, with restriction to human studies reported in English. Persistent pain was defined as pain that lasted at least 6 weeks after the delivery. Pain was considered at any location below the chest, except for isolated pelvic and pelvic girdle pain.

A manual inspection was subsequently performed through the reference lists of identified studies independently by two authors (RK and KA) using a data form generated by the authors, which specifically documents the following: region of study origin, study design, method of patient recruitment used, inclusion criteria, sample size, reported pain outcome, incidence of persistent pain, predictors for persistent pain, and ratio of patients with Caesarean delivery amongst included patients. Collected data have been aggregated into Supplementary Appendix Table 1.^1,6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25^,26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59 Disagreement was adjudicated by author PDF. RCTs, prospective cohort studies, retrospective cohort studies, and cross-sectional studies that contain information about persistent pain lasting at least 6 weeks were considered for inclusion. Only articles in which pain was described as a dichotomous or continuous variable, in locations at or below the chest, and that included evaluation of risk factors with statistical inference, were included in this review. Case reports, case series, narrative reviews, systematic reviews, letters to the editor, and editorials were excluded.

Results of literature search

The literature search using the MEDLINE database with the methods described previously, performed on September 23, 2018, resulted in 764 matches. Amongst the 764 matches, one duplication was found (one study published in two separate articles), reducing the records to 763. The reference lists of those records located 67 additional articles, which might have relevant information. After a review of titles and abstracts of a total of 830 records, we excluded 689 articles according to the aforementioned criteria and assessed the full text of 141 articles. After further exclusions according to the aforementioned criteria, 55 articles were retained and included in the current narrative review (Fig. 1). Information on study methodology, location, and cohort details relevant for consideration of applicability and level of evidence is provided in Supplementary Appendix Table 1.^1,6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25^,26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59

Fig 1.

Flow diagram of the search for relevant references. Eight hundred and thirty-one records were initially identified. Removal of one duplication reduced the records to 830. Of the 830 studies, 55 studies met the inclusion criteria and were included in the narrative review.

Fourteen studies included in this narrative review did not find any significant association between candidate risk factors and persistent pain outcomes.31, 32, 33, 34, 35, 36, 37^,49, 50, 51^,56, 57, 58, 59 Those consist of eight prospective cohort studies,31, 32, 33^{,35,37,50,56,57} five RCTs,^{34,49,51,58,59} and a non-randomised trial.³⁶ The types of pain investigated were low back pain,^31,33, 34, 35, 36, 37 Caesarean scar pain,49, 50, 51 perineal pain,57, 58, 59 and combined pain.^32,56

Nature of persistent post-partum pain

There are several significant issues that make identifying persistent post-partum pain difficult. The definitions of persistent pain in the literature, and indeed the terms used for persistent pain, are inconsistent. Other terms used to describe persistent pain, including prolonged, protracted, and chronic, were found in the included literature. Herein, we use the term ‘persistent pain’. The relevant time frames that have been customarily used to define persistent pain are 2 months,⁶⁰ 3–6 months, or longer depending on the studies.^61,62 We defined pain lasting 6 weeks or longer as persistent pain in this review. Post-partum women may have multiple different types of pain partially dependent on mode of delivery. Pregnancy itself is associated with a high prevalence of pain conditions that may continue into the post-partum period.⁶³ For example, 20–90% of women suffer from low back pain during pregnancy depending on the cohort studied and the methodology used.^64,65 As such, it is difficult to completely separate new onset of pain conditions that develop during or after the delivery from those that were pre-existing before pregnancy or delivery.

Persistent Caesarean scar pain

The common horizontal suprapubic scar resulting from Caesarean delivery can entrap the ilioinguinal or iliohypogastric nerves initiating diffuse persistent neuropathic pain. In a study with a sample size of 81, Bollag and colleagues⁵¹ demonstrated that 56%, 50%, and 26% of women described wound pain with predominantly neuropathic descriptors of the Short-Form McGill Pain Questionnaire 2 at 3, 6, and 12 months after Caesarean delivery, respectively. Further, there have been multiple reports of pain relief after nerve block, neurotomy, re-implantation of the peripheral nerve into muscle, and radio-frequency treatment of affected nerves in patients with persistent pain after Caesarean delivery.66, 67, 68, 69

Scar pain is the most commonly identified aetiology for persistent pain after Caesarean delivery. Amongst the studies that evaluated persistent Caesarean scar pain, the range of incidences reported has been large (Table 1). The reason underlying the wide range is likely multifactorial.^{38,40,43,45,46} The risk factors that have been evaluated for developing a painful scar after Caesarean delivery are summarised in Table 1.

