10 The long-term effects of analgesia in labour

10 The long-term effects of analgesia in labour

10 The long.term effects of analgesia in labour JACQUELINED U R B R I D G E MB, BS, FRCA Specialist Registrarin Anaesthesia ANITA HOLDCROFT Reader ...

1MB Sizes 0 Downloads 1 Views

10

The long.term effects of analgesia in labour JACQUELINED U R B R I D G E

MB, BS, FRCA

Specialist Registrarin Anaesthesia ANITA HOLDCROFT Reader in Anaesthesia

MD, FRCA

Department of Anaesthesia and Intensive Care, Imperial College of Science, Technology and Medicine, Hammersmith Hospital London W12 OHS, UK

Women frequently use a mixture of analgesics to gain relief from the distress of childbirth and antenatally require information on their effectiveness and side-effects. One such example would be the reported long-term neonatal behavioural changes following systemic opioids such as pethidine. The most frequently reported maternal effects of epidural or spinal analgesia are prolonged symptoms of headache, backache and neurological sequelae. Large retrospective studies of postpartum symptomatology have focused on correlations with regional nerve blockade rather than on other more commonly used analgesics. Post-dural puncture headache is a recognized long-term complication of epidural nerve blockade. However, prospective studies have not confirmed any causal relationship between epidural analgesia and backache and neurological complications are five times more common after childbirth itself than after regional nerve blockade. Postpartum symptomatology describes significant morbidity in the community but its relationship to analgesia in labour is still to be proved. Key words: analgesia; epidural analgesia; spinal analgesia; headache; backache; neurological problems; complications. Pharmacological pain relief in labour is frequently used and can be given by the inhalation, intramuscular, subcutaneous, intravenous, epidural and intrathecal routes. These methods are introduced at the time of painful contractions, but preparation for the discomfort of labour often begins earlier, with relaxation and breathing exercises. This is also a time when information about the usefulness and side-effects of the various analgesic methods can be discussed. In times past, the lack o f effective pain relief in labour m a y have precipitated operative delivery because of maternal distress. The long-term consequences of such stress can now be avoided. Maternal choice of pharmacological methods for analgesia in labour [email protected]'s Clinical Obstetrics and Gynaecology--

Vol. 12, No. 3, September1998 ISBN 0-7020-2470-8 0950--3552/98/030485 + 14 $12.00/00

485 Copyright © 1998, by Bailli&eTindall All rights of reproduction in any form reserved

486

J. DURBRIDGE AND A. HOLDCROFT

includes not only preferences for the route of drug delivery but also their efficacy and side-effects, both acute and long term for herself and her baby. In presenting such information to a mother, as summarized in Table 1, a physician would need to take into account additional clinical risk factors presented by the woman, for example previous back surgery or diabetes. Table 1. Risks of long-term effects after regional analgesia in obstetrics compared with expected risks in population. Expected in obstetric population postpartum

Identified by clinicians from full medical history

Headache Backache

22% (Grove, 1973) 40% (Grove, 1973)

Neurological

1 in 2530 (Holdcroft et al, 1995b)

1% (as post-dural puncture headache) 1 in 2000 (Crawford, 1985) 1 in I000 (Russell et al, 1993) 1 in I3 000 (Holdcroft et al, 1995b)

Closed claims (Chadwick, 1996) 21% 12% 13%

The long-term effects of analgesia in labour may be considered in a time scale which relates to either maternal perceptions or clinical relevance. Any headache in the mother after delivery which persists for even 12 hours could be considered too long because she needs to be mobile and to care for her baby. Clinically, if the headache is caused by a dural puncture at the time of epidural insertion, a period of about 24 hours will pass before the definitive treatment of a blood patch is given and this would not be recorded as a long-term complication. Neurological sensory losses of 48 hours duration are common and are considered transient because they are of limited duration. The most common long-term effects which are associated with epidural or spinal analgesia are headache, backache and sensory neuropathy. It is headache which causes the majority of litigation following obstetric anaesthesia in the United States (Chadwick, 1996), yet women antenatally are usually more concerned about long-term problems of paralysis and backache. Injury claims for nerve damage are higher in the non-obstetric anaesthetic population while those for headache and backache predominate in the obstetric population. In the closed claims analysis both maternal and fetal injuries feature. With regard to long-term maternal effects a period of more than 6 weeks duration is considered clinically significant. Neuraxial opioids have contributed to the reduction in local anaesthetic drug usage and significant long-term effects have not been described at routine drug dosages. Systemic opioids, although frequently used, have a poor record as analgesic drugs in labour. Pethidine has been most widely studied in this respect (Holdcroft and Morgan, 1974; Olofsson et al, 1996). If a drug lacks efficacy but has significant side-effects then a mother should be warned about these. In particular pethidine is sedative and delays stomach emptying. It can therefore contribute significantly to aspiration of stomach contents into the lungs with serious long-term sequelae. In addition, systemic opioid analgesics are considered to have long-term effects on the neonate especially when pethidine is used because delayed

