Buprenorphine

INTRODUCTION

Cocaine use in methadone maintenance programs continues to be a prevalent problem and an important factor interfering with successful treatment for opiate dependence (1). Opiate-dependent patients who used cocaine before treatment entry were three times more likely to drop out during the first month of methadone maintenance than patients not testing positive for cocaine (2). This combined dependence (cocaine and opiate) is complicated by the fact that depression is prevalent within this population and that untreated depression itself leads to poorer outcomes in substance abuse treatment programs (3). For instance, the lifetime prevalence of depression in drug-dependent patients attending treatment has been reported as 24% for major depression (4). Rates of major depressive disorder in clinical studies for stimulant-dependent patients have been reported to be 32%, compared with only 7% in a community sample (5,6). In addition, it appears that individuals with more intense depressive symptoms may experience an enhanced euphoric effect after cocaine administration (7,8). Thus, the higher risk of relapse in depressed patients may reflect not only the self-medication of their depressive symptoms but also this enhanced euphoria.

Although initial approaches using desipramine (DMI) to treat cocaine abuse appeared promising (9), later studies conducted with methadone patients have suggested otherwise (10,11). In addition to the well-established antidepressant effects of DMI, a human laboratory study showed that desipramine's anticraving effects after cocaine administration (12) make it a potentially appropriate intervention for depressed cocaine-dependent patients.

Treatment studies of the efficacy of DMI in cocaine-using patients with comorbid depression have had varied success. Ziedonis and Kosten (13) found that 60% of depressed methadone-maintained patients treated with DMI attained sustained abstinence, but only 35% of the nondepressed patients treated with DMI achieved sustained abstinence. Another study (14) found that DMI significantly reduced depression measured by Hamilton scores during cocaine withdrawal but the study did not evaluate cocaine use with urine toxicology. When comparing the efficacy of desipramine with psychotherapy interventions (relapse prevention and clinical management), Carroll et al. (15) found no significant differences between these interventions, but showed that depressed subjects had a greater reduction in cocaine use than nondepressed subjects. More recently however, Kosten et al. (16) found that not depressed cocaine-abusing patients on buprenorphine and DMI showed a greater increase in opiate-free urines than those who were depressed. Thus, depression has not been a consistent predictor of good treatment response to DMI in these drug abusers.

Despite the advances in medication developments for treatment of addictive behaviors, combining pharmacotherapy with some form of behavioral treatments offers the most promise (15). In this context, contingency management (CM) procedures, which are based on a behavioral perspective of drug abuse treatment, have had success in facilitating abstinence from cocaine alone, and from both opiates and cocaine in cocaine-abusing methadone-maintained patients (17-19). Contingency management also has been found to increase cocaine-free urines in cocaine-abusing patients that were maintained on the partial agonist buprenorphine (20). A recent meta-analysis of 30 contingency management studies in methadone treatment settings showed CM to be efficacious in reducing drug use, but the effect size (d = .25) was small (21). Furthermore, CM has shown some efficacy in treating depression when administered by significant others (22). Thus, depressed cocaine abusing patients may be particularly responsive to contingency management interventions.

In a 12-week, double-blind, placebo-controlled, randomized clinical trial that compared the efficacy of DMI with placebo (PL) and CM with a noncontingency management (NCM) control group, we found that the group that received both DMI and CM showed the greatest decrease in drug use (23). The present study examines whether depression status influenced the efficacy of these interventions. Our primary outcome was use of both cocaine and opiates, as determined by weekly illicit drug urine specimens, because contingency management procedures were targeting urines free of both drugs. Four hypotheses were examined. First, based on a previous finding (16), it was predicted that DMI would improve the rate of drug-free urines more for the never-depressed than for the depressed patients. Second, contingencies were expected to be more effective at reducing drug use among depressed patients (22). Third, DMI was predicted to reduce depressive symptoms. Fourth, we expected to replicate a previous finding that DMI plasma levels would be correlated with reductions in depressive symptoms and associated with an increase in drug-free urines (16).

