Predictors of attrition during the outpatient detoxification of opiate addicts


Attrition characteristics


Author: Carl D. CHAMBERS , Walter R. CUSKEY,, William F. WIELAND
Pages: 43 to 48
Creation Date: 1970/01/01

Predictors of attrition during the outpatient detoxification of opiate addicts

Director of Research, New York State Narcotic Addiction Control Commission, New York City, Ph.D. Walter R. CUSKEY,
Associate in Community Medicine, University of Pennsylvania, PhiladelphiaM.D. William F. WIELAND
Program Director, Narcotic Addict Rehabilitation Program, West Philadelphia Community Mental Health Consortium, Philadelphia


Both institutional and ambulatory techniques have been utilized in the detoxification of addicts from narcotics. This study was designed to provide empirical data on attrition from an outpatient programme for hard core heroin addicts. The background and treatment characteristics of 86 addict-patients were analysed. Education, marital status, and the abuse of other drugs concurrently with heroin were the only variables analysed to the extent that they determine whether or not the addict remains under treatment until the detoxification process is completed. A multivariant analysis isolated the concurrent abuse of other drugs with heroin as the most potent predictor for remaining in treatment.

Until quite recently the detoxification of opiate addicts in the United States was accomplished only in institutional or inpatient settings. In 1969, Wieland and Chambers [ 17, 18] reported the first results of a large scale attempt to detoxify addicts entirely as outpatients. Controversy about the relative efficacy of these two contrasting approaches has yet to be resolved.

Advocates of the inpatient technique point to the high incidence of concurrent psychiatric illness and the adverse social influences surrounding narcotic addicts as necessitating hospitalization. For example, Hekimian and Gershon [ 9] recently reported that 50 per cent of a sample of 22 heroin addicts at the Bellevue Psychiatric Hospital in New York City suffered from sociopathic personality disorders before addiction occurred and that the incidence of addiction among related family members-a potent recidivist force-was reported in 20 per cent of the cases. They went on to indicate that 50 per cent of their respondents reported that "a search for euphoria" was the primary reason for narcotic use. This was interpreted as evidence of chronic underlying clinical depression. In support of their argument of adverse social influences, the writers indicated that an additional 22 per cent of their subjects reportedly used narcotics because of "influence by friends and environment."

Others [ 4] , however, point out that institutions, at least in the United States, have too often been used as "dumping grounds" for social deviants and other "undesirables" and that utilization of inpatient facilities for the treatment of addiction can represent disengagement both on the part of the community, at the time of hospitalization, and on the part of the hospital staff, at the time of discharge. And despite the necessary preparation during hospitalization to utilize community services, to obtain productive employment, and to build the ego to resist negative social pressures, long-term hospitalization can produce an "institutional personality" less able to face the exigencies of the social setting. Writers such as Freeman and Simmons [ 7] [ 14] have well documented the difficulties in rehabilitating and reintegrating hospitalized psychiatric patients. One writer, Kurland [ 11] [ 12] even suggests that in view of the poor clinical and social results obtained from inpatient treatment, prolonged hospitalization is unwarranted. Results derived from numerous empirical follow-up studies of patients from inpatient treatment facilities would tend to reinforce Kurland's suggestion [ 3] [ 13] [ 15] [ 16] .

At the operational level, a major consideration of outpatient treatment modalities is that they are much less expensive to operate than inpatient units. When organized properly, outpatient clinics with proportionately smaller numbers of professional and clerical staff than required in an inpatient unit would have the capability of dispensing medication, providing counselling and giving medical, vocational and educational services to a larger number of patients on a daily basis [ 17] [ 18] .

The chief benefit of the outpatient modality for the patient is that even during the detoxification process, he can begin to learn new or re-establish old non-addict social patterns. This time can also be utilized to engage in ego-building experiences such as meaningful family and peer relationships and re-establishing himself in educational or occupational roles.

There is, however, a major clinical disadvantage in the outpatient detoxification modality. This modality, more than any other requires that each patient assume an active role and subsequently the largest share of responsibility for staying in treatment. The patient must come to clinic daily to receive medication and he must also follow the ingesting regimen without supervision.

With the increased incidence of opiate addiction in the United States as well as in other parts of the world, the advantages of having reliable outpatient modalities are obvious. At the present time, however, attrition prior to completing treatment occurs in almost one third of the cases [ 17] [ 18] . If the efficacy of outpatient detoxification is to increase, this premature attrition must be decreased.

This particular study was designed to ascertain what attributes of the addict or his background were associated with terminating treatment prior to attaining abstinence. This design presumes that if the attributes are known, clinicians can better predict which of their addict-patients will require a greater concentration of time and skill.

