Rapid Opioid Detoxification during General Anesthesia : A Review of 20 Patients

The Institutional Review Board of St. Elizabeth's Medical Center approved the protocol for the procedure, which included very detailed informed consent. Twenty patients, American Society of Anesthesiologists status I–II, addicted to various opioids were screened by St. Elizabeth's Comprehensive Addiction Program (SECAP) and referred to the Department of Anesthesiology for treatment. The patients presented to the hospital at least 1 day before the procedure and were evaluated by a physician who specialized in the practice of addiction medicine and by an anesthesiologist. Screening included a detailed clinical history, physical examination, blood analysis for complete blood count, serum chemistries, and liver function tests. Urine toxicology screening was performed for opioids, benzodiazepines, methadone, phencyclidine, cocaine, amphetamine, cannabinoids, barbiturates, propoxyphene, tricyclics, and alcohol. An electrocardiogram (ECG) and chest radiograph were also obtained. The patients were then admitted to St. Elizabeth's Comprehensive Addiction Program for observation and preparation.

All patients were at least 18 yr of age (range, 27–48 yr; mean, 37.6 yr). Eleven patients reported isolated use of more than one opioid. The additional opioids used and dosage and duration of use were extremely variable and would not be expected to influence withdrawal or detoxification. All patients are described herein according to the drug used most frequently, i.e. , by drug of choice (table 1).

Six patients reported a history of failure to withdraw from opioids on their own, and 14 patients were unsuccessful with previous detoxification techniques. The five most common symptoms were nausea, (18 patients), diarrhea (15 patients), diaphoresis (12 patients), abdominal cramping (12 patients), and muscle aches (11 patients). All patients reported that their symptoms were relieved by administration of opioid.

Two patients with previously diagnosed chronic pain syndromes were using high doses of opioid analgesics that were not approved by their treating physicians. One patient was using nasal butorphanol tartrate, 60–75 mg/day for 48 months, for low back pain that caused a reduction in employment status to approximately 20 h per week and limited his daily activities. The second patient was using meperidine hydrochloride, 300 mg/day, oxycodone, 60 mg/day, and hydromorphone, 24 mg/day, for 30 months for neck pain that resulted from a motor vehicle accident. This patient was not able to work and was severely limited in his daily activities. No formal pain assessment was performed before detoxification. The patients were referred for detoxification to attenuate tolerance to opioids.

Comorbid conditions included hepatitis A, B, or C (nine patients), depression (nine patients), bladder cancer (one patient, postresection), bulimia (one patient), and insulin-dependent diabetes (one patient). Eight other patients showed laboratory evidence of elevated liver enzymes without an established diagnosis. No patient had evidence of compromised hepatic synthetic function. Comorbid conditions were determined to be optimally managed and stabilized before acceptance into the program.

Eighteen patients reported abusing drugs other than opioids in various combinations (alcohol, 2 patients; marijuana, 4 patients; cocaine, 3 patients; benzodiazepines, 10 patients; propoxyphene, 1 patient; amphetamine, 1 patient). Nine patients were taking serotonin reuptake inhibitor antidepressants, as prescribed. Five patients were prescribed methadone. No patient was diagnosed as dependent on any drugs other than opioids or antidepressants. The Diagnostic and Statistical Manual of Mental Disorders, 12fourth edition, (DSM-IV) criteria were used to define “dependence” and “abuse.”

Results of qualitative urine toxicology screening for 11 drugs were available for 17 patients and are listed in table 2. These are compared with the number of patients that self-reported abusing the drug listed. Three patients were not tested or the results were lost. There were 14 positive test results for opioids, 9 positive test results for benzodiazepines, 4 positive test results each for methadone, tricyclics, and cannabinoids, 3 positive test results each for cocaine and barbiturates, and 1 positive test result each for amphetamine and propoxyphene. Test results for one patient were negative for any drug, for two patients were positive for opioids only, for five patients were positive for two drugs, for five patients were positive for three drugs, and for four patients were positive for four drugs.

The use of high doses of opioid antagonist or somato-statin in children, adolescents, pregnant women, or nursing mothers has not been studied. There are no reports of anesthesia-assisted opioid detoxification performed in these patient populations. Patients with severe hepatic disease are at risk for complications associated with the use of high doses of the hepatically metabolized drugs clonidine, propofol, nalmefene, and naltrexone. Patients with known hypersensitivity to any of the medications used would be at risk for allergic complications. Therefore, these patients were excluded from treatment.

