Persistent pain after caesarean section and vaginal birth: a cohort study

Introduction 

According to the definition of the International Association for the Study of Pain (IASP), pain is considered to be persistent (chronic) when it has lasted for two months. The incidence of persistent pain in the adult population varies from 2 to 40% depending on the study methods, populations studied, and risk factors (such as female gender, older age and lower income bracket).1 Persistent pain after surgery is not uncommon. In one study as many as 20% of patients in pain clinics suffered from persistent pain after some operations.2 The incidence of persistent pain is high especially after limb amputation, thoracotomy and mastectomy. After limb amputation, 60-80% of patients experience phantom pain.3 Following thoracotomy or mastectomy, up to 60% of patients may have persistent pain.[4], [5] After hysterectomy, the incidence of persistent pain has been reported to be 5-32%.6 Even minor surgery (such as inguinal hernia repair) has been found to be associated with persistent pain in 10-20% patients.[7], [8], [9] The risk factors for developing persistent post-surgical pain include the intensity of pre- or postoperative pain, type of surgery, and the subject’s genetic and psychosocial profile.10

Caesarean section is one of the commonest surgical procedures worldwide. Persistent pain after caesarean section has been investigated in only one study, which indicated an incidence of 12.3% at 10.2 months.11 In that study, the risk factors for persistent pain were caesarean section under general anaesthesia, as well as previous pain problems, and recall of severe acute postoperative pain. The incidence of persistent pain after vaginal birth has been found to be 4% at six months12 and 7% at six weeks.13 Assisted (vacuum extraction or forceps) vaginal birth was found to be associated with persisting perineal pain.12 The presence of a birth canal tear or episiotomy were found to be associated with perineal pain up to one week after vaginal birth, but not with persistent perineal pain.13

The aim of the present study was to determine whether the incidence of persistent pain is more common after caesarean section than after vaginal birth. The intensity and nature of persistent pain, and whether the persistent pain was related to previous pain problems, chronic diseases and some features of caesarean section and vaginal birth were examined.

Methods 

We studied 600 (300 undergoing caesarean section and 300 undergoing vaginal birth) consecutive Finnish-speaking patients at the Women’s Hospital, Helsinki University Central Hospital, beginning in April 2005, following approval by the hospital ethics committee. In 2005, our tertiary care hospital delivered 10 227 births of which 2044 (20%) were by caesarean section. Anaesthesia for caesarean section in 2005 was general anaesthesia (11%), combined spinal-epidural or spinal anaesthesia (54%), or epidural anaesthesia (35%).

There were 723 operative vaginal deliveries (7.1%) and all were performed with vacuum assistance; no forceps are used in our hospital. Epidural or spinal labour analgesia was used in 6714 of the vaginal births (65.6%). Epidural analgesia in our hospital is performed predominantly at the L2-3 interspace, using plain bupivacaine 20mg or ropivacaine 20mg with fentanyl 0.1mg or sufentanil 0.01mg. The dose is administered in two parts, the first part serving as a test dose. A midwife administers additional doses of the same solution when requested. Spinal analgesia for vaginal birth is provided using plain bupivacaine 2.5mg with fentanyl 0.025mg.

Caesarean sections are performed through a Pfannenstiel or vertical skin incision and a transverse lower uterine segment incision. Depending on haemostasis, the uterine wound is closed in one or two layers of continuous absorbable suture, the peritoneum and muscles are left open and the fascia is closed with a running continuous suture of absorbable material. The skin is closed with non-absorbable individual stitches that are removed on day five or six. The patients ambulate 8-10h after caesarean section. For elective caesarean section, a combined spinal-epidural technique is the most common form of anaesthesia, with plain bupivacaine 7.5-10mg, fentanyl 0.015mg and morphine 0.12-0.16mg. If the block is insufficient, 2% lidocaine with epinephrine is added via the epidural catheter. For emergency caesarean section, a similar anaesthetic technique is used, but if a labour epidural catheter is present, 18-20mL of 2% lidocaine with epinephrine and fentanyl 0.1mg is administered. If general anaesthesia is required, thiopental or propofol with succinylcholine is given for induction, with maintenance by sevoflurane or a propofol infusion combined with N2O. Rocuronium is given if additional muscle relaxation is needed.