LONG-TERM EFFECTS OF ANALGESIA

487

metabolism and excretion can have behavioural manifestations. The halflife of pethidine in the newborn is 18 hours compared with 3-4 hours in an adult (Caldwell and Notarianni, 1978) and maternal metabolites of pethidine, particularly norpethidine, still cross the placenta once plasma concentrations of pethidine have decreased (Belfrage et al, 1981). The depressant effect on the fetus appears to be maximal when the dosedelivery interval is between 2 and 3 hours (Shnider and Moya, 1964) and is least when delivery is within 1 hour (Belfrage et al, 1981) or when more than 6 hours have passed since the administration of the drug. The first extensive study of maternal complications after lumbar epidural analgesia for labour was conducted by Crawford (1985) who reviewed over 27 000 nerve blocks from one obstetric unit. He noted a spinal headache persisting for 5 months, localized backache with an incidence of 1 in 2000, sensory symptoms of moderate severity in three women and a 'number of cases in which the epidural has erroneously been identified as the cause of fleeting or permanent ill-health'. These women presented symptoms of numbness and/or pain but without any organic lesion. It was from the same unit that an investigation into long-term symptomatology after delivery was conducted in 1987 by MacArthur et al (1991) using a questionnaire distributed to 30 096 women and listing 25 symptoms. The mothers selected for the study were those delivering at Birmingham Maternity HospitM between the years 1978 and 1985 and only 11 701 (33%) replied. Recalled symptoms were included in the study results if they were new, persisted for more than 6 weeks and commenced up to 3 months after delivery and 46% of responders fulfilled these criteria. This starting time for symptoms is rather long after birth and would suggest that other contributory factors rather than just the delivery may have been operating. One of the most common symptoms was urinary stress incontinence in 15.2% of women and backache and tiredness each in more than 10% of women. The data from the questionnaire were then collated with both anaesthetic and obstetric data and analysed for predictors of symptomatology. Epidural analgesia was the main predictor of backache and headache. The authors acknowledged that they had not proved causality and that they did not measure the severity of the symptoms and stressed the need for further investigative studies. Researchers in another obstetric unit invited women who reported by questionnaire new-onset backache lasting for more than 6 months after delivery for consultation. A total of 156 women (of whom 109 had received epidural analgesia) were identified but on direct questioning 17% said that they no longer suffered from backache and 40% failed to attend a follow-up clinic (Russell et al, 1993). Again this was a retrospective survey and although it identified weaknesses in assessing symptom severity and possible recall bias for back pain after epidural analgesia (Reynolds and Russell, 1997) the significance of the maternal symptomatology to general health and well-being has not been explained. These studies present a considerable burden of long-term symptomatology in the community in women after epidural analgesia for labour, or even delivery itself, which appears to be difficult to define. It has also been observed in a regional study of neurological complications following

488

J. DURBRIDGE AND A. HOLDCROFT

delivery where medical practitioners rather than women themselves reported long-term problems. Out of 35 notifications received, 15 were symptomatic without a defined neurological lesion (Holdcroft et al, 1995b).

HEADACHE Headache as a symptom is common in the immediate postpartum period and can occur in up to one-quarter of parturients irrespective of the type of analgesia given (Grove, 1973). The small contribution of a dural puncture to the overall incidence of postpartum headache should be discussed with a woman prior to consent to an epidural or spinal injection because further interventions may be required. Where headache follows epidural analgesia the suspicion of a dural puncture must be excluded in the absence of an overt dural tap at the time of insertion of the epidural needle or catheter. If a dural puncture is recognized at the time of an epidural nerve block for pain relief in labour, the anaesthetist will usually resite the epidural and explain to the mother what has happened and all subsequent epidural injections will be given by an anaesthetist. This is to allow the recognition of intrathecal spread of the solution which could theoretically occur. Other management during labour need not change. A dural puncture has been considered to be an indication for an elective forceps delivery in order to prevent post-dural puncture headache (Okell and Sprigge, 1987) but with a lack of controlled studies and a difference in management for spontaneous rather than forceps delivery there is no evidence to support the practice of elective forceps delivery after an accidental dural tap. In a postal survey of all maternity units in the United Kingdom, half allowed a woman to push in the second stage of labour (Sajjad and Ryan, 1995). The decision to deliver by forceps may deprive a mother of the long-term satisfaction of spontaneous delivery. Initially it was thought that straining would increase the leakage of cerebrospinal fluid, but it also increases the pressure in the epidural space so that the pressure gradient across the dural membrane is unlikely to change. Where deliberate spinal analgesia is used, for example, the so-called 'mobile epidural' of the type which incorporates both a singleshot spinal and an epidural catheter, bearing down in the second stage of labour has not been associated with an increased incidence of post-dural puncture headaches. The low incidence of post-dural puncture headaches of 0.13% in 1565 women with a combined spinal-epidural technique using a 27 gauge atraumatic needle (Morgan and Kadim, t 994) was only achieved by experience. The choice of a small needle diameter, for example 27G, and use of a non-cutting, pencil-point tip will reduce long-term effects of spinal analgesia, which may be the fastest method to relieve very severe pain during labour, but even using larger 22G needles the incidence of spinal headaches did not increase when women were allowed to push at delivery (Ravindran et al, 1981). Common causes of headache after delivery include caffeine withdrawal, migraine, dehydration and sleep deprivation. These can be excluded on routine history taking. A typical woman after a dural tap will complain of