The selection process

Admission to the outpatient clinic was dependent on having been addicted to a narcotic and living within the geographic boundaries defined as the programme's catchment area. No other selection criteria were utilized. Addicts most frequently hear of the clinic's programme from informal communications on the street with the addicts who are already in treatment. Only a small proportion of referrals come from private physicians or agencies. The admission procedure, at the time of this study, consisted of an intake interview with a social worker or ex-addict counsellor. The counsellor talked with the addict and, when possible, with a member of his family, obtained a brief social and drug history and scheduled the patient for a complete physical cheek-up. Patients were then assigned either to a social worker or ex-addict counsellor, depending on which was available. Before the detoxification process with methadone was initiated, the staff psychiatrist evaluated each patient through a personal interview and a review of the social and drug history taken at intake. The psychiatrist then establishes the medicine regimen.

The treatment process

The detoxification attempt is conducted in the following manner. Patients normally begin treatment with methadone tablets at a dosage of 40 mg per day. Occasionally, patients with "smaller habits" start at 20-30 mg per day and patients with "larger habits" start with 50-60 mg per day. In general, however, most patients do well on a dosage of 40 mg per day.

It is usually recommended that each patient take two 20 mg tablets every 12 hours, although some patients prefer 10 mg tablets every 6 hours. New patients report to the clinic five days per week and see on every visit their designated counsellor who may be a social worker or ex-drug addict.

As soon as the patient reflects progress in the area of drug "cheating", criminality and normal social functioning, the daily dosage of methadone is decreased by a 10 mg increment. If the patient continues to show progress at the reduced dosage, the dosage is again reduced and so on until total abstinence of the drug has been maintained. The frequency of clinic visit may also be reduced to a minimum of one per week. In the event of relapse, the dosage and frequency of weekly visits may increase until re-stabilization occurs.

Besides individual counselling, the patients receive vocational and educational counselling, group therapy, and medical services where indicated.

The rate of detoxification is highly individual so that it may take as little as a week or as long as a year or more to complete the process. Treatment relapses are evaluated to determine whether they might benefit from a more intensive treatment, by temporary suspension from treatment, or by transfer to a separate methadone maintenance programme.


To better understand the addict-patients who, in spite of their avowed desire to detoxify, do terminate the detoxification process prior to its completion, two separate statistical analyses were planned. First, as the plan of the study was to ascertain, wherever possible, any significant differences between those who terminated the detoxification process before its completion and those who remained in treatment, a chi-square analysis was planned. A second technique, multi-variant analysis of low frequency events, was utilized with the same dependent variables in an attempt to isolate any predictors of attrition [ 10] .

To accomplish these planned statistical analyses, all addicts admitted to the clinic between 1 August and 1 November, 1968 were defined as the study sample. During this three month period, 86 addicts were admitted into treatment. A cut-off date of March 31, 1969 was chosen. This date provided an 8-month maximum and a 5-month minimum "at risk of attrition" period. Completing or active patients were defined as those who were still in treatment as of the cut-off date. Attriting or inactive patients were those who had terminated the detoxification process "against medical advice." This A.M.A. status was ascribed to any patient who quit coming to the clinic to receive his methadone medication and failed to respond to written or telephone suggestions that he continue in treatment.

Data isolating the variables utilized for these analyses were collected by a social worker at the initial treatment visit or during subsequent routine clinic visits.

Patient characteristics

As a group, the 86 addicts could be characterized as typically being Negro, male, under 30 years of age, school drop-outs, married and legally employed at the time they requested treatment. In contrast to an earlier study (6) of an addict population with comparable attributes, most of these addicts had not become addicted until after the age of 20. Except for their addiction problems, these patients did not have many of the social casualty characteristics typically associated with being hard core heroin addicts in the United States. For example, a majority did not report any prior formal detoxifications, they did not have histories of being on welfare, they had not been abusers of alcohol, they had not been convicted of any crimes, and they were not concurrently abusing other drugs with their heroin.


Characteristics of the addict-patients


Distribution among 86 addict-patients



Percent of total

1. Negro
67 77.9
2. Male
73 84.9
3. <="" />
51 59.3
4. >8 Years Education
64 74.4
5. Employed
47 54.7
6. Married
49 57.0
7. > Age 20 First Addicted
75 87.2
8. No History of Prior Treatment
69 80.2
9. No History of Criminal Conviction
48 55.8
10. No History of Welfare
64 74.4
11. No History of Alcohol Abuse
75 87.2
12. No History of Concurrent Drugs
52 60.5

Attrition characteristics

Of the 86 addict-patients, 59 or 68.6 per cent terminated the detoxification process against medical advice. During the study period, 10.2 per cent of the drop-outs terminated within the first month of treatment. The largest percentage of attrition occurred during the last two months of the study period with 55.9 per cent of all the terminations occurring within this period. Length of treatment ranged from less than one week to more than 30 weeks with a median of 24-26 weeks.


