How clinicians think about dual disorders: Insights from a case vignette survey

ABSTRACT

Background:

Dual disorders, involving co-occurring mental health and substance use conditions, present diagnostic challenges due to overlapping symptoms.

Aim:

To assess the diagnostic agreement of mental health professionals in evaluating dual diagnosis cases.

Methods:

This Cross-sectional study surveyed 73 mental health professionals using two clinical vignettes that depicted common dual disorder scenarios: psychotic symptoms co-occurring with cannabis use and depressive symptoms co-occurring with alcohol use. Participants chose a diagnosis, rated confidence (1–10), and identified influencing factors, along with alternative differentials.

Results:

For the first case vignette, majority (60.2%) favored schizophrenia with cannabis use over cannabis-induced psychosis, while small number endorsed other psychotic disorders. In the second case vignette, most (89.0%) clinicians preferred major depressive disorder with harmful alcohol use, rather than alcohol-induced mood disorder. Overall confidence agreement was high (average measure ICC = 0.986) despite differences in diagnosis.

Conclusion:

Although clinicians diagnostic reasoning varied but confidence ratings demonstrated excellent inter – rater agreement (Average ICC = 0.986).

INTRODUCTION

Mental diseases (MDs) and substance use disorders (SUDs) are serious public health concerns that impact individuals as well as society at large. The simultaneous existence of at least one MD and at least one SUD is known as dual diagnosis, or co-occurring addiction and mental disorders, and is not specifically classified in diagnostic systems such as the International Classification of Disease-10 (ICD-10).[1]

Dual disorder often present with significant diagnostic challenges like worsening symptoms of one or both conditions, greater illness severity, higher risk of suicide, increased relapse rates, poor quality of life, and reduced treatment adherence.[2] Patients with dual diagnoses are more challenging to treat because of severity of psychological, physical, and social impairments, which increases direct and indirect costs for patients, families, and healthcare systems.[3]

Differentiating between primary mental disease and substance-induced diseases allows clinicians to design focused, effective interventions and better long-term recovery. Misdiagnosis can lead to poor treatment response and raises risk of relapse.[4] Understanding clinicians’ perspectives on diagnostic decision-making can reveal factors influencing variability and improve clinical practice. Therefore, this study aims to assess diagnostic agreement among mental health professionals diagnosing dual disorder cases, and explore reasons that lead to their divergent clinical diagnoses.

METHODOLOGY

The current study focused on medical professionals employed in psychiatric field and used an online, quantitative, cross-sectional design. The Institutional Ethics Committee approved the study. A total of 73 professionals were recruited over a timeframe of 7 days from May 9, 2025 to May 15, 2025. Snowball sampling alongside digital communication tools such as WhatsApp was used. A preliminary section detailing study’s objective and extent, principal investigator’s identity, anticipated length of participation, data confidentiality, absence of foreseeable risks, lack of compensation, and voluntary nature of participation was used to obtain informed consent.

Participants who were actively involved in diagnosing MD and/or SUD, and gave informed consent via online platform to meet the inclusion requirements. The target was minimum of 49 participants. This was based on subsequent assumptions: 30 repetitions for each vignette, an intraclass correlation coefficient of 0.6, precision of 0.1, and a 95% confidence level (i.e. 1–alpha) (https://wnarifin.github.io/ssc/ssicc.html).

Two mental health professionals (MHP) took part in google form trial prior to open survey was made public. Their feedback contributed in making vignettes more realistic and diagnostically ambiguous, along with enhanced clarity and content. The Google Form link was then sent to MHP by WhatsApp without any incentive.

The google form consisted of three sections. In section one (demographics), gender (male or female), years of experience in care delivery (including residency), highest qualifications (MBBS, MD, DM, PhD, etc.), and current employment (government, private practice, private hospital, NGO, etc.) were included. There were two case vignettes in each of Sections two and three (presented in Supplementary material). In each case vignette, two pre-specified differential diagnoses were provided. After reading each case vignette participants were asked to rate their confidence in that diagnosis on a Likert scale of 1 to 10 and select relevant reasons for their differential diagnosis. Respondents were specifically asked if they want to list any other differential diagnoses, they thought were pertinent. In first case vignette, the two pre-specified differentials were “cannabis-induced psychosis” and “schizophrenia with harmful use of cannabis.” In second case vignette, the two pre-specified differentials were “alcohol-induced mood disorder” and “major depressive disorder with harmful use of alcohol.”