L O N G - T E R M EFFECTS OF A N A L G E S I A

489

bilateral continuous occipital or frontal pain which is often severe on sitting, coughing and sudden movements of the head but may disappear and certainly is relieved to a considerable extent when lying supine. Typically the headache presents on mobilizing but it can present a few days later. The postural headache is caused by traction on the pain-sensitive supporting structures such as the venous sinuses, cerebral vessels, the tentorium and falx cerebri by a reduction in cerebrospinal fluid pressure from a continous leak of cerebrospinal fluid from the hole in the dura. Painful sensations arising above the tentorium cerebri are transmitted via the trigeminal nerve to the frontal region and those arising below the tentorium are transmitted via the vagus and upper cervical nerves to the neck and occiput. Occasionally nausea and auditory symptoms can present, particularly changes in hearing acuity and tinnitus. The woman is miserable, tearful, depressed and bedridden. There should be no signs of meningitis, fever, hypertension or lateralization. A brief neurological examination should exclude spinal or intracranial pathology. In addition localized causes of pain should be sought such as spasm in the cervical muscles or sinusitis. Meningitis must be excluded if the dura~hasbeen punctured because the meningeal barrier has been breached. If the diagnosis of the headache is in doubt a neurological opinion should be sought. Recommended management after a dural tap is to treat the mother conservatively for no more than 24 hours with bedrest and oral fluids to satisfy thirst. Forced hydration has not been proved beneficial. Oral analgesics are necessary and morphine may be required if the pain is severe. If headache presents an anaesthetist must be informed so that if it persists for more than 24 hours a blood patch can be recommended. At this time it is most effective with a more than 95% immediate success rate (Loeser et al, 1978). It is not without risk or side-effects and the patient must be told that transient backache is common. Backache has been reported to occur in up to 35% of patients with an average duration of 27 days (Abouleish et al, 1975) and causes for this have been sought. After magnetic resonance imaging studies demonstrated extensive subcutaneous haematoma following blood patch (Beards et al, 1993) it has been recommended that flushing the blood from the shaft of the epidural needle before withdrawing it may help to prevent this long-term complication. The headache may reappear in about 10% of women (Loeser et al, 1978) and another blood patch may be necessary. Infection must be prevented by careful asepsis and blood cultures taken during the procedure with the patient being afebrile and having a normal white blood cell count. When a woman is septic or is human immunodeficiency virus (HIV) positive dextran 40 can be used as an alternative (Reyvoet et al, 1997). The need for prolonged hospital stay or follow-up at home, such as occurs after dural puncture, may be considered by the mother to be a restrictive long-term event. The rate of inadvertent dural puncture is 1.3% of obstetric epidurals (Stride and Cooper, 1993) but not all these women develop headache. Follow-up of these women is essential, if not by an anaesthetist, then by a physician who can recognize and treat any prolonged effects. The survey by MacArthur et al (1993) specifically identified that 23% of those women who had an accidental dural

490

J. DURBRIDGE AND A. HOLDCROFT

tap (74 out of 4700 epidurals) had a new headache (including 'migraine' but with no specific diagnostic criteria) or neckache starting within 3 months after delivery and lasting more than 6 weeks. There may be no causal relation between these symptoms and dural puncture, for example there was no specific history taken about the character of the headache and report bias may be a consideration. However, there are continuing general case reports of chronic subdural haematomas following a continued leakage of cerebrospinal fluid and reduction in cerebrospinal fluid pressure after dural tap in obstetric patients (Scott and Hibbard, 1990; Whitely et al, 1993). Each woman therefore needs to be assessed if symptomatology persists. The long-term follow-up of women who have received a blood patch also emphasizes a decreased success rate with time (Cooper and Stride, 1993). Non-anaesthetic medical staff should be aware of the importance of headache after inadvertent or suspected dural puncture and clearly should refer the woman back to a senior obstetric anaesthetist if headache recurs, does not improve or is atypical (see Table 2). Early discharge from hospital after vaginal delivery means that post-dural puncture headache may not present when a parturient is in hospital. Clear instructions should be given to a woman or her midwife that if she has had spinal or epidurat analgesia in labour and develops a postural headache she should be referred back quickly to a senior anaesthetist at the hospital where she delivered. Table 2. Differential diagnosis of headache after dural puncture. • • • • •

Low cerebrospinal fluid pressure Meningitis Subdural haematoma Migraine Caffeine withdrawal