Distribution of A.M.A. terminations by month of attrition


Distribution of attritions

Month of attrition


Per cent

1. August *
6 10.2
2. September *
3 5.1
3. October *
9 15.2
4. November
6 10.2
5. December
6. January
2 3.4
7. February
18 30.5
8. March
15 25.4
59 100.0

* Patients were admitted only during these months.

Chi square comparisons

Only two comparisons of demographic variables produced any significant differences in attrition-having left school prior to completing the eighth grade and being married at the time of admission for treatment. At the conclusion of the study period, 95.5 per cent (N = 21) of the 22 subjects with less than an eighth grade education had terminated A.M.A. This compared with only 59.4 per cent (N = 38) of the 64 subjects with more than that level of education. Attrition among the 49 married patients was 77.6 per cent but only 56.8 per cent among the 37 patients who were not married.

Only one statistical comparison for the social casualty variables was significant. Patients who were abusing another drug or drugs concurrently with their heroin were more likely to remain in treatment.

Multivariant analysis of low frequency events

Separating the patients into groups with either a low or high risk of A.M.A. termination through the multivariant technique, the following comparisons were produced. Those addict-patients with a low risk of A.M.A. termination were more likely to be the older patients averaging about 35 years and were more likely to be the patients who became addicted to heroin later in life having an average addiction onset age of about 27. In comparison, the average age for high risk addict-patients was about 26 and they became addicted at about 21 years of age. The low risk group was more frequently married at the time they began treatment but had completed fewer years of formal education. Having a history of a criminal conviction and concurrently abusing drugs with heroin were more frequently associated with the low risk patients. With the exception of the abusing of other drugs concurrently with heroin, none of the differences are statistically significant.


Comparisons of basic demographic variables by attrition

Basic demographic variables


Per cent of total 86

Per cent attrition

Significant differences

1. Race:
19 22.1 52.6
67 77.9 73.1  
2. Sex:
73 84.9 68.5
13 15.1 69.2  
3. Age:
51 59.3 72.5
35 40.7 62.9  
4. Education:
22 25.6 95.5
64 74.4 59.4
X2 Test Inappropriate
5. Employed at intake:
47 54.7 66.0
39 45.3 71.8  
6. Married at intake:
49 57.0 77.6
X2 = 4.232;
37 43.0 56.8
P = <.05


Comparisons of social casualty variables by attrition

Social casualty variables


Per cent of total 86

Per cent attrition

Significant differences

1. Age first addicted:
11 12.8 72.7
75 87.2 68.0  
2. Criminal conviction:
38 44.2 63.2
48 55.8 72.9  
3. Ever on welfare:
22 25.6 68.2
64 74.4 68.8  
4. Prior treatment:
17 19.8 76.5
69 80.2 66.7  
5. Ever abuse alcohol:
11 12.8 63.6
75 87.2 69.3  
6. Concurrent drug abuse:
34 39.5 35.3
X2 = 28.968;
52 60.5 90.4
P = <.001

To verify the significance of concurrent drug abuse with attrition, a further analysis combining the previously determined [ 1,2,5] inter-actional effects of age, a history of a criminal conviction and the concurrent abuse of drugs with heroin was performed. The influence of this latter factor, concurrent abuse, was clearly demonstrated.

Attrition was found to be the highest where the addict was abusing only one drug, heroin, regardless of age or having a criminal conviction. Conversely, the lowest attrition was found among those addict-patients concurrently abusing drugs with their heroin and this was also the case regardless of age or having a criminal conviction.

At least by these methods of analysis and among those factors analysed, concurrent or multiple drug abuse prior to the beginning of detoxification appears to be the most potent predictor for remaining in treatment until detoxification is realized.


This study was designed to assess the background characteristics of those addict-patients who either remained in or dropped out of an outpatient methadone detoxification treatment programme for hard core heroin addicts. Of 86 heroin addicts admitted for detoxification during the study period, 59 (68.6 per cent) terminated treatment against medical advice and prior to becoming detoxified.

The highest rate of attrition was found to occur between the sixth and eighth months of treatment.

Statistical comparisons between those who terminated A.M.A. and those remaining in treatment produced significant differences in only two demographic areas. First, the greater the amount of formal education, the greater the potential for remaining in treatment. Secondly, those addicts who were married at the time of initiating treatment were more likely to terminate treatment prior to detoxification. This first finding is, of course, compatible with similar studies relevant to the utilization of medical and psychiatric services. The second finding is meaningful only if one acknowledges the pressures on the family man to return to normal productive roles as soon as it is possible. In this case, it would be as soon as his addiction and/or detoxification became manageable.

