Patient-controlled epidural analgesia versus continuous infusion for labour analgesia: a meta-analysis

M. van der Vyver1, S. Halpern*,2 and G. Joseph2

1 Department of Anaesthesia, University of Stellenbosch and Tygerberg Academic Hospital Private Bag 3, Tygervallei 7505, Western Cape, South Africa. 2 Department of Anaesthesia, University of Toronto and Sunnybrook and Women’s Health Sciences Centre, Women’s College Campus, 76 Grenville Street, Toronto, Ontario, M5S 1B2 Canada *Corresponding author

Accepted for publication: April 11, 2002


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. Patient-controlled epidural analgesia (PCEA) is a relatively new method of maintaining labour analgesia. There have been many studies performed that have compared the efficacy of PCEA with continuous epidural infusion (CEI). The purpose of this systematic review is to compare the efficacy and safety of PCEA and CEI.

Methods. All randomized controlled trials that compared PCEA, without background infusion, with CEI were sought from the literature. These were rated for quality using a validated, five-point scale. The primary outcome was the number of patients who received anaesthetic interventions. Secondary outcomes included the dose of local anaesthetic, incidence of motor block, quality of analgesia, obstetric and safety outcomes. Where feasible, the data were combined using meta-analytical techniques. For dichotomous data, the risk difference (RD) and 95% confidence intervals (CI) were calculated. For continuous data, the weighted mean differences (WMD) were calculated. The differences were statistically significant when the 95% CI excluded 0.

Results. Nine studies comprised of 640 patients were found. There were fewer anaesthetic interventions in the PCEA group (RD, 27%; 95% CI, 18–36%; P<0.00001). This group also received less local anaesthetic (WMD, –3.92; 95% CI, –5.38 to –2.42; P<00001) and less motor block (RD, 18%; 95% CI, 6–31%; P=0.003). Both methods were safe for mother and newborn.

Conclusion. Patients who receive PCEA are less likely to require anaesthetic interventions, require lower doses of local anaesthetic and have less motor block than those who receive CEI. Future research should be directed at determining differences in maternal satisfaction and obstetric outcome.

Br J Anaesth 2002; 89: 459–65

Keywords: analgesia, obstetric; analgesia, patient-controlled; analgesic techniques, epidural; safety, techniques


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Epidural analgesia is the most effective way of providing pain relief in labour.1 A number of techniques have evolved to maintain the level of analgesia throughout labour. Historically, intermittent bolus dosing of local anaesthetic by the clinician (anaesthetist, nurse or midwife) was used. However, this technique had a number of drawbacks including inconsistent analgesia, potential toxicity, and concerns about sterility each time the clinician opened the system to administer a bolus.

Continuous epidural infusion (CEI) of local anaesthetics was introduced into common clinical practice in the 1980s. While the technique has circumvented a number of difficulties,2 it is not ideal. Many patients still required clinician-initiated top-ups and experienced unacceptably dense motor block in the lower extremities. Although many combinations of infusion rates and various concentrations of local anaesthetics and additives have been investigated, these problems persist.

Patient-controlled epidural analgesia (PCEA) for relief of labour pain was first described by Gambling in 1988.3 This technique allowed the patient to control the dose of epidural medication as labour and pain patterns changed. It also allowed for individualization of drug dosage by the patient, allowing her to trade off therapeutic effects (e.g. complete pain relief) and side-effects (e.g. motor block). However, the equipment needed for PCEA may be more expensive than CEI. Also, more time is required to educate both the patient and staff about the appropriate use of the medication and equipment.

Many studies have been performed comparing PCEA with CEI. The purpose of this systematic review is to compare the efficacy and safety of PCEA with that of CEI for maintenance of labour analgesia.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We sought randomized controlled trials (RCTs) that compared PCEA with CEI. Reports were identified by electronically searching Medline (January 1980 to September 2001), EMBase (January 1980 to September 2001), Science Citation index (1993–2001) and the Cochrane Library (June 2000, Issue 3).

We used the following search terms as key words and text words: ‘Patient-controlled’, ‘labour analgesia’, ‘anaesthesia’ (excluding anaesthesia for Caesarean section) and ‘pregnancy’. Alternative spellings of the search terms were also used. No language restriction was applied. The last search was done on the September 20, 2001. Additional reports were identified from reference lists of retrieved reports and review articles.