METHODS

Subjects

The data from 149 subjects (age range, 21-65 years) were selected from the 160 subjects enrolled in the 12-week clinical trial (23). The 149 subjects were predominately Caucasian (56%), males (66%), and unmarried (59%). The mean age was 37 years (SD = 7.9). Subjects were eligible for study participation if they were diagnosed with opiate dependence by a research psychiatrist or had a documented history of treatment in a methadone maintenance program. They also had to have reported use of cocaine (confirmed by urinalysis) within a month before study entry. Recruitees were excluded from the study if they had concurrent drug dependence or current alcohol dependence or any other substance use other than opiates, cocaine, or tobacco. A major cardiovascular, renal, endocrine, or hepatic disorder, a history of psychosis or schizophrenia, or with current suicidal ideation were other exclusion criteria. Pregnant or breast-feeding women were also excluded. Research subjects did not receive payment for their participation and received treatment at no cost.

Research Design

This article compares depressed (n = 53) with never depressed (n = 96) cocaine-abusing opioid-dependent individuals during a 12-week, doubleblind, placebo-controlled, randomized clinical trial. The two depressed groups (MDD and ND) were compared on treatment with DMI vs. PL and on treatment with CM vs. NCM. Cocaine-abusing, opioid-dependent patients were identified as depressed if they met DSM-IV criteria for lifetime diagnosis of major depressive disorder using the Structured Clinical Interview for DSM-IV (SCID) (24).

Participants were initially stabilized on buprenorphine during the first week before being randomly assigned to one of four treatment conditions: 1) DMI (150 rag/day) plus CM; 2) DMI (150 mg/day) without CM; 3) PL with CM; and 4) PL without CM (23). Participants in the CM group received a voucher with increasing monetary value for each consecutive drug-free urine sample. In comparison, individuals assigned to NCM groups also received monetary vouchers, but the vouchers were not contingent on drug-free urine samples. Participants assigned to the yoked-control group also received feedback concerning their urine results to keep the number and duration of contacts with the research staff consistent. Throughout the study, patients provided weekly self-reports concerning their drug use, adverse medication effects, and opiate withdrawal symptoms. Patients participated in weekly group coping skills/relapse prevention therapy and individual therapy sessions.

Medications

Buprenorphine (donated by Reckitt Coleman through the National Institute on Drug Abuse) was administered sublingually. The maintenance dose of buprenorphine hydrochloride (mean = 15 rag/day; range, 8-24 rag/ day) was selected because it was shown to be similar to 65 mg of methadone in a previous clinical trial (25). Desipramine (Marion Merrell Dow, Kansas City, Mo) and placebo desipramine tablets were placed in size 00 blue opaque capsules. The dose of DMI was selected to produce a plasma level range and side effect profile equivalent to that in our previous studies using DMI (26). Blood levels for DMI were obtained 24 hours after the last dose at week 6 and were evaluated by a nonblinded physician for any downward dose adjustment. No dose was increased above 150 mg per day and no patient needed a reduction in dosage.

All medications were administered with nursing supervision 6 days per week (Monday-Saturday) in a double-blind procedure. The dose of buprenorphine was not blind. All patients received tablets to hold under their tongue for 2 minutes and a set of three capsules that may or may not have contained active desipramine. On Saturday, patients were administered a double dose of buprenorphine and given a take-home bottle with capsules of study medications to take on Sunday. The pharmacist was not blind to the procedures. The principal investigator (T.R.K.) kept the medication assignment code in a sealed envelope for access in case of medical emergency.