BLADDER DISTURBANCE When local anaesthetics were administered epidurally in labour without opioids, a denser motor nerve block often occurred and bladder distension causing urinary retention was a recognized complication. More recent studies have confirmed that there is no increase in residual volume in women who have delivered following epidural analgesia in labour (Weissman et al, 1995; Chien et al, 1996). There has been no statistical evidence of a direct association of stress incontinence or frequency of micturition with analgesics used in labour (MacArthur et al, 1992). BACKACHE Low back pain is a common problem in women of childbearing age. Up to one-quarter of women will complain of back pain prior to ever becoming pregnant. A cross-sectional magnetic resonance imaging study has shown that one-third of symptomless women of childbearing age had lumbar disc

LONG-TERM EFFECTS OF ANALGESIA

491

degeneration and the prevalence increased with age (Powell et al, 1986). It is not surprising then that about one-half of women suffer from lower back pain during pregnancy (Mantle et al, 1977). In a study of 817 women (of whom 89% were responders) two-thirds of women immediately postdelivery complained of back pain and up to one-third had long-term back pain 3 months postnatally but this usually resolved (Ostgaard and Andersson, 1992). Factors at 1 year that correlated with back pain were the presence of back pain before pregnancy, the presence of back pain during pregnancy, heavy physical work and multiparity. Back pain and pregnancy has recently been reviewed (MacEvilly and Buggy, 1996). A report of a causal relationship between backache and epidural analgesia was made after a postal survey by MacArthur et al (1990) of women delivering between 1978 and 1985. The hypothesis was presented that the backache resulted from postural stresses during labour in combination with the muscle relaxation and lack of mobility induced by epidural analgesia, particularly because there was no excess backache after elective Caesarean section. However, no clinical assessment of the backache was made; the use of neither non-steroidal anti-inflammatory analgesics (which are commonly used to relieve back pain itself) nor other pain-relieving drugs after Caesarean section was examined and the survey covered a period of time when high concentrations of bupivacaine (0.25--0.5%) were in use for epidural analgesia. It is accepted that profound sensory and motor block could allow a woman to adopt a less than optimal position. An exaggerated lithotomy position or spinal and hip flexion can stretch sensitive structures during either general or regional anaesthesia. This study has made anaesthetists, obstetricians and midwives clearly aware of the importance of careful positioning. It has also generated prospective studies in relation to backache and epidural analgesia with less concentrated solutions of local anaesthetic drugs. The technical development of mixtures of drugs for epidural analgesia was made possible by basic science research in the late 1970s and early 1980s which demonstrated that when neuraxial opioid drugs were combined with a local anaesthetic drug the dosage of local anaesthetic could be reduced and at the same time adequate analgesia provided (Jungstrom et al, 1984). The results of prospective studies have not confirmed that epidural analgesia can increase backache. A prospective cohort study by MacArthur et al (1995) in 329 women using 0.25% bupivacaine in the epidural group recorded no long-term increased incidence in back pain over 6 weeks postpartum except for the first day when local trauma associated with needle insertion would be expected to be maximal. The study by Russell et al (1996) of 599 women randomized those requesting epidural analgesia into high- (0.125%) and low- (0.0625% with an opioid drug) dose bupivacaine infusion groups and also included women recruited at random who had laboured without epidural analgesia. The response rate to a postal questionnaire at 3 months after delivery was 75%. The only predictor of postpartum backache was a history of backache in the antenatal period. There was no correlation with motor block although the high-dose bupivacaine group produced significantly more leg weakness. A larger prospective study over

492

J. DURBRIDGE AND A. HOLDCROFT

a 6 month period in 1185 women who delivered a viable singleton infant surveyed back pain up to 2 months postpartum and follow-up was achieved in 88% (Breen et al, 1994). The standard epidural technique was a low-dose bupivacaine infusion of 0.04-0.125% combined with fentanyl and adrenaline. About half the women had epidural analgesia and 44% of all women reported back pain (44% in women who received epidural analgesia and 45% in women who did not). Backache was associated with a past history of back pain, younger age and greater weight but definitely not with epidural analgesia. Non-specific low back pain when it occurs in pregnancy tends to be more severe and disabling than when it occurs in the non-pregnant woman. There are various physiological changes which may predispose to it. Mechanically, there is an increasing lumbar lordosis, a loss of abdominal muscle support and a rise in the body centre of gravity, all of which may increase the likelihood of back pain. Relaxin is a hormone regulator of collagen which is secreted in increased quantities during pregnancy and contributes to a softening of ligaments. Sacroiliac dysfunction may cause quite severe pelvic pain and shows an association with increased levels of relaxin and increasing gestational age. Stimuli from viscera in the reproductive tract may also exacerbate back pain which is of a referred origin. Pregnancy may also exacerbate a pre-existing condition, for instance joint laxity may predispose to spondylolisthesis. Increased mechanical stress and work in the postnatal period and repetitive lifting in forward, bent and twisted positions may predispose to postpartum backache. It is important that all women with back pain associated with pregnancy are assessed to exclude serious spinal pathology. The risk of intervertebral disc prolapse, although rare (1 in 10 000 deliveries), is still present (Forster et al, 1996). All women with tenderness or backache should be examined and symptoms and signs of sensory or motor loss in the lower limbs or sphincter disturbance should be sought as summarized in Table 3 (Royal College of General Practitioners et al, 1996). The majority of women will have back pain in the lumbosacral region, buttocks and thighs which is mechanically induced. This pain can be classified into two types: lumbar pain and posterior pelvic pain which is associated with ligament laxity in the pelvis during pregnancy and in a minority of women also causes anterior pelvic discomfort (Ostgaard, 1996). Some may have nerve root pain associated with unilateral leg pain worse than back pain which radiates to the foot or toes. There may be numbness or paraesthesia in the same distribution and straight-leg raising cari reproduce the leg pain. These women need specialist referral and an even smaller number may present with cauda equina syndrome with gait and sphincter disturbances. These women require immediate neurosurgical referral. The management of parturient women with back pain should include the provision of information and general advice on back care. A history of back pain should be noted and those with sensory or motor nerve root symptoms or signs or reflex changes and those with previous lower back surgery should be referred to a consultant anaesthetist prior to labour and delivery. If an epidural or nerve block is provided for labour analgesia, the lowest