Patients with a dual addiction to opioids and alcohol are at increased risk for complications compared with patients with no other addictions. Studies suggest the involvement of the opioid system in alcohol dependence, 13and it is well-established that acute withdrawal from alcohol is associated with significant morbidity and mortality. Although precipitating a withdrawal syndrome using an opioid antagonist in these dually addicted patients has not been reported, these patients were also excluded from treatment.

On the morning of the procedure, the patients were transported to the postanesthesia care unit (PACU). Monitors included five-lead, two-channel electrocardiography (leads II and V), noninvasive blood pressure, pulse oximetry, train-of-four neuromuscular monitoring, end-tidal carbon dioxide monitoring, an esophageal temperature probe (°C), and a Bispectral Index monitor (BIS)(Aspect Medical Systems, Inc., Natick, MA). Lactated Ringer's solution was infused through a 20-gauge peripheral intravenous catheter. Fluid loading was standardized by assuming that each patient had an 8-h deficit, based on body weight, that was replaced by infusing fluid at three to four times the calculated maintenance rate until the deficit was replaced. 14The infusion was then slowed to the maintenance rate, also based on body weight, 15for the duration of the procedure. Baseline hemodynamic values and withdrawal scores were obtained before any medications were administered. All medications were administered intravenously. Withdrawal scores were determined using an adapted scoring system referred to as the Clinical Institute Narcotic Assessment (CINA) withdrawal scale 16(table 3). The CINA scale assigns a point value (0–6), depending on the specific parameter and according to severity, to a specific set of 11 withdrawal signs and symptoms. The range of scores is from 0 to 31.

Before induction, the patients were premedicated with 0.2–0.4 mg glycopyrrolate and 450–600 μg clonidine. Clonidine was administered to attenuate systemic effects of the withdrawal 17,18that would soon occur as a result of the administration of antagonist. The dose was titrated against decreases in heart rate and blood pressure to 80% of the baseline values. If this degree of change did not occur, the maximum dose administered was 600 μg. 19 

The patients were then preoxygenated with 100% oxygen and general anesthesia was induced with 2–2.5 mg/kg propofol and 0.1–0.2 mg/kg cisatracurium. After ablation of the twitch response to train-of-four stimulation, the trachea was intubated, and the patients were mechanically ventilated with oxygen and air (fractional inspired oxygen concentration, 0.21–0.35). During the preliminary phase of developing the protocol, general anesthesia for the first 10 patients was maintained by propofol infusion titrated to a Bispectral Index value of 40–60 and supplemental cisatracurium titrated to maintain a one-of-four twitch response. For subsequent patients, general anesthesia was maintained by propofol infusion combined with desflurane, 1–3%, titrated to a Bispectral Index value of 40–60, and cisatracurium as described. After induction, a urinary catheter, an orogastric tube, and an esophageal temperature probe were placed. Venodyne boots were applied to the patients’ lower extremities and a pillow was placed under their knees. The patients’ vital signs were recorded every 5 min and the Bispectral Index values and core temperature were recorded every 15 min.

Octreotide was then given intravenously as a 50-μg bolus dose followed by an infusion of 50 μg/h to a total dose of 300 μg to control gastrointestinal secretions. 20In addition to the cited references, information in the Physician's Desk Reference 21provided the basis for the dosing schedule.

After the patients were hemodynamically stable, a test dose of 0.4 mg intravenous naloxone was administered. 22–24To minimize risk, we administered a single dose of 0.4 mg intravenous naloxone after general anesthesia was established. This was intended to test the degree of physiologic homeostasis achieved by premedication. The patients were observed for signs of withdrawal (piloerection, movement, lacrimation, increases in heart rate or blood pressure) for 5 min. If any response was noted, 100–200 μg clonidine was administered as needed to control the response.

At this point, detoxification was continued by starting a nalmefene infusion. 25–27Based on published studies, we ultimately chose a dose of 4 mg intravenous nalmefene to achieve significant receptor blockade that would persist until maintenance therapy could be established after emergence and extubation by the administration of oral naltrexone.

