Original Authors: Paul Perry, Ph.D, BCPP, Bruce Alexander,
Pharm.D, BCPP
Latest Revisers: Paul Perry, Ph.D, BCPP, Brian C. Lund,
Pharm.D.
Creation Date: 1996
Last Revision Date: April 2004
Peer Review Status: Internally Peer Reviewed
INTRODUCTION
Discontinuation of benzodiazepines (BZD) can lead to relapse, rebound, and/or withdrawal symptoms (Noyes 1988). Relapse is defined as the gradual return of original symptoms of the anxiety or sleep disorder and is estimated to occur in 63 to 81% of patients abruptly withdrawn from therapeutic doses of BZD. Rebound is experiencing original symptoms at a level more severe than baseline. Withdrawal is the presentation of new symptoms related solely to discontinuation of the supratherapeutic BZD doses. Symptoms of withdrawal can include may delirium manifested by confusion, disorientation, and memory dysfunction as well as grand-mal seizures. Rebound symptoms, and withdrawal symptoms will gradually return to a baseline level following BZD discontinuation.
Withdrawal syndromes that occur with discontinuation of BZD require pharmacologic management for patient comfort and sometimes, for prevention of potentially serious medical consequences (Albeck 1987). These symptoms have been described for older anxiolytic/hypnotics (Isbell 1950, Wikler 1968).
BZD withdrawal can be classified several ways; minor vs major withdrawal and low- vs high-dose withdrawal (Albeck 1987). The minor/major convention follows the historical classification of ethanol withdrawal and emphasizes the severity of withdrawal signs and symptoms. The low/high-dose categorization will be used in this discussion since daily dose is usually the primary factor in assessing the medical risk of BZD withdrawal.
Typically, but not always, abrupt discontinuation of therapeutic doses (e.g., low-dose withdrawal) will result in minor withdrawal symptoms. For this discussion, low-doses of BZD will refer to the manufacturer's recommended daily doses. Likewise, major withdrawal symptoms will primarily be associated with high-dose use.
Two common clinical scenarios requiring management of BZD withdrawal exist. The first is the substance abuse patient without a primary psychiatric illness using high doses of a single BZD or multiple anxiolytic/hypnotics. The second is the Axis-1 anxiety disorder patient receiving therapeutic or high doses of a BZD. Different pharmacologic strategies are necessary to manage these patients. A brief review of the BZD withdrawal literature is presented to support our guidelines for management of BZD withdrawal.
LOW-DOSE WITHDRAWAL AND REBOUND SYMPTOMS
Presentation
Symptoms associated with low-dose withdrawal include nausea, vomiting, irritability, tremor, incoordination, insomnia, restlessness, blurred vision, sweating, anorexia, and/or weakness as well as anxiety (Noyes 1988). Depersonalization, heightened perception, especially to light and sound, and illusions are also commonly reported. Many of these symptoms have also been reported in unmedicated anxiety patients (Rodrigo 1986).
Incidence and Duration of Use
Because reports of BZD discontinuation may or may not have baseline measurements, the frequency of true withdrawal symptoms is difficult to determine. This contributes to a significant variability in the reported incidence of withdrawal.
Uncontrolled investigations report almost all patients experience withdrawal syndromes after discontinuation of therapeutic doses of BZD after long- term use (Noyes 1988).
In nine controlled short-term (mean 10 weeks, range 3 to 34 weeks) studies of therapeutic doses, seven described the occurrence of rebound/relapse symptoms (Noyes 1988). BZD were abruptly discontinued in seven. The actual incidence of rebound symptoms, was reported in 3 of 9 studies and ranged from 3% to 44%. These three studies are briefly summarized. When diazepam was discontinued after being administered for 6 weeks and 14 to 22 weeks in two groups of patients, rebound symptoms occurred in 3% and 18% of subjects, respectively. In two other studies, diazepam was administered for 4 weeks and alprazolam for 8 weeks. Discontinuation lead to rebound symptoms in 44% and 33% of subjects, respectively. The wide range of reported incidences in these three studies was due to differences in length of treatment, the drugs studied, and the definition of rebound symptoms. In all seven trials that reported rebound symptoms, withdrawal symptoms with BZD discontinuation were noted.
Nine controlled long-term (mean 3 years, range 1 to 16 years) studies examined the effect of BZD discontinuation (Noyes 1988). Six of the 9 studies abruptly discontinued the BZD, while the other 3 tapered the drug. Two of the studies reported no withdrawal symptoms upon abrupt discontinuation. Only five of nine studies reported the actual incidence of withdrawal. Though definitions of withdrawal varied, the mean incidence was 45% (range 0 to 100%). The two studies that reported the highest frequency of withdrawal reactions (82% and 100%) included patients who had previously experienced difficulty discontinuing BZD (Petursson and Lader 1981, Rickels 1986).
Rickels (1983) calculated the incidence of withdrawal after > or = 8 months (average 3 years) and < 8 months of daily BZD use. The incidence of withdrawal was 43% and 5%, respectively.
Overall, withdrawal symptoms occurred in approximately 50% of patients receiving a therapeutic dose of BZD for an average of 3 years (Noyes 1988). Most symptoms experienced were of mild to moderate intensity. Though discontinuation of a BZD taken daily for < 8 months carries a small risk of producing withdrawal symptoms, rebound symptoms are common.
BENZODIAZEPINE HALF-LIFE, WITHDRAWAL FREQUENCY AND TIME COURSE
Many reports suggest BZD withdrawal signs and symptoms differ with respect to the frequency, severity, and time course (e.g., onset, peak, and duration of symptoms) of withdrawal based on the half-life of the drug (Sellers 1978, Tyrer 1981, Morgan 1982, Fontaine 1984, Busto 1986, Rickels 1986). Table 1 classifies the BZDs as either short or long half-life drugs and gives their equivalent doses.
