Common complications and the estimated frequency with which they occur are shown in Table 4.5.
High dependency unit/intensive therapy unit admission
Maternal mortality is rare. In the UK it is 11.4/100,000 maternities,95 [evidence level 3] the direct maternal mortality rate from all causes is 1/20,000 maternities. The mortality rate for women who have vaginal deliveries is 16.9/million compared to 82.3 per million for women who have caesarean section (CS) (risk ratio [RR] 4.9, 95% confidence interval [CI] 2.96 to 7.97).95 However it was not possible to determine the proportion of the increased risk that is attributable to antecedent conditions or the procedure itself. The incidence of severe morbidity for women giving birth has been reported to be 12 per 1000 deliveries.292 A small proportion of women (0.1–0.9%) develop complications of pregnancy that require admission to an Intensive Therapy Unit (ITU)521. High dependency unit (HDU)/ITU admission was not reported as an outcome in any of the randomised controlled trials (RCTs).
In the NSCSA, 10% of women who had CS required special care postoperatively within a high dependency unit, 3.5% of these women were transferred to an intensive care unit.4 [evidence level 3]
shows the proportion of women who had CS and required admission to an intensive care unit according to the reason for the CS.
Admission to intensive care unit (ICU) according to reason for CS (n = 29,349).
We identified one case control study that examined risk factors associated with intensive care unit admission during hospital stay for childbirth among women in USA between 1984 and 1997 (n = 2046).522 The overall rate of admission to an intensive care unit (ICU) was 0.13%. The odds of admission to ICU was significantly higher for women who had CS compared with those who had vaginal birth, after adjustment for socio demographic factors (age and ethnicity) and type of hospital (OR 9.0, 95% CI 7.24 to 11.16). [evidence level 3] However it is not possible to disentangle the effect of CS from the reasons for CS when interpreting these results. A UK study that evaluated the risk of severe obstetric morbidity has not been included here because the comparison groups are between women who had unplanned CS to women who had either planned CS or vaginal births.292 [evidence level 3]
Recommendations
View in own window
| Number | Recommendation |
|---|
| 99 | Healthcare professionals caring for women after CS should be aware that, although it is rare for women to need intensive care following childbirth, this occurs more frequently after CS (about 9 per 1000). [B] [2004] |
9.1. Routine monitoring after CS
There were 3 deaths in the last CEMD triennium report in which poor postoperative care was a contributing factor. The importance of monitoring the patient adequately postoperatively was emphasised.95 [evidence level 3] Earlier triennial reports recommended electronic monitoring of oxygen saturation levels.523 [evidence level 3] UK obstetric anaesthesia guidelines suggest that the postoperative care of a CS patient should be in accordance with the care of any postoperative patient as laid out in guidelines for postanaesthetic recovery.311 [evidence level 4]
After CS, women should be observed on a one-to-one basis by an anaesthetist, recovery nurse, midwife or other properly trained member of staff until they have regained airway control and cardiorespiratory stability and are able to communicate. All recovery rooms must be staffed to a level which allows this to be routine practice. Women must be kept under clinical observation at all times and all measurements must be recorded. The introduction of automatic recording systems is encouraged. The frequency of recordings will depend on the stage of recovery and clinical condition of the patient. As a minimum non-invasive blood pressure, heart rate and rhythm, respiratory rate and continuous pulse oximetry every 5 minutes for the first 30 minutes in recovery (‘recovery’ refers to any area where the patient is cared for immediately postoperatively and is not limited to a specific recovery room. The following information should be recorded:
Other parameters depending on circumstances e.g. temperature, urinary output, central venous pressure, end tidal CO2, surgical drainage.
For all women, the name, hospital number, time of admission, time of discharge and destination should be recorded in a central register.
Women with epidural or intrathecal analgesia will need additional observations including pain and sedation scores, respiratory rate and mobility which should be laid out in individual hospital protocol. This recording will normally be continued after discharge from the recovery area. It is generally accepted that after discharge from the recovery area to the ward, observations (respiratory rate, heart rate, blood pressure, pain and sedation) should be continued every half hour for two hours and hourly thereafter provided that the observations are stable or satisfactory. If these observations are not stable, more frequent observations and medical review are recommended.
