Labor and delivery is a painful process. The treatment of acute pain during hospitalization has emerged as an important health care concern among both providers and patients.1 Assessment and treatment of pain in the opiate dependent patient is particularly challenging, given concerns of drug seeking, overmedication, and assumptions that the opioid agonist taken for treatment of opiate dependence is sufficient for pain control.2 Labor and delivery are rare clinical situations in which acute pain is nearly universal and characterized by reasonably defined onset and resolution.
Methadone maintenance during pregnancy for opiate-dependent women has been the standard of care since the 1960s, precipitated by case reports of intrauterine fetal death associated with rapid maternal withdrawal.3 Treatment of opiate dependence can reduce the use of other illicit drugs, increase delivery of prenatal care, and decrease risky behavior such as needle sharing.4 No studies specifically address the treatment of acute pain in labor and postpartum in women maintained on methadone for treatment of opiate dependence during pregnancy.
The goal of this study is to evaluate the pain medication needs during labor and postpartum in women maintained on methadone for opiate dependence during pregnancy as compared with control women not on methadone maintenance therapy. Our hypothesis is that women maintained on methadone require increased opiate use during labor and postpartum. The primary outcome measured was cumulative opiate use, excluding methadone maintenance, during labor and the puerperium. Secondary outcome variables included pain scores and intrapartum analgesia.
MATERIALS AND METHODS
We performed a historical cohort–control study of women maintained on methadone during pregnancy treated at Fletcher Allen Health Care from January 1999 until May 2005 (vaginal birth) or May 2006 (cesarean delivery). Patients treated with methadone for opiate dependence are cared for by a single substance abuse clinic within the institution. The interval for cesarean delivery was extended to acquire additional patients to achieve appropriate power. Institutional review board approval for the examination of medical records was obtained before the study. All patients were treated either under the institutional license to dispense methadone for opiate addiction during pregnancy or through the local opiate treatment center. Substance abuse counseling was recommended for all patients. Each patient was matched to a control with the following criteria: the next patient delivered matched by age (within 5 years), parity (nulliparous compared with multiparous), labor (none, induction, spontaneous labor), mode of delivery (vaginal, operative vaginal, or cesarean), gestational age (within 3 weeks), and smoking status. The institutional obstetric database was used to identify appropriate controls.
Data were extracted retrospectively from the antenatal and inpatient records. Subjects were assessed for symptoms of opiate withdrawal using a checklist in the prenatal record (craving, sweating, lacrimation, rhinorrhea, piloerection, yawning, abdominal pain, diarrhea) at each prenatal visit (at least every 2 weeks until 36 weeks, then weekly). One patient went into labor while still hospitalized for induction onto methadone, but had been stable on her dose for 24 hours. Opiate use during hospitalization was recorded from the medication records and anesthesia record. The initial pain score was performed within 2 hours of admission (and before any analgesia intervention) using a verbal pain score: the nurse asked the patient to describe pain on a scale of 0 to 10, with a score of 0 representing no pain and 10 the worst imaginable pain, and noted in the nursing record. Pain assessment was repeated before and after regional analgesia. If pain was noted by the nurse in subjective terms, it was scored as mild=2; moderate=5; severe=8. Postpartum, pain was assessed 2 hours after a pain intervention or with vital signs.
Labor analgesia was administered at patient request. When parenteral analgesia was requested, methadone-maintained women received morphine, and control women were offered either nalbuphine, a opiate agonist-antagonist, or morphine. The timing and type of regional analgesia (spinal, epidural, or combined spine and epidural) was determined by the patient and anesthesiologist. All patients receiving epidural analgesia had patient-controlled analgesia, with a basal rate, bolus volume, time of delay between bolus administration (in minutes), and a maximal amount of infusate volume per hour of a standardized epidural infusate solution (1/16% bupivicaine with 2 mcg of fentanyl per milliliter). The initial basal rate, bolus amount, delay between bolus doses, and maximal amount in 1 hour were all noted from the chart. The precise amount of infusate over the course of labor, ie, volume of bolused solution, was not available. The need for adjustment of the initial settings or additional bolus was noted from the anesthetic record. The epidural catheter was removed after the completion of the delivery and perineal repair after vaginal birth. For cesarean delivery under regional analgesia, a combination of local anesthetic with fentanyl was used. Long-acting regional opioid (morphine or meperidine) was used at the discretion of the anesthesiologist.
