Long-Term Outcomes of Adolescent Outpatient Treatment for Substance Use Problems: Exploring the Co-occurrence of Mental Health and Substance Use Problems

Data collection was conducted at Maria clinics in 12 Swedish cities. These clinics are specialized outpatient units for adolescents with alcohol and drug problems, operated in collaboration with social services and healthcare. They primarily serve individuals aged 15–21. The clinics provide various forms of individualized and/or manual-based treatment interventions. Additionally, medical and psychiatric interventions are offered through physicians and psychologists. The average intervention lasts for 4–6 months and the staff may include social workers, nurses, psychologists, and doctors (23). Specific details of the interventions that the participants have received during their treatment is not available. What we do know, from not yet published data derived from both survey and focus group interviews with the clinicians at the Maria clinics, is that they are working with multiple therapeutic methods and interventions. Such as, Motivational Interviewing (MI), Relapse Prevention (RP), Hashish withdrawal program (HAP)/Cannabis program for young people (CPU), Cognitive Behavioral Therapy (CBT), Functional family therapy (FFT), Adolescent Community Reinforcement Approach (ACRA).

Out of the adolescents aged 15 and above who initiated intervention at Maria clinics in 2016, 946 were invited to participate in the study, of which 469 chose to take part. For 14 individuals, no registry data were available due to incomplete personal identification numbers or migration from Sweden, while four adolescents had died (two from overdose and two from suicide) during the follow-up period. In total, 451 adolescents participated in the three-year follow-up reported in this study.

The conducted attrition analysis is based on a comparison between the adolescents who participated in the study (451 individuals) and those who declined participation (477 individuals). The study's sample consisted of 29% females, while the proportion of females in the dropout group was 22%. The median age was 17 in both groups. Concerning the primary drug of use, both groups reported similar usage patterns. For adolescents who participated in the study, the proportions were 77% for cannabis, 14% for alcohol, and 9% for other drugs. In the dropout group, the figures were 79% for cannabis, 13% for alcohol, and 8% for other drugs. However, there were significant differences in other substance-related variables, where the participating adolescents generally had more severe SUP compared to the dropout group, including a higher frequency of substance use (49 vs. 41%), a higher degree of polydrug use (38 vs. 26%), and a larger proportion with previous substance use treatment (31 vs. 20%). This result differs from earlier follow-up studies, in which, on the contrary, groups that opted not to participate often had more serious problems (55). It is likely that the differences can be partially explained by the somewhat larger proportion of girls – who generally have higher psychosocial loads – in the study group (see 23).

When the outpatient intervention commenced, the initial data collection began through admission interviews using UngDOK, which has been found to have satisfactory reliability and validity (56). The purpose of the interview method is to map problems, needs, and the current situation to create a basis for assessment, planning, and implementation of the intervention. The interview consists of a total of 75 questions covering the following life areas: housing and livelihood, occupation, alcohol and drugs, treatment history, criminality, upbringing environment, exposure to violence, family and relationships, as well as physical and mental health. These UngDOK-interviews were then used as baseline data in the present study. The ten risk and protective factors used in present study are a construction of the questions in the interview form UngDOK are: 1) Lack of occupation; means that the young person is unemployed and does not have daily employment, 2) Problems at school; problems that have led to deficiencies in attendance, performance and well-being, 3) Placement in foster care/residential home; The young person has previously been subject to community care, 4) Problems in childhood environment; The family/close relatives have had financial problems, substance use problems, mental illness and/or violence problems, 5) Early age at onset of drug and alcohol use; 12 years or younger for alcohol and 13 years or younger for other drugs, 6) Association with criminal or drug-abusing peers; a main interaction with other young people who have problems, crime and/or drug use, 7) Exposed to violence/abuse; earlier in life exposed to violence/abuse of a physical, psychological or sexual nature, 8) Depression; problem with depression in the last 30 days, 9) Aggressive behavior; serious problem with aggressiveness in the last 30 days and, 10) Traumatic events: The adolescent has been through a serious event, accident, violence or disaster that it is still affected through e.g., nightmares, vigilance, avoiding things related to the event.

