Fetal bradycardia due to intrathecal opioids for labour analgesia: a systematic review

BJOG: an International Journal of Obstetrics and Gynaecology March 2002, Vol. 109, pp. 274 –281

Fetal bradycardia due to intrathecal opioids for labour analgesia: a systematic review Chahe´ Mardirosoff a,*, Lionel Dumontb, Michel Boulvainc, Martin R. Trame`rb Objective To evaluate fetal and maternal adverse effects of intrathecal opioid analgesia during labour. Data sources A systematic search was performed, in Medline, Embase, the Cochrane Library, bibliographies, and personal contact with authors, in any language, up to February 2001. Study selection Full reports on randomised comparisons of any analgesia with intrathecal opioid (experimental group) with any non-intrathecal opioid regimen (control group) during labour. Data extraction Dichotomous data from 24 trials (3513 women). Results With intrathecal opioids, there was a significant increase in the risk of fetal bradycardia: odds ratio 1.8 (95% confidence interval 1.0 to 3.1), number-needed-to-harm 28. The risk of caesarean section due to fetal heart rate abnormalities was similar (6.0% versus 7.8%). The incidence of pruritus was significantly higher with intrathecal opioids: relative risk 29.6 (95% CI 13.6 to 64.6), number-needed-to-harm 1.7. Conclusions Intrathecal opioids for labour increase the risk of fetal bradycardia and maternal pruritus. The risk of subsequent caesarean section is not increased. INTRODUCTION In 1996, 13.5% of the French population received an anaesthetic; of those 9% were for labour1. Extrapolating these numbers to the UK indicates that 700,000 women receive an anaesthetic during labour each year. Intrathecal opioid analgesia, the injection of an opioid into the cerebrospinal fluid, has become a widely used technique for pain relief during labour. The main advantages of this method compared with the conventional epidural analgesia with a local anaesthetic are the lack of motor weakness enabling ambulation of the women, the faster onset of analgesia, and the reliability of the spinal injection2. Recently, fetal heart rate abnormalities including severe bradycardia have been reported following the administration of intrathecal opioids during the first stage of labour 3. It has been suggested that the risk of subsequent caesarean section may be increased4. Given the potentially higher incidence of side effects with intrathecal opioids, it may not be justified to use this technique as a routine analgesia in labour 5. However, several large series were unable to

a

Department of Anaesthesiology, Polyclinique de Savoie, Annemasse, France b Division of Anaesthesiology, Geneva University Hospitals, Geneva, Switzerland c Department of Gynaecology and Obstetrics, Geneva University Hospitals, Geneva, Switzerland * Correspondence: Dr C. Mardirosoff, De´partement d’Anesthe´sieRe´animation, Polyclinique de Savoie, 8 Rue Fernand-David, 74100 Annemasse Cedex, France. D RCOG 2002 BJOG: an International Journal of Obstetrics and Gynaecology PII: S 1 4 7 0 - 0 3 2 8 ( 0 2 ) 0 1 3 8 0 - 0

confirm an increased risk with intrathecal opioids6 – 8. A recent review of both randomised and observational studies suggested that fetal heart rate abnormalities ‘‘occurred with similar frequencies in association with any method of effective pain relief ’’9. For rational decision-making we need to know if there is a causal relationship between intrathecal opioids and fetal bradycardia, and if this adverse effect increases the risk of emergency caesarean section. We also need to quantify these risks. The objective of this systematic review was to evaluate the adverse effects related to the administration of intrathecal opioids for labour.

METHODS We included reports of randomised comparisons of any intrathecal opioid with or without a local anaesthetic (experimental group) with any analgesic regimen that excluded intrathecal opioids (control group) for labour analgesia. Relevant studies had to report dichotomous data on adverse effects. We searched systematically in Medline, Embase, and the Cochrane library (issue 1, 2001), using the free text terms combined spinal epidural, intrathecal, spinal, subarachnoid, labor, labour, opioid, randomized, randomised, controlled and, control, and combinations of the above. The last electronic search was performed on 25 January 2001. We hand searched locally available anaesthesia journals including the non-indexed International Journal of Obstetric Anaesthesia up to January 2001. We checked bibliographies of retrieved reports. No language restriction was applied. When there was ambiguity about data or when www.bjog-elsevier.com