We included published RCTs in healthy parturients that used the same local anaesthetic and additives for both the study and control group. Furthermore, we only included RCTs that had no continuous background infusion in the PCEA group. We excluded abstracts of scientific meetings, unpublished observations and correspondence.

Each report meeting the inclusion criteria, was read independently by each of the authors and assigned a quality score according to a previously published, validated scale.4 This scale consists of three items and has a maximum possible score of five. When the report was described as randomized, one point was given. If the randomization process was appropriate an extra point was given. The original point for randomization was lost if randomization was inappropriate (e.g. if it was based on date of birth). One point was given when a study was described as blinded. If blinding was inappropriate this point was lost, but an extra point was gained if blinding was appropriate. If all patients randomized were fully accounted for another point was given. The final quality score for each article was reached by the authors by means of consensus.

The primary outcome of the study was the number of patients who received anaesthetic interventions during the maintenance of labour analgesia. Secondary outcomes included the dose of local anaesthetic, the incidence of motor block, the quality of analgesia and maternal satisfaction. Obstetric and neonatal outcomes included the length of labour, rate of Caesarean section, the rate of instrumental deliveries and the incidence of low Apgar scores at 1 and 5 min. Other outcomes included the incidence of hypotension, high block, nausea, pruritus and shivering.

Data were recorded independently by two of the authors (M.V. and G.J.) and discrepancies resolved by reinspection of the original articles. The data were entered into the statistical program (S.H.H) and rechecked (M.V.).

Where feasible, meta-analytical techniques, using a random effects model were used (Metaview software, Revman 4.1, Cochrane Library, Oxford, UK). For dichotomous data, the risk difference (RD) and 95% confidence intervals (95% CI) were calculated. For continuous data, the weighted mean difference and 95% CI were calculated. We considered the results statistically significant if the 95% CI excluded 0. The {chi}2 test was used to calculate heterogeneity and a P-value of <0.05 was regarded as significant. Where meta-analysis was not appropriate, the results of the individual studies were reported. We also performed a sensitivity analysis on the primary outcome on the basis of the quality score (0–2 vs 3 or more).


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We identified a total of 12 RCTs. Two manuscripts were not available through the University of Toronto Library system and were excluded.5 6 The outcome of one of the studies was reported in two manuscripts.7 8 Therefore, nine studies (in ten manuscripts) described a total of 641 patients of whom 545 patients received bupivacaine714 and 96 patients received ropivacaine.15 16 The range of numbers of patients included in the studies was 30–159. The median quality score was 3. Five studies had quality scores of three or greater.1114 16

Table 1 details the quality score, the obstetric population studied, the number of patients involved in each group and drugs used for maintenance in PCEA and CEI groups. None of the studies reported the incidence of induced and spontaneous labour. All measured outcomes, as well as those that were significantly different, are shown.


View this table:
[in this window]
[in a new window]
 
Table 1 PCEA vs CEI study characteristics. B=bupivacaine; R=ropivacaine; epi=epinephrine; F=fentanyl; S=sufentanil; PCEA=patient-controlled epidural analgesia; CEI=continuous epidural infusion; UA=umbilical artery; UV=umbilical vein
 
Unscheduled anaesthetic interventions
In five of the studies, clinician top-up of the epidural for inadequate analgesia was reported as a separate outcome.911 14 In one study, this outcome and catheter manipulation or replacement were considered together.7 Figure 1 shows the combined data for the number of patients who needed no unscheduled anaesthetic interventions. Significantly fewer patients needed clinician top-ups in the PCEA group compared with CEI (RD, 27%; 95% CI, 18–36%; P<0.00001). This observation was also true when only studies with a quality score of three or greater were considered (RD, 35%; 95% CI, 19–51%; P<0.0001). Although four of the studies included patients of mixed parity,9 1113 there was no evidence of heterogeneity (P=0.36) among the studies for this outcome.