For the first patient, nalmefene was infused at 1 μg · kg⁻¹· h⁻¹for 1 h followed by infusion of the remainder of a 2-mg dose over the next 2 h. For the next 16 patients, the same schedule was followed, but the total dose was increased to 4 mg. For subsequent patients, the total dose of nalmefene (4 mg) was infused over 2 h. The dose was increased after the first patient because of the presence of moderate-to-severe withdrawal symptoms after the procedure. This suggested that the detoxification process was incomplete and therefore did not follow the time course suggested by previous studies. 4,11 

After the nalmefene infusion was complete, administration of cisatracurium was discontinued, allowing spontaneous recovery of the train-of-four twitch response over the next 1 to 2 h to avoid having to reverse neuromuscular blockade. The patient was monitored for signs of withdrawal for 30–60 min after discontinuation of the nalmefene infusion, and no patient required treatment. At this time, 8 mg intravenous ondansetron was administered, and the patient was observed for another 30–60 min. Approximately 1 h before emergence, blood was drawn and sent for analysis of serum electrolytes and osmolality. The anesthetic drugs were discontinued, and emergence followed very quickly. Before emergence and extubation, the orogastric tube, esophageal temperature probe, and urinary catheter were removed. Lidocaine jelly was placed into the patient's urethra after removal of the catheter to facilitate patient comfort. The patient's trachea was extubated when clinically indicated. After the patient was able to respond to simple questions, another withdrawal score was obtained, followed immediately by an injection of naloxone, 0.4 mg intravenous. After 5 min of observation, the withdrawal assessment was again repeated to assess any changes withdrawal signs or symptoms.

If no signs or symptoms of withdrawal were seen or if withdrawal scores were less than 7, the procedure was deemed complete, and the patient recovered in the PACU for 1 to 2 h as indicated. Withdrawal signs and symptoms were treated with adjunct medications: ketorolac, midazolam, or clonidine, as needed. When the patient was fully stabilized and oriented, 50 mg naltrexone was administered orally. The patient was then discharged to the medical floor at St. Elizabeth's Comprehensive Addiction Program. After a total hospital stay duration of approximately 24 h, patients began outpatient aftercare. This included naltrexone maintenance therapy, counseling, follow-up examinations, and participation in a 12-step program.

Follow-up information was obtained from each patient and corroborated by at least one immediate family member. This was done by telephone interviews using a standardized set of questions concerning perceptions about the procedure, current social situation, including family and work status, and current drug use. 28 

Values are reported as the mean values of the variable for the group ± SD. Differences in mean values of variables over time were tested by one-way repeated-measures analysis of variance followed by the paired Student t  test. Statistical significance was assumed for a P  value < 0.05.

All 20 patients successfully completed detoxification without any adverse medical or anesthetic events. Detoxification was verified by documentation of no significant withdrawal response to naloxone administered just after emergence from anesthesia (table 4). Furthermore, although no CINA score was obtained at the time, all patients tolerated 50 mg naltrexone without subjective changes in withdrawal symptoms before leaving the PACU. Changes in systolic arterial pressure and heart rate during the procedure were minimal. The only statistically significant change was a slight increase in the mean diastolic blood pressure after completion of the nalmefene infusion. Core temperatures remained at baseline values (± 1°C).

At completion of the procedure, the first patient had moderate-to-severe withdrawal symptoms. He was treated by titrating doses of clonidine and midazolam intravenously until the symptoms were controlled. CINA scores decreased from 12 of 20 to 4 of 20. Six of 20 patients showed no signs or symptoms of withdrawal, and 13 had mild symptoms suggested by withdrawal scores ranging from 1 of 20 to 4 of 20. Those patients were treated with clonidine in incremental doses ranging from 100 to 400 μg and with midazolam in incremental doses ranging from 0.5 to 4.0 mg. No patient received midazolam after leaving the PACU.

Table 5shows mean values for the total amounts of the primary anesthetic drugs used in the procedure and mean values for total intravenous fluid administration and urine output. Approximately 1 h before emergence from anesthesia, serum sodium, potassium, chloride, carbon dioxide, and plasma osmolality were all within normal ranges, suggesting that there is no clinically significant effect of detoxification on serum electrolytes, glucose, or osmolality.

No significant episodes of nausea, vomiting, or diarrhea occurred immediately after the procedure. Nausea developed in four patients in the early postprocedure period. Sixteen patients met all discharge criteria within 24 h after the procedure, of whom 6 were discharged. Patients were eligible for discharge when (1) oral intake of fluids was tolerated, (2) nausea was absent or controlled with therapy, (3) diarrhea was absent or controlled with therapy, (4) urine was voided without difficulty, (5) fever was absent, and (6) unassisted ambulation was possible. Fourteen patients remained in the hospital for more than 24 h: two for placement in a residential treatment facility, four to arrange home support, four because of a subjective sense of generalized malaise, and four for treatment of nausea. Seven patients were discharged within 24–48 h after the procedure, four patients were discharged within 48–72 h, and three patients were discharged within 72–96 h.

One patient, a 42-yr-old man, was found dead in bed at approximately 8:00 AM of the second day postprocedure, approximately 41 h after completion of the procedure. He was observed by the nursing staff 7 h before and had no complaints. At arrival of the “code team,” the patient was found supine in bed, unresponsive, cyanotic, cold, and with rigor mortis.