FETAL BRADYCARDIA DUE TO INTRATHECAL OPIOIDS

relevant data were missing in the original reports, we contacted authors and asked for clarification and also for unpublished data. We did not consider data from abstracts, letters, comments, editorials, or review articles. One author (C.M.) screened all retrieved reports. Studies that were not relevant for the purpose of this systematic review were excluded at this stage. Each potentially relevant report was then read by at least three authors independently to assess adequacy of randomisation, blinding and description of withdrawals, according to the validated three-item, 5-point Oxford scale10. Relevant data were extracted by one author (C.M.) and checked by the others. The primary endpoint was the incidence of one or more episode of fetal bradycardia, as defined by the authors, occurring within one hour after the injection of the study drugs and not related to maternal hypotension. This pre-stated definition was used to prevent misclassification of bradycardia not related to intrathecal opioids. Other endpoints were the incidence of caesarean section due to any fetal heart rate abnormalities and due to fetal bradycardia, spontaneous and instrumental delivery, use of oxytocin, Apgar scores, and maternal pruritus. We estimated the effect of intrathecal opioids calculating the relative risks and their 95% confidence interval11. A fixed effect model was used when there was no statistically significant heterogeneity ( P > 0.1); otherwise we used a random effects model12. If there were no events in a group, then 0.5 was added to all cells for that study, to calculate the relative risk13. As bradycardia was a very rare event, we calculated Peto odds ratios, which is a method more appropriate when too many groups show no events14. As an estimate of the clinical relevance of any difference between intrathecal opioids and control we calculated the number-needed-to-harm with 95% confidence interval15. When there were many groups with zero events, we calculated the number-needed-to-harm using the relative risk or the odds ratio and the average control event rate. Confidence intervals around the numbers-needed-to-harm are reported even when they contain negative numbers, and thus infinity16.

RESULTS Of 34 reports containing potentially relevant data for the purpose of this study, ten were subsequently excluded 4,6,8,17 – 23 (Fig. 1). Data from 24 trials published between 1991 and January 2001 were analysed5,24 – 46. We contacted authors of 14 reports for additional information. Eight responded to our inquiry and additional relevant data which could be included in our analyses were provided by four 26,33,35,46. The median quality score of the included studies was 3.5 (range 1 to 5). There were data on 2020 women receiving intrathecal opioids, and on 1493 controls. Median trial size was 73 women (range 24 to 1008). Most women were nulliparous. In most trials it was unclear if D RCOG 2002 Br J Obstet Gynaecol 109, pp. 274 – 281

275

Fig. 1. Retrieved, included and excluded reports of trials.

women were allowed to ambulate or not. We pooled data from ambulating and non-ambulating women in the only trial that made a clear distinction between these two groups40. Three intrathecal opioids were tested (sufentanil, fentanyl, morphine), with or without various doses of intrathecal or epidural bupivacaine (Table 1). In controls, different doses of epidural or intrathecal bupivacaine, epidural lidocaine, combinations of epidural bupivacaine and different doses of epidural sufentanil or fentanyl, and intravenous sufentanil were used. Seventeen trials (2081 women) reported on the incidence of fetal heart rate abnormalities. Abnormalities were defined as tachycardia, decelerations (mild, moderate or severe, late onset or variable) or bradycardia, occurring at any time during labour. There was no difference between intrathecal opioids and controls (Table 2). In nine trials (927 women), dichotomous data on fetal bradycardia within one hour after the intrathecal injections were reported. In three of these studies, there were no events in the experimental and the control group26,36,41 (Fig. 2). In the other six the risk was consistently increased with intrathecal opioids, although in none did the difference reach statistical significance. When the data from all trials were combined, the incidence of fetal bradycardia was significantly increased with intrathecal opioids, odds ratio 1.81 (95% CI 1.04 to 3.14). The number-needed-to-harm was 28, which means that for 28 women receiving an intrathecal opioid for

276 C. MARDIROSOFF ET AL.

Table 1. Intrathecal (IT) opioid for labour analgesia: included randomised controlled trials. Reference No.