View larger version (21K):
[in this window]
[in a new window]
 
Fig 1 The number of patients who required no unscheduled interventions by the clinician. The RD and 95% CI for each study and the pooled estimates are shown. The size of the box represents the relative weight of each study, the lines represent the 95% CI. The midpoint of the diamond is the pooled estimate of the RD. The diamond extends over the 95% CI of the pooled estimate.

 
Three studies reported the number of anaesthetic interventions in a manner that could not be combined in the meta-analysis above. One did not report the number of top-ups but noted the volume of local anaesthetic given per hour. In this study, the midwives administered less ropivacaine in the PCEA group compared with the CEI (5.4 vs 6.9 ml h–1; ns).16 Two studies did not measure the number of patients who required interventions, rather they reported the amount of bupivacaine used in each group.12 13 One of these studies reported that ‘similar numbers’ of patients required top-ups in each group.12 The other reported an increased bupivacaine requirement in the PCEA group compared with CEI related to clinician top-ups.13 This was statistically significant for first stage but not for the second stage or the total length of labour.

In addition to reporting the number of patients who required no interventions, three studies reported the total number of clinician top-ups in the CEI group compared with PCEA. Tan and colleagues8 found a significantly higher number in the CEI group than in the PCEA group (120 top-ups in 84 women compared with 49 in 75 women, respectively). Curry and colleagues11 reported 27 top-ups in 30 patients who received CEI compared with only two top-ups in 30 patients who received PCEA (P<0.001). Sia and colleagues15 reported six top-ups in 20 patients in the CEI group compared with three top-ups in 20 patients in the PCEA group during the first stage of labour (P=0.4).

Drug dose
The dose (mg h–1) of local anaesthetic administered to patients receiving CEI was significantly greater than those in the PCEA group (Table 2). All studies that measured this outcome reported the same result regardless of whether bupivacaine or ropivacaine was used. The addition of opioid to the maintenance solution (either fentanyl or sufentanil) did not make a difference to this outcome. Because different regimen were used, this outcome is heterogeneous (P<0.0001) even though all studies favoured the PCEA group (Fig. 2).


View this table:
[in this window]
[in a new window]
 
Table 2 Summary of study outcomes. PCEA=patient-controlled epidural analgesia; CEI=continuous epidural infusion; WMD=weighted mean difference; RD=risk difference, CI=confidence interval
 


View larger version (20K):
[in this window]
[in a new window]
 
Fig 2 Dose of local anaesthetic. The weighted mean difference and 95% CI and pooled estimates are shown. The size of the box represents the relative weight of each study, the lines represent the 95% CI. The midpoint of the diamond is the pooled estimate of the weighted mean difference. The diamond extends over the 95% CI of the pooled estimate.

 
Motor weakness
Motor power was measured in a number of different ways in each of the studies. Two studies reported ‘no motor weakness’ throughout the labour.9 14 Two additional studies described no or minimal weakness at 2-h intervals over time.10 15 We used the 2-h incidence of ‘no weakness’ in these studies for our analysis and found that this side-effect was significantly reduced in the PCEA group (Table 2). One study assessed motor power by asking the patient to sit up and then walk. However, the time at which this outcome was measured is not clear. The authors noted that all their patients, in both groups, could walk at some time during labour.16

Tan and colleagues8 reported that patients receiving PCEA were more likely to have ‘no motor weakness’ compared with those having CEI, based on repeated assessments of each patient (P<0.01). The incidence of complete immobility of the lower extremities was not different between groups (4.5 vs 2.9%; ns). Two other studies found no difference between groups in ‘dense motor block’ as defined by Bromage scores of 2 or 3.12 13

Analgesia
Labour analgesia was measured at intervals using a visual or verbal analogue score (VAS) in seven studies,915 and the number of distressing pain episodes in one.7 These results were either displayed in graphic form or described qualitatively. The VAS scores were not different between groups in any of the studies. The numbers of distressing pain episodes were also similar. Finally, pain scores, when classified as mild, moderate or severe, were the same in both groups.10

One study specifically reported VAS scores for pain in the second stage of labour and found no difference between groups.10 A second study reported that complete or good analgesia was also not significantly different between groups.7 However, in this study 16.9% of women suffered ‘unacceptable’ pain in the PCEA group compared with 2.8% in the CEI group (P=0.01).