Statistical comparisons were also made of a set of attributes associated with various states of social casualty. Only one of these comparisons produced a significant difference between those terminating A.M.A. and those remaining in treatment. The concurrent abuse of other drugs with heroin at time of initiating the detoxification process was more frequently associated with remaining in treatment than was the abuse of heroin alone.

A multivariant analysis was accomplished to isolate the predictors of attrition. The association between the number of drugs abused and attrition was reinforced through this level of analysis. To be abusing only heroin was the most potent predictor of attrition among these outpatients attempting to detoxify with the use of methadone. Although this finding cannot, of course, be conclusively interpreted, the authors surmise that addicts who concurrently abuse multiple drugs will correctly perceive of their detoxifications as being more difficult and requiring more time. Heroin-only addicts, on the other hand, would experience fewer physical and psychological changes and would perceive of their detoxifications as being less difficult and requiring less time. It is therefore probable that the heroin-only abuser can more readily ascertain when his addiction and/or detoxification becomes personally manageable and is more likely to judge for himself when treatment should terminate. Further studies, including repetition and comprehensive follow-up studies, are required, however, before complete interpretation is possible.



J. C. Ball, "Two Patterns of Opiate Addiction in the United States", J. of Criminal Law, Criminology and Police Science (1965) 56:203-211.


J. C. Ball, C. D. Chambers and M. J. Ball, "The Association of Marihuana Smoking with Opiate Addiction in the United States", J. of Criminal Law, Criminology and Police Science (1968) 59:171-182.


J. C. Ball and D. O. Pabon, "Locating and Interviewing Narcotic Addicts in Puerto Rico", Sociology and Social Research (1965) 49:401-411.


A. K. Berliner, "Bridging the Gap Between Institution and Community in the Treatment of Narcotic Addicts", Mental Hygiene (1968) 52:263-271.


C. D. Chambers, "Barbiturate-Sedative Abuse: A Study of Prevalence Among Narcotic Abusers", International J. of the Addictions (1969) 4:45-57.


C. D. Chambers, A. D. Moffett and J. P. Jones, "Demographic Factors Associated with Negro Opiate Addiction", International J. of the Addictions (1968) 3:329-343.


M. E. Freeman and O. G. Simmons, The Mental Patient Comes Home, New York, John Wiley & Sons, 1963.


D. L. Gerard and C. Kornetsky, "Adolescent Opiate Addiction: A Study of Control and Addict Subjects", Psychiatric Quarterly (1955) 29:457-486.


L. J. Hekimian and S. Gershon, "Characteristics of Drug Abusers Admitted to a Psychiatric Hospital", J.A.M.A. (1968) 205:125-130.


J. Ipsen, R. S. Paffenbarger, A. I. Wing and F. E. Ingenito, "Single and Multivariate Analysis of Early Predictors o Hypertension". Paper presented at the National Hear Association Meetings, November, 1969. Proceedings are forthcoming.


A. A. Kurland, "The Narcotic Addict: Some Reflections on Treatment", Maryland State Medical Journal (1966) March: 37-39.


A. A. Kurland, L. Wurmser, F. Kerman and R. Kokowski, "Controls in the Treatment of the Narcotic Addict", J. Amer. Psychiatric Assoc. (1966) 122:739-742.


J. A. O'Donnell, "A Follow-Up of Narcotic Addicts-Mortality Relapse and Abstinence", Amer. J. of Orthopsychiatry (1964) 34:948-954.


O. G. Simmons, Work and Mental Illness, New York, John Wiley & Sons, 1965.


G. E. Vaillant, "A 12-Year Follow-Up of New York Narcotic Addicts: III. Some Social and Psychiatric Characteristics", Archives of General Psychiatry (1966) 15:599-608.


G. E. Vaillant, "A 12-Year Follow-Up of New York Narcotic Addicts: IV. Some Characteristics and Determinants of Abstinence", Amer. J. of Psychiatry (1966) 123:573-584.


W. F. Wieland and C. D. Chambers, "Two Methods of Utilizing Methadone in the Outpatient Treatment of Narcotic Addicts", International Journal of the Addictions (1970-Forthcoming). Also presented at the Second Annual Methadone Conference in New York City on October 27, 1969.


W. F. Wieland and C. D. Chambers, "Narcotic Substitution Therapy", International Journal of Clinical Pharmacology Therapy and Toxicology (1970-Forthcoming). Also presented before the American Society of Clinical Pharmacology and Chemotherapy meetings in Philadelphia on November 26, 1969.