The themes of case vignettes were chosen to illustrate dual disorder situations that are clinically significant and commonly encountered. In particular, the two vignettes discussed depressive symptoms co-occurring with alcohol use and psychosis co-occurring with cannabis use. These cases were selected because they reflect scenarios where clinicians commonly face diagnostic ambiguity and must rely heavily on their clinical judgment and confidence.

SPSS version 26.0 was used to analyze responses, which were gathered in an Excel spreadsheet. Depending on how the data was distributed, descriptive statistics were presented as either mean (standard deviation) or median (interquartile range). The level of significance was set at P < 0.05, and appropriate statistical tests were employed based on characteristics of the data. The intraclass correlation coefficient (ICC) was computed to evaluate inter-rater agreement in confidence ratings. Both single-measure ICC (reliability of individual raters) and average-measure ICC (reliability of the mean rating of multiple raters) were reported.

RESULTS

A total of 73 participants took part in this study. Male and female genders were equally represented, and the median duration of experience in mental health care delivery was 4 years. Most of participants were working in government sector. Amongst participants, largest group consisted of those who had a postgraduate degree as their highest qualification.

For the first vignette presented in Table 1, diagnosis of cannabis induced psychosis was considered by 71, and diagnostic confidence ranged from 1 to 10, with median of 5. Temporal association was the most common reason for choosing this diagnosis. Schizophrenia with harmful cannabis use was considered by 73 participants, with range of 1 to 10, and median of 8. Symptomatology was given the most common reason in this group. For 71 participants who provided scores for both diagnoses, cannabis induced psychosis was the favored diagnosis in 21, schizophrenia with harmful cannabis use in 44, and both diagnoses received same score from six participants. Six participants preferred other diagnoses, that included schizophrenia, schizophreniform disorder, acute and transient psychotic disorder, schizoaffective disorder, and other primary psychotic disorders, unspecified.

Table 1.

Responses of 1^st Case Vignette

Clinical Diagnosis	Cannabis induced psychosis (n=71) [n or Median (Q1, Q3)]	Schizophrenia with cannabis harmful use (n=73) [n or Median (Q1, Q3)]
Diagnosis Favored by Participants	21 (29.5%)	44 (60.2%)
Confidence Level (1–10)	5 (2, 8)	8 (5, 9)
Possible reasons for above diagnosis
1. Temporality	60	40
2. Clinical course	34	43
3. Symptomatology	29	56
4. Standardized diagnostic criteria	21	39
5. Family history and other risk factors	8	11
Participants favored Other diagnosisa	6a

^aOther diagnoses included schizophrenia, schizophreniform disorder, acute and transient psychotic disorder, schizoaffective disorder, and other primary psychotic disorders, unspecified

For the second vignette presented in Table 2, alcohol induced mood disorder was considered as diagnosis by 70 participants, with diagnostic confidence ranging from 1 to 10 and median of 2. Temporality was the most common consideration for this diagnosis. Major depressive disorder with harmful use of alcohol was considered by all 73 participants. The confidence range was from 2 to 10. Clinical course was noted as the most common consideration for this diagnosis. Among the participants, who rated on both diagnosis, major depressive disorder and harmful alcohol use was the favored diagnosis in 65, three participants favored alcohol induced mood disorder and two participants gave similar ratings for both diagnoses. Four participants mentioned other diagnoses as a differential, including adjustment disorder or recurrent depressive disorder, alcohol use disorder, and major depressive disorder and alcohol dependence.

Table 2.