Postpartum pain medications were ordered as per standardized order sheets, with the addition of methadone for methadone-maintained women. The standard orders allow for the following pain medications: after vaginal delivery medications are ordered on an as needed basis: acetaminophen 650 mg every 4 hours; ibuprofen 400 mg every 4 hours; oxycodone 5–10 mg every 4 hours. After cesarean delivery, the following pain medications were ordered to be administered for 48 hours postoperatively (as opposed to as needed) with as needed dosing after the first 48 hours: acetaminophen 325 mg and oxycodone 5 mg as combination, one or two tablets (patient choice) every 4 hours and ibuprofen 400 mg every 4 hours. For the first 24 hours postoperatively, an additional order for morphine sulfate 2–5 mg intravenously (nurse discretion) every 2 hours was included for breakthrough pain on an as needed basis. Postpartum pain control was assessed by verbal communication with the nurse every 4 hours for the first 24 hours, then every 8 hours and noted in the medical record.
Total opiate use was calculated by combining intravenous and oral administration. Opiate use was standardized to oxycodone equivalents (intravenous, morphine 10 mg=fentanyl 0.1 mg=hydromorphone 1.5 mg=demerol 75 mg=oxycodone 20 mg; oral, oxycodone 20 mg=hydromorphone 7.5 mg).5 Cumulative medication use was calculated as the total amount (in milligrams) used divided by the duration from delivery to discharge (per 24 hours). Methadone used for the treatment of opiate dependence was not included in the total opiate use required in labor or postpartum, but is noted separately. Postpartum, women were maintained on the antepartum methadone dose and schedule (morning dosing by elixir).
Statistical anaysis was performed using NCSS/PASS 2000 Dawson Edition (NCSS, Kaysville, UT). Demographic data were compared using paired t test. Intrapartum analgesia data included only those patients in whom a trial of labor was attempted. Cumulative nonopiate and opiate use were calculated as the cumulative amount of medication administered divided by the duration of hospitalization after delivery up to 72 hours, expressed as medication per 24 hours. Cumulative pain scores were determined by the average score after delivery up to 72 hours postpartum. Based on preliminary control data, the estimated oxycodone equivalent use was 5 mg/d after vaginal birth (±5 mg) and 50 mg/d (±30 mg) after cesarean birth. For a clinically significant increase of doubling the oxycodone equivalent dose after vaginal birth (10 mg) and 60% increase after cesarean delivery (80 mg), 16 patients in each group would be needed for 80% power. Medication use was compared by paired t test. Paired frequencies were compared by McNemar test. Paired ordinal data (pain scores, pain score number) were compared using the Wilcoxon signed rank test. Nonpaired frequency analyses (regional analgesia) were compared with Fisher exact test; nonpaired nonparametric data were compared with Wilcoxon rank sum test; unpaired parametric data were compared with unpaired t test. Parametric data are shown as the mean±1 standard deviation from the mean. The median (25th, 75th percentiles) are noted for nonparametric data.
Demographics are presented in Table 1. Patients receiving methadone who underwent cesarean delivery were slightly younger than controls, but within the matching criteria. Birth weight was significantly less in women maintained on methadone despite the similar incidence of smoking.
The time intervals surrounding delivery are summarized in Table 2. Women maintained on methadone who delivered vaginally had a longer interval between admission and administration of regional analgesia, a trend toward less time between regional analgesia and delivery, but no difference in the time between admission and delivery. Women maintained on methadone who delivered by cesarean delivery after a trial of labor had similar admission to analgesia and delivery times to control patients. The duration from admission to discharge was increased in women receiving methadone regardless of mode of delivery, which likely reflects the longer neonatal stays for assessment of neonatal abstinence syndrome.