Patient satisfaction
Overall patient satisfaction with the method of analgesia, using a VAS score, was reported in five studies.911 14 15 The data from four of these could be combined (Table 2). One study reported maternal satisfaction using a VAS scale at various times during labour and reported the data in graphical form, but did not report an overall score.14 One study reported ‘patient satisfaction with analgesia’.7 These studies did not show a difference in maternal satisfaction between groups.

Obstetric and neonatal outcomes
The total length of labour (in minutes) was reported in three studies911 and there was no difference between the groups. Similarly there was no difference in the duration of first stage of labour12 13 or second one.7 12 13 15 None of the studies reported a difference in the incidence of Caesarean section or instrumental delivery between groups. The incidence of low Apgar scores at 1 and 5 min after birth was low and similar in each group (Table 2).

Other outcomes
There was no difference between groups in the incidence of hypotension or high block (as described by any of the authors). The incidence of pruritus, shivering and nausea were similar in each group (Table 2).


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
This systematic review shows that PCEA in labour consistently results in fewer clinician interventions when compared with CEI. This result was achieved using a lower dose of local anaesthetic but without any measurable change in pain relief or patient satisfaction. In addition, patients in most studies had better mobility in the PCEA group.

The need for unscheduled clinician interventions is an important outcome for a number of reasons, apart from simple convenience. Additional drug doses require opening the ‘closed system’ and may increase the risk of infection. Furthermore, it is usual for the clinician to use an increased drug dose compared with that available to the patient. If this dose is misplaced (either in the intravascular or intrathecal space), toxicity might result. Finally, because of the increased drug dose, motor block of the lower extremities may be more evident.

The use of less local anaesthetic is also an important advantage of PCEA. While this review did not show a difference in operative delivery between patients who received PCEA and those who received CEI, there is some evidence that drug dose may affect labour outcome. Recently, the COMET group17 in the United Kingdom have shown that there is an increased incidence of operative vaginal delivery when conventional doses of local anaesthetic (0.25% bupivacaine) are used in labour compared with low dose techniques. Furthermore, cumulative doses of opioids, which are often combined with local anaesthetics in the infusions, may result in mild neonatal depression.17 While none of the studies we have reviewed reported results from ‘high’ doses of local anaesthetics, it may still be an advantage to use less of the drugs.

The differences in patient mobility were also consistent between groups when this outcome was measured. This consistency was present despite the numerous ways in which motor weakness of the lower extremities was reported and was likely to have been a result of the observed increase in local anaesthetic dosage in the CEI group. Of note, in one study that used 0.1% ropivacaine and fentanyl 2 µg ml–1, all patients could walk. It is possible that by reducing the concentration of local anaesthetic sufficiently, the difference in motor block between PCEA and CEI may become insignificant.

Of interest, although many of the studies reported measures of maternal satisfaction, there was no consistent difference between PCEA and CEI. This may be due to a number of reasons. First, satisfaction with both forms of analgesia appeared to be extremely high. Therefore, even by combining the results of all studies, the sample size may be insufficient to demonstrate a difference. This is particularly true because of the numerous different methods used to measure satisfaction and the different times (during labour or the next day) at which the question was asked. Second, it is very difficult to accurately measure maternal satisfaction. Often satisfaction with a particular technique is confused with satisfaction with pain relief. However, satisfaction is multidimensional and other factors such as patient expectations, communication skills of the health care team and the obstetric outcome are far more important than analgesia per se.18 One study attempted to measure multiple aspects of satisfaction but still did not find any difference between groups.10

In this review, we found no difference between groups in the incidence of Caesarean section or instrumental vaginal delivery. Furthermore, there was no difference in the duration of either the first or second stages of labour. However, the studies were small and the total number of patients enrolled may have been insufficient. In addition, the combination of nulliparous and multiparous parturients in some of the studies would reduce the power to detect these outcomes. A single, large study with standardized obstetric management would be more likely to find a difference if it exists.

Both CEI and PCEA appear to be safe for the mother and neonate. The incidence of maternal side-effects and neonatal depression was extremely low in the studies that measured these outcomes. Maternal hypotension occurred in less than 3% of cases and a low 5 min Apgar score occurred in under 2% (Table 2). An extremely large series of patients would be required to determine if the two groups were different with respect to these outcomes.