For women who have had intrathecal opioids, there should be a minimum hourly observation of respiratory rate, sedation and pain scores for at least 12 hours for diamorphine and 24 hours for morphine. For epidural opioids and opioid PCA, there should be routine hourly monitoring of the latter throughout the duration of the treatment plus a further period of at least 2 hours after discontinuation.
An ECG, nerve stimulator, thermometer and capnograph should be readily available as well as facilities for resuscitation and emergencies. Women should only be discharged from the recovery area once they have been assessed by a trained recovery staff member and should be taken to the postoperative ward with all of their case notes. In addition no patient should be returned to a general ward unless control of emesis and postoperative pain is satisfactory. After the first 30 minutes postoperatively if the patient is stable then observations are carried out and documented half hourly, 2 hourly and then 4 hourly.324,524 [evidence level 4]
Recommendations
View in own window
| Number | Recommendation |
|---|
| 100 | After CS, women should be observed on a one-to-one basis by a properly trained member of staff until they have regained airway control and cardiorespiratory stability and are able to communicate. [D] [2004] |
| 101 | After recovery from anaesthesia, observations (respiratory rate, heart rate, blood pressure, pain and sedation) should be continued every half hour for 2 hours, and hourly thereafter provided that the observations are stable or satisfactory. If these observations are not stable, more frequent observations and medical review are recommended. [GPP] [2004] |
| 102 | For women who have had intrathecal opioids, there should be a minimum hourly observation of respiratory rate, sedation and pain scores for at least 12 hours for diamorphine and 24 hours for morphine. [GPP] [2004] |
| 103 | For women who have had epidural opioids or patient-controlled analgesia with opioids, there should be routine hourly monitoring of respiratory rate, sedation and pain scores throughout treatment and for at least 2 hours after discontinuation of treatment. [GPP] [2004] |
9.2. Pain management after CS
In the UK, intrathecal analgesia, patient controlled analgesia, local anaesthetic wound infiltration and nonsteroidal anti-inflammatory agents are commonly used for analgesia post-CS.
Intrathecal analgesia
Key issues related to intrathecal analgesia post-CS are which drug and dose to use as most side effects (particularly with morphine) are dose related.525 [evidence level 3] Morphine was commonly used until diamorphine was shown to be a useful alternative.526 [evidence level 1b] One RCT comparing intrathecal morphine with normal saline (n = 60) reported that the group given intrathecal morphine had less pain as measured by visual analogue scale (VAS) at 4 and 24 hours postoperative (p < 0.05) and morphine consumption was lower (p < 0.01).527 [evidence level 1b]
The documented side effects of intrathecal morphine include itching, nausea and vomiting. Alternative intrathecal opioids have been used more frequently because they have fewer reported side effects. One RCT (n = 40) comparing intrathecal diamorphine with intrathecal morphine reported no difference in VAS for pain or overall PCA morphine use. However VAS for itching and drowsiness were higher in the morphine group.528 [evidence level 1b]
Two RCTs have evaluated using intrathecal diamorphine in order to reduce the use of other analgesics. One RCT (n = 40) randomised women to either 0.3 mg of intrathecal diamorphine or normal saline (all women then had patient controlled analgesia via a morphine pump). Outcomes used were time to request for first analgesia and total morphine used. Both time to first analgesia (218 minutes versus 136.3 minutes, p < 0.05) and total morphine used were less in the group that received intrathecal diamorphine.529 [evidence level 1b] The second RCT (n = 40) used 0.3 mg of intrathecal diamorphine post spinal anaesthesia for CS. The women then used morphine PCA. The median amount of PCA morphine used over 24 hours was less in the group that received the intrathecal diamorphine (5 mg versus 45 mg, p < 0.05) and the time to request for first morphine dose was less (340 minutes versus 80 minutes, p = 0.0006).530 [evidence level 1b]
One RCT (n = 80) of women undergoing planned CS with spinal anaesthesia randomised to receive one of four doses of intrathecal diamorphine for post-CS analgesia (0.125 mg, 0.25 mg, 0.375 mg, or saline). The optimal intrathecal dose of diamorphine for intrathecal post-CS analgesia is reported to be between 0.25 mg and 0.375 mg. Nausea and pruritus increased as dose increased.531 [evidence level 1b] Higher doses than this have also been suggested because the minimum dose of intrathecal diamorphine required to prevent intraoperative supplementation of spinal anaesthesia for CS is 0.4 mg.532