LONG-TERM EFFECTS OF ANALGESIA

493

Table 3. Diagnostic triage for non-specific low back pain, nerve root pain and possible serious spinal pathology after delivery. 1. Simple backache: • lumbosacral, buttocks, thighs • mechanical • woman well 2. Nerve root pain: • unilateral leg pain worse than low back pain • radiation to foot or toes • sensory changes in the same distribution • straight leg raising reproduces leg pain • localized neurological signs 3. Serious pathology: • non-mechanical • unwell • widespread neurology • structural deformity • past medical history, e.g. HIV, steroids • cauda equina syndrome (requires immediate referral) • sphincter disturbance • gait disturbance • saddle anaesthesia

dose of local anaesthetic should be used to provide satisfactory pain relief with minimal motor block. However, occasionally significant loss of motor power still occurs. Although no cause and effect relation has been established between the degree of motor block and back pain, poor maternal positioning cannot be excluded from causative factors. Midwives and other staff involved should be trained to recognize and manage this correctly. Education is the key to effective prevention and treatment of most women with postpartum back pain (Russell and Reynolds, 1997). With up to twothirds of women complaining of back pain immediately post-delivery, an advice sheet on exercise and correct posture, particularly during lifting and sitting feeding the infant, should be available and physiotherapy readily offered postpartum by ward visits or a walk-in or telephone service without medical consultation. Localized trauma at the site of an epidural should be distinguished from muscular or sacroiliac strain, ff back pain persists, normal activity should be maintained while pain relief is provided with oral analgesics, particularly non-steroidal anti-inflammatory analgesics if not contraindicated. Specific physiotherapy treatments and an orthopaedic opinion may be necessary. The majority of back pain which develops postnatally is not severe. There is a very small minority of women who have prolonged disabling pain (Ostgaard, 1996). There is no study of these women and their mode of analgesia during labour, yet most obstetric physiotherapists can give anecdotal case reports. It is these women who require early identification and long-term physiotherapy or pain clinic management. Women can be reassured that the most likely cause of postpartum backache is a continuation of antenatal problems and that no prospective study of the use of regional analgesia in labour has shown an association with an increased risk of Chronic backache. However, occasional localization of

494

J. DURBRIDGE AND A. HOLDCROFT

pain to the site of insertion of the epidural needle does occur and has been documented in the large review of 27 000 lumbar epidurals for labour by Crawford (1985) as having an incidence of 1 in 2000 and in the study assessing long-term back pain by Russell et al (1993) one woman out of 612 who received epidural analgesia had localized tenderness over the site of epidural insertion and a diagnosis of resolving soft-tissue haematoma was made. Focal tissue damage after epidural analgesia has not been observed by magnetic resonance imaging within hours of catheter removal (Holdcroft et al, 1994, 1995a). Further studies of backache are required to examine the effects of randomization and use of standardized anaesthetic, obstetric and diagnostic criteria.

NEUROLOGICAL PROBLEMS Neurological sequelae after childbirth are five times more likely to be a result of the obstetric event than epidural analgesia (Holdcroft et al, 1995b). Central nervous system events can follow hypertensive episodes and peripheral nerve compression from obstetric events or from malpositioned intramuscular injections (e.g. pethidine analgesia in labour) can result in sciatic nerve palsy whether or not epidural analgesia was received (Silva et al, 1996). Local nerve compression in the epidural space can follow an epidural haematoma or abscess. An epidurat or spinal subdural haematoma is rare but case reports are continually published (Pryle et al, 1996) and a collection of fluid in the small subdural space may induce permanent nerve damage (McMenemin et al, 1992). Bleeding into the epidural space regularly occurs with a needle or catheter and, where there is a perceived increased risk such as in pre-eclampsia, a clotting screen and platelet count are checked. The incidence of bleeding on catheter removal after delivery is not known, but a similar check for factors which may increase bleeding should be made. A blood patch which is used to treat a dural puncture involves the insertion of up to 25 ml of blood into the epidural space and its effects dissipate within hours (Beards et al, 1993). The hazards of bleeding are increased by any form of anti-coagulation, for example in thromboembolic prophylaxis. It is a risk which has to be discussed with women who are receiving anti-platelet drugs such as aspirin or low molecular weight heparins such as enoxaparin which have a prolonged duration of action. Although the risk is small, the treatment of a haematoma which presents with nerve compression is a laminectomy to relieve the pressure in the epidural space. Therefore any woman who demands or in whom general anaesthesia is contraindicated and is given spinal or epidural analgesia in spite being informed of this risk needs hourly neurological follow-up postnatally in order for nerve compression to be detected. An epidural abscess takes longer than an epidural haematoma to present and, when women leave hospital early, it may not be detected quickly. Back pain is localized. The risk factors include those of a haematoma and also systemic infection and immunosuppression such as steroid therapy. Skin

