Signs Of Drug Abuse
Pretreatment attrition from couple therapy for male drug abusersShalonda Kelly Outpatient settings that treat substance abuse in the United States suffer from high pretreatment dropout rates and low treatment retention rates (1), especially compared to rates for treatment of other problems (2-4). Although between 50% and 85% of people with substance use disorders eventually seek treatment, they wait an average of 10 years or more from the onset of the disorder to initial treatment (5). Entering and engaging in treatment involves a series of decisions to continue (4); treatment outcome studies typically report at least four phases of contact with substance users in order to complete treatment. First, substance abusers must make an initial contact with the program. Second, they must undergo some form of intake assessment. Third, they begin the first treatment session, and fourth, they must complete the required number of sessions that the program offers. In some programs, clients also undergo a mandatory abstinence and/or stabilization period prior to starting treatment [e.g., Ref. (6).]
The bulk of research on attrition from drug abuse treatment focuses on the fourth phase, which involves treatment retention; relatively little information has been gathered regarding clients' difficulties in progressing to the first treatment session (1). Six recent studies have investigated attrition between the initial call and the intake appointment. They reported overall intake attendance ranging from 42% of the callers (7) to 60% (8). Two studies have investigated client dropout from intake to the initiation of substance abuse treatment. For these two studies, 62% to 76% of those who completed intakes initiated treatment within a period of 6 months (4,9). When the rates are combined from these eight studies, it can be expected that for every 100 callers seeking treatment, between 26 and 46 will actually attend the first treatment session. Two additional studies (3,6) systematically tracked cocaine abusers from initial call to treatment entry, and their data revealed even worse rates; among more than 3000 eligible callers seeking treatment across these two studies, only 24% actually entered treatment.
Several conclusions can be drawn about factors associated with pretreatment dropout. Regarding the progression from call to intake attendance, the most consistent finding is that better rates of attendance are associated with fewer delays in providing an appointment (1,7,8,10-12). For both the progression from call to intake and from intake to treatment, mixed findings have been reported regarding the significance of age, gender, race, cohabitation with a partner, education level, and employment status (3,4,6,7,9). In these studies, age, racial group, and some aspect of socioeconomic status appear to be implicated most consistently, such that older, Caucasian, more educated, and employed clients are more likely to enter treatment. Pattern, amount, and type of drug used, as well as history of successful drug treatment are associated with likelihood of entering treatment in complex ways (3,4,6,9). Finally, for users of drugs and alcohol who have completed intake sessions, coercion from employers, perceptions that their alcohol problems are extremely important to treat, and possibly coercion from the legal system may be more important than ultimatums from family and friends in leading them to begin treatment (4,9).
The studies reviewed here focus exclusively on individually oriented treatment. To date, no known studies have examined pretreatment dropout for those interested in couple therapy for substance abuse. Both alcohol and drug abusers in treatment commonly cite their spouses as a primary source of motivation for entering treatment (13,14), and the involvement of spouses and significant others increases the likelihood that participants will enter and complete treatment (2,15,16). In addition, significantly more clients in behavioral couple therapy (BCT) for substance abuse report reductions in substance abuse and improvements in their dyadic adjustment compared to those receiving individual therapy (17). Despite this empirical support, BCT's superior effects have been demonstrated primarily in studies using large numbers of court-referred clients [e.g., Ref. (18)], who may have strong external pressure to engage in treatment (9) or who have been on methadone maintenance (19), which typically yields higher entry and retention rates [e.g., Ref. (20)]. Therefore, there is currently a dearth of information on the rates at which substance abusing clients seek and engage in couple treatment in mostly voluntary samples. Also, most of the studies of pretreatment dropout focus upon users of a limited range of substances such as cocaine only.
In addition to clinical reasons to be concerned about pretreatment dropout, attrition also may compromise treatment research. In an examination of nine National Institute on Drug Abuse (NIDA)-funded clinical trials for substance abuse treatment, participant recruitment was found to be the biggest problem with implementing the studies, to the extent of project failure or seriously compromised findings in more than one trial (21). Knowledge of typical rates of client progression through the pretreatment phases of a study and the factors related to their continuation can help researchers improve their planning to overcome recruitment obstacles (20,21). Such knowledge is crucial for both treatment development studies and randomized clinical trials (21).
