Epidurals and labour |
||
|
|
Review
The review had a thorough search strategy for randomised studies comparing epidural analgesia with parenteral opioids. It used MEDLINE, the Cochrane Library, and hand searched non-abstracted journals and meeting abstracts. It obtained information from authors of abstracts and papers where needed. The primary outcome measure was the Caesarean section rate. There were many secondary outcomes sought.
Results
Ten studies met the inclusion criteria, with 1614 nulliparous and 755 multiparous patients randomised. All but one study provided information on an intention-to-treat basis. The parenteral opioid was predominantly pethidine (meperidine) given by intramuscular or intravenous injection; one study used intravenous fentanyl and another intravenous butorphanol.
The Caesarean section rate was 8.2% with epidural and 5.6% with parenteral opioid (Figure). They found no significant difference using a random-effects model for odds ratios. The implication is that there may be one additional Caesarean section for every 40 women given an epidural rather than parenteral opioid. Caesarean section rates were no different with epidural or parenteral opioid for nulliparous women (about 8%) or multiparous women (about 2.5%).
Figure: Caesarian section rates with epidural and parenteral opioid analgesia. Each point is one trial (filled circles are trials in original review).
|
|
There were a number of secondary outcomes for mothers and babies (Table). For mothers, epidurals resulted in much less pain in the first and second stages of labour, and much less dissatisfaction. However, the cost was a higher rate of instrumental delivery, longer labours, more use of oxytocin and more episodes of hypotension and elevated temperature. For newborns, epidurals resulted in fewer babies with low APGAR scores or low umbilical pH, and fewer needing naloxone. |
| Maternal outcome | Studies/ Patients | Result: NNT or NNH (95%CI) |
Better outcomes with epidural |
||
| First stage labour pain | 6/ 2031 | Average 40 mm lower with epidural (95%CI 38 to 42 mm) on 100 mm scale |
| Second stage labour pain | 5/ 1062 | Average 29 mm lower with epidural (95%CI 21 to 38 mm) on 100 mm scale |
| Dissatisfaction | 5/ 1581 | 15% with epidural, 43% with parenteral opioid. NNT 3.6 (3.1 to 4.3) |
Worse outcomes with epidural |
||
| Instrumental delivery | 9/ 2319 | 15% with epidural, 9% with parenteral opioid. NNH 15 (11 to 26) |
| First stage labour duration | 5/ 1079 | Average 42 minutes longer with epidural (95%CI 17-68 minutes) |
| Second stage labour duration | 6/ 1190 | Average 14 minutes longer with epidural (95%CI 5-23 minutes) |
| Use of oxytocin after analgesia | 4/ 1001 | 45% with epidural, 32% with parenteral opioid. NNH 7.9 (5.4 to 15) |
| Temperature >38 °C | 2/ 1371 | 23% with epidural, 5% with parenteral opioid. NNH 5.6 (4.7 to 7.0) |
| Hypotension | 3/ 1684 | 37% with epidural, 0% with parenteral opioid. NNH 2.7 (2.5 to 3.0) |
No difference between epidural and parenteral opioid |
||
| Nausea | 5/ 835 | 7% with epidural, 6% with parenteral opioid. No significant difference |
| Instrumental delivery for dystocia | 2/ 211 | 12% with epidural, 17% with parenteral opioid. No significant difference |
| Neonatal outcome | Studies/ Patients | Result: NNT (95%CI) |
Better outcomes with epidural |
||
| APGAR <7 at 1 minute | 6/ 2015 | 1% with epidural, 2% with parenteral opioid. NNT 68 (40 to 231) |
| APGAR <7 at 5 minutes | 7/ 2176 | 4% with epidural, 7% with parenteral opioid. NNT 35 (21 to 112) |
| Low umbilical artery pH | 6/ 2034 | 14% with epidural, 17% with parenteral opioid. NNT 27 (15 to 220) |
| Need for naloxone in the newborn | 2/ 815 | 1% with epidural, 3% with parenteral opioid. NNT 37 (22 to135) |
No difference between epidural and parenteral opioid |
||
| Foetal heart rate abnormal, or intrapartum meconium | 3/ 1126 | 19% with epidural, 20% with parenteral opioid. No significant difference |
| Severe asphyxia (umbilical artery pH <6.99) | 5/ 1715 | <1% on both. No significant difference |
CommentThe bulk of the information in this review was published after that featured in Bandolier 34 (and published in 1994) was completed. There is now information on over 10 times as many randomised patients than in the earlier review. It shows clearly that the earlier trials over-estimated any increased Caesarean section rate with epidural analgesia in labour. Reference:
|