Busto (1986) studied these relationships in a double-blind protocol. They recruited 40 patients who had been ingesting an average diazepam equivalent dose of 15 ± 10 mg/d for 9-240 months. The drugs were discontinued by abrupt placebo substitution or diazepam tapering doses of 1 - 5 mg/week over five- to six-weeks. The placebo group experienced more withdrawal symptoms (p<0.0001), more severe symptoms (p<0.0001), and an earlier onset (p<0.01) of symptoms that occurred in the diazepam-taper group. In the placebo group, the onset of symptoms occurred within one day, after the discontinuation of the short-acting BZD. However, onset of symptoms occurred five days after discontinuation of the long-acting BZD. Peak withdrawal symptoms with short- and long-acting BZD occurred at 1.0 ± 0.5 days and 9.6 ± 5.1 days, respectively. Withdrawal symptoms usually disappeared within two to four weeks. Thus any tapering BZD tapering schedule in an anxiety disorder patient ought to be spread out over a 2-4 week period. The patients who used short-acting BZD prior to the study reported more severe withdrawal symptoms that those patients using long-acting BZDs. Though this finding was not significant, it is noteworthy that all seven patients who refused to continue taking the "study drug", i.e., placebo, were receiving a short-acting BZD prior to the study.
Rickels et al (1990) contrasted the effect of abrupt discontinuation of therapeutic doses of short half-life (SHL) BZDs to long half-life BZDs (LHL) in 57 benzodiazepine-dependent patients (daily use > 1 yr). BZD intake was stabilized was three weeks prior to randomization to either the placebo substitution group (n=47) or the BZD continuation control group (n=10). Using the Physician Withdrawal Checklist, it was discovered that the only difference between the SHL BZDs (lorazepam and alprazolam) and the LHL BZDs (diazepam and clorazepate) was that the adrenergic symptom cluster (nervousness, agitation, nausea, diaphoresis) presented with greater severity for the SHL benzodiazepines. Severity of withdrawal was also related to higher BZD doses, neuroticism and dependency personality traits, less education, and more anxiety and depressive symptoms. Depending on the outcome criteria used, it was estimated that between 58-100% of the placebo substituted group experienced a withdrawal reaction with the SHL BZDs symptoms peaking at 2 days and the LHL benzodiazepines peaking at 4 to 7 days. Relapse occurred in 57% of the SHL BZDs patients and 27% of the LHL BZDs patients. Predictors of relapse included a nonpanic diagnosis, higher BZD doses, and higher levels of neuroticism. Patients who remained BZD-free for greater than five weeks experienced lower levels of anxiety than contrasted to their baseline scores. Thus successful withdrawal from low dose levels of BZDs is enhanced if LHL BZDs are used by the patient.
High-dose withdrawal syndrome
Discontinuation of high-doses of BZDs may produce severe symptoms (e.g., seizures, disorientation, psychosis, depression) as well as the minor withdrawal symptoms previously described. Several early studies examined the effect of abrupt discontinuation of high-dose, short-term BZD administration. Minor withdrawal symptoms occurred in 46% (6/13) and seizures in 8% of subjects receiving diazepam 120 mg/d for 21 consecutive days (Hollister 1963). Hollister (1961) also switched 11 patients to placebo after administering chlordiazepoxide 300 to 600 mg/d for two to six months. Six patients had received 600 mg/d and 10/11 patients had been treated for > or = 5 months. Ten patients experienced withdrawal symptoms, including agitation, insomnia, anorexia, nausea, and depression. Two patients (18%) developed major motor seizures 7 and 8 days, respectively, after chlordiazepoxide discontinuation. However, one study reported the severity of withdrawal reactions did not differ between discontinuation of diazepam 135 mg/d and 20 mg/d (Hallstrom 1981).
These results contrast with the relative low risk of severe withdrawal reactions occurring after discontinuation of therapeutic doses of BZD in controlled trials. None of the 494 patients included in discontinuation studies after short-term administration (average 10 weeks, range 3 to 34 weeks) of therapeutic doses of a BZD experienced severe withdrawal symptoms (Fruensgaard 1976, Kahn 1980, Olajide 1984, Noyes 1985, Noyes 1988).
Of 271 patients receiving therapeutic doses an average of three years, 14 (5%) developed severe reactions. One report contributed 10 of these 14 cases (Petursson 1981). Their 10 patients had prior difficulty discontinuing BZD. Excluding their report, 4/255 (1.6%) patients experienced severe reactions.
Today, because of obvious ethical considerations prohibiting prospective studies of abrupt discontinuation of high-dose BZD, case reports are the major source of information regarding severe BZD withdrawal reactions. A 1986 review of these cases indicated that seizures occur more frequently following the abrupt discontinuation of short-acting than long-acting BZD (Perry, 1986). (See Table 1 for BZD half-life classification.) Similar numbers of BZD-induced withdrawal seizures were reported for the short-acting BZD, lorazepam, oxazepam, and alprazolam, as compared to the far more commonly prescribed long-acting BZD, diazepam, chlordiazepoxide, and clorazepate. Twenty-three cases have been reported for the short-acting group (Quitkin 1972, DeBuck 1973, Tyrer 1979, Tyrer 1980, Einarson 1980, Howe 1980, de la Fuente 1980, Khan 1980, Barton 1981, Einarson 1981, Lennane 1982a, Lennane 1982b, Moore 1982, Ananth 1983, Levy 1984,) while 17 cases of withdrawal seizures have been described for the long-acting group (Hollister 1961, Hollister 1963, Aivazian 1964, Barten 1965, Clare 1971, Vyas 1975, Woody 1975, deBard 1975, Rifkin 1976, Preskorn 1977, Chandora 1980, Winokur 1980, Robinson 1982, Jeffries 1983, Simon 1983). A total of 14 in each group specifically involved diazepam and lorazepam. These cases are briefly reviewed. More specifically, seizures following discontinuation of the short-acting BZD, lorazepam, occurred after 2-12 mg/d (median 7.5 mg/d) administered for three weeks to four months (median 3 months). Only three of the 14 patients were receiving doses outside the recommended therapeutic range of 1-10 mg/d. The 14 cases of seizures following the abrupt discontinuation of the long-acting BZD, diazepam, occurred over a dose range of 20 to 170 mg/d (median 77.5 mg/d) taken for three weeks to 12 years (median 7.5 months). Nine of the 14 taken for three weeks to 12 years (median 7.5 months). Nine of the 14 patients were receiving dosages outside the recommended therapeutic range of 4-40 mg/d.