Intrapartum analysis included all patient pairs that had a trial of labor (n=45; 10 eventually required cesarean delivery). Intrapartum pain interventions and responses were similar. Methadone-maintained and control women in spontaneous labor presented for admission with a moderate and similar level of pain (pain score: methadone, 5.0 [1.5–8.5]; control, 6.5 [3.7–9.0]), P=.41). A similar number of methadone maintained and control women delivered without labor analgesia (methadone, 4 of 45 [8.9%]; control, 5 of 45 [11.1%], P=.68). Three patients receiving methadone and two control patients received intravenous opiate intrapartum (P=1.0). Of the patients on methadone who received intravenous opiate, one patient refused regional analgesia and requested patient-controlled anesthesia with intravenous morphine; the other two patients required intravenous opiate medication to facilitate placement of regional analgesia. There was no difference in the proportion of women choosing neuraxial analgesia (methadone, 36 of 45 [80%]; control, 35 of 45 [77.8%], P=1.0), in the pain score before or after placement of the analgesic agent (pain score before: methadone, 9 [(8–10)]; control, 9 [(7.5–10)], P=.86; after epidural: methadone, 1 [0–3.3]; control, 1.3 [0–2], P=.77). The initial patient-controlled epidural analgesia settings were similar (mean±1 standard deviation; methadone, basal 11.7±1.7 mL/h; delay 6.6±1.9 minutes; bolus 8.0±2.8 mL; 1-hour maximal volume, 34.6±1.6 mL; control: basal, 10.6±1.6 mL/h; delay 6.1±1.7 minutes; bolus 8.0±2.5 mL; 1-hour maximal volume, 34.0±3.0 mL; P=.19; P=.32; P=.96; P=.38, respectively). Although the exact amount of solution administered was not available, 11 of 36 (30.6%) of women receiving methadone required additional epidural bolus or an increase in infusion or both compared with 4 of 35 (11.4%) of control patients (P=.08). Two cesarean deliveries occurred under general anesthesia, both in women who were receiving methadone: one for failed intrathecal analgesia (inability to achieve a level to T12) and the other due to emergent delivery because of acute abruption. Overall, the data suggest that there is little difference regarding intrapartum pain perception, response to interventions for pain control, and management of intrapartum pain in women receiving methadone maintenance for opiate dependence compared with control patients.
Cumulative pain medication use normalized to duration of hospitalization and pain scores are summarized by mode of delivery in Table 3. Pain scores after vaginal birth were elevated in women maintained on methadone, but there was no difference in the amount of opiate or the frequency of opiate use (methadone 14 of 35 [40%]; control 14 of 35 [40%], P=.80). In contrast, after cesarean delivery, women maintained on methadone had higher postoperative pain scores and required a 70% increase in opiate use compared with control women. The increased pain and opiate use was more apparent immediately after surgery (within 4 hours postpartum) compared with throughout the initial postoperative day (data not shown). Pain and opiate use remained increased in methadone-maintained women even after the immediate postoperative period (Table 4). Within each group, opiate use decreased by approximately 30% during hospitalization, despite little change in pain score. The frequency of pain assessment during hospitalization was similar (number of pain scores recorded per 24 hours of hospitalization: vaginal delivery: methadone, 3.5 [2.4–5.5]; control, 3.0 [2.5–4.0], P=.19; cesarean delivery: methadone, 7.3 [4.4–6.9]; control, 6.0 [3.8–6.7], P=.66). There was little correlation between methadone dose and opiate use per day in those women who delivered by cesarean delivery (r=0.004).
The standardized orders allow for up to 120 mg of oxycodone equivalents during the first postoperative 24 hours and 60 mg of oxycodone equivalents for the rest of the hospitalization. Opiate dosing or potency in excess of the standardized orders were required more frequently by women maintained on methadone compared with controls (methadone 12 of 33 [36.4%]; control 3 of 33 [9.1%], P=.03). The change in orders occurred within the first 24 hours after cesarean delivery in all but two patients (one methadone maintained and one control patient).
Long-acting intrathecal opiate was administered at cesarean delivery at the discretion of the anesthesia team equally in methadone-maintained and control women (methadone 9 of 33 [27.2%]; control 7 of 33 [21.2%], P=.61). This did not alter the group results as above, because systemic use of opiate in the first 24 hours in this subgroup remained increased in methadone-maintained women (methadone 103±57; control 44±26 mg, P=.02, unpaired). In contrast, patient-controlled epidural analgesia was quite effective in reducing the opiate use for the first 24 hours postoperatively in both groups (methadone [n=3], two required no opiate during patient-controlled epidural analgesia infusion; one required 20 mg oxycodone; control [n=1], required 30 mg oxycodone). One methadone-maintained patient who required 140 mg oxycodone equivalents in the first 4 hours postoperatively required no more opiate analgesic for the next 32 hours after patient-controlled epidural analgesia placement.