LONG-TERM EFFECTS OF ANALGESIA

495

preparation for epidural nerve blockade should be optimal in such patients because staphylococci may persist in hair follicles (Sata et al, 1996). Spinal or epidural nerve blockade should only be performed in an infected patient if the woman has a positive response to appropriate antibiotics because haematogenous spread could cause meningitis. Skin contamination during a combined spinal-epidural nerve block for analgesia in labour has been reported (Harding et al, 1994). Meningitis is the more generalized form of central nervous system inflammation but, unlike an abscess which compresses nerves, long-term complications from appropriately treated postpartum meningitis are not a problem. Chemical irritation in the cerebrospinal fluid may result from systemic drug administration, such as the non-steroidal anti-inflammatory analgesics (Burke and Wildsmith, 1997) or from direct contamination during regional nerve blockade. Spinal analgesia allows the introduction of drugs and other irritant agents into the intrathecal space. Intrathecal nerve roots are more vulnerable to damage because they lack the protective dural coat which covers nerves as they traverse from the intrathecal to the epidural space. Drug dosage and concentration matter, particularly hyperbaric 5% lignocaine and transient radicular irritation has been reported with its spinal use (Salmela et al, 1996). More long-term disability with the development of the cauda equina syndrome has followed the use of intrathecal catheters and hyperbaric 5% lignocaine (Rigler et al, 1991). These techniques and drugs are not used in routine obstetric analgesia. Peripheral single-nerve damage is the commonest non-fatal complication following regional nerve blockade for analgesia in labour (Scott and Hibbard, 1990). Its incidence was 38% of all transient and long-term adverse events examined retrospectively from 1982 to 1986. In a second prospective survey during the years 1991-1992 the incidence was 36% (Scott and Tunstall, 1995). It can occur as a result of trauma by the needle or catheter and is most likely to affect a sensory nerve root because of their posterior distribution. It is well known that in diseases such as diabetes mellitus nerves may be more vulnerable to damage because of alterations in microvascular blood supply (Chambers, 1992). Trauma may damage blood supply or induce oedema and subsequent inflammatory reactions can lead to degenerative changes in the nerve with dysfunction. Neurological complications respond poorly, if at all, to therapies, so at all costs they must be prevented. Epidural and spinal needles are now used which divide rather than cut tissues and softer epidural catheters are being manufactured which induce less paraesthesia on insertion. Within the general obstetric population, however, prolonged nerve damage has been reported to have an incidence of 1 in 2600 deliveries, mainly related to the lumbosacral plexus (Hill, 1961). The most common neuropathy of pregnancy is foot drop, followed by femoral and obturator neuropathy. Diagnosis is based on the clinical signs summarized in Table 4. In our recent survey of an obstetric population of 48 066 women, t in 2530 experienced prolonged neurological complications as defined by a neurologist (Holdcroft et al, 1995b). Epidural anaesthesia was only considered contributory to such neurological

496

J. DURBRIDGE AND A. HOLDCROFT

Table 4. The recognition of (1) nerve root damage from sensory function and (2) mixed nerve damage from motor nerve function. Roots Sensory loss

L2 Upper thigh

Nerves (associated motor loss)

~

L3 Lower thigh

L4 (knee jerk) L5 Medial leg Dorsum foot

S 1 (ankle jerk) Behind lateral malleolus

Obturator (weak adduction of thigh) -~ Femoral (no knee jerk) -,11---Sciatic (peroneal nerve damage > tibial)---tl~ (foot drop)

complications in t in 13 007 patients. This survey also identified that the majority of complications were not identified by hospital staff but by practitioners in the community. This is important to recognize in study design and in the routine audit of obstetric morbidity. SUMMARY