Measures

Baseline sociodemographic characteristics were collected at intake on forms modified from the NIMH Collaborative Study on Treatment of Depression. The SCID was completed for DSM-IV psychiatric diagnoses, which included major depression and substance use disorders (24). Substance abuse-related problems and psychosocial functioning were assessed at intake and 12 weeks by using the Addiction Severity Index (ASI) (27,28). Depression symptoms were measured with the Center for Epidemiologic Studies Depression Inventory (CES-D) (29) and Hamilton Depression Rating Scale (30) at baseline. The CES-D was also administered monthly. Urine samples were collected thrice weekly (Monday, Wednesday, Friday) for opiates, cocaine metabolite (benzoylecgonine) and other drugs of abuse (e.g., benzodiazepines, barbiturates). The cut-off for a negative urine was less than 300 ng/mL for cocaine and benzodiazepine with their metabolites, and less than 200 ng/mL for opiates. The mean proportion of negative urines was used as the aggregated outcome measure for weekly urine toxicology. Desipramine blood levels were collected at week 6 and were analyzed using reverse-phase high-pressure liquid chromatography (31). Because of sample loss, blood levels of DMI were not available on 30 of 160 subjects, and 20 other subjects had dropped out before the week 6 bloods were drawn.

Data Analysis

The principal analyses used the modified intent-to-treat sample, which included only the 149 subjects who completed the first week of the trial and thereby received at least one dose of DMI. Analyses proceeded in several stages. First, the group diagnosed with major depressive disorder (MDD) was compared with the never depressed subgroup on sociodemographic and baseline clinical characteristics using chi-square tests for categorical variables and t-tests for continuous variables. Significant baseline differences were included as covariates in the analyses described below. Second, survival analysis using Kaplan-Meier estimates (32) of mean retention times were used to compare retention among the groups by using the Mantel-Cox log rank test (33). Third, we compared trial completers with patients who dropped out before the 12th week using chi-square tests for categorical variables and t-tests for continuous variables. Fourth, hierarchical linear models (HLM) (34,35) were conducted to examine whether change in drug use over time varied as a function of the treatment factors, and whether a lifetime diagnosis of MDD mediated the effect of treatment while controlling for baseline differences. Fifth, HLM analyses were conducted to evaluate the change over time of depressive symptoms (as measured by the CES-D), and to determine whether these changes in depression outcomes varied as a function of treatment factors.

RESULTS

Demographics and Baseline Clinical Measures

At baseline, the MDD subsample (n = 53) did not differ from the never depressed (ND) subsample (n = 96) on most sociodemographic characteristics and clinical measures including the severity of illicit drug use. However, the MDD had a larger proportion of females (45% vs. 28%, P = 0.02) and was more likely to be married (13% vs. 7%) than those ND. As expected, the depressive symptoms measured with HAM-D and CES-D were significantly higher among the MDD patients than the ND patients. Furthermore, the depressed and never depressed groups were equally distributed in the four different treatment groups, and they had a comparable baseline mean proportion of cocaine and opiate free urines (0.12 [+ or -] 0.23 vs. 0.11 [+ or -) 0.21, respectively) (Table 1).

Treatment Retention

Survival analysis on the 149 patients showed no significant difference of the mean survival time between the MDD group (mean = 9.41 weeks, 95% CI=8.74, 10.08) and the ND group (mean=9.44 weeks, 95% CI=8.64, 10.26; Kaplan Meier, log rank=0.18, P=0.6). Furthermore, there were no significant differences in retention between the MDD and ND groups that were treated with desipramine or placebo (MDD + DMI = 50%; ND + DMI = 50%; MDD + PL = 48% and ND + PL = 57%; Kaplan-Meier, log rank= 1.4, P=0.6). However, there was a significant retention rate difference between depressed and ND subgroups that were treated with and without contingency management (MDD + CM = 32%; ND + CM = 67%; MDD + NCM = 61% and ND + NCM = 40%, Kaplan-Meier, log rank 9.4, P=0.02). Trial completion rates were 32% for the MDD group that received contingency management and 40% for the ND group that did not receive contingency management compared with an overall 50% rate. Those who did not complete 12 weeks of study participation were significantly younger (35 years vs. 38 years, P = 0.02), but did not differ from study completers on diagnoses of MDD, or on the severity of depressive symptoms, as measured by baseline HAM-D.