Assuming lorazepam 2 mg is approximately equivalent to diazepam 10 mg, the following observation can be made (Smith 1983, Busto 1986). Lorazepam, a short-acting BZD, appears to provoke withdrawal syndromes in less than half the duration of daily use and at half the equivalent dose of diazepam. The seizures following lorazepam withdrawal occurred at one to four days (median 64 hours) whereas diazepam withdrawal seizures occurred between two and 27 days (median 5 days).
The mean half-life values for lorazepam, diazepam, and its major metabolite, desmethyldiazepam, are 14, 31, and 51 hours, respectively (Sellers 1978). The onset of withdrawal symptoms are expected to occur when the drug has been nearly eliminated. Therefore, a short-acting BZD with a half-life that is 25% to 50% of a long-acting BZD results in seizures occurring in half the time of a long-acting drug.
Alprazolam deserves special mention because of its prescribing popularity. The FDA instructed the manufacturer to include a patient package insert warning patients of possible withdrawal symptoms occurring with discontinuation of this drug. Alprazolam withdrawal case reports that appear in the literature through 1985 were reviewed (Browne 1986). Eight patients received alprazolam 1.5-10 mg/d from two to eighteen months. Six of the eight patients were receiving alprazolam doses above the manufacturer's recommended therapeutic dose of 4 mg/d. However, it is important to note that doses of 4-10 mg/d have been clinically recommended for treatment of panic disorders. Withdrawal seizures occurred in four of eight cases with the median dose being 6 mg/d. Delirium/psychosis occurred in the remaining four cases. In all cases, withdrawal signs and symptoms were evident within 72 hours of abstinence.
SUMMARY
The correlation of BZD dose and duration of use to the incidence and severity of BZD withdrawal remains to be precisely quantified. However, these two factors appear to be primary considerations in predicting the likelihood of withdrawal. A review of the controlled studies of long-term therapeutic doses of BZD indicates that nearly 50% of patients ingesting a BZD for an average of three years will experience a minor withdrawal syndrome when the drug is discontinued. Eight months or more of daily use may produce a 43% withdrawal incidence, while use less than 8 months significantly reduces the incidence of withdrawal. Minor withdrawal reactions usually begin within one and five days after discontinuation of the short- and long-acting BZD, respectively. Symptoms gradually disappear over two to four weeks. The rate of severe withdrawal symptoms in this group was two to five percent. Studies of high-dose use indicate a significant risk of severe withdrawal reactions upon abrupt discontinuation of the BZD. Case reports indicate signs and symptoms occur earlier and are more intense with short-acting than long-acting drugs. Duration of severe withdrawal symptoms is generally one and two weeks for short- and long-acting BZDs, respectively.
MANAGEMENT OF THE BZD WITHDRAWAL PATIENT
Withdrawal protocols need to address the previously described clinical situations. Protocol design considerations include BZD metabolism (e.g., short- vs long-acting), dose and duration of use, patient diagnostic status (e.g., concurrent psychiatric illness vs no active diagnosis), time-course of withdrawal (e.g., onset, peak, and duration of withdrawal symptoms), and concurrent cross-tolerant drug use (e.g., ethanol) (Albeck 1987). The effect of age on success of a BZD withdrawal protocol was recently addressed (Schweizer 1989). The authors concluded a tapered BZD withdrawal was not associated with greater risks in elderly patients.
Management of BZD withdrawal will be divided into low- and high-dose categories. A patient with a preexisting (and maybe concurrent) anxiety state adds the confounding variable of relapse to the possibility of rebound and withdrawal.
TREATMENT OF LOW-DOSE WITHDRAWAL SYNDROME
Benzodiazepine Taper Methods. The necessity of tapering a BZD in someone using therapeutic doses daily for less than one month is questionable. However, the conservative approach would be to taper the drug over a one to two week period of time.
It is recommended patients who have received manufacturer-recommended doses of BZD on a daily basis for greater than one month should not have the drug abruptly discontinued (Petursson 1981, Laughren 1982, Tyrer 1983, Harrison 1984, Fontaine 1984). Because severe withdrawal symptoms are not expected, gradual tapering of the BZD on an outpatient basis should be attempted. Since withdrawal symptoms are most intense with the last few doses of drug, some patients might not be able to complete the protocol as an outpatient. Thus, they may require inpatient management.
If the patient is receiving a long-acting BZD (see Table 1), the patient can remain on the originally prescribed drug. The weekly tapering rate is determined by dividing the total daily dose of BZD by five and rounding the number to a dose attainable with available dosage forms. Each week the dose would be reduced by the tapering dose calculated above. Thus a patient receiving diazepam 20 mg/d would be tapered at a rate of 16 mg/d during week 1, 12 mg/d during week 2, 8 mg/d during week 3, and 4 mg/d during week 4. This withdrawal protocol covers the period of time when withdrawal symptoms typically occur (Busto, 1986). If necessary, the tapering schedule can be slowed toward the end of the protocol.