Table 1.

Incidence and risk factors for Caesarean scar pain. *Pain at 12 months after delivery. ^†The risk factors for combined pain of multiple body parts were reported, and the risk factors specific for Caesarean scar pain were not reported. ^‡Pain at 8 months after delivery. ^¶Pain at 14.5 months after delivery. ^§Pain at 26 months after delivery. ^||Pain at 10.1 months after delivery. ^#Studies in which evaluated risk factors for Caesarean scar pain failed to demonstrate statistical significance. CD, Caesarean delivery; SF-36, Short Form 36 Health Survey.

Study	Sample size		Time points after delivery
			8 weeks	3 months	6 months	8 months or greater
Wang and colleagues38	786	Incidence		6.9%	1.9%	0.3%*
		Risk factors		†
Jin and colleagues40	502	Incidence		8.2%	3.7%	1.7%*
		Risk factors		†	†	†
Moriyama and colleagues41	225	Incidence		30.7%
		Risk factors		Lighter weight and omission of intrathecal morphine
Niklasson and colleagues42	231	Incidence		22.4%	16.8%	5.5%*
		Risk factors		†	†
Richez and colleagues43	268	Incidence		28%	19%
		Risk factors		Pain during current pregnancy, history of miscarriage, and lower mental component summary score of SF-36	Postoperative complications and lower mental component summary score of SF-36
Ortner and colleagues44	335	Incidence	11.0%		3.0%	0.6%*
		Risk factors	Higher pain score at 24 h after operation
de Brito Cancado and colleagues45	402	Incidence		11.4%
		Risk factors		Higher pain level at 12 h after operation and postoperative Day 2, lower intrathecal bupivacaine dose, and omission of perioperative NSAIDs
Shahin and Osman46	325	Incidence				4.3%‡
		Risk factors				†
Sng and colleagues47	857	Incidence		Pain lasting more than 3 months: 9.2%		6.0%¶
		Risk factors		Higher pain scores immediately after operation, pain present elsewhere, and non-private insurance status
Loos and colleagues48	690	Incidence				32.3%§
		Risk factors				>2 previous Pfannenstiel incisions, presence of scar numbness, and emergency CD
Nikolajsen and colleagues1	220	Incidence		Pain lasting more than 3 months: 18.6%		12.3%||
		Risk factors		General anaesthesia, pain problem elsewhere, and higher recall of severe postoperative pain
Mackeen and colleagues49#	589	Incidence
		Risk factors	Skin closure with staples (vs suture)
Liu and colleagues50#	469	Incidence	14.6%			4.2%*
		Risk factors	Elective CD (vs emergency CD), repeat CD (vs primary CD), severe acute postoperative pain, general anaesthesia, and wound infection
Bollag and colleagues51#	81	Incidence
		Risk factors		Transversus abdominis plane block with placebo (vs bupivacaine vs bupivacaine+clonidine)	Transversus abdominis plane block with placebo (vs bupivacaine vs bupivacaine+clonidine)	Transversus abdominis plane block with placebo (vs bupivacaine vs bupivacaine+clonidine)

All studies that investigated Caesarean scar pain are prospective studies. The sample sizes are comparable between studies in which candidate risk factors predicted the pain outcome with statistical significance (positive studies; median: 297 [inter-quartile range: 236–599]), and those in which candidate risk factors failed to predict the pain outcome with statistical significance (negative studies; 469 [275–529]). Further, in one negative study, a power analysis was performed to detect an a priori determined effect size of a candidate risk factor.⁴⁹ Therefore, the negative study has at least a comparable quality as that of positive studies. Both positive and negative studies combined, the risk factors studied in more than two studies are coexisting persistent pain problems in other parts of the body,^1,43,47 and acute postoperative pain.^1,47,50 All studies that assessed persistent pain problems in other parts of the body reported the effect sizes as odds ratios of 1.9–2.8^43,47 or a relative risk of 3.3.¹ As the effect size is consistently in the same direction in all the studies, persistent pain problem in other parts of the body can be considered a robust risk factor. Acute postoperative pain has been assessed using different metrics between the studies.^1,47,50 Sng and colleagues⁴⁷ utilised a 0–10 numeric rating scale, and reported the effect size in adjusted odds ratio of 1.4 (odds of having persistent pain increases by 1.4 with every 1 point increase in numeric rating scale of acute postoperative pain), whereas Nikolajsen and colleagues¹ reported a relative risk of 1.5 in the patients who had recall of severe acute postoperative pain vs no recall of severe acute postoperative pain. Liu and colleagues⁵⁰ reported a relative risk of 1.1 in those who experienced moderate-to-severe acute postoperative pain vs none-to-mild pain, which failed to attain statistical significance. Quantification of the magnitude of effect size of acute pain is difficult because of the use of different metrics in the studies. However, as the direction of the effect size is consistent in all studies, it is likely that the acute postoperative pain burden is associated with the development of persistent Caesarean scar pain.