Epidural diamorphine 2.5 mg to 5 mg is an alternative to intrathecal diamorphine as a significant proportion of unplanned CS (34%) are carried out using epidural anaesthesia.312 [evidence level 3] This has been evaluated in 2 RCTs. One RCT (n = 50) showed no difference in the duration and quality of analgesia between intrathecal and epidural diamorphine. There was no difference in the incidence of pruritus between the two groups but there was a higher incidence of nausea and vomiting in the epidural group (24% vs. 4%, p < 0.05).533 [evidence level 1b] The other RCT (n = 53) comparing epidural with intrathecal diamorphine reported that time to first request for morphine and side effects were similar between the two groups but VAS pain scores and additional morphine consumption was higher in the intrathecal group (p = 0.03 and p = 0.03 respectively).534 [evidence level 1b]
Another RCT compared intramuscular administration of diamorphine and four epidural regimens for the administration of diamorphine. Time to next analgesia was shorter in the intramuscular group when compared to any of the epidural groups (3.53 hours vs. 5.7, p = 0.007).535 [evidence level 1b]
A small American cost-effectiveness study was identified that evaluated the addition of intrathecal morphine to a regimen of oral analgesia. The effectiveness data was gathered retrospectively for 55 patients. The comparator was patient controlled analgesia. There was no synthesis of costs and benefits. The authors reported that the mean intrathecal morphine cost US$15 (1997 prices) compared with US$35 for patient controlled analgesia. Nursing time was also significantly reduced for intrathecal analgesia patients. Since there were no reported differences in pain control or side-effects, the study concluded that the addition of intrathecal morphine was a less expensive and less time consuming and therefore the more cost-effective option.536
Patient-controlled analgesia
Patient-controlled analgesia (PCA) is either epidural patient-controlled analgesia (EPCA) or via an infusion pump device. In the UK EPCA is not common practice post-CS and hence is not considered here
We did not identify any RCTs that evaluate the effectiveness of PCA compared to other forms of analgesia after CS. We identified two RCTs that compared different drugs for PCA after CS using a pump device. One RCT (n = 77) compared morphine alone to morphine combined with alfentanil for PCA. The group with alfentanil and morphine scored higher on a written questionnaire in terms of speed of onset of effectiveness of analgesia but there were no differences in terms of grading for duration of analgesia or overall satisfaction.537 [evidence level 1b] The other RCT compared morphine to fentanyl and found no difference in patient satisfaction or provision of effective analgesia over 37 hours.538 [evidence level 1b]
Wound infiltration with local anaesthetic
Three RCTs evaluated the use of wound infiltration and nerve blocks for post-CS analgesia specifically. One RCT (n = 45) used 20 ml of 0.1% bupivacaine infiltrated into the CS wound. They randomised the women into three groups: one group had general anaesthetic and wound infiltration; one group regional anaesthetic and wound infiltration and one group general anaesthetic only. They reported that the two groups that had wound infiltration did not use any pethidine in the first 6 hours postoperatively compared to the group with no infiltration in which all the women needed at least one dose of pethidine within the first 6 hours.539 [evidence level 1b]
Another RCT (n = 62) compared the effectiveness of bilateral ilioinguinal nerve block and wound infiltration with 0.5% bupivacaine for postoperative analgesia after CS. Mean VAS scores and mean papaveretum (morphine derivative) requirements were compared at 4, 8, 12, 16, 20 and 24 hours post-CS. Mean VAS scores for the ilioinguinal block group were reduced compared to control at 4, 8, 12, 20 and 24 hours and papaveretum requirements were less at 4, 8, 12 and 20 hours. Mean VAS scores for the wound infiltration group were reduced compared with the control group at 4 and 12 hours and papaveretum requirements less at 4, 8 and 12 hours (p < 0.05).540 [evidence level 1b]
A review of 26 RCTs, (n = 1211) evaluating the effectiveness of wound infiltration with local anaesthetic in a range of general surgical abdominal operations. Outcome measures were pain scores, supplementary analgesics and time to first analgesic requirements. Overall the study did not find any effect of local anaesthetic wound infiltration for postoperative pain.541 [evidence level 1a]
Non-steroidal anti-inflammatory analgesia
Non-steroidal anti-inflammatory drugs (NSAIDs) are used together with other modalities of pain relief after CS mainly to reduce the need for morphine based analgesics. We considered evidence on NSAID preparations available in the UK.