Quality of care relating to analgesia in labour may be assessed not only by the efficacy of the drugs and methods used but also by their long term effects. Individual unit data are often not available because obstetric clinical audit often ends with delivery of the baby and there is no duty to record postpartum morbidity, particularly when women return home within hours after delivery. A pregnancy-based record should continue into the postpartum period in order to provide data for research into links between pregnancy care and maternal health (Drife, 1995). Such monitoring by both hospital staff and general practitioners can only be encouraged so that identified medical complications and the symptoms that women perceive as relating to delivery can be minimized. Prevention begins antenatally in the community and continues throughout labour with patient advice and coordinated primary and secondary professional care. Epidural analgesia is the most effective method of analgesia for labour pains. It is an invasive method and, before its use is considered by a mother for pain in the absence of any medical indications, information on potential side-effects such as headache from a dural puncture and backache as a rare cause of a commonly presenting symptom is essential. Other forms of analgesia have not been investigated in the diverse long-term symptomatology following childbirth. The role of pethidine in neonatal behavioural effects coupled with a lack of effective analgesia from pethidine and the enhancement of sedation and vomiting does warrant more information being given on the side-effects of other popular analgesic drugs. REFERENCES Abouleish E, de la Vega S, Blendingre I & Tit TO (1975) Long term follow up of epidural blood patch. Anesthesia and Analgesia 54: 459-463.

LONG-TERM EFFECTS OF ANALGESIA

497

Beards SC, Jackson A, Griffiths AG & Horsman EL (1993) Magnetic resonance imaging of extradural blood patches: appearances from 30 min to 18 h. British Journal of Anaesthesia 71: 182-188. Belfrage P, Boreas LO, Hartvig P, Irestedt L & Raabe N (1981) Neonatal depression after obstetrical analgesia with pethidine. Acta Obstetrica et Gynecologia Scandinavica 60: 43-49. *Breen TW, Ransil BJ, Groves PA & Oriol NE (1994) Factors associated with back pain after childbirth. Anesthesiology 81: 29-34. Burke D & Wildsmith JAW (1997) Meningitis after spinal analgesia. British Journal of Anaesthesia 78: 635-636. Caldwell J & Notarianni /.,2 (1978) Disposition of pethidine in childbirth. British Journal of Anaesthesia 50: 307-308. *Chadwick HS (1996) An analysis of obstetric anesthesia cases from the American Society of Anesthesiologists closed claims project database. International Journal of Obstetric Anesthesia 5: 258-263. Chambers WA (1992) Peripheral nerve damage and regional anaesthesia. British Journal of Anaesthesia 69: 429-430. Chien PFW, Khan KS, Augustsson P & Patel NB (1996) The determinants of residual bladder volume following spontaneous vaginal delivery. Journal of Obstetrics and Gynecology 16:146-150. Cooper GM & Stride PC (1993) Blood patch not always benign. British Medical Journal 306: 1339. Crawford JS (1985) Some maternal complications of epidural analgesia for labour. Anaesthesia 40: 1219-1225. *Drife JO (1995) Assessing the consequences of changing childbirth: better data are needed. British Medical Journal 310: 144. Forster MR, Nimmo GR & Brown AG (1996) Prolapsed intervertebral disc after epidural analgesia in labour. Anaesthesia 51: 773-775. Grove LH (1973) Backache, headache and bladder dysfunction after delivery. British Journal of Anaesthesia 45:1147-1149. Harding SA, Collis RE & Morgan BM (1994) Meningitis after combined spinal-extradural anaesthesia in obstetrics. British Journal of Anaesthesia 73: 545-547. Hill EC (1961) Maternal obstetric paralysis. American Journal of Obstetrics and Gynecology 83: 1452-1460. Holdcroft A & Morgan M (1974) An assessment of the analgesic effect in labour of pethidine and 50 per cent nitrous oxide in oxygen (entonox). Journal of Obstetrics and Gynaecology of the British Commonwealth 81: 603-607. Holdcroft A, Samsoon G, Fernando RA & Baudouin C (1994) Acute tissue damage following epidural cannulation: a comparison between the midline and paramedian approach in obstetric patients. International Journal of Obstetric Anesthesia 3: 35-38. *Holdcroft A, Baudouin C, Fernando R et al (1995a) Post partum magnetic resonance imaging: lumbar tissue changes are unrelated to epidural analgesia or mode of delivery. International Journal of Obstetric Anesthesia 4: 201-206. *Holdcroft A, Gibberd FB, Hargrove RL et al (1995b) Neurological complications associated with pregnancy. British Journal of Anaesthesia 75: 522-526. Jungstrom E, Eastwood D, Patel H et al (1984) Epidural fentanyl and bupivacaine in labour: doubleblind study. Anesthesiology 61" A414. Loeser EA, Hill GE, Bennett GM & Sederberg JH (1978) Time vs. success rate for epidural blood patch. Anesthesiology 49: 147-148. MacArthur C, Lewis M, Knox EG & Crawford JS (1990) Epidural anaesthesia and long term backache after childbirth. British Medical Journal 301: 9-12. MacArthur C, Lewis M & Knox EG (1991) Health after Childbirth. London: HMSO. MacArthur C, Lewis M & Knox EG (1992) An investigation of long-term problems after obstetric epidural anaesthesia. British Medical Journal 304: 1279-1282. MacArthur C, Lewis M & Know EG (1993) Accidental dural puncture in obstetric patients and long term symptoms. British Medical Journal 306: 883-885. *MacArthur A, MacArthur C & Weeks S (1995) Epidural anaesthesia and low back pain after delivery: a prospective cohort study. British Medical Journal 311: 1336-1339. *MacEvilly M & Buggy D (1996) Back pain and pregnancy: a review. Pain 64: 405-414. Mantle MJ, Greenwood RM & Currey HLF (1977) Backache in pregnancy. Rheumatology and Rehabilitation 16: 95-101. McMenemin IM, Sissons GRJ & Brownridge P (1992) Accidental subdural catheterization: radiological evidence of a possible mechanism for spinal cord damage. British Journal of Anaesthesia 69: 417-419.