Potential interventions to decrease acute postoperative pain burden are aggressive use of labour epidural analgesia to decrease the need for general anaesthesia for urgent intrapartum Caesarean delivery, and increasing the dose of intrathecal local anaesthetics and other adjuvants for the cases in which intrathecal morphine is contraindicated.⁴⁵

Persistent vaginal and perineal pain

The nature of persistent vaginal and perineal pain is less well characterised. As it has been speculated that tissue damage and inflammation are potential mechanisms for vaginal and perineal pain, it is considered predominantly nociceptive pain.⁷⁰ Neuropathic pain attributable to obstetric trauma, such as pudendal neuropathy, is also a possible contributor.

Perineal and vaginal pain after vaginal delivery has been associated with tissue trauma related to operative vaginal delivery, perineal laceration, and use of episiotomy,^15,16,18, 19, 20^,22,26,53, 54, 55 although some studies also report its occurrence after Caesarean delivery with and without labour.^20,22,26,52 Incidences of vaginal and perineal pain at different time points after the delivery and the risk factors are summarised in Table 2.

Table 2.

Incidence and risk factors for perineal pain and vaginal pain. *Pain at 12 months after delivery. ^†Pain at 18 months after delivery. ^‡Pain at 8 weeks after delivery. ^¶The risk factors for combined pain of multiple body parts were reported, and the risk factors specific for perineal and vaginal pain were not reported. ^§Both vaginal and Caesarean deliveries were included in the study, but only incidences for vaginal delivery were reported. ^||The incidences for the entire study population were not reported, and only the incidence for each delivery type was reported. ^#Pain at 6 weeks after delivery. **Studies in which evaluated risk factors for vaginal and perineal pain failed to demonstrate statistical significance.

Study	Sample size		Time points after delivery
			6–8 weeks	3 months	4 months	6 months	12–18 months
Glowacka and colleagues15	150	Incidence		27.3%
		Risk factors		Non-genitopelvic pain prior to pregnancy
Woolhouse and colleagues16	1431	Incidence		29.1%		6.4%	2.2%,* 1.8%†
		Risk factors
Paterson and colleagues18	114	Incidence		30.2%
		Risk factors		Previous diagnosis of non-genital chronic pain
Thompson and colleagues22	1193	Incidence	22.1% for 0–8 weeks		6.5% for 9–16 weeks	4.2% for 17–24 weeks
		Risk factors	Primiparity and operative vaginal delivery (vs spontaneous vaginal delivery)		Primiparity and operative vaginal delivery (vs spontaneous vaginal delivery)	Primiparity and operative vaginal delivery (vs spontaneous vaginal delivery)
Eisenach and colleagues20	939	Incidence	1.3%‡
		Risk factors	¶
Glazener and colleagues26	438	Incidence	22% between discharge home and 8 weeks				10% between 2 and 18 months
		Risk factors	Primiparity and operative vaginal delivery (vs spontaneous vaginal delivery)				Primiparity and operative vaginal delivery (vs spontaneous vaginal delivery)
Declercq and colleagues19	1573	Incidence	48% for vaginal delivery‡§			2% for vaginal delivery§
		Risk factors	Mode of delivery and episiotomy			Mode of delivery and primiparity
Klein and colleagues55	697	Incidence		55.7%
		Risk factors		Degree of perineal trauma
Sartore and colleagues54	519	Incidence		4.4%
		Risk factors		Episiotomy
Williams and colleagues52	482	Incidence					Caesarean delivery: 0%; normal vaginal delivery: 30.0%; operative vaginal delivery: 36.4%*||
		Risk factors					Asian race (vs Caucasian) and forceps delivery
Soares and colleagues6	55	Incidence	14.5%‡
		Risk factors	Catastrophising assessed by Pain-Related Self-Statements Scale during admission of delivery, and perineal trauma
Macarthur and Macarthur53	447	Incidence	7.1%#
		Risk factors	Primiparity
Fodstad and colleagues57**	179	Incidence		24.6%
		Risk factors		Episiotomy type (midline vs mediolateral vs lateral vs non-classifiable), episiotomy length, and spontaneous vaginal delivery (vs operative vaginal delivery)
Leeman and colleagues58**	444	Incidence		9.7%
		Risk factors		Degree of perineal trauma
Albers and colleagues59**	5404	Incidence		6.8%
		Risk factors		Degree of perineal trauma