A questionnaire (Appendix: see web) with a pre-stamped return envelope was sent to parturients within five days of the first anniversary of giving birth. A reminder notice was sent if a reply was not received within one month. The women were asked if pain still existed at the surgical site or in the birth canal. They were asked about the intensity and frequency of pain, its impact on daily life, and what activities worsened it. A visual analogue scale (VAS) was used for grading the pain, with a 10-cm line indicating no pain (0) and the worst imaginable pain (10). A verbal scale (VS) was also used, grading the pain as mild, moderate, intense and unbearable. All subjects were asked about the duration of post-partum pain, and their overall use of pain medication and other means of treating pain. The questionnaire also asked about general health status, obstetric history, previous abdominal operations, pain problems elsewhere, and details of the caesarean section or vaginal birth. The subjects were asked to indicate the site and nature of their pain on an illustration.

A group size of 184 patients was determined to detect an increase in persistent pain from 5% after vaginal birth to 15% after caesarean section at a 0.05 significance level with 90% power. To account for non-responders, questionnaires were sent to 600 patients (300 caesarean section and 300 vaginal birth cases). Two study nurses not involved in the analysis transferred the questionnaire responses into a database. Data were analysed with SPSS Data Editor, version 13.0 (Chicago, IL). For nominal data, cross-tabulation was performed by χ2 test, numeric data using the Mann-Whitney test and logistic regression analysis for parameter associations.

Results 

Responses were received from 229 (76%) and 209 (70%) women who had undergone caesarean section and vaginal birth, respectively. A few questions were left unanswered by some individuals. Demographic data were similar in the two groups, except the mean age of the women in the caesarean section group was higher (34 vs. 31 years, P<0.001), and they had undergone more caesarean sections (P<0.001). Demographic data and obstetric characteristics are recorded in Table 1. Logistic regression analysis was also performed for the demographic data, with the same results as cross-tabulation, except that parity was significantly different (P<0.001) between the groups (vs. cross-tabulation, P=0.093). The groups did not differ in the incidence of certain clinical factors (previous pain and back pain, and presence of chronic disease).

Table 1. Patient demographic data and characteristics of the caesarean sections and vaginal births

		Caesarean section	Vaginal birth
	Response rate	229 (76%)	209 (70%)
	Age (years)	34 (20-49)	31 (19-47)⁎
	Primipara	141 (62%)	112 (54%)
	Multipara	88 (38%)	97 (46%)
	Smoker	23 (11%)	33 (19%)
	No chronic diseases	177 (78%)	156 (75%)
	Previous caesarean section	67 (29%)	11 (5%)⁎
	Previous abdominal operations	53 (23%)	38 (18%)
	Previous pain problem	56 (24%)	39 (19%)
	Previous back pain	19 (8%)	12 (6%)
	General anaesthesia	31 (14%)
	Trial vaginal delivery	108 (47%)
	Planned operation	84 (37%)
	Vertical incision	24 (10%)
	Horizontal incision	205 (90%)
	Any complication	37 (16%)
	Wound infection	17 (7%)
	Epidural or spinal analgesia		138 (66%)
	Episiotomy		78 (37%)
	Any complication		35 (17%)
	Tear		11 (5%)

⁎ P<0.001.

Post-partum pain lasted significantly longer (P=0.003) after caesarean section than after vaginal birth (Table 2). The pain had resolved by two months in 70% and 83% of women after caesarean section and vaginal birth, respectively. One year after giving birth, 18% (42/229) and 10% (20/209) of the mothers still experienced pain after caesarean section and vaginal birth, respectively (P=0.011; logistic regression P=0.012). The odds ratio for this result was 2.1 (CI 1.2-3.7), with a number needed to harm of 11.

Table 2. Duration of pain after delivery

		Caesarean section⁎	Vaginal birth
	Less than 3 weeks	84 (37%)	107 (51%)
	3 weeks – 2 months	75 (33%)	67 (32%)
	2 – 5 months	20 (9%)	20 (9%)
	Over 5 months, no longer	7 (3%)	3 (1%)
	Pain at 1 year	42 (18%)	20 (10%)†
	No response	1 (0%)	3 (1%)

⁎ P<0.003 overall duration of pain, caesarean section vs. vaginal birth. † P=0.011 caesarean section vs. vaginal birth.