THE CURRENT STUDY
The purpose of the current study was to determine which variables were associated with treatment entry for drug users who made a preliminary inquiry about couple therapy for males' substance abuse. It was hypothesized that men who inquired about vs. men who actually initiated couple treatment for substance abuse would differ on demographic, partner, substance use, and access to service variables. Specifically, it was hypothesized that older, Caucasian men of higher socioeconomic status would be more likely to enter treatment than men without these characteristics. Second, it was expected that the men whose female partners called seeking treatment would be just as likely to enter treatment as those who called to seek treatment for themselves. Third, it was expected that the partners of those entering treatment would be more likely to abstain from substances or use them in moderation compared to female partners of those who
did not enter treatment. Fourth, it was hypothesized that a higher percentage of men who had previous substance abuse treatment and who reported greater use of alcohol would enter treatment compared to those who did not enter treatment. Finally, in terms of access-related variables, it was expected that among the eligible couples, those who entered treatment would have had fewer days between the date of their treatment seeking call and the first scheduling of an intake appointment, more external inducements to enter treatment, and a greater likelihood of having been referred to treatment by a clinician versus having been recruited via newspaper advertisements, compared to those who did not enter treatment.
METHOD
Participants
Participants were 120 adults in central New Jersey who made telephone inquiries from March 2000 to June 2001 about an outpatient couple treatment and research program for male drug users and their female partners. Participants were recruited through university-affiliated and community addictions clinics, direct advertising to the community, and word of mouth by local physicians and therapists. Data from eligible callers were used for all analyses in this study; data from participants who did not meet study screening criteria were not included. Telephone screen inclusion criteria were: both partners aged 18-75; the male used drugs within 60 days prior to the initial call; his partner was a female; and the couple was either married, living together for at least 6 months, or separated with hope of reconciliation. In-person intake criteria were: the male met criteria for a current amphetamine, cocaine, sedative, cannabis, non-IV opiate or polysubstance abuse or dependence diagnosis; the male was medically cleared to participate in the study; and both partners consented to baseline assessments, treatment, and follow-up assessments. At the intake, couples were excluded if either partner had a history or showed current signs of psychosis, either partner showed signs of organic brain syndrome, either partner had fears about participating in couple therapy due to severe domestic violence in the past year, or the couple had concurrent involvement in other marital treatment. Because the larger study focused upon treatment development, some changes to the above criteria were made after the second cohort, such that IV-opiate users were included (yet given referrals and encouragement for additional substance use services), and the window of pretreatment drug use was extended to 90 days.
Procedures
The current study is derived from research designed to adapt a cognitive-behavioral model of couple therapy for alcoholism to couples in which the male is abusing or dependent on drugs. Potential clients and their significant others were told that the treatment consisted of up to 24 weekly therapy sessions over a period of 6 months, designed to help the male partner abstain from drug use and to improve the couple relationship. Participants were screened initially by telephone. Interested couples who met the initial screening criteria were mailed questionnaires and scheduled for an in-person clinical intake, during which they were interviewed and given more specifics about the study. Afterwards, they were given a second packet of questionnaires to complete at home and were scheduled for a baseline assessment interview. Subsequent to the baseline assessment, all interested and eligible participants entered a manual-guided cognitive-behavioral couple treatment program for substance abuse. Entry into treatment was defined as attending the first therapy session.
Measures
Telephone Screen
Callers were screened by telephone, using a Telephone Screen form designed for this study to determine initial eligibility criteria and to obtain demographic information (age, education level, employment status, and marital status). Questions also were included about the relationship of the caller to the client, whether or not the woman used drugs or alcohol, whether or not her use was problematic, how often the man used alcohol, the number of years he used drugs, the number of years his drug use has been problematic, and the method by which they learned of the program (referral or advertisement). Finally, the first scheduled date of the intake session was recorded.
Clinical Intake Interview
To assess eligibility for the treatment, intake clinicians used a semistructured intake interview to assess the clients' mental status, medical clearance to participate, and use of concurrent treatment. Level of couple violence was assessed with the Conflict Tactics Scale (22), and alcohol and drug use diagnoses were assessed using the Structured Clinical Interview for Diagnosis [SCID; e.g., Ref. (23)]. Clients also were asked about their income level and ethnicity.
RESULTS
The focus of this study was on pretreatment attrition rates across the telephone screen, clinical intake, and the first therapy session, and on identifying factors that distinguished those who did not enter treatment from those who attended at least one treatment session. Figure 1 presents a flow chart of recruitment and participation across the phases of the study. As shown in Fig. 1, of the 120 callers, 36 (30%) were either ineligible (n = 31) or provided insufficient information to determine eligibility (n = 5). Of the 84 couples that were potentially eligible, 43 (51%) did not attend an intake session because they were not interested in continuing with the study, as evidenced by no-shows and cancellations (n = 12) and failures to schedule the clinical screen (n = 31). Of the remaining 41 couples who completed the clinical screen intake (49% of potentially eligible callers), nine (11% of eligible callers) showed no further interest, two (2%) did not come to the first treatment session, and six (7%) were excluded. Thus, 24 couples (29%) of the 84 potentially eligible callers entered into treatment.