Both studies, in which candidate risk factors predicted the pain outcome with statistical significance (positive studies) and those in which candidate risk factors failed to predict the pain outcome with statistical significance (negative studies), are mostly prospective studies, and sample sizes are comparable between positive studies (median: 501 [inter-quartile range: 445–821]) and negative studies (444 [227–1684]). As none of the studies performed a priori sample size calculation aiming for the detection of a specific effect size of candidate risk factors, it is possible that potential risk factors with clinically relevant effect size might have been missed in the negative studies because of lack of power. For positive and negative studies, the risk factors assessed in more than two studies are the mode of delivery^22,26,52,57 and magnitude of perineal trauma.^6,19,54, 55, 56^,58,59 Operative vaginal delivery (vs spontaneous vaginal delivery) was consistently reported as a risk factor, and in the studies where effect sizes were reported in the form of odds ratio, the magnitude ranges from 1.6 to 3.6.²² Even in the negative study, pain level assessed on 0–10 visual analogue scale tended to be higher with operative vaginal delivery.⁵⁷ As the effect is in the same direction in all studies, operative vaginal delivery is likely to be a robust predictor for persistent perineal pain. Studies of the magnitude of perineal trauma vary in comparison groups. Some compared episiotomy vs no episiotomy,^19,56 others compared perineal trauma vs no perineal trauma,^6,55,58,59 or multilevel comparisons between episiotomy and different degrees of perineal laceration.⁵⁴ This makes assessment of effect size on the effect of perineal trauma on persistent pain difficult. However, even in the negative studies, reported pain intensity of perineal pain tended to be higher in women with episiotomy than those without episiotomy,⁵⁶ with a trend of increased incidence of persistent pain with increased tissue trauma burden (dose–response relation).⁵⁹ It is likely that perineal trauma burden has some effect on the development of persistent perineal pain, and the mechanism has face validity. A potential intervention that may address to decrease perineal trauma burden is the judicious use of episiotomy.

Other persistent pain syndromes after vaginal and Caesarean delivery

Aside from pain in specific body locations, many studies have evaluated risk factors for persistent general back pain,^10,13,21,23, 24, 25^,27, 28, 29, 30 persistent diffuse pain, or any pain after childbirth.^{7,9,11,12,17,20,28,39,40,42,44,45}

Persistent back pain has been extensively studied. Many potential risk factors have been evaluated. Our literature search identified the following risk factors for persistent back pain: history of back pain prior to pregnancy,23, 24, 25^,27,29,30 back pain during index or previous pregnancy,^21,25,27 earlier onset, longer duration,²¹ more severe and disabling back pain during pregnancy,^21,23 sick leave attributable to back pain,²⁹ pain in other body locations during pregnancy,²⁸ unmarried status,²⁸ heavier weight,^13,27 larger weight gain during the post-partum period,²¹ heavy physical work,²⁹ Caesarean delivery (vs vaginal delivery),^10,16 use of epidural labour analgesia,²⁸ induced labour as opposed to spontaneous labour,¹³ no use or short-term use of oral contraceptives (vs long-term use),²³ and urinary tract infection and renal calculi.¹⁰ Older age^10,21,24,27 and primiparity^22,24 have been reported both as protective and risk factors. Candidate variables from negative studies include the use of epidural for labour analgesia,³⁷ Caesarean delivery (vs nulliparity), vaginal delivery (vs nulliparity),³³ Caesarean delivery (vs vaginal delivery),³⁵ operative vaginal delivery (vs spontaneous vaginal delivery),³⁵ post-surgical complication (such as diastasis recti abdominis),³¹ a physical therapy programme developed by the authors,³⁴ and a home exercise programme developed by the authors.³⁶

