Substance Use and Somatic Care in Forensic Psychiatry

In This Post

Featured Article
Article Title
Authors
Abstract
Keywords
Summary of Research
Translating Research into Practice
Other Interesting Tidbits for Researchers and Clinicians

Featured Article

Nordic Journal of Psychiatry | 2026, p.1-7

Article Title

Substance use disorders and use of somatic health services among forensic psychiatric patients: a nationwide register-based cohort study

Authors

Christian Haurdahl Jentz; Department of Forensic Psychiatry, Aarhus University Hospital Psychiatry, Aarhus, Denmark; Department of Clinical Medicine, Health, Aarhus University, Aarhus, Denmark

Harry Kennedy; Department of Forensic Psychiatry, Aarhus University Hospital Psychiatry, Aarhus, Denmark; Department of Clinical Medicine, Health, Aarhus University, Aarhus, Denmark; Department of Psychiatry, DUNDRUM Centre for Forensic Excellence, Trinity College Dublin, Dublin, Ireland

Morten Deleuran Terkildsen; Department of Forensic Psychiatry, Aarhus University Hospital Psychiatry, Aarhus, Denmark; Department of Clinical Medicine, Health, Aarhus University, Aarhus, Denmark

Annelli Sandbæk; Steno Diabetes Center Aarhus, Aarhus University Hospital, Aarhus, Denmark; Department of Public Health, Aarhus University, Aarhus, Denmark

Anette Andersen; Steno Diabetes Center Aarhus, Aarhus University Hospital, Aarhus, Denmark

Lisbeth Uhrskov Sørensena; Department of Forensic Psychiatry, Aarhus University Hospital Psychiatry, Aarhus, Denmark; Department of Clinical Medicine, Health, Aarhus University, Aarhus, Denmark

Abstract

Background: Substance use disorders (SUD) are prevalent among forensic psychiatric patients and contribute to a complex mental and somatic health burden. While barriers to healthcare access are well documented, little is known about how individuals with comorbid SUD use general practitioner (GP) and acute somatic hospital services before entering forensic psychiatric treatment.

Aims: To examine the association between SUD and use of GP and acute outpatient hospital services.

Methods: We identified all individuals with a forensic psychiatric measure initiated in 2021–2022 from the Danish National Patient Register. SUD was defined as any ICD-10 F1 diagnosis, excluding acute intoxication, recorded during 2013–2017. Outcomes were GP consultations and acute outpatient hospital contacts during 2018. Associations were examined using negative binomial regression adjusted for age, sex, region, and prior-year service use.

Results: The cohort included 829 patients (39% with SUD). Almost all had ≥1 GP contact in 2018 (90% vs 89%). SUD was not associated with GP utilization (adjusted IRR = 0.99, 95% CI: 0.88–1.12). Acute outpatient contacts were more frequent among patients with SUD (45% vs 32%), with higher rates (IRR = 1.43, 95% CI: 1.04–1.97), attenuating after adjustment for prior-year use (aIRR = 1.28, 95% CI: 0.97–1.68).

Conclusions: Forensic psychiatric patients with SUD had greater use of acute outpatient hospital services, but similar GP utilization, compared with those without SUD, suggesting reliance on crisis-driven rather than preventive care.

Keywords

Substance use disorders; primary care utilization; register-based research; severe mental disorder

Summary of Research

Substance use disorders (SUDs) are described as “common in forensic psychiatric populations” and as complicating “psychiatric treatment, rehabilitation, and long-term outcomes.” High proportions of forensic patients present with comorbid SUD, “often identified as a key criminogenic factor.” At the same time, patients with severe mental illness “face multiple barriers to accessing somatic healthcare,” including cognitive deficits, diagnostic overshadowing, stigma, and system-level fragmentation between psychiatric and somatic care. These challenges are “amplified in forensic psychiatric settings,” where patients may experience “the combined stigma of mental illness, substance use, and criminal justice involvement.” SUD further contributes to “excess somatic illness,” while patients with SUD “tend to seek care differently, often delaying help-seeking and relying more on acute rather than preventive contacts.” Because patterns of health service use can provide insight into how well somatic needs are met, this study aimed “to compare the use of GP services and acute outpatient hospital contacts prior to forensic psychiatric treatment among patients with and without SUD” (p. 1-2).

This retrospective cohort study used Danish nationwide registers, leveraging the CPR number to enable “accurate linkage of individual-level data across registers.” The study population included all individuals with a forensic psychiatric trajectory marker initiated between January 1, 2021, and December 31, 2022. To reduce institutional influence on outcomes, service use was measured in 2018, prior to the forensic measure, and exposure to SUD was defined using ICD-10 F1 diagnoses recorded between 2013 and 2017, excluding acute intoxication codes. Outcomes were defined as (1) “acute outpatient hospital contacts,” corresponding primarily to emergency department visits and other urgent somatic encounters, and (2) GP contacts obtained from the National Health Service Register. Differences between patients with and without SUD were analyzed using negative binomial regression, reporting rate ratios with 95% confidence intervals, with models adjusted for age, sex, region, and prior-year service use (p. 2-3).