Two RCTs looked at the analgesic sparing effect of rectal NSAIDs suppository (diclofenac) administered immediately post-CS. In one RCT (n = 50) there was no difference in the VAS scores but the time to request for first analgesia was prolonged with rectal NSAID from 13 hours 45 minutes in the placebo group to 18 hours 58 minutes in the study group (p < 0.03).542 [evidence level 1b] The other RCT (n = 45) used the amount of PCEA as an outcome measure as well as VAS scores of pain. Women who received the rectal NSAID used less PCEA local anaesthetic solution (52.8 ml) compared to the control group (74 ml). There was no difference in VAS pain scores.543 [evidence level 1b]
Another RCT (n = 50) administered the NSAID (75 mg diclofenac) intramuscularly to women who were using morphine based PCA post-CS. The women who had the NSAID consumed less morphine via the PCA than the control group (mean at 18 hours post-CS was 61.4 mg compared with 91.4 mg).544 [evidence level 1b]
Complications following regional anaesthesia
In England and Wales 77% of unplanned and 91% of planned CS are performed with regional anaesthetic (spinal or epidural).4 [evidence level 3] Information on anaesthetic complications in the UK is not routinely collected other than serious complications resulting in mortality.95 A prospective multi disciplinary audit in the UK reported that epidural analgesia contributes to a neurological complication in 1/13,007 women.545 [evidence level 3] The National Obstetric Anaesthetic Database reported incidence of headache ranged from 1.1% to 1.9% between all anaesthetic techniques and increased to 11% for women receiving multiple regional anaesthetics.546 [evidence level 3] Unpublished data from a UK audit of 517,455 deliveries including 135,546 epidurals for analgesia and anaesthesia described complications rates associated with regional anaesthesia. 1/5000 (0.02%) epidural catheters are sited in the epidural vein; 1/3000 (0.034%) are sited in the intrathecal space; total spinal block occurs in 1/20,000 (0.005%) epidurals and 1/4000 (0.025%) subdural bleeds occur.547 [evidence level 3]
An audit of epidural related complications from Australia reports rates of complications for regional anaesthesia as follows: need for re insertion of epidural catheter 4.7%; hypotension after epidural for CS 28%; inadequate block 1.7%; conversion to general anaesthetic 0.5%. Serious complications are relatively rare: unexpected high block 0.07%; high block requiring intubation 0.02% respiratory depression 0.06% and local anaesthetic toxicity 0.04%.548 [evidence level 3]
Recommendations
View in own window
| Number | Recommendation |
|---|
| 104 | Women should be offered diamorphine (0.3–0.4 mg intrathecally) for intra- and postoperative analgesia because it reduces the need for supplemental analgesia after a CS. Epidural diamorphine (2.5–5 mg) is a suitable alternative. [A] [2004] |
| 105 | Patient-controlled analgesia using opioid analgesics should be offered after CS because it improves pain relief. [GPP] [2004] |
| 106 | Providing there is no contraindication, non-steroidal anti-inflammatory drugs should be offered post-CS as an adjunct to other analgesics, because they reduce the need for opioids. [A] [2004] |
View in own window
| Number | Research recommendation |
|---|
| RR 43 | Further research is needed to determine the effect of wound infiltration with local anaesthetic at CS on the need for post-CS analgesia. |
9.3. Early eating and drinking after CS
A systematic review compared early with delayed oral fluids and food after CS and included 6 RCTs. Three RCTs were limited to CS with regional anaesthesia; the other 3 RCTs included both regional and general anaesthesia. The intervention groups varied (either allowing immediate access to fluids and food within 6–8 hours if the woman was hungry or thirsty). The comparison groups delayed oral intake for a minimum of 12 hours to 24 hours, or to the presence of bowel sounds and graduated intake. Early eating and drinking was associated with reduced time to return of bowel sounds (1 RCT, n = 118; weighted mean difference of −4.3 hours, 95% CI −6.78 hours to −1.82 hours) and reduced postoperative hospital stay (2 RCTs, n = 220). There was no difference between the intervention and control groups with respect to nausea, vomiting, abdominal distension, time to bowel action, paralytic ileus and number of analgesic doses.549 [evidence level 1a]