Both studies, in which candidate risk factors predicted the pain outcome with statistical significance (positive studies) and those in which candidate risk factors failed to predict the pain outcome with statistical significance (negative studies), are largely prospective studies, and sample sizes are comparable between positive studies (median: 916 [inter-quartile range: 385–1253]) and negative studies (185 [95–6921]). Because none of the studies included an a priori sample size calculation aiming for the detection of a specific effect size of candidate risk factors, it is possible that potential risk factors with clinically relevant effect sizes might have been missed in the negative studies because of lack of power. For positive and negative studies, the risk factors studied in more than two studies are obesity and pregnancy-related weight gain,^13,21,27 pregnancy-related back pain,^{21,25,27,29,30} history of non-pregnant back pain,^21,23, 24, 25^,29 Caesarean delivery (vs vaginal delivery),^10,16,35 and age.^{10,21,24,27,28} As the reported effect size is consistently in the same direction^13,21,27 and there is a mechanistic rationale, obesity and pregnancy-related weight gain are likely to be associated with the development of persistent back pain. The effect size is not readily interpretable in many of these studies as an odds ratio or relative risk is not reported. For the studies of pregnancy-related back pain, in which effect sizes were reported in interpretable form, the magnitude odds ratios range from 1.7 to 5.9.^21,25,27 Odds ratios for history of pre-pregnancy back pain range from 4.4 to 7.5.^21,23,25 Given the large effect size that is consistently in the same direction, both pregnant and non-pregnant back pain are robust risk factors for persistent back pain. In contrast, the effect size for age has been rarely reported.¹⁰ The association between Caesarean delivery (vs vaginal delivery) and the development of persistent back pain was reported to be small (odds ratios and relative risks between 0.9 and 1.4).^10,16,35

Pregnancy-related hormonal changes and persistent pain

After surgical (Caesarean) delivery, a lower incidence of persistent pain has been reported in comparison with that after some general surgical procedures in similar locations in women.⁷¹ Because of this observation, it has been speculated that the high oestrogen and progesterone concentrations found during pregnancy may result in pregnancy-induced analgesia. Animal models of central and peripheral nerve injury have demonstrated neuroprotective effects for oestrogen and progesterone via multiple mechanisms. In human experimental models, results have been mixed related to the modality of pain.72, 73, 74, 75 Preclinical studies that have demonstrated an anti-nociceptive effect from oxytocin are robust.^76,77 At this time, the results of human clinical trials are sparse.

Genetic factors and pain related to childbirth

The physiology that underlies pain perception is multifactorial, complex, and also beyond the purview of this review.⁷⁸ Briefly discussed next, genetic factors have been considered as potential contributing factors in the setting of acute pain,⁷⁹ persistent pain,⁸⁰ and in some studies of acute post-partum pain.81, 82, 83 Although few studies have evaluated the effects of gene polymorphisms specifically on persistent post-partum pain, a few selected gene polymorphisms that have been widely implicated in other pain phenotypes have been evaluated for association with acute post-partum pain. As acute post-partum pain is a risk factor for persistent post-partum pain, these associations may be relevant.

The μ-opioid receptor gene (OPRM-1) encodes the μ-opioid receptor, a primary site of opioid action. A common single-nucleotide substitution at position 118 of the OPRM-1 gene, adenine by guanine (A118G), is associated with changes in receptor glycosylation, decreased binding affinity for exogenous and endogenous opioids, and increased pressure pain threshold in humans.⁸⁴ Whilst this variant is strongly associated with higher opioid requirements in the perioperative setting, the effect size is relatively small.⁸⁵ With regard to acute post-partum pain, several studies have evaluated the effect of OPRM-1 A118G polymorphism on systemic opioid requirements after Caesarean delivery performed under spinal anaesthesia.81, 82, 83 Results have been mixed according to the population studied and perhaps the relative frequency of the polymorphism. In a study of a North American cohort, systemic opioid requirements during 3 post-partum days were not different between parturients who carry the G118 allele and those with AA118 homozygosity.⁸³ In Chinese Singaporean patients, post-partum 24 h i.v. patient-controlled opioid analgesia consumption was higher in women carrying G118 allele than those with AA118 homozygosity.⁸¹ The same group showed an association between higher 24 h i.v. morphine requirements and the G118 allele in a separate cohort of mixed ethnicities.⁸² As higher opioid dose requirements are likely associated with higher acute post-partum pain burden, a predictor for persistent pain,^17,20 it is possible that OPRM-1 A118G polymorphism might have some effects on developing persistent pain after the delivery, and may be worth further investigation.