Analgesic protocol IT opioids

24 25 26 27

D RCOG 2002 Br J Obstet Gynaecol 109, pp. 274 – 281

IT morphine 0.2 mg þ/ epidural bupivacaine 0.125% 10ml (40) IT fentanyl 25 mg þ/ bupivacaine 2.5 mg (42) IT fentanyl 25 mg þ morphine 250 mg (26) IT sufentanil 10 mg þ intravenous or epidural saline (9)

Randomisation and concealment of the allocation

Blinding

Control (No. IT opioids) Epidural bupivacaine 0.125% 10 ml (22)

Not described

None

None

IT bupivacaine 2.5 mg (34)

Not described

Observer

Technical problems (4)

Epidural bupivacaine 0.25% þ sufentanil 10 mg (33) IT þ intravenous saline þ epidural sufentanil 10 mg or IT þ epidural saline þ intravenous sufentanil 10 mg (15) Epidural bupivacaine: 5 to 25 mg (34)

Not described

None

None

Random numbers, sequentially numbered opaque envelopes Not described

Drugs prepared by anesthesiologist not involved in the study Drugs prepared by anaesthesiologist not involve in data collection Solutions prepared by anaesthesiologist not involve in the study None None

None

28

IT sufentanil (2 mg to 10 mg) þ/ epidural bupivacaine (2.5 to 12.5 mg) (66)

29

IT sufentanil 10 mg þ/ bupivacaine 2.5 mg (29)

IT bupivacaine 2.5 mg (14)

Computer-generated

30 31

IT fentanyl 25 mg (12) IT fentanyl 25 mg þ bupivacaine 2.5 mg (98) IT sufentanil 10 mg þ epidural saline 12 ml (25)

Epidural lidocaine 1.5% 10 ml (12) Epidural bupivacaine 0.25% 10 ml (99)

Not described Not described

IT saline þ epidural bupivacaine 0.25% 12 ml (25)

Table of random numbers

33

IT fentanyl 25 mg þ bupivacaine 2.5 mg (524)

Epidural bupivacaine 0.1% 20 ml þ fentanyl 2 mg/ml (484)

Sealed envelopes

Solutions prepared by anaesthesiologist not involve in the study Observer

34

IT sufentanil 10 mg (35)

Epidural lidocaine 1.5% 3 ml with epinephrine (test dose) þ sufentanil 40 mg (34)

Not described

Observer

32

Exclusions (n)

None

Technical problems (4)

None None Technical problems (7), early deliveries (5) Removal of consent (8), early labour (9), second stage labour (9), rapid analgesia requirement (19), other causes (5) No analgesia achieved (1)

D RCOG 2002 Br J Obstet Gynaecol 109, pp. 274 – 281

35

IT sufentanil 10 mg þ epidural saline 8 ml (30)

IT saline þ epidural bupivacaine 0.25% 8 ml (28)

Random numbers

36

IT fentanyl 25 mg þ bupivacaine 2.5 mg (26) IT sufentanil 10 mg þ bupivacaine 1 mg þ epinephrine 25 mg þ Epidural saline 10 ml (32) IT sufentanil 10 mg or IT fentanyl 10 mg þ bupivacaine 2.5 mg (39)

Epidural bupivacaine 0.0625% þ fentanyl 2 mg/mg (24) Epidural bupivacaine 12.5 mg þ sufentanil 10 mg þ epinephrine 12.5 mg (31)

37

38

39

40

5 41 42 43 44

46

IT fentanyl 25 mg þ bupivacaine 2.5 mg (69) IT sufentanil 10 mg þ epidural saline 12 ml (20) IT fentanyl 25 mg þ bupivacaine 2.5 mg (45) IT sufentanil 10 mg (39) IT sufentanil 10 mg þ bupivacaine 2.5 mg (50) IT sufentanil 1.5 mg þ bupivacaine 2.5 mg þ epinephrine 2.5 mg þ epidural saline 10 ml (55) IT sufentanil 2.5 mg to 10 mg þ IT bupivacaine 2.5 mg (136)