[FIGURE 1 OMITTED]
Table 1 presents the analyses comparing the demographic characteristics of callers, those who attended the intake, and those who entered treatment. Because of the relatively small sample sizes (n = 84, n per group ranges from 17 to 43), effect sizes are presented in addition to significance levels so that potentially important contrasts may be detected. Consistent with the first hypothesis pertaining to participant demographics, men who participated in treatment had significantly higher incomes from employment, t(37)= -1.99, p=0.02, and total household incomes, t(37)= -2.44, p = 0.05, compared to those who only came to the intake session. Education and employment status, two other socioeconomic indicators, were not associated with the probability of progressing to the in-person intake or treatment. In addition, although the finding was nonsignificant, there was a large effect size (0.42) for the contrast between the ages of men who entered treatment and the ages of those who attended an intake session only. Specifically, the men who progressed to treatment were older, consistent with the first hypothesis. Similarly, there was a nonsignificant trend (p=0.08) for the female partners' ages to be associated with initiation of treatment, but large effect sizes characterized the contrasts between the ages of women who attended an intake session and both those who only called (0.57) and those who progressed to treatment (0.70). The women who only attended an intake session were younger on average than women in the other two groups.
Both of the hypotheses regarding partner related variables were supported, as shown in Table 2. Consistent with the second hypothesis, the proportion of men whose female partners called seeking treatment was not significantly different from the proportion of men who called for themselves for any of the three groups. Consistent with the third hypothesis, female partners' use of substances was significantly associated with the probability of progressing from the phone call to intake, and intake to treatment, [chi square](4, n = 80) = 20.48, p [less than or equal to] 0.00. Follow-up Fisher's Exact Tests revealed two significant contrasts: 1) in comparison to the intake only group, female partners of the group who called but never came to an intake session were significantly more likely to abstain from substance use than use moderately, [chi square](1, n = 42) = 4.92, p [less than or equal to] 0.05, or to engage in significantly more problematic than moderate substance use, [chi square](1, n = 25) = 6.77, p [less than or equal to] 0.05, 2) in comparison to the intake only group, female partners who entered treatment were significantly more likely to abstain from use than to use moderately, [chi square](1, n = 39) = 5.59, p [less than or equal to] 0.05.
The fourth hypothesis, that characteristics of the men's substance abuse would be associated with rates of progression to treatment, received marginal support, as shown in Table 3. None of the men's substance use variables were significantly associated with entering treatment, but there was a large effect size (0.39) associated with the contrast between those who only called and those who attended an intake session, with men who only called reporting more years of problem drug use than men who attended an intake. This contrast was consistent with hypothesis number four.
The fifth hypothesis predicted that access-related variables would be associated with rates of progression to treatment. Table 4 presents these data. There were no significant differences among those who called, attended an intake session, or entered treatment in regard to the length of time from call to scheduled date of the intake interview, or external motivation to seek treatment. Still, large effect sizes were associated with the nonsignificant contrast that was contrary to hypothesis five; there were more days on average between the call seeking treatment and the first scheduled intake appointment for those who entered treatment than for those who only called (0.75) or attended the intake session (0.55). Significantly different proportions of the couples who saw an advertisement about the program progressed to treatment compared to those who were referred by local programs and practitioners, [chi square](84) = 7.32, p = 0.03. Follow-up Fisher's Exact Tests revealed significant differences consistent with hypothesis five: for those who only called, very high proportions became aware of the program through advertisements, while close to half of those who entered treatment were referred to the program, [chi square](67) = 6.67, p [less than or equal to] 0.05. A similar trend existed between those who only attended an intake session and those who entered treatment (p [less than or equal to] 0.10). Overall, hypothesis five was supported partially.