Among 829 forensic psychiatric patients, 39% had a registered SUD diagnosis. The majority were men (76%), and psychotic disorders were the most frequent primary diagnosis (68%), significantly more common among patients with SUD. GP use was similar across groups. “Almost all patients had at least one GP contact during the year,” with a median of seven contacts in both groups. In regression analyses, SUD “was not associated with the number of GP contacts,” neither in unadjusted nor adjusted models. In contrast, acute outpatient hospital use differed. Overall, 63% had no acute contacts, but fewer patients with SUD had zero contacts compared with those without SUD. In regression analyses, SUD was associated with “a significantly increased rate of acute contacts” (IRR = 1.43), although this association was attenuated after adjustment for prior-year use and was no longer statistically significant (aIRR = 1.28) (p. 3-4).

The main findings indicate that there were “no differences in GP utilization between groups,” whereas patients with SUD had “more acute somatic contacts,” although this association was diminished after adjusting for prior service use. Comparable GP attendance suggests that “within healthcare systems with well-established primary care, access in terms of contact frequency is not the limiting issue.” Instead, the concern is whether such contacts provide “adequate preventive and coordinated care.” The higher reliance on acute services among patients with SUD may reflect “unmet or unaddressed health needs,” as well as “clinical instability and possibly unmet needs.” The attenuation after adjusting for prior-year use suggests that patients with SUD “already belonged to a group with persistently high healthcare use, rather than representing a new surge of contacts.” Within broader models of health service utilization, the increased use of acute somatic services may be interpreted as “an indicator of unmet or unaddressed health needs,” despite Denmark’s universal healthcare access (p. 4-5).

Translating Research into Practice

“Clinically, the results highlight that forensic psychiatric patients are in frequent contact with GPs regardless of SUD status, yet those with SUD still rely more on emergency and other unplanned outpatient services. These findings point to potential gaps in somatic healthcare for this population, despite frequent primary care contact. Future studies should examine subgroups more closely, including patients with no GP contact and those with very high levels of service use, as these groups may reflect distinct patterns of unmet need not captured by average utilization measures. Similarly, care pathways and barriers to preventive somatic healthcare should be explored, as well as strategies for improving coordination between psychiatric and somatic services, including on-site or visiting somatic clinicians, structured referral pathways, and care coordination. Such work could benefit from repeated measures and longer look-back periods to capture trajectories of healthcare use over time. Future research is also needed to investigate somatic health outcomes, including both perceived and objective health improvements as well as patient satisfaction, as these may help describe the impact of unmet health needs. From a methodological perspective, this study demonstrates how Danish national registers can be used effectively to study forensic psychiatric patients and their somatic healthcare contact” (p. 5-6).

Other Interesting Tidbits for Researchers and Clinicians

“The study has several strengths. It is based on Danish nationwide registers with complete population coverage and virtually no loss to follow-up, enabling a large, representative cohort. Forensic psychiatric patients were identified using validated trajectory markers, ensuring a transparent and reproducible case definition. The study design ensured a clear temporal ordering of exposures and subsequent outcomes. Outcomes were chosen for their clinical relevance, capturing both continuous care (GP contacts) and crisis-driven care (acute outpatient contacts). Finally, SUD was defined using all relevant hospital diagnoses, including both primary and secondary codes, increasing sensitivity. Several limitations must also be acknowledged. GP contacts were recorded at the week level, as remuneration is organized per week, which introduces some imprecision in the exact number of consultations. However, this misclassification is unlikely to differ systematically between patients with and without SUD and would therefore tend to attenuate observed associations rather than inflate them. We restricted our definition of GP contacts to consultations with a physician in general practice, excluding municipal acute services, which may underestimate broader first-line care. Our definition of acute hospital contacts included only outpatient encounters flagged as acute in the DNPR. Acute somatic admissions were not included, which may underestimate total acute hospital use but ensured comparability with GP contacts as front-door services. The Statistics Denmark health registers do not cover diagnoses made in primary care or private practice, meaning that only hospital-recorded conditions were captured. This likely underestimates the true prevalence of SUD, as individuals receiving treatment solely in primary care or specialized addiction services would not be identified. The relatively low proportion of people with SUD (39%) compared with prior studies suggests misclassification of some patients with SUD as non-SUD, which would bias results toward the null for both GP and acute somatic service use outcomes. Because SUD was coded as a binary variable, variation in severity or type of substance use could not be captured, and residual confounding may therefore remain. Furthermore, the study relied on a single year of follow-up, which may not capture longer-term utilization patterns. Because sampling occurred in 2021/22 while outcomes were measured in 2018, the cohort includes patients who survived and remained in care until 2021/22; this could reduce generalizability to prevalent forensic patients in 2018. Similarly, the exclusion of patients with earlier forensic trajectory markers may limit generalizability to patients with recurrent or long-term forensic measures, as the findings primarily reflect healthcare use among individuals entering forensic psychiatric treatment for the first time. Although this study was not designed to test a specific hypothesis, incorporating a broader theoretical framework could strengthen interpretation of the findings and inform more specific hypotheses for future research. Finally, residual confounding from factors such as socioeconomic status or type of index offense cannot be excluded” (p. 5).