If the patient is receiving a short-acting BZD, consideration is given to slowly tapering the dose with the prescribed BZD using a similar schedule described above for long-acting BZDs or substituting a cross-tolerant, long- acting BZD. The rationale for the substitution method is the observation of less severe withdrawal symptoms with long-acting compounds. Though several reports suggest this is an acceptable approach to patients receiving a short-acting drug, this method has not been extensively studied (Perry 1981, Harrison 1984, Busto 1986, Albeck 1987). To accomplish the switch from a short-acting to long-acting BZD, approximate equivalent doses for BZD have been suggested (Smith 1983, Busto 1986). (See Table 1) However, inaccuracy in suggested BZD dose equivalents may affect the success of this recommendation. Though, theoretically, any long-acting BZD could be used, diazepam has been the most widely studied (Perry 1981, Harrison 1984, Busto 1986). Recently, clonazepam has been reported to be useful. Herman et al (1987) substituted on average 1 mg of clonazepam for every 2 mg of alprazolam that the 48 panic disorder patients were taking. Thirty-nine (82%) of the patients preferred clonazepam to alprazolam because of decreased interdose anxiety and fewer doses being required per day, e.g., twice daily dosing versus four times daily dosing.
Non-benzodiazepine Management. Non-BZD agents have been suggested for management of BZD withdrawal. The rationale stems from the desire to remove the patient from the reinforcing pharmacologic effects of the BZD. This approach has not been extensively investigated. More research is required to determine the role of the following agents in the management of BZD withdrawal.
Two-weeks of propranolol, a b-adrenergic blocker, in doses of 60-120 mg/d reduced the severity, but not the incidence of withdrawal symptoms in one study (Tyrer 1981).
Clonidine, an alpha-2-adrenergic agonist, 0.3 to 0.6 mg/d PO for up to 3 months has been utilized to treat several cases of low-dose BZD withdrawal with mixed results (Kesharan 1985, Vinogradov 1986, Goodman 1986).
Carbamazepine has been reported to be of value as an adjunctive agent in benzodiazepine withdrawal in both an anecdotal report (Klein 1986) and an open study (Ries 1989). Schweizer (1991) evaluated the effect of carbamazepine on benzodiazepine low dose (diazepam 15-25 mg/d or equivalent) withdrawal severity and outcome in a controlled trial of 40 patients who had a history of difficulty discontinuing long-term use of the drugs. Forty patients were randomly treated with either carbamazepine 200-800 mg/d (mean = 432 mg/d, 6.1 mcg/ml) or placebo during their benzodiazepine taper (25% per week). Five weeks following the end of the benzodiazepine tapering period, more patients (95%) who were tapered with the carbamazepine remained benzodiazepine free than those who were tapered with placebo (62%). This was especially true of patients taking more than 20 mg/d of diazepam or its equivalent. Five weeks after the end of the taper, 5 (63%) of the 8 patients taking greater than 20 mg of diazepam or its equivalent were unable to continue to do without their benzodiazepine while only three of the less than diazepam 20 mg/d group (23%) of 13 patients receiving placebo during the taper were unable to do without their benzodiazepine. However, this difference disappeared at 12 weeks. There was only a trend to suggest that the carbamazepine reduced the severity of the withdrawal symptoms. The authors concluded that carbamazepine was most useful in facilitating the withdrawal process in patients taking more than 20 mg/d of diazepam because this group had such a poor outcome on follow-up.
Antidepressants as an adjunct treatment in the management of BZD withdrawal deserves further research. Several patients were successfully withdrawn from a BZD while receiving only antidepressants (Tyrer 1985). Both tricyclic antidepressants (TCA) and monoamine oxidase inhibitors (MAOI) are effective antipanic drugs. Studies suggest TCA and MAOI can be as useful in treating generalized anxiety disorder as BZD. These agents might be especially attractive for the patient with a preexisting (and possible concurrent) anxiety disorder. The second generation antidepressants amoxapine, bupropion, fluoxetine, maprotiline, and trazodone have not been extensively investigated in panic disorder or generalized anxiety disorder. Their utility in managing BZD withdrawal is unknown. It is recommended to place patients whose original symptoms are likely to reemerge on imipramine > or = 150 mg/d at least two weeks before BZD withdrawal is initiated (Tyrer, 1985). Though the sedating TCAs of doxepin, imipramine, or amitriptyline have been recommended, research demonstrating differences in efficacy based on relative sedation has not been performed.
Pharmacologic studies demonstrate no cross-tolerance between buspirone and the BZD. Thus it is not surprising that controlled studies have shown the drug to be of no value in the treatment of BZD withdrawal (Schweizer 1986, Lader 1987).
TREATMENT OF HIGH-DOSE WITHDRAWAL WITHOUT TOLERANCE TESTING
Many cases of high-dose BZD use are described in substance abuse patients. Typically, these patients present a difficult management picture because of a questionable dose/duration history or mixed anxiolytic/hypnotic abuse. The following sections discuss several management options for BZD discontinuation. This method assumes a reasonably accurate history is obtained. The use of a BZD as compared to a barbiturate derivative for this method decreases the likelihood the patient will suffer significant untoward effects if the dose administered is too high compared to the actual amount of drug ingested. Administering too high a dose will usually lead to slight intoxication within one to three doses of the schedule. Over-sedation is quickly reversed by withholding or decreasing successive doses. Using too low a dose will often lead to minor withdrawal signs and symptoms that will indicate a need for a higher dose.
Short-Acting Benzodiazepines. A patient using high doses of a short-acting BZD might be withdrawn using the same drug (Albeck 1987). Most experience with short-acting BZD withdrawal procedures involves alprazolam. The manufacturer originally recommended that patients be tapered from the drug at the rate of 1 mg every three days. However, a study reported 14 of 18 patients treated with recommended doses for > or = 12 weeks could not complete this protocol (Fyer 1987). A similar experience was noted in three additional case reports of alprazolam withdrawal (Browne 1986). The manufacturer has updated the package insert recommending the drug not be tapered faster than 0.5 mg every three days. The success of this recommendation, especially with higher doses, has not been reported. This method is not usually practical for tapering an inpatient taking high doses of alprazolam because of the extensive tapering time required to achieve abstinence.