The catechol-O-methyltransferase (COMT) enzyme breaks down catecholamines and regulates oestrogen metabolism. A single-nucleotide polymorphism in the COMT gene at position 472 (p. 472 G/A) results in amino acid substitution Val158Met, which is associated with lower COMT enzyme activity and higher baseline pain sensitivity.^86,87 In an experimental setting, subjects homozygous for the Met158 allele demonstrate increased pain sensitivity⁸⁸ and greater opioid-induced hyperalgesia.⁸⁹ This polymorphism has been put forward as a risk factor for the development of persistent pain⁹⁰ and severity of acute pain.⁹¹ In non-pregnant surgical patients, Met158 allele has been associated with higher numerical rating scores to various acute noxious stimuli,⁹¹ and i.v. fentanyl was less effective for the treatment of early-stage labour pain in women homozygous for Met at COMT158.⁹² Thus far, one Chinese study has evaluated the effect of COMT Val158Met polymorphism on systemic analgesic requirements during the acute period after Caesarean delivery and incidence of persistent pain at 3 months; this found that the polymorphism is not associated with either outcome.⁹³ However, an a priori sample size calculation to detect a specific effect size of the polymorphism on the incidence of persistent pain was not performed in that study.⁹³

Possible preventive interventions

Intraoperative management that limits acute postoperative pain has been put forward as being potentially preventative for persistent pain. This seems reasonable because of the association between the severity of acute postoperative pain and persistent pain.^1,47 Neuraxial analgesia is the most common anaesthetic technique used for elective Caesarean delivery in the USA and in Europe,^94,95 often using a combination of local anaesthetic with lipophilic opioids, such as fentanyl, and more hydrophilic, longer-lasting opioids, such as morphine. Neuraxial analgesia is increasingly supplemented with a multimodal postoperative analgesic regimen consisting of non-steroidal anti-inflammatory agents, acetaminophen, and oral opioids.⁹⁴ In an attempt to reduce postoperative pain burden and opioid utilisation, multiple analgesic adjuncts have been evaluated and found effective in reducing acute pain. The efficacy for the prevention of persistent pain after surgical delivery has been studied less commonly and with mixed results.

Clonidine is a prototypical α2-adrenergic agonist. Neuraxially administered clonidine mediates its analgesic effect by activating α₂-receptors located in the dorsal horn of the spinal cord and postsynaptic activation of descending inhibitory neurones. The efficacy of neuraxial clonidine for post-Caesarean pain and opioid requirement has been evaluated.96, 97, 98, 99 A meta-analysis published in 2018 showed that clonidine prolongs neuraxial anaesthesia and reduces opioid consumption more than 24 h after Caesarean delivery.¹⁰⁰ Only a few studies have evaluated the effect of clonidine on more persistent pain.^51,97 Lavand'homme and colleagues⁹⁷ demonstrated that high-dose (150 μg) intrathecal clonidine decreased the incidence and area of peri-incisional hyperalgesia at 48 h after operation, which has been shown to be a marker of central sensitisation and a predictor of persistent pain in a non-pregnant surgical population.¹⁰¹ Bollag and colleagues⁵¹ studied clonidine's activity in the peripheral nervous system. Clonidine 150 μg, administered into the transversus abdominal plane combined with a local anaesthetic (bupivacaine), did not have a significant effect on the incidence or area of peri-incisional hyperalgesia at 24 and 48 h after Caesarean delivery.⁵¹ Neither study found a decreased incidence of persistent pain at 6 months,⁹⁷ or decrease in persistent pain descriptors evaluated using the Short-Form McGill Pain Questionnaire at 3–12 months in clonidine-treated patients.⁵¹ However, those studies may have been underpowered to detect a beneficial effect of clonidine on chronic pain after Caesarean delivery likely because of a floor effect. At this time, the efficacy, optimal dose, route of administration, and target population have yet to be confirmed for α2-agonists.