498

J. DURBRIDGE AND A. HOLDCROFT

Morgan BM & Kadim MY (1994) Mobile regional analgesia in labour. British Journal of Obstetrics and Gynaecology 101: 839-841. Okell RW & Sprigge JS (1987) Unintentional dural puncture. A survey of recognition and management.Anaesthesia 42: 1110-1113. Olofsson Ch, Ekblom A, Ekman-Ordeberg G e t al (1996) Lack of analgesic effect of systemically administered morphine or pethidine on labour pain. British Journal of Obstetrics and Gynaecotogy 103: 968-972. Ostgaard HC (1996) Assessment and treatment of low back pain in working pregnant women. Seminars in Perinatology 20: 61-69. Ostgaard HC & Andersson GBJ (1992) Postpartum low-back pain. Spine 17: 53-55. Powell MC, Szypryt P, Wilson M e t al (1986) Prevalence of lumbar disc degeneration observed by magnetic resonance in symptomless women. Lancet 2: 1366-1367. Pryle BJ, Carter JA & Cadoux-Hudson T (1996) Delayed paraplegia following spinal anaesthesia. Spinal subdural hematoma following dural puncture with a 25 gauge pencil point needle at T12-L1 in a patient taking aspirin. Anaesthesia 51: 263-265. Ravindran RS, Viegas OJ, Tasch MD et al (1981) Bearing down at the time of delivery and the incidence of spinal headache in parturients. Anesthesia and Analgesia 60: 524-526. Reynolds F & Russell R (1997) Epidural anaesthesia does not cause long term backache. British Medical Journal 316: 69-70. Reyvoet MEJ, Cosaert PAJM, Desmet MFR & Plaaschaert SM (1997) Epidural dextran 40 patch for postdural puncture headache. Anaesthesia 52: 884-895. Rigler ML, Drasner K, Trejcie TC et al (1991) Cauda equina syndrome after continuous spinal anaesthesia. Anesthesia and Analgesia 72: 275-281. Royal College of General Practitioners, Chartered Society of Physiotherapists, Osteopathic Association of Great Britain, British Chiropractic Association and National Back Pain Association (1996) Clinical Guidelines for the Management of Acute Low Back Pain. London: Royal College of General Practitioners. *Russell R & Reynolds F (1997) Back pain, pregnancy, and childbirth: postpartum pain is most likely to be a continuation of antepartum pain. British Medical Journal 314: 1062-1063. Russell R, Groves P, Taub N e t al (1993) Assessing long-term backache after childbirth. British Medical Journal 306: 1299-1303. *Russell R, Dundas R & Reynolds F (1996) Long term backache after childbirth: a prospective search for causative factors. British Medical Journal 312: 1384-I 388. Sajjad T & Ryan TDR (1995) Current management of inadvertent dural taps occurring during the siting of epidurals for pain relief in labour: a survey of maternity units in the United Kingdom. Anaesthesia 50: 156-161. Salmela L, Aromaa U & Cozanitis DA (1996) Leg and back pain after spinal anaesthesia involving hyperbaric 5% lignocaine. Anaesthesia 51: 391-393. Sara S, Sakuragi T & Dan K (1996) Human skin flora as a potential source of epidural abscess. Anesthesiology 85: 1276-1282. Scott DB & Hibbard BM (1990) Serious non-fatal complications associated with extmdural block in obstetric practice. British Journal of Anaesthesia 64: 537-541. *Scott DB & Tunstall ME (1995) Serious complications associated with extradural/spinal blockade in obstetrics: a two-year prospective study. International Journal of Obstetric Anesthesia 4: 133-139. Shnider SM & Moya F (1964) Effects of meperidine on the newborn infant. American Journal of Obstetrics and Gynecology 89: 1009-1015. Silva M, Malinson C & Reynolds F (1996) Sciatic nerve palsy following childbirth. Anaesthesia 51: 1144-1148. Stride PC & Cooper GM (1993) Dural taps revisited. A 20-year survey from Birmingham Maternity Hospital. Anaesthesia 48: 247-255. Whitely SM, Murphy PG, Kirollos RW & Swindells SR (1993) Headache after dural puncture. British Medical Journal 306: 917-918. Weissman A, Grisaru D, Shenhav Met al (1995) Postpartum surveillance of urinary retention by ultrasonography; the effect of epidural analgesia. Ultrasound in Obstetrics and Gynecology 6: 130-134.