Harrison et al (1984) withdrew 16 of 23 inpatients from high-dose benzodiazepines ranging from 40 to 500 mg diazepam equivalents (median = 150 mg/d) without any complications. Of the remaining 7 patients, 6 experienced symptoms of minor withdrawal while one patient experienced major withdrawal symptoms of paranoia and confusion. Diazepam was substituted at a dose of 40% of the reported daily consumption and tapered at a rate of 10% per day. The advantage of utilizing diazepam originates from its long elimination half-life as well as its primary metabolite, desmethyldiazepam. The diazepam dose should be divided and administered every six hours.
It has been suggested the triazolobenzodiazepines, alprazolam and triazolam, may not be cross-tolerant with other BZDs. In two cases diazepam substitution did not alleviate alprazolam's withdrawal syndrome once it occurred (Browne 1986). However, diazepam's doses used in these two cases may have been too low. One case of triazolam withdrawal psychosis responded to lorazepam (Heritch 1987) while a patient using triazolam 10 mg/d still had a seizure 13 hours after discontinuing the drug despite receiving chlordiazepoxide 600 mg PO and diazepam 75 mg IV prior to the seizure (Schneider 1987).
Long-Acting Benzodiazepines. Patients using long-acting BZDs prior to admission can be withdrawn with the same drug using the procedure outline above for short-acting BZDs (e.g., start with 40% of total daily dose on day 1 and taper at 10% per day using diazepam). The use of carbamazepine might be considered using the protocol outlined under low-dose withdrawal.
WITHDRAWAL PROCEDURE WITH TOLERANCE TESTING
Background. While identification and treatment of the dependent person are necessary in order to avoid symptomatic withdrawal, the patient may be unable or unwilling to give an accurate account of drug intake. Also mixed anxiolytic/hypnotic abuse would indicate tolerance testing to determine a total tolerant BZD dose.
Historically, the pentobarbital tolerance test has been used to establish dependence, primarily in patients abusing barbiturates (Fraser 1958, Ewing 1967). As BZD are more extensively prescribed today, the need for a BZD tolerance test was apparent (Perry 1981). Though barbiturates and benzodiazepines might be cross-tolerant, diazepam is preferred over pentobarbital in tolerance testing. Diazepam can be dosed at a rate of 20 mg po q2h until the endpoint of intoxication is reached, while a maximum pentobarbital dose of 1000 mg can be used because of the risk of respiratory depression, even in patients tolerant to the hypnotic effects at this large dose. The analogous diazepam tolerance test has been successfully utilized but in a small number patients (Perry 1981).
The experience of tolerance testing alprazolam patients is limited. One case report described a man treated with alprazolam 4 mg/d for 18 months, which he increased to 10 mg/d for 2 weeks prior to being tolerance tested with diazepam (Votalato 1987). This patient tolerated a diazepam dose more than 2.5 times the predicted base on the reported equivalent dose. See Table 1. Only at 260 mg/d did the patient demonstrate slight ataxia. He was successfully withdrawn with diazepam over an unspecified period of time, only experiencing significant insomnia for the first 5 days of the withdrawal protocol.
Stabilization Procedure. The patient cannot be tolerance tested until he shows neither signs of drug withdrawal nor drug intoxication. Because the patients are often admitted in the middle of the day or at night, tolerance testing usually cannot be conveniently commenced until the following day. Thus, if a patient is lucid with signs of neither withdrawal nor intoxication when initially seen, an order for diazepam 20 mg orally at bedtime may be given and the test begun at 8 a.m. the following morning. If the patient exhibits signs of withdrawal when initially seen, he should be given diazepam 20 mg every two hours until asleep and then no medication until the test commences at 8 a.m. the following morning. If the patient shows signs of intoxication (e.g., lateral gaze nystagmus, minimal ataxia, slurred speech, positive Romberg sign) when initially seen, then no medication is necessary until the test dose begins at 8 a.m. the following morning. While the patient is being tested, he should be NPO except for fruit juices.
Tolerance Testing. The procedure for conducting the tolerance test utilizes repeated doses of diazepam 20 mg administered orally to clinically unintoxicated patients who have been NPO except for fruit juices for two hours.
Normally the initial dose of diazepam is administered at 8 a.m. If the patient exhibits any signs of intoxication such as ataxia, dysarthria, coarse nystagmus, rombergism, or sleep, between 8 a.m. and 10 a.m., he is considered nontolerant and no further medication is necessary. If alert, he is given additional 20 mg doses every two hours until signs of intoxication are exhibited anytime during two hours after the dose. Patients who are intoxicated with one or two doses (e.g., up to diazepam 40 mg) are considered not at great risk of experiencing severe withdrawal symptoms and may not, based on the judgment of the clinician, require further hospitalization. Conservatively, the patient might be placed on a low-dose withdrawal protocol.
Withdrawal Protocol. If the tolerance test is positive, the patient is withdrawn with diazepam. The daily requirement of diazepam is decreased by 10% per day from an initial level equal to the intoxicating dose. The tapering schedule (e.g., Day 1) begins at midnight of the day following tolerance testing. The total dose of Day 1 is 10% less than the tolerance test dose.
A divided dosage schedule of every 6 hours is recommended rather than four-times-a-day dosing. Dosing every 6 hours means one or two doses would be scheduled at times when a patient would be expected to be sleeping. A patient experiencing withdrawal typically does not sleep well. Therefore, if the person is sleeping those doses should be held and not "made up". This is the "self- titrating" part of the schedule.
Monitoring Procedures. Staff should observe the patient for nausea, diaphoresis, and tremor, and take the patient's blood pressure, temperature, and pulse prior to each dose. The number of hours of sleep per night and any symptoms are recorded in the nursing notes. These nursing observations are used to adjust the withdrawal schedule should the patient develop the signs and symptoms of abstinence or intoxication.