A more detailed analysis of persistent pain one year after giving birth is shown in Table 3. Women with persistent pain had significantly more pain during coughing (P=0.023) and deep breathing (P=0.040) after caesarean section. However, among those women who still had pain at one year, sexual intercourse aggravated the pain significantly less after caesarean section than after vaginal birth (P=0.003).

Table 3. Analysis of subjects with pain at one year after delivery

		Caesarean section (n=42)	Vaginal birth (n=20)	P
	Pain			NS
	mild	23 (55%)	11 (55%)
	moderate	13 (31%)	2 (10%)
	intense	4 (10%)	5 (25%)
	unbearable	0	1 (5%)
	no response	2 (5%)	1 (5%)
	Pain at rest, VAS			NS
	0-2	33 (84%)	16 (80%)
	3-5	4 (10%)	1 (5%)
	6-	2 (5%)	0
	no response	3 (7%)	3 (15%)
	Pain during deep breathing, VAS			0.040
	0-2	36 (86%)	17 (85%)
	3-5	3 (7%)	0
	6-	0	0
	no response	3 (7%)	3 (15%)
	Pain when coughing, VAS			0.023
	0-2	35 (83%)	16 (80%)
	3-5	3 (7%)	1 (5%)
	6-	1 (2%)	0
	no response	3 (7%)	3 (15%)
	Occurrence of pain			NS
	constant	2 (5%)	0
	daily	8 (19%)	2 (10%)
	weekly	3 (7%)	4 (20%)
	less often than weekly	9 (21%)	2 (10%)
	in special situations	17 (40%)	11 (55%)
	no response	3 (7%)	1 (5%)
	Pain hinders daily activities	6 (14%)	3 (15%)	NS
	Pain disturbs sleep	4 (10%)	2 (10%)	NS
	Pain has worsened	2 (5%)	1 (5%)	NS
	Intercourse aggravates pain	9 (23%)	12 (65%)	0.003

Values are number (%).

The association between patient history and occurrence of persistent pain one year after delivery is shown in Table 4. Persistent pain was significantly more common in women with a chronic disease (P=0.016), previous pain problems (P=0.013) and back pain (P=0.016), when cross-tabulation was used for the analysis. After logistic regression analysis, the corresponding P values were 0.047, 0.024 and 0.029, respectively. Those who had pain at one year recalled having significantly more pain on the day after birth in both the caesarean section (P=0.004) and vaginal birth (P=0.001) groups than those who did not have persistent pain. The median VAS pain scores were 4 with persistent pain and 2 without persistent pain one day after vaginal birth, and 7 and 5 one day after caesarean section, respectively. VAS pain scores during vaginal birth or caesarean section were not different in patients with or without persistent pain.

Table 4. Occurrence of persistent pain in relation to some diseases and conditions, after vaginal birth and caesarean section

		Yes		No
		n	Pain	n	Pain
	Any chronic disease	95	21 (22%)	334	41 (12%)
	Depression	10	3 (30%)	422	59 (14%)
	Previous pain problem	94	21 (22%)	388	41 (12%)
	Back pain	31	9 (29%)	401	53 (13%)
	Smoker	55	9 (16%)	375	52 (14%)
	Previous abdominal surgery	89	12 (13%)	343	50 (15%)
	Primipara	250	36 (14%)	182	26 (14%)
	Present pain medication	13	8 (62%)	374	48 (13%)
	Other pain treatment	5	2 (40%)	417	60 (14%)

The features of caesarean section and vaginal birth associated with the development of persistent pain are shown in Table 5, Table 6. When all complications were pooled together, an association with persistent pain was found only in the caesarean section group (P=0.004; logistic regression analysis P=0.016). Wound infection was the most common complication after caesarean section (17/229). Endometritis and excessive bleeding occurred in six partiturients but other complications were rare after caesarean section. The incidence of persistent pain after caesarean section was similar with and without trial vaginal delivery.

Table 5. Association of some features of caesarean section with incidence of persistent pain

		Yes		No
		n	Pain	n	Pain
	Trial of vaginal birth	108	20 (19%)	115	21 (18%)
	General anaesthesia	31	7 (23%)	194	35 (18%)
	Elective caesarean section	83	15 (18%)	133	22 (17%)
	Vertical incision	24	6 (25%)	196	36 (18%)
	Any complication⁎	37	13 (35%)	191	29 (15%)
	Wound infection	17	5 (29%)	211	37 (18%)

⁎ P=0.004’.