Illicit Drug Use Outcomes

As shown in Table 2, all treatment groups had an increase in their mean proportion of cocaine and opiate free-urines from baseline to week 12. The never depressed group treated with desipramine (ND + DMI) had a significant increase in the mean proportion of drug-free urines from 0.12 at baseline to 0.62 at week 12, which represents a 5-fold increase. This increase in the mean proportion of drug-free urines was significantly superior to that of the never depressed group treated with placebo (ND+PL), which had an increase from 0.10 to 0.31 at week 12, and to the depressed group treated with placebo (MDD +PL), which had an increase from 0.13 to 0.52, which represented a 3-fold and 4-fold increase, respectively. Figure 1 shows the change in the mean proportion of drug-free urines over the 12-week trial in which the DMI treated ND group showed more improvement than either of the other groups treated with DMI or placebo (Z=-2.89, P<0.003). Figure 2 illustrates the HML trends for drug free urines over the 12 weeks for the four groups displayed in Figure 1, and clearly shows that the ND+DMI group improved more than the other groups. Thus, treatment with desipramine in the never depressed group appears to significantly facilitate cocaine and opiate free urines.

[FIGURES 1-2 OMITTED]

As shown in Table 2, the depressed group that received contingency management (MDD + CM) had a significant increase in the mean proportion of drug-free urines from 0.11 at baseline to 0.71 at week 12, which represents a 6.7-fold increase from baseline, compared with the depressed and ND group that did not participate in contingency management (MDD+NCM and ND+NCM) (Z=2.44, P<0.01). The MDD+NCM group went from a mean proportion of drug-free urines equal to 0.14 at baseline to that equal to 0.44 at week 12, and the ND + NCM group went from 0.10 at baseline to 0.35 at week 12, which represents a 3.5-fold and a 3. 1-fold increase, respectively. Although the MDD + CM group appeared to have a significant increase of drug-free urines from baseline to week 12, this particular group had a significantly lower retention rate of 31%. Figure 3 shows that both the MDD and ND groups that received contingency management procedures had superior improvement in cocaine and opiate-free urines than those not getting contingency management, but the MDD group showed a greater slope on the HLM plots than the never depressed group (Z = 2.44, P = 0.01; Figure 4). Therefore, contingency management appeared to be more effective in increasing both cocaine and opiate free urines among MDD patients.

[FIGURE 4 OMITTED]

Depression Outcomes

The MDD group had higher average scores than the ND on depressive symptoms at baseline as measured by both the HAM-D and CES-D. Monthly comparisons showed the group with MDD to have significantly higher CES-D scores. In general, all the subjects had a reduction in depressive symptoms over time (Z = - 12.58, P< 0.0001). However, there were no significant treatment effects of DMI or CM in changes of depressive symptoms as measured by CES-D scores when we compared never depressed with depressed subjects (Z = -0.5, P = 0.5).

Desipramine Levels

The treatment groups that received desipramine 150 mg/day had a mean blood level of 125 [+ or -] 135 ng/mL that was determined at week 6 of the clinical trial. There were no significant differences between depressed and ND groups on DMI levels, and there were no correlations of these DMI levels with reduction in depressive symptoms or increase in drugfree urines.

DISCUSSION

These findings support our prediction that the antidepressant desipramine improved the rates of drug-free urines more for the ND than for depressed patients. The depressed patients did poorly, as measured by cocaine and opioid-free urines, when treated with DMI. Also, the contingencies were effective for the depressed and ND patients, with an advantage for the depressed over the ND patients, which supported our initial prediction. Our third and fourth predictions were not supported by our results. Depressive symptoms were not improved more by DMI than by placebo. Also, the DMI plasma levels had a relatively low therapeutic level (mean = 125 ng/mL) and we found no correlation of DMI levels with reductions in depressive symptoms or associated with increases in drug free urines.