In summary, PCEA for labour pain has several advantages when compared with CEI. These include a reduction in the number of unscheduled clinician top-ups, a reduction in the amount of drug used and a reduction in the amount of motor block of the lower extremities. Both methods are safe for the mother and newborn. More information is needed to determine which method of analgesia maintenance leads to a reduced need for obstetric intervention and better patient satisfaction.


    Acknowledgement
 
We would like to thank Dr Barbara Leighton, Professor of Clinical Anesthesiology, Weill Medical College of Cornell University, New York, USA for her editorial assistance.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Glosten B. Epidural and spinal analgesia/anesthesia. In: Chestnut DH, ed. Obstetric Anesthesia: Principles and practice, 2nd edition. St Louis: Mosby, 1999; 360–86

2 Taylor HJ. Clinical experience with continuous epidural infusion of bupivacaine at 6 ml per hour in obstetrics. Can Anaesth Soc J 1983; 30: 277–85

3 Gambling DR, Yu P, Cole C, McMorland GH, Palmer L. A comparative study of patient controlled epidural analgesia (PCEA) and continuous infusion epidural analgesia (CIEA) during labour. Can J Anaesth 1988; 35: 249–54

4 Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996; 17: 1–12

5 Turhanoglu S, Ozyilmaz MA, Olmez G, Tok D, Bayhan N. Comparison of continuous epidural infusion and patient-controlled epidural analgesia in labour. Agri 1999; 11: 23–8

6 Uckunkaya N, Bilgin H, Sahin S, Ozarac O. Comparison of continuous infusion and patient-controlled epidural analgesia during labor. Turk Anesteziyoloji Ve Reanimasyon 1996; 24: 19–23

7 Purdie J, Reid J, Thorburn J, Asbury AJ. Continuous extradural analgesia: Comparison of midwife top-ups, continuous infusions and patient controlled administration. Br J Anaesth 1992; 68: 580–4

8 Tan S, Reid J, Thorburn J. Extradural analgesia in labour: complications of three techniques of administration. Br J Anaesth 1994; 73: 619–23

9 Boutros A, Blary S, Bronchard R, Bonnet F. Comparison of intermittent epidural bolus, continuous epidural infusion and patient controlled-epidural analgesia during labor. Int J Obstet Anesth 1999; 8: 236–41

10 Collis RE, Plaat FS, Morgan BM. Comparison of midwife top-ups, continuous infusion and patient-controlled epidural analgesia for maintaining mobility after a low-dose combined spinal-epidural. Br J Anaesth 1999; 82: 233–6

11 Curry PD, Pacsoo C, Heap DG. Patient-controlled epidural analgesia in obstetric anaesthetic practice. Pain 1994; 57: 125–7

12 Ferrante FM, Lu L, Jamison SB, Datta S. Patient-controlled epidural analgesia: demand dosing. Anesth Analg 1991; 73: 547–552

13 Ferrante FM, Rosinia FA, Gordon C, Datta S. The role of continuous background infusions in patient-controlled epidural analgesia for labor and delivery. Anesth Analg 1994; 79: 80–4

14 Gambling DR, Huber CJ, Berkowitz J, et al. Patient-controlled epidural analgesia in labour: varying bolus dose and lockout interval. Can J Anaesth 1993; 40: 211–7

15 Sia AT, Chong JL. Epidural 0.2% ropivacaine for labour analgesia: parturient-controlled or continuous infusion? Anaesth Intensive Care 1999; 27: 154–8

16 Smedvig JP, Soreide E, Gjessing L. Ropivacaine 1 mg/ml, plus fentanyl 2 microg/ml for epidural analgesia during labour. Is mode of administration important? Acta Anasthesiol Scand 2001; 45: 595–9

17 Comparative Obstetric Mobile Epidural Trial (COMET) Study Group UK. Effect of low dose mobile versus traditional epidural techniques on mode of delivery: a randomized controlled trial. Lancet 2001; 358: 19–23

18 Drew N, Salmon P, Webb L. Mothers’, midwives’ and obstetricians’ views on the features of obstetric care which influence satisfaction with childbirth. Br J Obstet Gynaecol 1989; 96: 1084–8