At the time of this report, 3 of 17 patients (18%) who were expected to remain abstinent since treatment have done so (≤ 18 months). Follow-up for all patients is ongoing. Another four patients are “clean,” but experienced relapse within 1 month to 1 yr after discharge. Two patients have been lost to follow-up, three transferred to methadone maintenance programs, four relapsed to active drug use, and one died of an accidental overdose of heroin in January 1999.

The two patients treated for chronic pain are continuing opioid analgesic therapy as planned. The patient previously taking 60–75 mg/day butorphanol tartrate is being treated with 30 mg/day. By patient report, this provides adequate analgesia to facilitate full-time employment and independence in daily activities. The other patient, previously treated with moderate to high doses of three agents, is treated with controlled-release morphine sulfate on a dosing schedule of 120 mg, 4 times per day. By patient report, this provides adequate to good relief of pain and facilitates independence in activities of daily living and caring for his disabled spouse. He is not able to work.

This is the first publication to report a detailed protocol for anesthesia-assisted opioid detoxification. Other studies have described the essential characteristics of precipitating and attenuating an acute withdrawal syndrome in patients addicted to opioids. 29–31Clinical 4and animal studies 11show that withdrawal symptoms peak and return to near baseline within 4–6 h, thus establishing the expected time course for the procedure.

Patient reports of drug abuse and results of urine toxicology screening obtained at admission compare well (table 2). It is not clear why three patients showed negative screening results for any opioid. Perhaps highly motivated patients abstained long enough before testing that the amount of drug remaining in the blood was below minimum detectable levels. The rates reported for false-negatives in our laboratory for these tests are 0% for specimens containing 300 ng/ml or more of the drug.

Although electrocardiography or chest radiography findings for any patient before admission were not significant, they might be useful to establish baseline information.

Prophylaxis for gastrointestinal symptoms of withdrawal included H²-receptor antagonist therapy administered the night before and again on the morning of the procedure. Antidiarrheal therapy was never prescribed. No patient experienced adverse gastrointestinal events in the interval between admission and the time of the procedure.

A major concern about this procedure has been the safety of general anesthesia when combined with opioid addiction and comorbid conditions. 32No data specifically address the influence of the conditions described in this report of the extent of withdrawal precipitated by an opioid antagonist. However, patients with comorbid depression can be detoxified with opioid antagonists during general anesthesia without adverse outcomes or specific drug interactions. 33,34Profound increases in plasma catecholamine levels during anesthesia-assisted opioid detoxification associated with significant cardiovascular changes have been shown. 35Yet, we saw no significant changes in systolic arterial pressure or heart rate. Although this does not exclude the possibility of cardiovascular stimulation, any of the indices based on heart rate or systolic arterial pressure should show a lesser degree of stimulation than that observed in the study referenced previously. We did not analyze plasma samples for catecholamine levels, but using our protocol, no end-organ effects were seen. Presumably, this was caused in part by the noradrenergic stabilizing effects of clonidine and the use of different anesthetics.

A recent study showed significant increases in respiratory rate, minute ventilation, oxygen consumption, and carbon dioxide production in patients undergoing anesthesia-assisted opioid detoxification who were allowed to breathe spontaneously. 36We maintained deep anesthesia, muscle relaxation, and controlled ventilation during the procedure to hold end-tidal carbon dioxide at 28–38 mmHg associated with oxygen saturation measured by pulse oximetry at more than 97%. Whether the patient benefits from this approach is not clear and necessitates further investigation. In our study group, as muscle relaxation diminished, several patients exhibited spontaneous and random-appearing muscle activity, perhaps reflecting effects of the antagonist on receptors in the spinal cord. 37,38 

We believe detoxification was complete because CINA withdrawal scores after emergence were always low for all patients, suggesting a significant decrease in withdrawal physiology. Furthermore, the scores did not change significantly after administration of a final intravenous test dose of 0.4 mg naloxone (table 4). Finally, all patients tolerated 50 mg naltrexone before leaving the PACU without subjective changes in withdrawal symptoms, a clinical end point in previous studies of antagonist-induced opioid detoxification. 5,39,40 

Patients were administered appropriate doses of adjunct medications as indicated, such as ketorolac, midazolam, or clonidine. For two patients, subsequent doses of naltrexone were unable to be administered until nausea was controlled (up to 3 days). This allowed blood levels to decrease and minimized receptor blockade, increasing the risk for immediate relapse. In response, the initial dose was increased for the twentieth patient (and all subsequent patients) to 150 mg, providing a sufficient level of receptor blockade for 3 days. 41 

The occurrence of mild-to-moderate withdrawal symptoms for 3 to 4 days after a rapid detoxification procedure is expected. 4,11Studies of anesthesia-assisted detoxification published in the peer-reviewed literature report the severity and character of withdrawal symptoms as relatively minor after this type of treatment. 9,10,29,31Our results support these conclusions.