Gabapentin and pregabalin are anticonvulsants widely used for the treatment of neuropathic pain conditions. They decrease excitatory neurotransmitter activity by binding to voltage-gated calcium channels,¹⁰² thereby attenuating afferent nociceptive signalling. Several placebo-controlled trials have evaluated the efficacy of perioperative gabapentin on acute post-partum pain after Caesarean delivery with spinal anaesthesia.103, 104, 105, 106, 107, 108 The incidence of persistent pain at 3 months post-partum was evaluated in two studies.^103,104 The study by Moore and colleagues¹⁰³ used a single dose of gabapentin at 600 mg, whereas the study by Short and colleagues¹⁰⁴ used a single dose of gabapentin at 300 or 600 mg, 1 h before operation. Those studies failed to show a statistically significant difference in the incidence of persistent pain between gabapentin and placebo. However, those studies similarly suffered from a floor effect with a few patients in the placebo group suffering from severe or persistent pain. Gabapentin may require a larger dose, longer duration, or an enriched population to demonstrate a clinically meaningful effect for the prevention of persistent pain after Caesarean delivery.

Ketamine is an N-methyl-D-aspartate (NMDA) receptor antagonist, which has analgesic properties at a sub-anaesthetic dose, and has been used for the prevention and treatment of acute and persistent pain. It prevents hyperalgesia and central sensitisation caused by noxious stimuli via NMDA inhibition,^109,110 and may also prevent acute opioid tolerance.¹¹¹ A meta-analysis, published in 2015, aggregated the results of 12 RCTs, in which i.v. ketamine administered during surgery (0.15–1 mg kg⁻¹), placebo, or other interventions were compared with respect to the incidence of post-Caesarean pain.¹¹² The analyses demonstrated that time to first analgesic request was significantly longer, and the visual analogue scale pain sores at rest 2 h after surgery performed under spinal anaesthesia were significantly lower in ketamine-treated women. No such effects were observed in the studies, in which general anaesthesia was used.¹¹² Very few studies have investigated the effect of ketamine on persistent pain after Caesarean delivery.¹¹³ One RCT failed to show the efficacy of ketamine on persistent pain at 1, 6, and 12 months (0.25–1 mg kg⁻¹ in conjunction with general anaesthesia).¹¹³ Another RCT, in which S-ketamine 0.5 mg kg⁻¹ was given intramuscularly, followed by i.v. infusion at 2 μg kg⁻¹ min⁻¹ for 12 h in conjunction with spinal anaesthesia found no effect on persistent pain at 3 yr after operation.¹¹⁴ In contrast, a study, in which ketamine 10 mg was given after Caesarean delivery performed with spinal anaesthesia, lower pain scores were reported in women who received ketamine 2 weeks post-partum despite no measurable analgesic benefit of ketamine was observed during 24 post-partum hours.¹¹⁵ S-ketamine did attenuate hyperalgesia at the dermatomes corresponding to Caesarean incision at 12 and 24 h after operation,¹¹⁴ which suggests the potential of ketamine for the prevention of persistent neuropathic pain.

In the majority of the studies of the efficacy of clonidine,^51,97, 98, 99 gabapentin,103, 104, 105^,107 and ketamine described previously,113, 114, 115 multimodal analgesia was used for post-Caesarean analgesia. One study of clonidine⁹⁶ and two studies of gabapentin^106,108 did not document whether other agents were used in their studies and were therefore of lower quality. Currently, there are few studies with mixed findings on the efficacy of perioperative analgesics on persistent pain after Caesarean delivery, and even less information regarding preventive measures of persistent pain after vaginal delivery.