Two separate logistic regressions were conducted to examine combinations of variables predicting callers' progression to the in-person intake or to treatment for the eligible participants. As discussed by Cohen (24) the expected medium effect sizes detectable at the 0.05 level require a larger sample size than is available if more than four predictors are used. Because there were 14 potential predictor variables from the telephone screen and five potential predictor variables from the intake interview, hypotheses regarding each of the individual predictors relative to each other could not be fully evaluated. Thus, logistic regression analyses were conducted only with the variables that were significant in the univariate analyses, including the female partner's use of drugs and alcohol and whether or not the couple saw a newspaper advertisement or were referred to the program. Because the chi-square analyses indicated significant differences between the categories of no female use, some female use, and problematic female use of drugs and/or alcohol, all three levels of this variable were considered, and both significant variables were entered simultaneously into the regression predicting intake attendance for the eligible participants. This model was significant, [chi square](3, n = 80) = 13.36, p [less than or equal to] 0.01. As shown in Table 5, when the female partners did not use drugs or alcohol, couples were more than six times as likely to attend the intake session than those whose female partners had problems with drugs and/or alcohol. Similarly, when the women used drugs or alcohol but did not consider their use to be problematic, the couple was more than 13 times as likely to attend an intake session as when the female's use was problematic.
Next, in predicting callers' progression from in-person intake to the first treatment session, significant variables from the univariate analyses were the men's yearly income, their household income, the partner's degree of drug/alcohol use, and whether they were referred to treatment or responded to a newspaper ad. After simultaneously entering these variables into the regression, the overall model achieved significance, [chi square](5, n = 39) = 10.85, p [less than or equal to] 0.05. None of the individual variables in the model were significant, although there was only enough power to detect a large effect when the intake variables derived from the intake and treatment subsamples were included in the model. Thus, another logistic regression was run to predict treatment entry, using only the significant variables available at the telephone screen, so that medium effect sizes could be detected. Significant telephone screen variables included the female partner's use of drugs and alcohol and whether or not the couple saw a newspaper ad or were referred to the program. The overall model achieved significance, [chi square](3, n = 80) = 18.41, p [less than or equal to] 0.00. As shown in Table 5, although the women's use of substances was not significant in predicting entry to treatment when being referred was considered, odds ratios show that couples who responded to a newspaper advertisement about the program were significantly less likely to progress to treatment compared to those who received a referral.
DISCUSSION
Results partially supported predictions regarding the association of demographic, significant other, substance use, and access related variables with pretreatment attrition for couples with a drug abusing male partner. As predicted, men who had higher incomes were more likely to enter into outpatient treatment, and though nonsignificant, large effect sizes indicated that the men who entered into treatment also tended to be older. As predicted, treatment entry was similar for men whose female partners called the program and men who called seeking treatment for themselves. Consistent with the third hypothesis, males who entered treatment were more likely than men who only attended the in-person intake to have female partners that abstained from drug and/or alcohol use. Significance tests failed to support the fourth hypothesis, but large effect sizes suggested that men in the intake only group tended to report more years of problematic substance abuse compared to the telephone only group. Consistent with hypothesis five, those who were referred to the study were more likely to enter treatment than those who responded to newspaper advertisements. Contrary to hypothesis five, large effect sizes supported the finding that those who had more delays between the call and the scheduled appointment tended to enter treatment.
Additional analyses investigated the strength of relationships between phone screen variables and progression to treatment to determine combinations of variables to predict entry into treatment. When the females' use of substances and the method by which the couple learned of the treatment program were considered together, regressions supported hypotheses three and five, that couples whose partners did not use substances or who used in moderation were more likely to attend the intake session, and couples who received referrals were more likely to enter treatment than those who responded to a newspaper ad.
Overall, results largely were consistent with previous findings, and also suggest new ways to consider entry into couple therapy for drug users. Similar to the two large-scale studies that tracked treatment seekers from their initial call to intake to treatment (3,6), only 29% of eligible callers in the current study actually entered treatment and 35% were ineligible for treatment. In the present study, couples entering treatment were more likely to be socially stable; the men were older, had a higher personal and household income, and/or an abstinent or moderate-using female partner. Similarly, although specific predictors vary across studies, the literature commonly reports indices of social stability (e.g., age, indices of socioeconomic status, being in a stable relationship) as predictors of treatment entry [e.g., Ref. (6).] The findings suggest that the female partners may have influenced entry into treatment indirectly by providing stability and support, rather than directly pressuring the men into treatment. Consistent with this interpretation, other studies have found that substance-abusing men in treatment often cite their partners as sources of motivation to change (13,14) but pressure from their female partners is not an explicit motivator for them to enter treatment (4). Moreover, in one study, "warnings by a spouse or another key person," unlike the other highly endorsed reasons for seeking treatment, was not predictive of entering or completing substance abuse treatment (13). Finally, the lack of finding of a relationship between treatment entry and the number of years that the men's use was problematic was consistent with literature indicating that substance abusers who entered treatment reported more consequences from substance use, but not greater chronicity and/or severity of use in comparison to a matched, untreated sample (15).