The vital signs used as indices of an abstinence syndrome are hyperthermia and a poor cardiovascular response to orthostasis. Hyperthermia is defined as any temperature above 37.5C. A poor cardiovascular response to orthostasis is considered to be an increase in heart rate of 20 beats/minute or a fall in systolic blood pressure of 20 mm Hg. Additionally, the withdrawal schedule is slowed if symptoms of withdrawal occur (e.g., tremor, nausea, and diaphoresis) or if the patient is observed to sleep less than four hours per night. If significant withdrawal symptoms occur between scheduled PO doses, diazepam 10 mg or lorazepam 2 mg IV can be administered every 20 minutes as necessary to eliminate objective signs of withdrawal. This amount should be added to the scheduled BZD dose and the tapering dose readjusted. If withdrawal symptoms are not that severe and the patient is able to take oral medications at the time, then diazepam 20 every two hours should be administered until withdrawal symptoms cease. Again, the doses should be added to the calculated daily dose and the tapering schedule readjusted.
Patients are also monitored throughout the tapering schedule for signs of drug intoxication as indications of tolerance test inaccuracy (e.g., false- positive tests or overestimation of drug tolerance). If prior to a scheduled dose, the patient presents with ataxia, dysarthria, coarse nystagmus, rombergism, or persistent drowsiness, doses should be held until these clear and the daily dose is reduced by one day's requirement.
Busto et al developed a 22-item clinical scale to assess
benzodiazepine withdrawal that can be used to assess and monitor
benzodiazepine or benzodiazepine-like drug withdrawal. The authors
utilized for their data base a group of 23 high-dose BZD abusers
(mean diazepam equivalent dose = 150 mg/d, range 40-500 mg/d) and 40
low-dose BZP abusers (mean diazepam equivalent dose = 15 mg/d, range
5-40 mg/d, mean use = 72 months, range = 6-240 months). The authors
measured 104 symptoms and found 22 that could distinguish withdrawal
from prewithdrawal and therefore help the physician control the rate
of tapering of anxiolytic medications. The instrument is presented in
Table 2.
|
TABLE 1: BENZODIAZEPINE DOSE EQUIVALENT AND HALF-LIFE CLASSIFICATION |
||
|---|---|---|
|
DRUG |
|
|
|
alprazolam (Xanax) |
|
|
|
chlordiazepoxide (Librium) |
|
|
|
clonazepam (Klonopin) |
|
|
|
clorazepated(Tranxene) |
|
|
|
flurazepamd (Dalmane) |
|
|
|
halazepam (Paxipam) |
|
|
|
lorazepamd (Ativan) |
|
|
|
oxazepamd (Serax) |
|
|
|
prazepam (Centrax) |
|
|
|
temazepam (Restoril) |
|
|
|
triazolam (Halcion) |
|
|
|
Table 2. Benzodiazepine withdrawal schedule rating scale devised by the Addiction Research Foundation of Toronto (Busto, 1989). |
||||||||
|
BENZODIAZEPINE WITHDRAWAL SCHEDULE RATING SCALE |
||||||||
|
1. Observe behavior for restlessness and agitation. |
||||||||
|
0 none |
1 |
2 restless |
3 |
4 |
||||
|
2. Ask patient to extend arms with fingers apart, observe tremor: |
||||||||
|
0 |
1 |
2 |
3 |
4 |
||||
|
3. Observe for sweating, feel palms |
||||||||
|
0 no tremor |
1 not visible, can be felt in fingers |
2 visible but mild |
3 moderate with arms extended |
4 severe, with arms not extended |
||||
|
For each of the following items, please estimate by number the severity of your problem based on the scale below |
||||||||
|
4. Do you feel irritable? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
5. Do you feel fatigued? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
6. Do you feel tense? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
7. Do you feel you have difficulties concentrating? |
||||||||
|
0 no difficulty |
1 |
2 |
3 |
4 unable to concentrate |
||||
|
8. Do you have loss of appetite? |
||||||||
|
0 no loss |
1 |
2 |
3 |
4 no appetite, unable to eat |
||||
|
9. Have you any numbness or burning sensation on your face, hands or feet? |
||||||||
|
0 no numbness |
1 |
2 |
3 |
4 intense burning or numbness |
||||
|
10. Do you feel your heart racing (palpitations)? |
||||||||
|
0 no disturbance |
1 |
2 |
3 |
4 constant racing |
||||
|
11. Does you head feel full or achy? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 severe headache |
||||
|
12. Do you feel muscle aches or stiffness? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 severe stiffness or pain |
||||
|
13. Do you feel anxious, nervous, or jittery? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
14. Do you feel upset? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
15. How restful was your sleep last night? |
||||||||
|
0 very restful |
1 |
2 |
3 |
4 not at all |
||||
|
16. Do you feel weak? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
17. Do you think you had enough sleep last night? |
||||||||
|
0 yes, very much so |
1 |
2 |
3 |
4 not at all |
||||
|
18. Do you have any visual disturbances (sensitivity to light, blurred vision)? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very sensitive to light, blurred vision |
||||
|
19. Are you fearful? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
20. Have you been worrying about possible misfortunes lately? |
||||||||
|
0 not at all |
1 |
2 |
3 |
4 very much so |
||||
|
21. How many hours of sleep do you think you had last night?________ |
||||||||
|
22. How many minutes do you think it took you to fall asleep last night?________ |
||||||||
REFERENCES
Aivazian, G.H. (1964). Clinical evaluation of diazepam. Diseases of the Nervous System. 25, 491-496.
Albeck, J.H. (1987). Withdrawal and detoxification from benzodiazepine dependence: A potential role for clonazepam. Journal of Clinical Psychiatry, 48(supplement 10), 48-49.
Ananth, J. (1983). Abstinence syndrome from therapeutic doses of oxazepam. Canadian Journal of Psychiatry, 28, 592.