We found that women maintained on methadone for opiate dependence during pregnancy were not different from controls during intrapartum presentation and pain management. After delivery, women maintained on methadone had increased pain and required up to 70% more oxycodone equivalents after cesarean delivery. Much of the additional opiate was required during the first few postoperative hours, with both methadone-maintained women and controls decreasing opiate use by 30% over the course of hospitalization. Overall, we saw no indication that methadone-maintained patients presented earlier in labor or used delivery as a reason for increased opiate use. The use of opiate medication by 40% of the patients after vaginal birth supports routine availability of opiate analgesics for all postpartum patients.
The challenges associated with the treatment of acute pain in the opioid dependence patient have been the subject of two recent reviews.2,6 There are no randomized trials to specifically guide treatment for pain in methadone-maintained individuals, with both reviews relying on consensus and expert opinion in their final recommendations. The variability of cause and resolution of acute pain has hindered the study of analgesic needs in the methadone-treated population. A strength of this study is the ability to match both the cause and treatment of acute pain to a control group.
This study is limited by the retrospective nature of the study and the knowledge of the nursing and medical staff regarding methadone maintenance during pregnancy. One benefit of the retrospective design is that no patient knew that their opiate use was being measured, which could have altered medication requests. The pain score is subjective and reported by the patient in numerical terms. Although the policy is to describe the numerical score to all patients in the same manner, the retrospective design did not allow for assessment of this interaction. The pain scoring protocol used in this study has been found to be as effective as other pain scoring methods.7 We did not have baseline information regarding pain before pregnancy, although no patient was receiving medication for this indication. Given the concordance of observations of increased pain with increased opiate use, it is most likely the increased opiate use is a reflection of the patient’s pain score and the appropriate response of the nursing staff. Use of other illicit medication could have been missed in either the methadone-maintained or control patients. Urine drug screens are routinely performed in the office setting in women maintained on methadone, but not in the control population. Excluding use of marijuana, there were few positive screens or additional opiate use in pregnancy in the methadone-maintained women. The extended length of postpartum stay likely reflects the longer neonatal stays required for evaluation for neonatal abstinence syndrome in women who received methadone.
Neuraxial analgesia during labor with an infusate combining a short-acting opiate (fentanyl) and local anesthetic was equally effective in women maintained on methadone and controls, which raises the potential of this approach for postoperative pain. The few patients who received patient-controlled epidural analgesia using a combined opiate with local anesthetic for postoperative pain markedly decreased, but did not eliminate, systemic opiate use. Reliance on intrathecal opiate analgesics alone (without local anesthetic) may not be optimal, because sustained spinal administration of mu-opioid agonists in rats can induce both opioid tolerance, opiate-induced hyperalgesia, and abnormal pain that is similar to neuropathic pain.8 Our findings support the use of patient-controlled epidural analgesia for those women with poorly controlled postoperative pain.
The results for this report provide general guidelines regarding the degree of pain and opiate use that can be expected during delivery of women enrolled in a methadone maintenance program for opiate dependence. Prospective trials comparing postoperative pain interventions are necessary before specific recommendations can be made. Women who are untreated or noncompliant in their treatment program may require a different approach to postoperative pain.
1. Gordon DB, Dahl JL, Miaskowski C, McCarberg B, Todd KH, Paice JA, et al. American Pain Society recommendations for improving the quality of acute and cancer pain management. American Pain Society Quality of Care Task Force. Arch Intern Med 2005;165:1574–80.
2. Alford DP, Compton P, Samet JH. Acute pain management for patients receiving maintenance methadone or buprenorphine therapy [published erratum appears in Ann Intern Med 2006;144:460]. Ann Intern Med 2006;144:127–34.
3. Rayburn WF, Bogenschutz MP. Pharmacotherapy for pregnant women with addictions Am J Obstet Gynecol 2004;191:1885–97.
4. Fischer G. Treatment of opioid dependence in pregnant women. Addiction 2000;95:1141–4.
5. Principles of analgesic use in the treatment of acute pain and cancer pain. 5th ed. Glenview (IL): American Pain Society; 2003. p. 16–7.
6. Mehta V, Langford RM. Acute pain management for opioid dependent patients. Anaesthesia 2006;61:269–76.
7. Williamson A, Hoggart B. Pain: a review of three commonly used pain rating scales. J Clin Nurs 2005;14:798–804.
© 2007 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
8. Devulder J. Hyperalgesia induced by high-dose intrathecal sufentanil in neuropathic pain. J Neurosurg Anesthesiol 1997;9:146–8.