Data from access-related variables both support and contradict prior findings. First, previous findings that those who are referred are more likely to enter treatment (6) were strongly supported in this study. In contrast to previous findings, large effect size differences suggested that couples who entered treatment may have longer appointment delays in comparison to those who only called or attended an intake session. Other research on substance abusers seeking individual treatment suggests that the largest reductions in pretreatment attrition occur when appointments are made within 24 hours of the initial call to seek treatment (7,12), ostensibly because the clients are in crisis and have temporarily overcome all obstacles to making the first appointment (7). For substance abusing men seeking couple treatment, the female partners may either prolong the sense of crisis or support the men through it, so that briefer appointment delays do not matter. In addition, perhaps longer appointment delays allowed the callers time needed to verify their partners' interest, coordinate both schedules, and work together on increasing the man's commitment to stop using, thereby increasing the likelihood that they entered treatment.
Limitations of the current study should be noted. Across this and similar studies, eligibility requirements vary widely regarding such variables as primary substance(s) used or exclusion criteria, which complicates comparison of findings. Eligibility and characteristics associated with pretreatment attrition also are difficult to determine in studies utilizing data obtained by phone, because some callers only seek information, and others are reluctant to provide personal data at the calling stage. Small sample sizes in this and other substance abuse studies (13,20,21) limit generalizability and stem from difficulties with recruitment, ineligibility, and significant attrition. As a treatment development study, the current sample was limited by design. Thus, for most variables, univariate analyses could only detect large effects (24). Similarly, only the regressions that contained the two significant variables obtained from the telephone screen had adequate power to detect medium effects. Ironically, these limitations make the significant findings that much more important, as do the large effect sizes for some of the contrasts involving nonsignificant variables.
Despite the limitations, the findings suggest implications for both treatment and research. High pretreatment attrition rates suggest the importance of identifying and targeting the characteristics associated with pretreatment attrition. Findings regarding the women's substance use suggest that it may be important for one partner to be more distant from the problem in order for the men to overcome their ambivalence about treatment. Perhaps this is why studies that demonstrate the success of BCT for couples with a male substance abusing partner tend to exclude couples with female partners that have a substance use disorder (18,19). Referral findings have outreach and research design implications. First, substance abuse programs can network with existing agencies that encounter substance abusers in order to increase participation in their programs. One study shows that almost 90% of injection drug users and smokers of crack cocaine use a variety of human services, such as homeless shelters, food pantries, and medical services (25), which in addition to the legal system, are likely organizations from which drug users in need of treatment may be recruited. Second, for the current study, referral sources prescreened the clients so that referred callers all were eligible, which removes problems with eligibility and may increase motivation for BCT for drug abuse. Third, if such screenings were done by providers of individual treatment, potential clients could then benefit from both individual and couple treatment. Currently, only 27% of community-based outpatient programs in the United States actually provide couple-based treatment (26). Given that married or cohabiting couples with a male partner who abuses substances are also a relatively small proportion of treatment seekers (18,19), these programs may welcome the complement of a couple treatment program for male substance abuse, and screen their clients for eligibility in such a program.
Future studies of couple treatment for men who abuse substances should capitalize on findings of pretreatment attrition studies. First, future studies should list eligibility requirements, track pretreatment attrition across each phase of contact with potential clients, and discuss confidentiality early with potential clients in the hopes of obtaining more complete pre-intake information. Second, towards developing larger-scale studies, well-designed outreach components should include plans and provisions for networking with local agencies and service programs, target the female partners of substance abusers, and provide reminder and follow-up calls. Conversely, some use pretreatment attrition results to suggest a different approach, wherein future smaller scale studies with fewer resources make it harder for potential clients to enter treatment, and/or provide treatment role-inductions, such that only those with a demonstrated commitment remain at the point at which the program would begin investing time and resources (12). Finally, meta-analytic studies that can aggregate findings across a number of small studies are warranted to reveal strong and consistent predictors of pretreatment dropout for couples seeking substance abuse treatment.
Table 1. Demographic characteristics of eligible
participants in each phase of the study.