Table 6. Association of some features of vaginal birth with incidence of persistent pain

		Yes		No
		n	Pain	n	Pain
	Epidural or CSE for labour pain	135	16 (12%)	65	3 (4%)
	Episiotomy	75	8 (11%)	127	10 (8%)
	Any complication	35	4 (11%)	169	16 (10%)
	A tear	11	3 (27%)	193	17 (9%)

CSE: combined spinal-epidural analgesia. No significant differences between groups’.

An evaluation of the pain illustrations noted that in most of the caesarean section patients, the pain, ache or tenderness appeared at the site of the scar. Most women undergoing caesarean section indicated numbness around the incision/scar site, including those who did not have persistent pain. In women undergoing vaginal birth, the site of the pain, tenderness or ache was typically around the vulva or anus.

Discussion 

Our main result is that persistent pain is more common one year after caesarean section than after vaginal birth. In both groups, however, more than half the women who still experienced pain after one year reported that it was mild; few women experienced intense or unbearable pain. The pain disturbed the daily life of 14% (caesarean section) and 15% (vaginal birth) of the women, respectively. Of all the parturients, 4% after caesarean section and 1% after vaginal birth still suffered from constant or daily pain one year after delivery.

The incidence of persistent pain after caesarean section was higher in our study than in the Danish study by Nikolajsen et al.11 They reported that 12.3% of the parturients experienced persistent pain at the end of a follow-up period ranging from 6 to 18 months, whereas in our study the follow-up time was standardized at 12 months for all parturients. The incidence of persistent pain in our study is similar to that in other studies on persistent pain after minor surgery.[6], [7], [8], [9] The incidence of persistent pain after vaginal birth was more frequent than in two other studies.[12], [13] Macarthur et al.13 found that 7% of parturients suffered from perineal pain after six weeks and Thompson et al.12 reported that 4% of parturients had perineal pain at six months.

As in the study by Nikolajsen et al.,11 no correlation was found between the type of incision and persistent pain. However, the number of patients, especially in the vertical incision group, was too small to detect a significant difference. Unlike Nikolajsen et al.,11 we did not find a significant correlation between general anaesthesia and persistent pain, although this element may also be underpowered, as only 14% of our patients had undergone general anaesthesia. Patients with persistent pain had more complications after caesarean section, with the most common complication being wound infection; the occurrence of complications did not differ statistically between the groups.

The number of patients who experienced persistent pain following vaginal birth (n=20) was too small to find any strong correlations. Macarthur et al.13 reported that 424 patients had more pain eight days following vaginal birth with birth canal tears or episiotomy procedures, but no difference was found six weeks after vaginal birth. Thompson et al.12 found perineal pain to be more common in patients with assisted vaginal birth (forceps or vacuum extraction) than those with unassisted vaginal birth. In our hospital, all assisted vaginal births are performed with vacuum extraction, but mode of vaginal birth was not recorded in our study.

Persistent pain was aggravated by sexual intercourse significantly more often after vaginal birth (65%) than after caesarean section (23%; P=0.003), although the actual number of subjects experiencing persistent pain was small (vaginal birth=12, caesarean section=9), and did not differ significantly between delivery types. Thompson et al.12 observed that parturients who had assisted vaginal birth had more sexual problems than parturients who had undergone unassisted vaginal birth or caesarean section.

Women with persistent pain at one year had a higher recall of pain on the first postpartum day after both vaginal birth (P=0.001) and caesarean section (P=0.004). This result may reflect the phenomenon confirmed by some investigators4 that persistent pain is associated with acute or postoperative pain. However, this finding may also reflect recall bias by subjects experiencing persistent pain, as previously noted in patients following breast surgery.14 Regardless, a history of pain, including previous back pain, was associated with persistent pain. This may reflect a genetic,[10], [15], [16], [17] psychological, or personal susceptibility to persistent pain.[10], [18], [19] Although patients with persistent pain commonly reported depression, this was not statistically significant.

We conclude that persistent pain is more common after caesarean section than vaginal birth, although the pain was usually mild. Persistent pain was associated with a history of previous pain, chronic disease and pain after delivery. A more extensive prospective study is needed to examine risk factors for persistent pain after caesarean section and vaginal birth.