The group with MDD may not have decreased their use of opiates and cocaine for several reasons. First, the depressed patients generally have poorer outcomes (3). Second, their depressive symptoms throughout the trial may have been associated with higher levels of cocaine craving after using cocaine, which increased the likelihood of relapsing to using cocaine again (7,8). Third, the mean blood levels of desipramine (125 ng/mL, SD = 135) were in the lower therapeutic antidepressant range of 125-300 ng/mL that is used to treat depression. The dose of DMI 150 mg/day was chosen based on our previous experience with patients maintained on methadone, which is associated with higher levels of DMI because of inhibition of DMI metabolism (36). The low levels of DMI in this study suggest that there was no inhibition of DMI metabolism with buprenorphine. At these low levels, DMI may have produced anticholinergic side effects and stimulated craving among the depressed subjects without sufficient antidepressant effects (37).

Nevertheless, consistent with our previous study (16), the ND group who received DMI significantly increased their cocaine and opiate-free urines over time. As shown in Table 2 and Figure 3, the ND group that was treated with DMI improved more than the other groups. The ND group may have increased their drug-free urines because a lower blood level is required for the anticraving compared to antidepressant effects of DMI (Gawin, personnal communication).

[FIGURE 3 OMITTED]

The results for CM procedures were as expected. This intervention was aimed at both cocaine and opiate-free urines, and it appears that the MDD group responded better to contingencies than the ND group. This finding is limited, however, by the fact that only 31% of these depressed subjects were retained. Although the HLM analyses controlled for missing data, the increased rate of drug-free urines of the depressed over the ND may still be related to the significant attrition in that group. Thus CM may have a modest impact in drug-free urines in the depressed group, but the severe attrition raises questions about the implementation of this intervention in treating depressed patients because of the detrimental effect on retention.

The differences that we detected between the MDD and the ND groups could have arisen from limitations of this study. One methodological limitation of this study is that the treatment groups were not stratified by depression. We addressed this problem by including measures that were significantly different across depressed groups as covariates in HLM analyses, although none of these factors distinguished the four treatment groups. Thus the association between depression and treatment was not biased. A second limitation of our study is the 50% retention rate of the entire sample at 12 weeks that restricted the power to analyze our primary outcome. Because of this attrition, we used HLM as the analytical technique for this type of repeated measures. HLM allows for inclusion of data from patients who do not complete treatment, have varying assessment times and may have different number of assessments per subject.

In summary, our results indicate that cocaine-abusing buprenorphine-maintained subjects with lifetime MDD may be more likely to respond to CM procedures but are less likely to respond to treatment with DMI, whereas those who were ND were significantly more likely improve when treated with DMI. These results did not appear to be determined by a simple attrition, because we adjusted for missing data by using HLM analyses, nor did drug abstinence seem to be influenced by gender differences as we controlled for this factor in our analyses. We also found no effect of DMI or CM in reducing depressive symptoms in depressed opiate-dependent patients. Thus the response to DMI in the ND group suggests that this group may be more likely to respond to the anticraving effects of DMI than the depressed subgroup, and therefore, there may be some therapeutic benefit from adding this medication in treatment of those ND cocaine abusers.

ACKNOWLEDGMENTS

This research was supported by the National Institute on Drug Abuse grants IK23DA14331-01(GG), K05DA00454(TRK), R01-DA05626, P50DA12762 and Veterans Administration Mental Illness Research, Education and Clinical Center (MIRECC).

REFERENCES

(1.) Rosenblum A, Magura S, Palij M, Foote J, Handelsman L, Stimmel B. Enhanced treatment outcomes for cocaine-using methadone patients. Drug Alcohol Depend 5-3-1999; 54(3):207 218.

(2.) DeMaria PA Jr, Sterling R, Weinstein SP. The effect of stimulant and sedative use on treatment outcome of patients admitted to methadone maintenance treatment. Am J Addict 2000; 9(2):145-153.

(3.) Rounsaville B J, Tierney T, Crits-Christoph K, Weissman MM, Kleber HD. Predictors of outcome in treatment of opiate addicts: evidence for the multidimensional nature of addicts' problems. Compr Psychiatry 1982; 23(5):462-478.

(4.) Compton WM III, Cottler LB, Ben Abdallah A, Phelps DL, Spitznagel EL, Horton JC. Substance dependence and other psychiatric disorders among drug dependent subjects: race and gender correlates. Am J Addict 2000; 9(2):113 125.