Not described

Observer

None

IT saline þ bupivacaine 2.5 mg (20)

Not described

Women and obstetricians

IT bupivacaine 1.25 mg þ epinephrine 25 mg (23)

Table of random numbers

Anaesthesiologist and observer

Abruptio placentae with emergency cesarean section (1) Protocol violation (1), technical problem (1)

Epidural bupivacaine 0.25% 5 ml (test Not described dose) þ bupivacaine 0.25% 6 ml þ fentanyl 50 mg (256) Epidural bupivacaine 12.5 mg þ fentanyl 50 mg (73) IT saline þ epidural bupivacaine 0.25% 12 ml (20) Epidural bupivacaine 0.1% 15 ml þ fentanyl 75 mg (48) Epidural sufentanil 20 mg þ bupivacaine 0.25% 6 to 8 ml (40) Epidural bupivacaine 0.25% 12 ml (50)

Not described

None

Incomplete data (14)

Not described

Women and midwife

None

Not described

Anaesthesiologist and observer Women and midwife

None

Epidural bupivacaine 0.125% 10 ml þ sufentanil 0.75 mg/ml þ epinephrine 1.25 mg/ml (55) IT bupivacaine 2.5 mg (19)

Sealed and numbered envelopes Table of random Numbers Opaque, shuffled envelopes Not described

Computer-generated table

Observer

Refusal or incomplete data (7) Technical problem (1)

Observer

None

Women and observer

None

Anaesthesiologist and observer

None

FETAL BRADYCARDIA DUE TO INTRATHECAL OPIOIDS

45

IT sufentanil 2.5 or 5 mg þ bupivacaine 1.25 mg þ epinephrine 25 mg (68) IT sufentanil 10 mg (505)

None

Not described

Solutions prepared by anaesthesiologist not involve in the study Observer

277

278 C. MARDIROSOFF ET AL.

Table 2. Obstetric outcomes and adverse effects of intrathecal (IT) opioids for labour analgesia. Outcome

Event rate with IT opioids n/n (%)

Event rate in controls n/n (%)

Fetal heart rate abnormalities

97/1264 (7.7)

55/817 (6.7)

1.17 (0.87 to 1.57)

75 (22 to 97)

Fetal bradycardia within 1 h

39/535 (7.3)

19/392 (4.8)

1.81 (1.04 to 3.14)**

Caesarean section Any indication Fetal heart rate abnormalities

286/1682 (17.0) 52/871 (6.0)

211/1272 (16.6) 54/689 (7.8)

Instrumental delivery

393/1600 (24.6)

Spontaneous delivery Oxytocin

D RCOG 2002 Br J Obstet Gynaecol 109, pp. 274 – 281

Apgar < 7 at 5 min Pruritus Opioids in controls No opioids in controls

No. of studies

References

0.99

17

5,25,26,28 – 31,34,36 – 41,43,45,46

28 (11 to 594)

0.96

9

1.03 (0.87 to 1.21) 0.86 (0.60 to 1.23)

208 (29 to 47) 87 (31 to 54)

0.98 0.99

17 8

24,27,29,31 – 40,42 – 45 24,33,34,36 – 38,43,46

312/1231 (25.3)

0.94 (0.83 to 1.07)

66 (22 to 54)

0.74

15

29,31 – 40,42 – 45

983/1666 (59.0)

757/1286 (58.9)

1.01 (0.95 to 1.07)

204 (24 to 33)

0.31

17

24,26,29,31 – 40,42 – 45

462/764 (60.5)

250/473 (52.9)

1.06 (0.96 to 1.16)

34 (12 to 48)

0.54

7

10/964 (1.0)

6/659 (0.9)

1.17 (0.44 to 3.11)**

623 (50 to 192)

0.99

11

24,26,31,35 – 37,39,40,42,43,45

402/772 (52.0) 355/539 (65.9)

158/536 (29.5) 2/316 (0.6)

1.71 (0.97 to 3.02)* 29.6 (13.6 to 64.6)

4.7 (1.7 to 125) 1.7 (1.0 to 3.9)