Responses of 2^nd Case Vignette

Clinical Diagnosis	Alcohol induced mood disorder (n=70) [n or Median (Q1, Q3)]	Major Depressive Disorder with harmful use of alcohol (n=73) [n or Median (Q1, Q3)]
Diagnosis Favored by Participants	3 (4.2%)	65 (89.0%)
Confidence Level (1–10)	2 (1, 3)	9 (8, 10)
Possible reasons for above diagnosis
1. Temporality	46	52
2. Clinical course	29	54
3. Symptomatology	28	53
4. Standardized diagnostic criteria	18	48
5. Family history and other risk factors	19	45
Participants favored Other diagnosisb	4b

^bOther diagnoses included adjustment disorder or recurrent depressive disorder, alcohol use disorder, and major depressive disorder and alcohol

Inter-rater agreement for diagnostic confidence was assessed using the ICC. For case vignette 1, the single-measure ICC was 0.188 (95% CI: 0.033–0.996), while the average-measure ICC was 0.943 (95% CI: 0.706–1.000). For case vignette 2, the single-measure ICC was 0.847 (95% CI: 0.516–1.000) and the average-measure ICC was 0.997 (95% CI: 0.987–1.000). When confidence ratings across both vignettes were combined, the single-measure ICC was 0.514 (95% CI: 0.244–0.937) and the average-measure ICC was 0.986 (95% CI: 0.957–0.999), indicating high agreement when ratings from multiple clinicians were averaged.

DISCUSSION

This study explored how MHP approach diagnostic reasoning and reflect to their confidence while reviewing clinical vignettes depicting dual disorder cases. Findings showed divergence in clinicians’ diagnostic conclusion, highlighting difficulty in distinguishing between substance-induced psychiatric conditions, and primary psychiatric disorders with comorbid substance use.

In first vignette, participants frequently considered both cannabis-induced psychosis and schizophrenia with cannabis use, but predominantly favored the latter. This finding indicates a diagnostic inclination to interpret psychotic symptoms as a manifestation of a pre-existing mental disorder, with cannabis use viewed as an exacerbating rather than a causative factor. Such reasoning aligns with existing studies indicating that greater weight is given to symptom severity, duration, and longitudinal course over temporal proximity of substance use by clinicians when diagnosing psychosis.[2,5,6] This inclination to under-recognize substance-induced psychosis might be influenced by growing evidence linking cannabis consumption to onset or worsening of schizophrenia, as seen in studies demonstrating cannabis’s ability to produce both positive and negative psychotic symptoms[7] and strong association with emergence of first-episode psychosis.[8,9]

In second vignette, nearly all participants diagnosed major depressive disorder (MDD) alongside harmful alcohol use, while relatively few considered an alcohol-induced mood disorder. The high diagnostic confidence for MDD suggests clinicians viewed the depressive symptoms as primary or likely persistent. This observation echoes with findings from a large national survey from United States showing that mood and anxiety disorders frequently overlap with alcohol use disorder (AUD), with depressive disorders as the most frequent psychiatric comorbidity.[10] Additionally, observational data suggest that depressive symptoms initially seen as substance-induced often get reclassified as primary mood disorders when beyond a period of abstinence highlighting diagnostic complexity.[11] However, other studies indicates that alcohol-induced depression can be just as severe and often goes underdiagnosed, even though it is linked to higher relapse rates and a greater risk of suicide.[12,13]

The inter-rater agreement in confidence ratings was high, particularly when ratings from multiple clinicians were averaged. While the average-measure ICC indicated excellent agreement, the single-measure ICC values were comparatively lower, reflecting variability in confidence ratings among individual clinicians. Such differences are expected in studies involving multiple raters, as averaging ratings tends to reduce individual variability and improves overall reliability.[14]

Key strengths of this study are that it focused on practical diagnostic reasoning. The case vignette method recorded clinicians’ confidence levels and thought processes while standardizing diagnosis. However, some limitations exist. The hypothetical vignettes might not accurately capture intricacy of real-world clinical settings, where diagnosis is influenced by history, follow-up, and collateral information. Generalizability may be limited by selection bias introduced by the cross-sectional design and snowball sampling. Additionally, online survey lacked timing data, consistency checks, and protections against duplicate responses. Longitudinal studies could look at how diagnoses change with treatment and abstinence, and future research should evaluate diagnostic decision-making in clinical practice.