Recommendations
View in own window
| Number | Recommendation |
|---|
| 107 | Women who are recovering well after CS and who do not have complications can eat and drink when they feel hungry or thirsty. [A] [2004] |
9.4. Urinary catheter removal after CS
Urinary bladder catheters are commonly used during CS to prevent damage to the bladder during surgery. The effect of urinary bladder catheterisation at CS on has been described in a prospective survey (n = 8402) as a risk factor for postpartum urinary retention.550 [evidence level 3] Evidence to determine timing of removal of the urinary catheter and the value of routine indwelling catheterisation is currently under review.551 We identified two RCTs on this topic. One RCT compared immediate catheter removal to removal of an indwelling catheter the next day in women who had a CS under general anaesthetic (n = 107). They report no difference in incidence of urinary tract infection (RR 1.64, 95% CI 0.80 to 3.34) but more instances of urinary retention with intermittent catheterization (39% vs. 0%).552 [evidence level 1b] A small RCT compared urinary retention after CS with a general anaesthetic to urinary retention after CS with an epidural anaesthetic and found no difference.553 [evidence level 1b]
Another RCT (n = 78) compared removal of the urinary bladder catheter immediately post-operatively with removal the next day in women undergoing gynaecological (pelvic) surgery, 29 who had CS. They found no difference in the incidence of urinary tract infection, urinary retention or fever but the method of randomisation is unclear and data given in the paper is incomplete.554 [evidence level]
Recommendations
View in own window
| Number | Recommendation |
|---|
| 108 | Removal of the urinary bladder catheter should be carried out once a woman is mobile after a regional anaesthetic and not sooner than 12 hours after the last epidural ‘top up’ dose. [D] [2004] |
9.5. Respiratory physiotherapy after CS
One RCT (n = 120) has evaluated the effect of respiratory physiotherapy after CS under general anaesthesia. The RCT did not detect a difference between the intervention group who had post-CS respiratory physiotherapy and the control group for coughing, phlegm, body temperature, chest palpation and auscultation.555 [evidence level 1b]
Recommendations
View in own window
| Number | Recommendation |
|---|
| 109 | Routine respiratory physiotherapy does not need to be offered to women after a CS under general anaesthesia, because it does not improve respiratory outcomes such as coughing, phlegm, body temperature, chest palpation and auscultatory changes. [A] [2004] |
View in own window
| Number | Research recommendation |
|---|
| RR 44 | Research is needed to establish the effect of non-respiratory physiotherapy for women following CS on post-CS recovery. |
9.6. Debriefing for women after CS
A longitudinal study, based in Australia suggests that a high level of obstetric intervention during childbirth, such as unplanned CS is associated with the development of acute traumatic symptoms in women postnatally.556 [evidence level 3] Midwife led debriefing has been proposed to be of value in reducing the incidence of depression and anxiety after birth. A systematic review (11 RCTs) evaluating the effect of psychological debriefing on the prevention of post traumatic stress disorder (PTSD) in the general population reported that single session individual debriefing did not affect the incidence of PTSD at 3 to 5 months (6 RCT, n = 387, OR 1.22, 95% CI 0.60 to 2.46) and increased the likelihood of long term PTSD (after one year 2 RCTs, n = 238, OR 2.04, 95% CI 0.92 to 4.53).557 [evidence level 1a] Only two of the included studies were in an obstetric setting.558,559 Of these two trials, one was UK based (n = 129) and included primigravid women who had a normal vaginal birth.