Opioid use after childbirth

Despite the fact that childbirth is a common entry point for opioid use amongst a large number of young women, the issue had not gathered attention until two large population-based studies evaluated persistent opioid use after Caesarean delivery in 2016.^2,116 Those studies identified younger age, cocaine and other illicit substance, tobacco, antidepressant, benzodiazepine use, and persistent pain history (back pain and migraine) as risk factors for persistent opioid use.^2,116

Several other studies have evaluated the extent of opioid use during the acute phase after Caesarean^117,118 and vaginal delivery in the USA.^119,120 Most women (estimated 83% in one study¹¹⁷) in the USA used opioids after discharge from a hospital after Caesarean delivery. The median duration of opioid use after discharge amongst all women included in this one study was 8 days.¹¹⁷ The investigators found that smoking during pregnancy, public insurance, and higher opioid dose requirement during post-partum hospital stay were associated with greater opioid use after discharge.¹¹⁷ They subsequently performed an RCT comparing the standard prescription of 30 oxycodone 5 mg tablets at hospital discharge with an individualised discharge opioid prescription based on inpatient opioid requirement for each patient. The group that received an opioid prescription based on their in-hospital use had a 50% reduction in post-partum outpatient opioid dose without compromising the quality of pain control.¹²¹ Similarly, Bateman and colleagues¹¹⁸ found that 85% of women filled an opioid prescription after Caesarean delivery, and used a median 20 tablets of oxycodone or hydrocodone 5 mg. Women who were prescribed a larger amount of opioid also used more medication, whilst pain reports and patient satisfaction were not different according to opioid prescription.¹¹⁸ This suggests that the amount of opioids the patients take after Caesarean delivery is influenced by the number of pills/doses prescribed at discharge from the hospital.

In the setting of vaginal delivery, single-institution data showed that only a minority of women required opioids after discharge.¹²² Twenty per cent of American women who delivered vaginally filled an opioid prescription within 5 days of delivery.¹¹⁹ The risk factors for filling opioid prescription were tobacco use and mental health condition.¹¹⁹ In another study of vaginal delivery, multiparous women used more opioids than primiparous women.¹²⁰

Limitations

The sample sizes of many studies included in this narrative review did not have a priori statistical plan and were not powered to detect a priori determined effect size of candidate risk factors for persistent pain. In small studies with inadequate power, there is a significant risk of a Type 2 error. Further statistical adjustments for confounding variables for persistent pain were often not performed. Pain present during pregnancy and new pain developed after the delivery were often not differentiated in the included studies. Therefore, the risk of bias could be substantial, and caution is warranted for interpretation of the results of many individual studies included in this narrative review.

Future directions

The reported incidence of persistent pain after childbirth is quite variable depending on the location and type of pain syndrome, duration studied, and study methodology. The good news is that most women have an uneventful childbirth and require analgesic medication for a short duration, and do not suffer from persistent pain. Most studies evaluated demographic, obstetrical, surgical, anaesthetic, and some psychological attributes as candidate predictors for persistent pain. Few included a thorough evaluation of genetic or psychological factors using current, validated instruments.

Studies involving larger populations with pre-specified analysis plans and sample size calculations will help to identify robust generalisable predictor–outcome relationships. In future prospective studies, clear pre-pain phenotypes and documentation of the timing of onset of pain are essential. To improve comparability between studies, and to develop normative data specific to parturients, assessment using validated instruments, such as the ones recommended by the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials, will be needed.¹²³ Furthermore, bidirectional predictor–outcome relationships for some phenotypes, such as depression, are likely where the interaction between predictor and outcome happens over time.^40,124 As such, acquisition of longitudinal data for both predictor and outcome is warranted to allow for identification of specific type and timing of preventive interventions to be tested. Further, studies focused on functional outcome in relation to pain⁴⁰ and those focused on women with vaginal delivery are important, as the incidence and severity of persistent pain after vaginal delivery are not as different from Caesarean delivery as one might suspect,²⁰ and information about functional outcome and the differences in persistent pain outcomes between Caesarean and vaginal delivery is presently scarce. Amongst the consistently reported risk factors for persistent pain, acute post-partum pain is unique in that it is easily recognisable and is potentially modifiable. Strategies most effective for women at risk are only now developing and require rigorous evaluation and testing. Further multidisciplinary interventional studies designed to evaluate the efficacy and effectiveness of interventions in women at risk for persistent post-partum pain are warranted.

Authors' contributions

Study design: all authors

Data collection: RK, KA

Data interpretation: all authors

Drafting/final approval of article: all authors

Declaration of interest

The authors declare that they have no conflicts of interest.

Funding

Department of Anaesthesiology and Pain Medicine, University of Washington.

Handling editor: Jonathan Hardman

Appendix A. Supplementary data

The following is the supplementary data to this article: