Perceived advantages and disadvantages of substance use in a dual diagnosis population with severe mental disorders and severe substance use disorder. Considering the self-medication hypothesis

Abstract

Aim

Based on a large cohort of dual diagnosis patients, the aim of this study was to quantify the patient-perceived problems and advantages of their substance use and relate the quantity of problems to the substance type and psychiatric diagnosis.

Material

Data comes from a naturalistic cohort admitted to an in-patient facility in Denmark specialized in integrated dual diagnosis treatment. We included 1076 patients at their first admission to the facility from 2010 to 2017. Participants completed 607 DrugCheck and 130 DUDIT-E questionnaires.

Method

we analyzed the questionnaires and included admission diagnosis by use of t-test and ANOVA to depict the patterns in substance use in relation to psychiatric diagnosis.

Results

The three most common substance related problems according to the DrugCheck questionnaire were: feeling depressed, financial problems, and losing interest in daily activities. From DUDIT-E, the highest-ranking negative substance related effects were financial ruin, deterioration of health, and problems at work. Effects on social life relationships were also evident with more than 40% of participants. The top three positive substance related effects reported were relaxation, improved sleep, and control over negative emotions. The number of problems listed varied significantly with the type of preferred substance. Patients using pain medication, sedatives, central stimulants, and alcohol reported most problems. Diagnosis did not differentiate the problems experienced. Results partially support the broad self-medication hypothesis for patients with severe mental illness, but also points out that patients are well aware of negative effects.

Keywords:

• Dual diagnosis
• substance use
• integrated treatment
• DUDIT-E
• Drugcheck Questionnaire

Introduction

It is well-established that there is a high level of comorbidity between mental illness and substance use disorder, also referred to as ‘dual diagnosis’ [1–3]. The causal links between the two conditions have been debated since the phenomenon of dual diagnosis came to attention in the 1980s [4]. A highly influential article by [5] suggested four models for the causal relationship between severe mental illness and substance use disorders: Common factor models, secondary substance use disorder models, secondary psychiatric disorder models, and bidirectional models [5]. The models were, however, primarily hypothetical and with vague empirical support [6].

The secondary substance use disorder model often also referred to as the ‘self-medication hypothesis’, has attracted much attention [7–10]. In brief, originally coined by Khantzian in 1985, the hypothesis claimed that people with a specific mental illness preferred specific substances based on the unique pharmacological properties of the drugs (known as the specific self-medication hypothesis) [11]. Khantzian later revised the concept and abandoned the idea of specificity to claim a broader self-medication hypothesis: A wide range of substances were used to relieve a dysphoric state [12]. The role of self-medication in severe mental illness, such as schizophrenia and bipolar disorder, has not been confirmed in naturalistic cohorts. Hence, the aim of achieving relief of symptoms or unpleasant emotions, versus the consequences, also negative, remains understudied. Not much research has been done on which negative effects people with dual diagnosis themselves report, how they perceive the substance related challenges, and certainly not by the time they are admitted to a psychiatric ward; often at a time of severe destabilization since there is need for admission.

Dual diagnosis patients, a not rigidly bordered concept describing the combination of severe mental illness and substance use disorders, are generally understudied, and patient reported data are scarce, as they are most often excluded from both psychiatric, and substance use, research. All data available in cohorts of dual diagnosis patients are therefore essential at this point to decode what strategies could be helpful in this particular patient group, especially when they are confirmed clinical cases as in this study.

This following study has several aims. The first aim is to document which positive and negative effects of substance use are experienced by patients with dual diagnosis based on a large material of questionnaires collected during their first admission to a specialized dual diagnosis in-patient facility. The patients in this study are patients within the spectrum of psychosis and affective disorders, and a small population of emotionally unstable personality disorders, referred to in-patient treatment. The second aim is to investigate the relation between number of reported problems related to the use of substance and the specific substance type or psychiatric diagnosis as well as to investigate the relationship between number of problems reported and number of advantages. The third aim is to use the findings to reflect on the self-medication hypothesis and how this may or may not be useful when treating this type of dual diagnosis patients. It is thus not the main aim for this study to test the self-medication hypothesis against the other three hypothesis mentioned in the beginning.

Methods

Setting and patient population

The analysis is based on questionnaire data from at large cohort of patients with severe mental illness and co-morbid substance disorder admitted to an in-patient facility in Denmark with an integrated dual diagnosis treatment program – the REDDPAC cohort. In a Danish context, in-patient treatment is a proxy for severity as admittance to an in-patient setting only takes place when treatment cannot be carried out successfully in an out-patient setting. To our knowledge this cohort is one of the largest clinical cohorts of dual diagnosis inpatients with both severe mental and substance use disorder in the world [13–15].

The REDDPAC cohort is a naturalistic and prospective cohort [1]. The cohort is made up of patients admitted to a specialized treatment facility. Due to the naturalistic set up, the data collection has not been as rigorous as would have been the case in a designed research project. Patients were admitted to in-patient integrated treatment in a facility dedicated to dual diagnosis patients if remission has not been possible in the primary sector or general psychiatric out-patient settings.

Patients had to have an SUD diagnosis as well as a main diagnosis of schizophrenia spectrum, major affective disorder, or personality disorder with major impact on daily functioning, to be referred to the integrated treatment facility.

The treatment manual involved three months of integrated in-patient treatment (target time), including manualized transdiagnostic group and individual cognitive behavioral therapy sessions, consultations with a physician to assess diagnoses and psychopharmacological treatment, psychological evaluation, and social worker assessment and, if necessary, intervention. The treatment approach was combined psychopharmacological interventions, group sessions of cognitive behavioral therapy, and cognitive milieu therapy (CMT) by a multidisciplinary team. CMT is a structured psychotherapeutic intervention grounded in cognitive behavioral therapy principles [16,17], performed during the full length of the admission, using all possible everyday situations to promote coping strategies and behavioral changes.

During their admission to the treatment facility, the patients completed questionnaires regarding substance and alcohol use, assisted by the staff, all trained to assist as a semi-structured process, allowing the patients to elaborate in cases of doubt. We included data from 2010 to 2017. During this period management decided to use DrugCheck Questionnaire [18], a clinically relevant, but at the time not validated, questionnaire, and later switched to use the DUDIT-E [19], as it became validated.

Patient responses were only included at their first admission (n = 1076), and during the period they answered 607 DrugCheck and 130 DUDIT-E questionnaires.

The approval for collecting the data was given by the Regional Data Protection Agency permit number RHP-2017-019 and the collection of data without consent from patients was approved by the Danish Patient Safety Authority permit number 3-3013-2017/1.

DrugCheck and DUDIT-E questionnaires

As mentioned, at the beginning of an admission, all patients were asked to answer either the DrugCheck questionnaire or the DUDIT-E questionnaire, both designed to quantify and identify substance type and consumption patterns, as well as perceived benefits and downsides of consumption. From 2010 to 2015, the DrugCheck questionnaire was used at the facility, but was replaced from 2016 and onwards with the DUDIT-E questionnaire, as DUDIT-E has become the preferred questionnaire in a Scandinavian context. The two questionnaires have substantial overlap, an overview is provided below (Figure 1).

Figure 1. Comparison DUDIT-E and DrugCheck items on domains affected and type of problems.

From the DrugCheck questionnaire, we used information on substance type use and frequency of usage. To use valuable data collected carefully by highly trained staff, we decided to standardize the responses to gain full knowledge from these patient-reported-data. The responses in this questionnaire were given in free-text form and were thus re-categorized to be comparable to the DUDIT-E questionnaire category by a trained psychiatrist, the first author, who evaluated the value of each individual questionnaire for equivalence to DUDIT-E value, and e.g. decided which frequency category they could be assigned to, and substance type categorization.

From the DrugCheck questionnaire, we also used information from the so called ‘problem list’. We report in this paper on specific problems listed in the ‘problem list’ if the patient answered that this was a problem ‘some’ or ‘a lot of the time’ (options in the DrugCheck questionnaire), indicating severity to the patient. DrugCheck does not, as DUDIT-E does, include a list of positive effects.

Likewise, from the DUDIT-E questionnaire we used information on substance type and frequency of use. For both the DUDIT-E and the DrugCheck questionnaire, the frequency of use-related questions was dichotomized. Patients were only grouped as using a specific substance if the patient used this on a weekly basis, or with a higher frequency than weekly. The DUDIT-E, in addition to highlighting negative effects of substance use, also reports on positive effects, thus both are included in results. The negative effects were only counted if the patient reported these to be present ‘every week or daily/almost daily’ (scale 1) and the positive effects were only vaunted if the patient reports them to be ‘very much or absolutely present’ (scale 2). In Figure 1 the two scales are compared, and the figure illustrates the similarities.

Other key variables included

Diagnosis of psychiatric illness and substance use disorder

We used the main psychiatric diagnosis given in the first treatment plan made by a senior physician in the medical records when available, otherwise we used information from the admission chart. We applied the hierarchical nature of the ICD-10 diagnostic system to categorize the patients. We grouped the patients in the following categories: DF2 (schizophrenia spectrum), DF3 (affective disorder spectrum), DF4 (nervous disorder spectrum), DF6 (personality disorder spectrum), and ‘other’ which also includes the eight patients with a missing diagnosis. With regards to the substance use disorder, we used the ICD-10 diagnosis of DF10-19 (second cipher identifying substance type) to define disorder- and substance type. We only included diagnosis of harmful use of substance (F1x1) or substance dependence (F1x2) (the x denotes the substance type). Other key variables included: sex, age grouped: years 18–29, 30–39, 40–49, and 50 or above, and year of admission grouped in two-year bands.

Analyses

We report on the group of patients reporting ‘a problem’, defined by DrugCheck questionnaire as well as negative and positive effects from DUDIT-E. In addition, we computed the average number of problems reported on the DrugCheck questionnaire and tested whether the number of problems reported differed according to substances used (self-reported), number of substances used (self-reported), substance diagnosis and psychiatric diagnosis. For substance types, we used t-test and tested against not using this specific substance, and for number of substances used, substance use diagnosis and psychiatric diagnosis we used ANOVA and preceded with a t-test testing the individual groups against the remaining if the p-value was below 0.05 or borderline. We further plotted the number of negative effects against number of positive effects reported on the DUDIT for the 124 patients who had filled in both.

Results

Table 1 shows the characteristics of the patient population according to whether they completed DUDIT or the DrugCheck questionnaire or neither. We found no major differences between the three groups except for patients reporting using cannabis more frequently among those who filled in the DUDIT-E compared to the patients who filled in the DrugCheck questionnaire. Furthermore, patients who filled in either of the two questionnaires had longer stay compared to those who did not complete the questionnaires. In all three groups, patients with F2 diagnosis were the most common diagnosis group. Men comprised approximately two thirds, and the most common substance was alcohol.

Table 1. Demographics and descriptive features of the population by DUDIT-E or DrugCheck questionnaire respectively.

		Total	Questionnaire filled in
			None		DrugCheck		DUDIT-E
		N	N	%	N	%	N	%
Total		1076	339		607		130
Main diagnostic group	F2 Schizophrenia, schizotypal and delusional disorders	432	150	44.2	223	36.7	59	45.4
	F3 Mood [affective] disorders	228	65	19.2	138	22.7	25	19.2
	F4 Neurotic, stress-related and somatoform disorders	115	36	10.6	73	12.0	6	4.6
	F6 Disorders of adult personality and behaviour	134	36	10.6	74	12.2	24	18.5
	Other diagnosis	167	52	15.3	99	16.3	16	12.3
Gender	Male	726	236	69.6	403	66.4	87	66.9
	Female	350	103	30.4	204	33.6	43	33.1
Age group	18–29 years	328	106	31.3	177	29.2	45	34.6
	30–39 years	305	90	26.5	185	30.5	30	23.1
	40–49 years	282	90	26.5	159	26.2	33	25.4
	50 Years or above	161	53	15.6	86	14.2	22	16.9
Year of admission	2010–2011	323	140	41.3	183	30.1	.
	2012–2013	294	81	23.9	213	35.1	.
	2014–2015	258	59	17.4	198	32.6
	2016–2017	201	59	17.4	13	2.1	129	99.2
Length of stay	<30 days	235	146	43.1	74	12.2	15	11.5
	30–90 days	296	85	25.1	167	27.5	44	33.8
	90 + days	545	108	31.9	366	60.3	71	54.6
Substances used on a weekly basis according to DrugCheck or DUDIT-E	Alcohol	276	.		217	35.7	59	45.4
	Opiates	81	.		70	11.5	11	8.5
	Cannabis	207	.		150	24.7	57	43.8
	Cocaine	85	.		68	11.2	17	13.1
	Hallucinogenes	3	.		3	0.5	.
	Solvents	6	.		5	0.8
	Sleeping/tranquilizers	214	.		177	29.2	37	28.5
	Pain	86	.		70	11.5	16	12.3
Substance-related diagnosis	Alcohol	548	157	46.3	325	53.5	66	50.8
	Opiates	216	63	18.6	140	23.1	13	10.0
	Cannabis	526	157	46.3	304	50.1	65	50.0
	Hypnotics	193	55	16.2	125	20.6	13	10.0
	Cocaine	195	44	13.0	118	19.4	33	25.4
	Other stimulants	158	53	15.6	89	14.7	16	12.3
	Hallucinogenics	18	9	2.7	9	1.5	.
	Solvents	4	–	–	–	–	–	–
	Multiple	79	26	7.7	43	7.1	10	7.7

We found that the three most common drug related problems reported on the DrugCheck questionnaire (Table 2) by the patients were feeling depressed (64.3%), having financial problems (55.1%), and losing interest in other daily activities (54.5%). However, experiencing problems with family or at home as well as having strange or odd thoughts was also reported by more than 50% of patients. The most prevalent negative drug related effects reported on the DUDIT-E questionnaire (Table 3) were: experiencing financial ruin (67.8%), deterioration of health (52.5%), and problems at work (50.8%). Effects on social well-being and relationship were also evident with more than 40% responding that they had experienced negative consequences related to destruction of family life, passivity, avoiding social contact, and seeing fewer friends. 40% reported that they perceived everything to be chaotic. The main positive effects of using substances (Table 4) included: increased relaxation (67.5%), improved sleep (48.8%), and better control over negative emotions (43.9%). For 30.9% or a third of the sample, the use of the illegal substances made them ‘feel normal’. When plotting the number of negative effects against number of positive there was no linear relationship between these two (reg.coef.= 0.04, p = 0.60) and no relationship between scoring above average in one and scoring above/below average in the other (chisq = 0.56, p = 0.99).

Table 2. Ranking of problems listed on the DrugCheck questionnaire according to percentage answering ‘some’ or ‘a lot’.

N = range 566–585 depending on question (not all questions answered in all cases) ‘Have you, within the last three months….’	Some or a lot (%)
Felt depressed?	64.3
Did (substance) cause any money problems for you?	55.1
Been uninterested in your usual activities?	54.5
Were there problems at home or with your family because of (substance)?	53.9
Had strange thoughts?	53.2
Been suspicious or distrustful of others?	47.6
Did (substance) make you have problems at work, or at school (College/University/training courses)?	41.7
Has (substance) caused any health problems or injuries?	38.0
Did you have any arguments or fights because of (substance)?	29.9
Did you have housing problems because of (substance)?	28.0
Forgotten to take your medication?*	25.9
Have you done anything ‘risky’ or ‘outrageous’ after using (substance) (e.g. driving under the influence; unprotected sex; sharing needles)?	23.1
Has (substance) caused any trouble with the law, or the police?	18.1

*Add-on question specifying for dual diagnosis patients.

Table 3. Ranking of experienced negative consequences reported on the DUDIT-E questionnaire (two different scales) according to most often experienced to the least often experienced.

Question (n = 125–128 depending on question) What is negative for you	Every week or daily or almost daily (%)	Very/absolutely
In the past year, I’ve had trouble at work due to use of substances	50.8
In the past year, I’ve been to the doctor or hospital due to use of substances	14.8
In the past year, I’ve argued or been violent due to use of substances	9.0
In the past year, I’ve had problems with the police due to use of substances	1.6
Do you experience ruined financial situation?		67.8
Do you experience decreased health?		52.5
Do you experience destruction of family life?		50.0
Do you experience passiveness?		49.2
Do you experience difficulty with concentration?		45.8
Do you experience avoiding social contact?		44.9
Do you experience less contact with friends?		44.9
Do you experience that everything is chaotic?		40.7
Do you experience anxiety?		32.2
Do you experience being inconsiderate?		27.1
Do you experience decreased sexual interest?		22.9
Do you experience suicidal thoughts?		17.8
Do you experience headache, nausea, or vomiting?		14.4

Table 4. Ranking of positive effects of using preferred substance reported on the DUDIT-E questionnaire.

Positive effects of using preferred substance (n = 126–129 depending on question)	Very/absolutely (%)
Relaxation	67.5
Improved sleep	48.8
Control of negative emotions (anger depression)	43.9
Feeling happy	37.4
On the substance I function better socially	31.7
Feeling normal	30.9
Decreased pain	30.1
Life without substance is dull	27.6
Increased creativity	26.0
Increased activity (e.g cleaning, work)	26.0
Increased self-confidence	26.0
Getting a sensation everything will be alright	24.4
Feeling strong	23.6
Better contact with others	20.3
On substance I feel part of group	17.1
I love everybody in the world	14.6
I get more out of life	11.4

The number of problems (Figure 2), the average reported was 5.3. The patients with a diagnosis of ‘other disease’ or F4 (anxiety disorder) had the highest number of problems, 6.9. However, there were no significant differences between the main psychiatric diagnosis. Regarding substance use diagnosis patients who had a diagnosis related to using cocaine or central stimulants reported significantly more problems compared to the patients that did not have these diagnoses. The number of problems reported increase with number of different substances being used with patients only reporting using one substance weekly, reporting on average 4.9 problems and patients using more than 4 different averaging above 8 problems reported. We additionally found that patients who used pain medication, sedatives, and central stimulants as well as patients using alcohol frequently reported significantly more problems than patients not using this specific substance. The highest average number of problems, 7, was reported by patients using amphetamine or cocaine. The lowest number, 4.7, by patients who had not answered this question or reported ‘other drugs’ as their main substance.

Figure 2. Average number of problems listed on the Drugcheck questionnaire according to primary diagnosis and substances used weekly. *T-test against the remainder. **Test of trend.

Discussion and implication for practice

This study shows patterns of self-reported positive and negative effects (‘problems’) of using substances in a population of dual diagnosis patients referred to a highly specialized facility to receive in-patient integrated treatment for DD, although naturalistic, these data originates from a group of patients who is most often excluded from both clinical studies in psychiatry and from substance use studies, and can provide information directly relatable to clinical practice.

To our knowledge, this is the first study providing larger-scale questionnaire data from this important and rarely studied patient group. The number of problems listed varies significantly with the type of preferred substance, and number of substances used, but not with the psychiatric diagnosis. The differences in number of problems reported according to specific substances used could be due to characteristics of the specific substance, for instance some substances can be expensive and illegal, as in the case of cocaine [20], or abundant, as in the case of alcohol [21], as well as the different side effects of substances on somatic health [22–24] and mental health e.g. excerbazation of psychosis related to cannabis use [25–27]. Further, the number of problems is higher for individuals using prescription medicine: sleep medication and opioids. This could be due to specific problems related to obtaining or using these substances, for example in relation to communication with health staff and medical doctors, and challenges affecting the clinical process of tapering off such as high pre-tapering anxiety levels [28].

There was no relationship between number of problems and number of positive effects of using substance reported thus patients can experience both that the substance use was creating problems for them, but also experience positive effects, e.g. reduction of stress and improved sleep. In a clinical practice this co-existence of both problems and the experience of positive effects often result in ambivalence or fluctuating motivation towards treatment [29–31]. In an older review of studies of reasons for substance use in psychosis, five main reasons seem to have been identified [6]: (1) Intoxication effects (to get high, to feel good); (2) Social reasons (to facilitate social interactions/to fit in); (3) Dysphoria relief (to relieve anxiety/depression/boredom; to relax); (4) Psychotic symptoms (to alleviate/cope with hallucinations/feelings of suspicion/paranoia); and (5) Medication side effects (cope with) (p 501–502). Our study supports these findings: Some of the most frequent answers in our study were (1) Relaxation; (2) Improve sleep; (3) Control over negative emotions; and (4) Feel normal. However, we do not find ‘to get high’ as a prominent answer in our material – maybe because the respondents in our study are more burdened. The implication for treatment is providing alternative methods/activities that can produce the same or similar positive effects without drug use for instance to activate other areas of interest, motivation for social life, economic independence or other, to promote substance use reduction or discontinuation, rather than focusing on the (already known) negative aspects [32,33].

Further, from clinical practice in psychotherapeutic treatment, mainly cognitive behavioral therapy [34–36] of patients with substance use or dual diagnosis, various reasons are mentioned for the continued use of substances (positive effects), and even though there are as many reasons as there are patients, a certain clinical pattern arises when working clinically with substance use, namely ambivalence [29,37,38]. One could argue that positive effects to some extent support the self-medication hypothesis, though in a broader sense than the originally presented form, namely the first version of a tight relation between a certain substance and certain diagnosis [11].

For instance, from the field of combined posttraumatic stress disorder (PTSD) and substance use disorders it has been discussed whether self-medication can be an explanation of the high rate of substance use problems in groups suffering from PTSD. Two studies of non-clinical populations came to two different results. Haller and Chassin [8] found the strongest support for the self-medication hypothesis among the four hypotheses tested, where Read [10] found that trait vulnerability explained the relationship best. However, one should be careful with the comparison between schizophrenia and bipolar disorder on the one hand and PTSD on the other, as the psychopathology is rather different.

According to our data positive effects can be interpreted as self-medication context, and it seems that symptom treatment is part of the pattern in certain cases, thus lending some support to the broader self-medication hypothesis, that is the vaguer modification that demonstrates the relation between the need for relief of symptoms and intake of substances for that aim. Beneficial effects do pose an important part of substance use to the user [39], it has been initiated for a reason, sometimes for recreational purposes and in other cases due to bothersome feelings, e.g. boredom or loneliness.

For some substances, self-medication might still be a relevant concept, for others further research is needed to clarify the relation between e.g. pleasure or addiction to understand why substance use continues. According to our data a complex condition is at hand, posing ambivalence as the main clinical picture. So, even though DD patients are admitted voluntarily to a ward for integrated treatment, it is not only a matter of just deciding to cease intake of a substance, but also a matter of enduring going through withdrawal, or having lost the one thing that relieves a problem [40].

Our results show that in many cases, patients do acknowledge various problems related to SUD and their current consumption pattern, and that they are in fact closely related to clinical features of dependency according to diagnostic criteria [41]. Interestingly, there is a certain discrepancy between the realization of existing problems, and the experience of negative effects, and then reversely, a report of beneficial effects in the context where the same compound is considered helpful in certain contexts [19,42]. This could support biological theories of habit and repetitive behavior, despite psychological determination [27,43,44]

Acute effects of substance cessation, like withdrawal, are unpleasant for the patient, and require medical attention in many cases, for example alcohol, benzodiazepine, and opioid withdrawal [28,45,46]. But after the acute phase has passed, other emotions and sensations are emerging: Craving [47] and a different view at the world in a non-intoxicated state, are in some cases emotionally distressing, and often the patient is on their own during these phases, often with relapse as a result [48]. Ambivalence towards positive and negative effects of ending consumption is clinically obvious, and can, in essence, affect decision-making [49–51]. The positive effects of substance use will often relate to complicated or problematic aspects of the patient’s life, to which they seek a solution, in line with the broader self-medication hypothesis. Sleep disorders or anxiety may be acutely reduced after intake of relevant substances, that is substances that affect certain transmitter systems and cause sedation, and other ways of managing symptoms may seem inferior to this instant solution. Moreover, even after accepting admission, ambivalence towards cessation or reduction can be present, and must be elucidated and used in the treatment strategy, both therapeutically and guiding choice of psychopharmacological treatment

It is crucial in the treatment of dual diagnosis patients that staff is aware of both the more biological effects of the individual substances and how they influence the mood and cognitive functioning of the patients as well as the presence of ambivalence as a natural constituent of addiction, and awareness of these common traits of appeal of continued substance use. In the implementation of integrated services the complexity of dual diagnosis and the need for the staff to have solid training, these data support that ambivalence is a main target for treatment interventions, and to some extent, for instance implementation of motivational interviewing [52,53] and open dialogue [54,55] are perhaps more constructive in addressing the ambivalence in the individual patient, although these well-established methods may not be as effective in dual diagnosis patients, as solid evidence is lacking.

Study strengths and limitations

The study of this naturalistic cohort has several strengths which include the inclusion of questionnaire data on a large cohort of dual diagnosis patients, who can be difficult to engage in completing questionnaires. Another strength was the staff of the in-patient facility being highly trained, and familiar with addressing substance-related issues, as well as experienced in assisting with filling in the questionnaires, contributing to recruitment and completion of questionnaires in a valid manner.

A limitation of the study is choices of positive and negative responses were limited by the questionnaires. Thus, the full scope breadth of positive and negative aspects of substance use may not have been covered. Likewise, questionnaires have limitations when it comes to capturing the nuanced aspects of human experiences, and a deeper understanding of reasons for using substances should be investigated using more in-depth qualitative studies. Further we did not include secondary psychiatric diagnosis that could have been of relevance, but instead assigned one main diagnosis according to the hierarchical nature of the ICD-10.

In conclusion

This study found that most dual diagnosis patients are fully aware of the negative consequences of substance consumption at the time of admission to an integrated treatment facility, and they have lived experiences with advantages and disadvantages of continued use of substances. Motivation can possibly be found and identified between accepting both the lack of beneficial effects from using a substance and a goal of retrieving beneficial future goals after ceasing consumption. The self-medication hypothesis is not a complete explanation for the continued substance consumption in this particular group of patients. The paradoxes are obvious in the questionnaires, but aspects of relief from symptoms are definitely present in this data and must be addressed when working with the ambivalence towards treatment or cessation of substance use. These data can inform clinicians working with dual diagnosis and nuance the ideation of the self-medication hypothesis in a psychotherapeutic context when working with dual diagnosis in the spectrum of psychosis or affective disorders.

Acknowledgements

The authors wish to acknowledge Jakob Krarup, MD, chief physician.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study received financial support from Helsefonden and Muusfeldt Foundation.