The measures used to analyze treatment outcomes are based on experiences from previous studies (see, e.g., 57) and aim to provide a reliable picture of the adolescents' development. Information indicating continued issues with alcohol and drug use was gathered from several national registers. The occurrence of substance use disorder diagnosis in outpatient and inpatient care within somatic care, psychiatric care, and addiction care (diagnosis code according to ICD 10) was obtained from the National Patient Register of the Swedish National Board of Health and Welfare. Information about medication for SUP was found in the National Board of Health and Welfare's Prescription Drug Register. The occurrence of involuntary care for SUP was retrieved from the National Board of Health and Welfare's Compulsory Care Register. Substance use-related criminality, such as drug offenses or driving under the influence, was found in the Conviction Register through the Swedish National Council for Crime Prevention. Indications of psychiatric conditions, both outpatient and inpatient psychiatric care, were obtained (through diagnosis code according to ICD 10) from the National Patient Register of the Swedish National Board of Health and Welfare, as well as the prescription of medication for psychiatric issues in the Prescription Drug Register. In addition, to examine outcomes for specific groups, four different categories were created: 1) No indication of substance use problems or mental health problems (no SUP/no MHP), 2) Sole indication of Substance Use Problems (SUP/no MHP), 3) Sole indication of Psychiatric Condition (no SUP/MHP), or 4) Co-occurring condition (SUP/MHP) (see Table 1).

Given that an individual could have appeared in the included registers for SUP and/or MHP, four outcome groups were formed in a first step based on relevant background data. Descriptive statistics and gender distribution of the four outcome groups: SUP and MHP three years after initiated treatment are presented descriptively with percentage distribution (see Table 1). Logistic regressions were conducted to separately describe the predictive value of the risk factors, with indications of psychiatric conditions, substance use, and co-occurring conditions as outcomes. Odds ratios were reported for each independent variable, along with Nagelkerke's pseudo R2 as a measure of explained variance. This was done with and without control for gender, age, and drug use frequency (of the primary drug). Furthermore, logistic regression analysis was conducted in a similar manner to examine the effect of cumulative risk burden at treatment initiation. To reduce the risk of false-positive results (i.e., type 1 error), p-values at 5% should be interpreted with caution. SPSS 26 was used for all statistical analyses.

Table 4 presents specific risk factors predicting psychiatric conditions at the third-year follow-up for the outcome groups MHP and SUP/MHP. The outcome group MHP consists of adolescents who only have indications of MHP at the third-year follow-up (i.e., indications of SUP are not present). Conducted analyses showed that the risk factors school problems and depression had significant predictive values in both Model 1 without control variables and in Model 2 with included control variables. The two control variables, gender (female) and low frequency of use for the primary drug (i.e., one day per week or less) at the start of treatment, also had significant values in Model 2. These mentioned models are significant but have a relatively low proportion of explained variance in the model.

The outcome group SUP/MHP consists of the specific group of adolescents who, at the three-year follow-up, have indications of both SUP and cooccurring MHP conditions. From the conducted analyses, it appears that placement in foster care or an institution, as an individual risk factor, has a predictive effect in Model 3 without control variables and in Model 4, which included control variables. The control variables did not show any significant effect in these analyses.

A high cumulative risk indicates that an adolescent has an accumulation of risk factors, while a lower cumulative risk means that the youth has only a few risk factors. In the following conducted analyses, the cumulative effect is tested in relation to the same outcome groups as in the previous analyses of individual risk factors, see Table 5.

The results also revealed a significant and substantial cumulative effect for the MHP group (indication of MHP) in Model 5 without control variables and in Model 6 with control variables. Furthermore, the analysis showed that the control variables gender (female) and low frequency of use for the primary drug had significant values. Model 6 is significant, but it also has a relatively low value in terms of explained variance. For the outcome group with a co-occurring condition (SUP/MHP) at the three-year follow-up, a significant cumulative effect was found in Model 7 for adolescents with 6–10 risk factors when the treatment contact began. In Model 8, where the control variables were added, there is no significant effect for the cumulative risk groups. The control variables also did not show any significant values.

The findings of the present study indicate that the risk factors diverge when distinguishing adolescents with MHP from those with both MHP and SUP three years post-treatment. Indeed, experience of school problems and depression, as well as being female and reporting low frequency of use for the primary drug increased the likelihood of exhibiting MHP even when SUP are not evident (i.e., no SUP/MHP group). The association between school problems, SUP, and MHP is well-known in previous research (65,66,67,68). Self-reported depression is prevalent across the entire study group and serves as a risk factor that can impact the outcome of continued mental health issues. Depression often has a protracted course of illness and may involve long-term treatment with antidepressant medications (69). This is noteworthy, as previous longitudinal studies have shown that youth with depression are at an increased risk of both suicide and poor labor market integration later in life (70). As the prevalence of mental health issues is generally higher for women than men in Sweden (2,71) it is not surprising that also gender seems to be a significant risk factor for MHP three years after outpatient treatment. Furthermore, the results show that less severe SUP (primary drug use frequency at treatment start) are also linked to the MHP group and continued MHP. This could be discussed considering the self-medication hypothesis (72), which suggests individuals may use substances like alcohol or drugs to alleviate symptoms of mental illness or emotional distress. Interestingly, the only predictive risk factor for co-occurring MHP and SUP three years after outpatient treatment (i.e., SUP/MHP group) was placement in foster care or in an institution. Placement in residential care indicates a significant degree of severity concerning basic care deficiencies and social vulnerability. Children and young placed in residential care are at considerably greater risk of experiencing both mental and social problems later in life (57).