Barten, H.H. (1965). Toxic psychosis with transient dysmnestic syndrome following withdrawal from Valium. American Journal of Psychiatry, 121, 1210-1211.
Barton, D.F. (1981). More on lorazepam withdrawal. Drug Intelligence and Clinical Pharmacy, 15, 134.
Browne, J.L., Hauge, K.J. (1986). A review of alprazolam withdrawal. Drug Intelligence and Clinical Pharmacy, 20, 837-841.
Busto, U., Sellers, E.M., Naranjo, C.A., et al. (1986). Withdrawal reaction after long-term use of benzodiazepines. New England Journal of Medicine, 315, 854-859.
Busto, U., Sykora K., Sellers, E.M. (1989) A clinical scale to assess benzodiazepine withdrawal. Journal of Clinical Psychopharmacology, 9:412-6
Chandora, D.B. (1980). Delayed diazepam withdrawal syndrome: A case of auditory and visual hallucinations and seizures. Journal of the Medical Association of Georgia, 69, 769-770.
Clare, A.W. (1971). Diazepam, alcohol, and barbiturate abuse. British Medical Journal, 4, 340.
de Bard, M.L. (1979). Diazepam withdrawal syndrome: A case with psychosis, seizures and coma. American Journal of Psychiatry 136, 104-105.
de la Fuente, J.R., Rosenbaum, A.H., Martin, H.R., et al. (1980). Lorazepam-related withdrawal seizures. Mayo Clinic Proceedings 55, 190-92.
deBuck, R. (1973). Clinical experience with lorazepam in the treatment of neurotic patients. Current Medical Research and Opinion 1, 291-295.
Einarson, T.R. (1980). Lorazepam withdrawal seizures. Lancet 1, 151.
Einarson, T.R. (1981). Oxazepam withdrawal convulsions. Drug Intelligence and Clinical Pharmacy 15, 487-88.
Ewing, J.A., Bakewell, W.E. (1967). Diagnosis and management of depressant drug dependence. American Journal of Psychiatry 123, 909-917.
Fontaine, R., Chouinard, G., Annable, L. (1984). Rebound anxiety in anxious patients after abrupt withdrawal of benzodiazepine treatment. American Journal of Psychiatry, 141, 848-852.
Fraser, H.F., Wiler, A., Essig, C.F., et al. (1958). Degree of physical dependence induced by secobarbital or pentobarbital. Journal of the American Medical Association, 166, 126-129.
Fruensgaard, K. (1976). Withdrawal psychosis: A study of 30 consecutive cases. Acta Psychiatrica Scandinavica, 53, 105-118.
Fyer, A.J., Liebowitz, M.R., Gorman, J.M. (1987). Discontinuation of alprazolam treatment in panic patients. American Journal of Psychiatry, 144, 303-308.
Goodman W.K., et al (1986). Ineffectiveness of clonidine in the treatment of the benzodiazepine withdrawal syndrome: Report of three cases. American Journal of Psychiatry, 143, 900-903.
Hallstrom, C., Lader, M. (1981). Benzodiazepine withdrawal phenomena. International Pharmacopsychiatry, 16, 235-244.
Harrison, M., Busto, U., Naranjo, C.A., et al. (1984). Diazepam tapering in detoxification for high-dose benzodiazepine abuse. Clinical Pharmacology and Therapeutics, 36, 527-533.
Heritch, A.J., Capwell, R., Roy-Byrne, P.P. (1987). A case of psychosis and delirium following withdrawal from triazolam. Journal of Clinical Psychiatry, 48, 168-169.
Herman, J.B., Rosenbaum, J.R., Brotman, A.W. (1987). The alprazolam to clonazepam switch for the treatment of panic disorder. Journal of Clinical Psychopharmacology, 7, 175-178.
Hollister, L.E., Motzenbecker, F.P., Degan, R.O. (1961). Withdrawal reactions from chlordiazepoxide. Psychopharmacologia 2, 63-68.
Hollister, L.E., Bennett, J.L., Kimbell, I., et al (1963). Diazepam in newly admitted schizophrenics. Disease of the Nervous System 24, 746-748.
Howe, J.G. (1980). Lorazepam withdrawal seizures. British Medical Journal 280, 1163-1164.
Isbell, H. (1950). Manifestations and treatment of addiction to narcotic drugs and barbiturates. Medical Clinics of North America, 34, 425-438.
Jeffries, J.J., Rosenblat, H. (1983). Diazepam withdrawal. Canadian Journal of Psychiatry 28, 235.
Kahn, A., Joyce, P., Jones, A.V. (1980). Benzodiazepine withdrawal syndromes. New Zealand Medical Journal 92, 94-96.
Kesharan, M.S., Crammer, J.L. (1985). Clonidine in benzodiazepine withdrawal. Lancet 1, 1325-1326.
Klein, E., Uhde, T.W., Post, R.M. (1986). Preliminary evidence for the utility of carbamazepine in alprazolam withdrawal. American Journal of Psychiatry, 143, 235-236.
Lader, M., Olajide, K.D. (1987). A comparison of buspirone and placebo in relieving benzodiazepine withdrawal symptoms. Journal of Clinical Psychopharmacology, 7, 11-15.
Laughren, T.P., Battey, Y., Greenblatt, D.J., et al. (1982). A controlled trial of diazepam withdrawal in chronically anxious outpatients. Acta Psychiatrica Scandinavica, 65, 171-179.
Lennane, K.J. (1982a). Oxazepam withdrawal syndrome. Medical Journal of Australia, 1, 287-88.
Lennane, K.J. (1982b). Oxazepam withdrawal syndrome. Medical Journal of Australia 1, 545.
Levy, A.B. (1984). Delirium and seizures due to abrupt alprazolam withdrawal: case report. Journal of Clinical Psychiatry 45, 38-9.
Moore, C. Oxazepam withdrawal syndrome. Medical Journal of Australia 2, 220.