Phone only Intake only
(n = 43) (n = 17)
Age [n = 78; M (SD)]
Men 37.03 (8.09) 35.53 (7.55)
Women partners 36.68 (7.55) 32.06 (7.16)
Education level (n = 70)
Less than high school 3.4% 11.8%
High school or GED 89.7% 76.5%
Beyond high school 6.9% 11.8%
Employment status (n = 78)
Full/part time work 65.8% 62.5%
Unemployed 28.9% 31.3%
Other 5.3% 6.3%
Income [n = 24; M (SD)]
Employment N/A $19,550 ($15,300)
Total for household N/A $39,159 ($27,200)
Race [n = 39; Men (Women)]
African-American N/A 26.7% (26.7%)
Caucasian N/A 40.0% (66.7%)
Latino N/A 20.0% (6.7%)
Other N/A 13.3%
Treatment Analysis
(n = 24)
Age [n = 78; M (SD)]
Men 38.88 (8.21) ANOVA
Women partners 37.71 (9.22) ANOVA
Education level (n = 70) [chi square]
Less than high school 12.5%
High school or GED 66.7%
Beyond high school 20.8%
Employment status (n = 78) [chi square]
Full/part time work 66.7%
Unemployed 25.0%
Other 8.3%
Income [n = 24; M (SD)]
Employment $32,667 ($22,400) t-test
Total for household $65,047 ($35,000) t-test
Race [n = 39; Men (Women)] [chi square]
African-American 45.8% (45.8%)
Caucasian 33.3% (41.7%)
Latino 20.8% (12.5%)
Other 0%
Significance Effect
level size
Age [n = 78; M (SD)]
Men NS 0.19/0.23/0.42 (a)
Women partners NS 0.57/0.13/0.70 (a)
Education level (n = 70) NS 0.18
Less than high school
High school or GED
Beyond high school
Employment status (n = 78) NS 0.05
Full/part time work
Unemployed
Other
Income [n = 24; M (SD)]
Employment 0.05 0.47
Total for household 0.02 0.52
Race [n = 39; Men (Women)] NS (NS) 0.33 (0.24)
African-American
Caucasian
Latino
Other
Note: Means and standard deviations or percentages are given
based upon available participant data for each variable, as
provided in the table. Some data were not available for those
who did not progress to the intake session.
(a) Three effect sizes (ES) were calculated for these data.
The first number presents the ES comparing the phone and intake
data, the second ES compares the phone and treatment data,
and the third ES compares the intake and treatment data.
Table 2. Characteristics of female partners of eligible
participants in each phase of the study.
Phone Intake
only only Treatment
(n = 43) (n = 17) (n = 24)
Marital status
(n = 84)
Married 46.5% 47.1% 66.7%
Separated 7.0% 5.9% 4.2%
Cohabiting 34.9% 29.4% 20.8%
Committed 11.6% 17.6% 8.3%
Relationship of caller
to client (n = 83)
Client 45.2% 52.9% 58.3%
Partner 50.0% 47.1% 41.7%
Other 4.8% 0% 0%
Partner's drug/alcohol use
(n = 80)
None 59.0% 47.1% 87.5%
Some 10.3% 41.2% 12.5%
Problematic use 30.8% 11.8% 0%
Significance Effect
Analysis level size
Marital status [chi square] NS 0.14
(n = 84)
Married
Separated
Cohabiting
Committed
Relationship of caller [chi square] NS 0.13
to client (n = 83)
Client
Partner
Other
Partner's drug/alcohol use [chi square] 0.00 0.34
(n = 80)
None
Some
Problematic use
Note: Percentages are given based upon available participant
data for each variable, as provided in the table.
Table 3. Substance related characteristics of
eligible participants in each phase of the study.
Phone Intake
only only
(n = 43) (n = 17)
Alcohol use (n = 74)
None 33.3% 5.9%
Weekly or less 36.4% 41.2%
More than weekly 30.3% 52.9%
Number of years
[n = 73 and 72, respectively;
M (SD))
Client has used drug 17.75 (9.15) 15.35 (8.98)
Use has been problematic 15.21 (23.38) 8.71 (7.41)
Prior drug treatment (n = 79)
No 12.8% 18.8%
Yes 87.2% 81.3%
Positive urine at intake (n = 36)
No N/A 25.0%
Yes N/A 75.0%
Treatment
(n = 24) Analysis
Alcohol use (n = 74) [chi square]
None 29.2%
Weekly or less 50.0%
More than weekly 20.8%
Number of years
[n = 73 and 72, respectively;
M (SD))
Client has used drug 17.83 (9.20) ANOVA
Use has been problematic 10.57 (7.76) ANOVA
Prior drug treatment (n = 79) [chi square]
No 29.2%
Yes 70.8%
Positive urine at intake (n = 36) [chi square]
No 41.7%
Yes 58.3%
Significance Effect
level size
Alcohol use (n = 74) NS 0.22
None
Weekly or less
More than weekly
Number of years
[n = 73 and 72, respectively;
M (SD))
Client has used drug NS 0.26/0.01/0.27 (a)
Use has been problematic NS 0.39/0.28/-0.22 (a)
Prior drug treatment (n = 79) NS 0.18
No
Yes
Positive urine at intake (n = 36) NS 0.16
No
Yes
Note: Means and standard deviations or percentages are given
based upon available participant data for each variable, as
provided in the table. Some data were not available for those
who did not progress to the intake session.