(5.) Myers JK, Weissman MM, Tischler GL, Holzer CE III, Leaf P J, Orvaschel H, Anthony JC, Boyd JH, Burke JD Jr, Kramer M. Six-month prevalence of psychiatric disorders in three communities 1980 to 1982. Arch Gen Psychiatry 1984; 41(10):959 967.

(6.) Rounsaville B J, Anton SF, Carroll K, Budde D, Prusoff BA, Gawin F. Psychiatric diagnoses of treatment-seeking cocaine abusers. Arch Gen Psychiatry 1991; 48(1):43 51.

(7.) Uslaner J, Kalechstein A, Richter T, Ling W, Newton T. Association of depressive symptoms during abstinence with the subjective high produced by cocaine. Am J Psychiatry 1999; 156(9):1444- 1446.

(8.) Sofuoglu M, Brown S, Babb DA, Hatsukami DK. Depressive symptoms modulate the subjective and physiological response to cocaine in humans. Drug Alcohol Depend 7-1-2001; 63(2): 131-137.

(9.) Gawin FH, Kleber HD, Byck R, Rounsaville B J, Kosten TR, Jatlow PI, Morgan C. Desipramine facilitation of initial cocaine abstinence. Arch Gen Psychiatry 1989; 46(2):117 121.

(10.) Arndt I, Dorozynsky L, Woody G, McLellan AT, O'Brien CP. Desipramine treatment of cocaine abuse in methadone maintenance patients. NIDA Res Monogr 1989; 95:322-323.

(11.) Kosten TR, Schumann B, Wright D, Carney MK, Gawin FH. A preliminary study of desipramine in the treatment of cocaine abuse in methadone maintenance patients. J Clin Psychiatry 1987; 48(11):442 444.

(12.) Kosten T, Gawin FH, Silverman DG, Fleming J, Compton M, Jatlow P, Byck R. Intravenous cocaine challenges during desipramine maintenance. Neuropsychopharmacology 1992; 7(3): 169- 176.

(13.) Ziedonis DM, Kosten TR. Pharmacotherapy improves treatment outcome in depressed cocaine addicts. J Psychoact Drugs 1991; 23(4):417-425.

(14.) Giannini AJ, Malone DA, Giannini MC, Price WA, Loiselle RH. Treatment of depression in chronic cocaine and phencyclidine abuse with desipramine. J Clin Pharmacol 1986; 26(3):211-214.

(15.) Carroll KM. Integrating psychotherapy and pharmacotherapy to improve drug abuse outcomes. [Review] [70 refs.]. Addict Behav 1997; 22(2):233-245.

(16.) Kosten TR, Falcioni J, Oliveto A, Feingold A. Depression predicts higher rates of heroin use on desipramine with buprenorphine than with methadone. Amer J on Addict (in press).

(17.) Silverman K, Higgins ST, Brooner RK, Montoya ID, Cone EJ, Schuster CR, Preston KL. Sustained cocaine abstinence in methadone maintenance patients through voucher-based reinforcement therapy. Arch Gen Psychiatry 1996; 53(5):409 415.

(18.) Silverman K, Chutuape MA, Bigelow GE, Stitzer ML. Voucher-based reinforcement of cocaine abstinence in treatment-resistant methadone patients: effects of reinforcement magnitude. Psychopharmacologia 1999; 146(2):128-138.

(19.) Dallery J, Silverman K, Chutuape MA, Bigelow GE, Stitzer ML. Voucher-based reinforcement of opiate plus cocaine abstinence in treatment-resistant methadone patients: effects of reinforcer magnitude. Exp Clin Psychopharmacol 2001; 9(3):317-325.

(20.) Downey KK, Helmus TC, Schuster CR. Treatment of heroin-dependent poly-drug abusers with contingency management and buprenorphine maintenance. Exp Clin Psychopharmacol 2000; 8(2):176 184.