CONCLUSION

In conclusion, mental health professional consistently favored primary psychiatric diagnoses over substance-induced conditions in dual disorder vignettes, exhibiting divergent diagnostic reasoning. Despite varying attributions, strong clinical certainty is highlighted by the excellent inter-rater agreement in confidence ratings.

Acknowledgement Declaration regarding the use of generative AI

The authors attest that there was no use of generative artificial intelligence (AI) technology in the generation of text, figures, or other informational content of this manuscript.

Conflicts of interest

There are no conflicts of interest.

Supplementary Material: Case Vignettes.

1st case vignette

Mr. S, a 24-year-old college student, is brought to the emergency department by his roommates due to increasingly bizarre behaviour. The roommates report that over the past 6 weeks, Mr. S has become progressively paranoid, claiming that his professors are “inserting thoughts” into his mind and that government agencies are monitoring his thoughts through his electronic devices. He has been observed muttering to himself when alone, laughing or crying inappropriately without apparent reason, and expressing beliefs that strangers on the street can “read his thoughts”. When specifically questioned about the timeline of his symptoms, Mr. S hesitantly admits that he began experiencing “odd feelings” and suspiciousness approximately 3 months ago but initially attributed these to academic stress. He acknowledges that he increased his cannabis use in an attempt to manage these experiences. His academic performance and self-care have declined significantly. A detailed review of his electronic health records reveals a documented visit to a general practitioner 10 weeks ago where he reported feeling “increasingly anxious” and having “strange thoughts that everyone might be watching me.” There is no known family history of psychotic disorders.

Substance Use History - Mr. S began using cannabis recreationally at age 19. His use was occasional (1-2 times monthly) until approximately 8 weeks ago, when he increased his consumption to near-daily use, smoking 2-3 cannabis joints per day. When questioned about this escalation, Mr. S states, “I started feeling anxious for upcoming exams and thought cannabis would help me calm down.” He reports that he completely stopped using cannabis 16 days ago because he felt “it started making the voices worse instead of helping.”

MSE - During assessment, Mr. S appears dishevelled with poor personal hygiene. He exhibits tangentiality with monotonous speech. He expresses paranoid delusions about thought broadcasting and being monitored. He reports hearing voices that comment on his actions and sometimes commands him to “be careful” around other people. His affect is notably constricted, and he demonstrates limited emotional reactivity during the interview. He acknowledges having difficulty concentrating and experiencing reduced motivation for previously enjoyed activities.

2nd case vignette

Ms. J, a 42 year old marketing executive, presents to your psychiatric outpatient clinic following referral from her primary care physician for persistent low mood, sleep disturbances, and increasing alcohol consumption over the past 4 months. She reports feeling “increasingly feeling lonely” and states that “life has lost its meaning.” Her initial presentation includes significant distress about her inability to function at work and deteriorating personal relationships. She describes a pervasive sense of sadness that is present almost all day, nearly every day. She reports having lost interest in previously enjoyable activities, including her weekly book club and tennis matches, which she has now abandoned entirely. She states, “I don't see the point in anything anymore.” She also reports of weight loss, difficulty falling asleep, fatigue and feeling of letting everyone down.

Family History - Ms. J's mother was treated for what she describes as “clinical depression” in her 50s and responded well to antidepressant medication. There is no known family history of alcohol or substance use disorders.

Substance Use History - Ms. J reports a history of social drinking (2-3 glasses of wine weekly) for many years without significant problems. However, over the past 3 months, her alcohol consumption has increased markedly to approximately 1-2 bottles of wine daily. She states, “I started drinking more to help me sleep and to numb these feelings, but it's not working anymore.” She describes drinking alone at home, often starting in the early evening. When asked about this sequence, Ms. J states, “I felt so low that I started drinking more to cope, but it just made everything worse.” Ms. J's primary care physician conducted comprehensive laboratory tests two weeks prior to her psychiatric evaluation, including liver function tests, which showed mild elevation of GGT and AST consistent with recent alcohol consumption. TSH was within normal limits.

Funding Statement

Nil.