(a) Three effect sizes (ES) were calculated for these data.
The first number presents the ES comparing the phone and intake
data, the second ES compares the phone and treatment data, and
the third ES compares the intake and treatment data.
Table 4. Variables related to treatment access for
eligible participants in each phase of the study.
Phone Intake
only only
(n = 43) (n = 17)
Days between call and
initial appointment date
[n = 51; M (SD)] 6.25 (4.35) 8.35 (7.11)
External treatment motivation (n = 37)
None N/A 76.9%
Partner mandate N/A 23.1%
Other mandate N/A 0%
How learned of program (n = 84)
Advertisement 83.7% 82.4%
Referral 16.3% 17.6%
Treatment
(n = 24) Analysis
Days between call and
initial appointment date
[n = 51; M (SD)] 14.18 (14.45) ANOVA
External treatment motivation (n = 37) [chi square]
None 50.0%
Partner mandate 41.7%
Other mandate 8.3%
How learned of program (n = 84) [chi square]
Advertisement 54.2%
Referral 45.8%
Significance Effect
level size
Days between call and
initial appointment date
[n = 51; M (SD)] NS 0.20/0.75/
0.55 (a)
External treatment motivation (n = 37) NS 0.28
None
Partner mandate
Other mandate
How learned of program (n = 84) 0.03 0.31
Advertisement
Referral
Note: Means and standard deviations or percentages are given
based upon available participant data for each variable, as
provided in the table. Some data were not available for those
who did not progress to the intake session.
(a) Three effect sizes (ES) were calculated for these data.
The first number presents the ES comparing the phone and intake
data, the second ES compares the phone and treatment data, and
the third ES compares the intake and treatment data.
Table 5. Logistic regressions predicting intake attendance
and treatment entry from significant telephone screen variables.
95% Confidence
Predictor variables Odds ratio interval
Phone screen to intake attendance
Partner's drug/alcohol use
None vs. problematic use 6.51 (b) 1.30-32.58
Some vs. problematic use 13.18 (c) 1.95-89.02
How learned of program
Advertisement vs. referral 0.44 0.14-1.36
Phone screen to treatment entry
Partner's drug/alcohol use
None vs. problematic use 5265.51 (a) 0.00-1.34E + 26
Some vs. problematic use 1963.04 (a) 0.00-5.06E + 25
How learned of program
Advertisement vs. referral 0.27 (b) 0.08-0.83
(a) No couples having females with acknowledged substance
problems entered into treatment (see Table 1). This empty
cell caused the odds of the female having less than
problematic drug/alcohol use as compared to those having
problematic use to be exponentially high, despite the lack
of significance for this variable.
(b) p < 0.05.
(c) p < 0.01.
REFERENCES
(1.) Stasiewicz PR, Stalker R. A comparison of three "interventions" on pretreatment dropout rates in an outpatient substance abuse clinic. Addict Behav 1999; 24(4):579-582.
(2.) Onken LS, Blaine JD, Boren JJ. Treatment for drug addiction: it won't work if they don't receive it. NIDA Res Monogr 1997; 165:1-3.
(3.) Siqueland L, Crits-Christoph P, Frank A, Daley D, Weiss R, Chittams J, Blaine J, Luborsky L. Predictors of dropout from psychosocial treatment of cocaine dependence. Drug Alcohol Depend 1998; 52:1-13.
(4.) Weisner C. Factors affecting the initiation of substance abuse treatment in managed care. Addiction 2001; 96(5):705-716.
(5.) Kessler RC, Aguilar-Gaxiola S, Berglund PA, Caraveo-Anduaga JJ, DeWit DJ, Greenfield SF, Kolody B, Olfson M, Vega WA. Patterns and predictors of treatment seeking after onset of a substance use disorder. Arch Gen Psychiatry 2001; 58:1065-1071.
(6.) Siqueland L, Crits-Christoph P, Gallop B, Gastfriend D, Lis J, Frank A, Griffin M, Blaine J, Luborsky L. Who starts treatment: engaging in the NIDA collaborative cocaine treatment study. Am J Addict 2002; 11:10-23.