Women who received midwife debriefing were less likely to have high anxiety and depression scores after birth than women who did not (anxiety score OR 13.5, 95% CI 0.41 to 56.9; depression OR 8.5, 95% CI 2.8 to 30.9).559 [evidence level 1b] The second RCT was from Australia (n = 1041) looked at the effect of midwife-led debriefing on maternal depression after operative childbirth. No difference was detected in depression scores (OR 1.24, 95% CI 0.87 to 1.77) or in the proportion of women who reported that depression had been a problem at six months after the birth (OR1.37, 95% CI 1.00, 1.86)558 [evidence level 1b]
Subsequently a further two RCTs have been published. One RCT (n = 103) tested opportunity to debrief at an initial postnatal interview (less than 72 hours postpartum) and 4–6 weeks postpartum to usual care. The RCT reported a high baseline prevalence of post-traumatic stress disorder (9.6% of women at 4 to 6 weeks postpartum). No difference was detected in the prevalence of symptom profile for PTSD immediately following debriefing or at 3 months. (RR 1.06, 95% CI 0.61 to 1.84). This is RCT is underpowered to detect a 2% difference in prevalence of symptoms of post-traumatic stress disorder.560 [evidence level 1b] A recently published RCT (n = 1745) compared a midwife debriefing session within 72 hours of birth to usual care. No differences were detected between the groups for either stress disorders or depression (assessed EPDS and report of depressive illness).561 [evidence level 1b]
Recommendations
View in own window
| Number | Recommendation |
|---|
| The recommendation for this review has been amended and updated following a new (2011) review. The new recommendation can be found in section 11.2 |
View in own window
| Number | Research recommendation |
|---|
| RR 45 | More RCT evidence is required to determine the effect of midwifery-led debriefing following CS. |
9.7. Length of hospital stay and readmission to hospital
Length of hospital stay after childbirth is declining; recent routine national statistics for England562 suggest that women who have a spontaneous vaginal delivery spend on average 1 day in hospital, women who have an instrumental delivery spend 1 or 2 days in hospital and women who have a CS spend 3 or 4 days in hospital.
In one RCT48 that compared planned CS with planned vaginal birth, the median length of hospital stay for women in the planned CS group was 4 days (5th centile 1.7 days, 95th centile 7.4 days). For women in the planned vaginal birth group it was 2.8 days (0.8, 6.9 days). The median length of stay reported in this RCT48 is compatible with routine maternity statistics for the U.K. In 3 RCTs38,39,42 the length of hospital stay was reported as either greater or less than 10 days. On pooling these results, the relative risk of length of hospital stay greater than 10 days for women in the planned CS group was 1.27 (95% CI 0.35 to 4.65). [evidence level 1a]
Readmission to hospital
Infection and bleeding constitute the main reasons for readmission to hospital following birth.563 Two surveys of women in the postpartum period have estimated about 3% are readmitted to hospital for reasons related to their own health.563,564
Readmission to hospital was not included as an outcome measure in the RCTs of planned CS versus planned vaginal birth.
One prospective cohort study in Australia564 examined rates of readmission to hospital within 8 weeks of birth. A higher proportion of women who had CS (5.3%) reported readmission to hospital compared to women who had vaginal birth (2.2%) (OR 2.46, 95% CI 1.11 to 5.43). [evidence level 2b] Similar findings were reported in a retrospective cohort study conducted in Washington USA565 (n = 256,795). The age adjusted relative risk for rehospitalisation among women who had CS compared to those who had vaginal birth was increased (RR 1.8, 95% CI 1.6 to 1.9). [evidence level 2b]
Discharge from hospital after CS usually occurs on day 3.562 [evidence level 3] A systematic review of early post natal discharge from hospital included eight RCTs but only two RCTs included women who had caesarean births, one of which is ongoing.566 [evidence level 1a] The RCT (n = 61) randomised women having CS to either early hospital discharge and home follow up or usual hospital discharge (requires the woman to be ambulatory, voiding, tolerating a normal diet, passing flatus, normal uterine involution, afebrile for 24 hours, uncomplicated wound healing, removal of skin sutures or stapes and an adequate blood count). Women in the intervention group were discharged when they met the same criteria other than afebrile for 24 hours and staple or suture removal. They report no difference in maternal or infant rehospitalisations, maternal affect or overall maternal functional status. Women in the early discharge group were more satisfied with care and had a 29% reduction in health care requests.567 [evidence level 1b]
Recommendations
View in own window
| Number | Recommendation |
|---|
| 110 | Length of hospital stay is likely to be longer after a CS (an average of 3–4 days) than after a vaginal birth (average 1–2 days). However, women who are recovering well, are apyrexial and do not have complications following CS should be offered early discharge (after 24 hours) from hospital and follow-up at home, because this is not associated with more infant or maternal readmissions. [A] [2004] |