Morgan, K., Oswald, I. (1982). Anxiety caused by a short-life hypnotic. British Journal of Psychiatry, 284, 942.
Noyes, R., Clancy, J., Coryell, W.H., et al. (1985). A withdrawal syndrome after abrupt discontinuation of alprazolam. American Journal of Psychiatry, 142, 114-116.
Noyes, R., Garvey, M.J., Cook, B.L. et al. (1988). Benzodiazepine withdrawal: a review of the evidence. Journal of Clinical Psychiatry, 49, 382-389.
Olajide, D., Lader, M. (1984). Depression following withdrawal from long-term benzodiazepine use: a report of four cases. Psychological Medicine, 14, 937-940.
Perry, P.J., Stambaugh, R.L., Tsuang, M.T., et al. (1981). Sedative- hypnotic tolerance testing and withdrawal comparing diazepam to barbiturates. Journal of Clinical Psychopharmacology, 1, 289-96.
Perry, P.J., Alexander, B. (1986). Sedative/hypnotic dependence: patient stabilization, tolerance testing and withdrawal. Drug Intelligence and Clinical Pharmacy, 20, 532-537.
Petursson, H., Lader, M.H. (1981). Withdrawal from long-term benzodiazepine treatment. British Medical Journal, 283, 643-645.
Preskorn, S.H., Denner, L.J. (1977). Benzodiazepines and withdrawal psychosis: Report of three cases. Journal of the American Medical Association, 237, 36-38.
Quitkin, F.J., Rifkin, A., Kaplan, J., et al. (1972). Phobic anxiety syndrome complicated by drug dependence and addiction. Archives of General Psychiatry, 27, 159-62.
Rickels, K., Case, G., Downing, R.W., et al. (1983). Long-term diazepam therapy and clinical outcome. Journal of the American Medical Association 250, 767-771.
Rickels, K., Case, W.G., Schweizer, E.E., et al. (1986). Low-dose dependence in chronic benzodiazepine users: A preliminary report on 119 patients. Psychopharmacology Bulletin, 22, 407-415.
Rickels, K., Schweizer, E.E., Case, G., et al. (1990). Long-term therapeutic use of benzodiazepines. I. Effects of abrupt discontinuation. Archives of General Psychiatry 47, 899-907.
Ries, R.K., Roy-Byrne, P.P., Ward, N.G., et al. (1989). Carbamazepine treatment for benzodiazepine withdrawal. American Journal of Psychiatry, 146, 536-537.
Rifkin A., Quitkin F., Klein, D.F. (1976). Withdrawal reaction to diazepam. Journal of the American Medical Association 236, 2172-2173.
Robinson, G.M., Sellers, E.M. (1982). Diazepam withdrawal seizures. Canadian Medical Association Journal, 126, 944-945.
Rodrigo, E.K., Williams, P. (1986). Frequency of self-reported "anxiolytic withdrawal" symptoms in a group of female students experiencing anxiety. Psychological Medicine, 16, 467-472.
Schneider, L.S., Syapin, P.J., Pawluczyk, S. (1987). Seizure following triazolam withdrawal despite benzodiazepine treatment. Journal of Clinical Psychiatry 48, 418-419.
Schweizer, E., Rickels, K. (1986),. Failure of buspirone to manage benzodiazepine withdrawal. American Journal of Psychiatry, 143, 1590-1592.
Schweizer, E., Case, G., Rickels, K. (1989). Benzodiazepine dependence and withdrawal in elderly patients. Journal of Clinical Psychiatry 146, 529-531.
Schweizer, E., Rickels, K., Case, G. (1991). Carbamazepine treatment in patients discontinuing long-term benzodiazepine therapy: effects on withdrawal severity and outcome. Archives of General Psychiatry 48,448-452.
Sellers, E.M. (1978). Clinical pharmacology and therapeutics of benzodiazepines. Canadian Medical Association Journal 118, 1533-1538.
Simon, P. (1983). Benzodiazepine withdrawal seizures. Biologic Psychiatry 18, 517.
Smith, D.E., Wesson, D.R. (1983). Benzodiazepine dependency syndromes. Journal of Psychoactive Drugs, 15, 85-95.
Tyrer, P., Steinberg, B., Watson, B. (1979). Possible epileptogenic effect of mianserin. Lancet, 2, 798-799.
Tyrer, P. (1980). Lorazepam withdrawal seizures. Lancet 1, 151.
Tyrer, R., Rutherford, D., Huggett, T. (1981). Benzodiazepine withdrawal symptoms and propranolol. Lancet, 1, 520-522.
Tyrer, P., Owen, R., Dawling, S. (1983). Gradual withdrawal of diazepam after long-term therapy. Lancet 1, 1402-1406.
Tyrer, P. (1985). Clinical management of benzodiazepine dependence. British Medical Journal, 291, 1507.
Vinogradov, S., Reiss, A.L., Csernansky, J.G., et al (1986). Clonidine therapy in withdrawal from high dose alprazolam treatment. American Journal of Psychiatry, 143, 1188.
Votolato, N.A., Botcha, K.J., Olson, S.C. (1987). Comment: Alprazolam withdrawal. Drug Intelligence and Clinical Pharmacy, 21, 753-754.
Vyas, I., Carney, M.W.P. Diazepam withdrawal fits. British Medical Jo, 4, 44.
Wikler, A., (1968). Diagnosis and treatment of drug dependence of the barbiturate type. American Journal of Psychiatry, 125, 758-765.
Winokur, A., Rickels, K., Greenblatt, D.J., et al. (1980). Withdrawal reaction from long-term low-dosage administration of diazepam. Archives of General Psychiatry 37, 101-105.
Woody, G.E., O'Brien, C.P., Greenstein, R. Misuse and abuse of diazepam. An increasingly common medical problem. International Journal of the Addictions, 10, 843-48.