(7.) Festinger DS, Lamb RJ, Kountz MR, Kirby KC, Marlowe D. Pretreatment dropout as a function of treatment delay and client variables. Addict Behav 1995; 20(1): 111-115.
(8.) Gariti P, Alterman AI, Holub-Beyer E, Volpicelli JR, Prentice N, O'Brien CP. Effects of an appointment reminder call on patient show rates. J Subst Abuse 1995; 12(3):207-212.
(9.) Hser Y, Maglione M, Polinsky ML, Anglin MD. Predicting drug treatment entry among treatment-seeking individuals. J Subst Abuse 1998; 15(3):213-220.
(10.) Festinger DS, Lamb RJ, Marlowe DB, Kirby KC. From telephone to office: intake attendance as a function of appointment delay. Addict Behav 2002; 27:131-137.
(11.) Gottheil E, Sterling RC, Weinstein SP. Outreach engagement efforts: are they worth the effort? Am J Drug Alcohol Abuse 1997; 23(1):61-66.
(12.) Stark MJ, Campbell BK, Brinkerhoff CV. "Hello, may we help you?": a study of attrition prevention at the time of the first phone contact with substance-abusing clients. Am J Drug Alcohol Abuse 1990; 16(1 & 2):67-76.
(13.) Cunningham JA, Sobell LC, Sobell MB, Gaskin J. Alcohol and drug abusers' reasons for seeking treatment. Addict Behav 1994; 19(6):691-696.
(14.) Steinberg ML, Epstein EE, McCrady BS, Hirsch LS. Sources of motivation in a couples outpatient alcoholism treatment program. Am J Drug Alcohol Abuse 1997; 23(2): 191-205.
(15.) Carroll KM, Rounsaville BJ. Contrast of treatment-seeking and untreated cocaine abusers. Arch Gen Psychiatry 1992; 49(6):464-471.
(16.) Grella C, Joshi V. Gender differences in drug treatment careers among clients in the National Drug Abuse Treatment Outcome study. Am J Drug Alcohol Abuse 1991; 25:385-406.
(17.) Fals-Stewart W, O'Farrell T, Feehan M, Birchler GR, Tiller S, McFarlin SK. Behavioral couples therapy versus individual-based treatment for male substance-abusing patients: an evaluation of significant individual change and comparison of improvement rates. J Subst Abuse 2000; 18:249-254.
(18.) Fals-Stewart W, Birchler GR, O'Farrell T. Behavioral couples therapy for male substance-abusing patients: effects on relationship adjustment and drug-using behavior. J Consult Clin 1996; 64(5):959-972.
(19.) Fals-Stewart W, O'Farrell T, Birchler GR. Behavioral couples therapy for male methadone maintenance patients: effects on drug-using behavior and relationship adjustment. Behav Ther 2001; 32:391-411.
(20.) Baekeland F, Lundwall L. Dropping out of treatment: a critical review. Psychol Bull. 1975; 82(5):738-783.
(21.) Ashery RS, McAuliffe WE. Implementation issues and techniques in randomized trials of outpatient psychosocial treatments for drug abusers: recruitment of subjects. Am J Drug Alcohol Abuse 1992; 18(3):305-329.
(22.) Straus MA. Measuring intrafamily conflict and violence: the conflict tactics scales. J Marriage 1979; 41(4):75-88.
(23.) Kranzler HR, Kadden RM, Babor TF, Tennen H. Validity of the SCID in substance abuse patients. Addiction 1996; 91:859-868.
(24.) Cohen J. A power primer. Psychol Bull. 1992; 112(1):155-159.
(25.) Falck RS, Ashery RS, Carlson RG, Wang J. Injection drug users, crack smokers, and the use of human services. Soc Work Res 1995; 19(3):164-173.
(26.) Fals-Stewart W, Birchler GR. A national survey of the use of couples therapy in substance abuse treatment. J Subst Abuse 2001; 20(4):277-283.
Shalonda Kelly, * Elizabeth E. Epstein, and Barbara S. McCrady
* Correspondence: Shalonda Kelly, Graduate School of Applied and Professional Psychology, Rutgers, The State University of New Jersey, 152 Frelinghuysen Road, Piscataway, NJ 08854-8085, USA; Fax: (732) 445-4888; E-mail: skelly@rci. rutgers.edu.
COPYRIGHT 2004 Taylor & Francis Ltd.
COPYRIGHT 2004 Gale Group
|