(21.) Griffith JD, Rowan-Szal GA, Roark RR, Simpson DD. Contingency management in outpatient methadone treatment: a meta-analysis. Drug Alcohol Depend 2-1-2000; 58(1-2):55-66.

(22.) Brannon SE, Nelson RO. Contingency management treatment of outpatient unipolar depression: a comparison of reinforcement and extinction. J Consult Clin Psychol 1987; 55(1): 117 119.

(23.) Kosten TR, Oliveto A, Feingold A, Poling J, Sevarino K, McCanceKatz E, Stine S, Gonzalez G, Gonsai K. Desipramine and contigency management for cocaine and opiate dependence in buprenorphine mantained patients. Drug Alcohol Depend 2003; 70:315-325.

(24.) First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV. Washington, DC: American Psychiatric Association Press, 1995.

(25.) Kosten TR, Schottenfeld R, Ziedonis D, Falcioni J. Buprenorphine versus methadone maintenance for opioid dependence. [see comments]. J Nerv Ment Dis 1993; 181(6):358 364.

(26.) Kosten TR, Morgan CM, Falcione J, Schottenfeld RS. Pharmacotherapy for cocaine-abusing methadone-maintained patients using amantadine or desipramine. Arch Gen Psychiatry 1992; 49(11):894-898.

(27.) McLellan AT, Kushner H, Metzger D, Peters R, Smith I, Grissom G, Pettinati H, Argeriou M. The fifth edition of the addiction severity index. J Subst Abuse Treat 1992; 9(3):199-213.

(28.) McLellan AT, Luborsky L, Woody GE, O'Brien CP. An improved diagnostic evaluation instrument for substance abuse patients. The addiction severity index. J Nerv Ment Dis 1980; 168(1):26-33.

(29.) Radloff LS. A CES-D Scale: a self-reported depression scale for reaserch in the general population. Appl Psychol Meas 1977; 1:385 401.

(30.) Hamilton MA. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960; 23:56-62.

(31.) Proelss HF, Lohmann HJ, Miles DG. High-performance liquid-chromatographic simultaneous determination of commonly used tricyclic antidepressants. Clin Chem 1978; 24(11):1948- 1953.

(32.) Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958:53:457 481.

(33.) Peto R, Peto J. Asymptomatically efficient rank invariant test procedures. J R Stat Soc [A] 1972; 135:185-207.

(34.) Bryk AS, Raudenbush SW. Application of heirarchical linear models to assessing change. Psychol Bull 1987; 101:147 158.

(35.) Gibbons RD, Hedeker D, Elkin I, Waternaux C, Krarmer HC, Greenhouse JB, Shea MT, Imber SD, Sotsky SM, Watkins JT. Some conceptual and statistical issues in analyses of longitudinal psychiatric data. Arch Gen Psychiatry 1993; 50:739-750.

(36.) Kosten TR, Gawin FH, Morgan C, Nelson JC, Jatlow P. Evidence for altered desipramine disposition in methadone-maintained patients treated for cocaine abuse. Am J Drug Alcohol Abuse 1990; 16(3-4): 329 336.

(37.) Weiss RD. Relapse to cocaine abuse after initiating desipramine treatment. JAMA 11-4-1988; 260(17):2545 2546.

Gerardo Gonzalez, M.D.,* Alan Feingold, Ph.D., (#) Alison Oliveto, Ph.D., Kishor Gonsai, M.D., and Thomas R. Kosten, M.D.

Department of Psychiatry, Yale University School of Medicine, and VA Connecticut Healthcare System, West Haven, Connecticut, USA

* Correspondence: Gerardo Gonzalez, M.D., Department of Psychiatry, Yale University School of Medicine, VA Connecticut Healthcare System, 950 Campbell Avenue, West Haven, CT 06516, USA; Fax: (203) 937-3478; E-mail: gerardo. gonzalez-haddad@yale.edu.

(#) Current address: Alan Feingold, Ph.D., Private practice, New York, New York, USA.

COPYRIGHT 2003 Marcel Dekker, Inc.
COPYRIGHT 2003 Gale Group