The cumulative effect predicting MHP at the three-year follow-up was evident for the no SUP/MHP group. Six or more risk factors, out of 10 in total, were linked to almost a three-time risk of MHP over time. These results support the cumulative risk assumption with higher psychosocial risk load at treatment start predicts more or continued MHP. This was especially evident for females and, interestingly, higher primary drug use frequency at intake seems to reduce the risk of MHP over time. How this quite surprisingly finding could be explained might relate to the heterogeneity of the group and complexity when SUP, and MHP are studied parallel. The primary drug use frequency might be more linked to a severe SUP and substance dependency solely. More research is indeed needed.

Furthermore, regarding the cumulative effect and co-occurring problems, a corresponding effect is not found for the SUP/MHP group. The lack of cumulative effect on SUP only has been found previously in the project in (73) and even though prior studies on the target population have indicated a cumulative effect for risk factors on continued SUP before (23,74). The weak predictive effect could have various possible explanations, such as the prediction strength diminishing over time or other risk factors not captured during treatment admission being more critical than those analyzed. It may also be that the predictive effect gradually diminishes due to what is known as regression to the mean—in other words, some young individuals might be in the early stages of SUP when treatment begins, while others with more extensive SUP may make greater progress over time (42).

The attrition analyses show that our study group has a larger proportion of girls and a somewhat larger psychosocial burden and serious SUP (23). This is quite uncommon as participants with more serious problems tend to opt out (55). The positive results should be interpreted with some caution since the current registries do not capture adolescents who may “fly under the radar” and possibly continue their problematic substance use without having any contact with the relevant clinics, healthcare, or the legal system. On the other hand, this type of information also includes a degree of overestimation, as occasional visits to, for example, outpatient care may indicate ongoing SUP or MHP, even if the youth seek care for a temporary setback or crisis where certain support has been resumed. Registry data can thus be an indication of both the need for new care for a psychiatric condition and the need for ongoing care, such as follow-up support and/or medication. In this context, the indication can be seen as both positive and negative.

The current study is part of a research project on outpatient treatment of adolescents with SUP in a naturalistic context, where longitudinal follow-up has occurred over three years. Combining data from structured interviews at baseline and multiple registry sources at follow-up provides reliable data and can be an innovative methodology to address the attrition issues that are otherwise common. However, a limitation of registry follow-up is that certain key variables are not found in registries, such as frequencies of continued drug use, something that is difficult to capture in longitudinal studies in general (41). However, a strength is that the adolescents represent Maria clinics from several different cities, contributing to increased generalizability concerning adolescents in Swedish outpatient care.

We see implications of the study for practice and research concerning both preventive efforts and treatment content. As evident from the study's results, Maria clinics appear to address a significant portion of adolescents' needs for assistance and support with their SUP. However, for a relatively large group, MHP persist. Since females are overrepresented among adolescents with ongoing MHP, and generally bear a greater burden of psychosocial risk factors than men, they are likely to need multidimensional and more extensive treatment interventions that span a longer period (cf. 22,75). It is particularly important to consider challenging upbringing conditions, traumas, and the MHP that many of them carry (75,76,77). Adolescents with experience of foster care or institutional care should receive extra attention since their social network is expected to be vulnerable, and the MHP are generally more widespread and complex in this group (57).

Adolescents with co-occurring conditions are at risk of being “passed around” between different healthcare providers without receiving relevant help for their problems (18). This group is particularly highlighted regarding deficiencies in coordination between social services, school health, and healthcare (63). Integrated treatment generally has strong scientific support for addressing this co-occurring issue (14,17,62). However, this type of treatment is offered to an insufficient extent in Sweden (68). The Maria clinics are an example of a facility that can offer integrated treatment. However, there is a need for more facilities of this kind (64,78).

Given that a large proportion of the female patients at the Maria clinics have faced difficulties in school and have been subjects of previous contacts with child and adolescent psychiatry due to MHP, it should also be possible to address and identify SUP in these arenas to offer more relevant support at an earlier stage. Early interventions through student health services, as well as social and educational support in schools to enhance the well-being of children and adolescents and prevent serious mental health issues, are crucial for positive development (79).

It is also crucial that the adolescents are provided with opportunities for social inclusion and support for meaningful engagement in activities during their leisure time, in addition to assistance in addressing both SUP and MHP. This becomes evident since the majority of the youth in present study have or have had issues with school. The transition to adult roles and opportunities for further education or employment, along with changes in social belonging, is also associated with reduced MHP and SUP (19).