This protocol provides another alternative to traditional techniques for detoxification from opioids. Long-term abstinence rates (i.e. , 12–18 months) appear to be comparable with rates for conventional therapies, which range from 20–30%. 42Assuming similar long-term abstinence rates, the major value in the anesthesia-assisted procedure is the 100% success rate in achieving detoxification. Therefore, a larger absolute number of patients will begin rehabilitation and achieve long-term abstinence.

Overall function improved with lower effective doses after return to opioid analgesic therapy for both patients with chronic pain, who received high doses of single or multiple agents. However, no data specifically address the effects of detoxification on tolerance to opioids in patients with chronic pain.

The death of one patient on the second day postprocedure shows a possible risk of rapid detoxification from opioids. At the bedside, no information was apparent about the cause of death, and no autopsy was performed, at the family's insistence. Analysis of blood samples showed the presence of medications, prescribed according to protocol, in concentrations reflecting appropriate dosing. The presence of illicit drugs was not found in the analysis. Preoperative evaluation revealed only a remote history of hypertension that was not being treated at the time of admission. Blood pressure was not elevated before the procedure. In the absence of any further information, we cannot know whether this death was a result of the detoxification procedure. Certainly, it shows the need for continuous monitoring of these patients during their hospital course and may suggest the need to observe selected patients beyond the first 24 h after the procedure.

Patients paid approximately $6,000 for professional and hospital services. This is a comprehensive price and includes preadmission testing, a bed on the addiction medicine service ward, the detoxification procedure, and all follow-up rehabilitation services for 6 months. Other programs charge $2,500–$7,500, 9,42which often does not include comprehensive rehabilitation.

Anesthesia-assisted opioid detoxification followed by naltrexone maintenance may be more cost-effective than long-term methadone maintenance. 43A recent survey reported the average annual cost for “standard methadone treatment” to be between $3,750 and $4,400, 44not including counseling services. This cost is cumulative. Naltrexone costs $4.50–$5.00 per 50-mg tablet, or $1,642–$1,825 per yr for a 50-mg/day dosing regimen. Combined with the $6,000 cost of the detoxification procedure, the cost for the first year would be between $7,642 and $7,825. However, naltrexone maintenance is discontinued as the patient adopts behavior consistent with long-term abstinence. Therefore, the cost decreases progressively.

As the role for anesthesia-assisted opioid detoxification evolves, experts in addiction medicine debate issues of safety, effectiveness, and cost-effectiveness. Essentially, all agree on the importance of establishing comprehensive psychologic and psychosocial aftercare programs and long-term outcome data. 31,32,42 

In summary, anesthesia-assisted opioid detoxification is an alternative to conventional techniques. Using an appropriate protocol of premedication and supportive treatment during the procedure, we observed no significant hemodynamic or physiologic derangement. The initial detoxification rate is 100%. Eighteen percent of patients remained abstinent since discharge (≤ 18 months), which is comparable with 20–30% for traditional methods of detoxification. 42Because all patients are successfully detoxified by the procedure, the absolute number of patients beginning rehabilitation is greater, compared with traditional techniques. Assuming similar rates of achieving a specific long-term abstinence end point, compared with traditional detoxification techniques, the absolute number of patients achieving that end point also is greater. This feature may prove to be a significant advantage in the treatment of opioid-addicted patients. The procedure may prove to be cost-effective as well. Long-term outcome studies and large, multicenter studies using standardized protocols and appropriate controls will further define the value of anesthesia-assisted opioid detoxification.

The authors thank Magda Bishai, M.D., Richard Eng, M.D., Kenneth MacDonnell, M.D., Andrew Menard, Esq., Harry Barnett, Esq., Kathy Kiley, R.N., and the staff of the postanesthesia care unit and St. Elizabeth's Comprehensive Addiction Program at St. Elizabeth's Medical Center. Editor's Note:  The use of nalmefene for rapid narcotic detoxification as described in the preceeding article may be covered entirely or in part by US Patents 5,783,583 (April 12, 1996) and 5,922,705 (April 13, 1998) assigned to David Simon and Intensive Narcotic Detoxification Centers of America, LLC (Toland, CT).