Healthcare costs of multimorbidity among those with mental health diagnoses

Mental disorders such as personality disorders, major depressive disorder, and anxiety disorders are associated with significant personal and familial difficulties (Christensen 2020), an increased risk of early death (Plana-Ripoll 2019), as well as being diagnosed with a comorbid physical illness (Firth et al. 2019). Given the extent of multimorbidity among people with mental disorders, it is hypothesised that the complexities of treatment for this group contributes significantly to the national health care cost (Christensen 2022).

This paper (Christensen et al, 2022) investigates healthcare cost by number of comorbid mental and somatic disorders in people diagnosed with a mental disorder using Danish registry data. Christensen et al. hypothesised that, firstly, individual cost of health care would increase with the a greater number of comorbid diagnoses, and secondly, that somatic health care services would constitute a considerable proportion of the cost of care by number of comorbid somatic diagnoses.

Christensen et al. (2022) investigate healthcare cost by number of comorbid mental and somatic disorders in people diagnosed with a mental disorder using Danish psychiatric registry data.

Methods

Data from nearly 450,000 people living in Denmark and diagnosed with a mental disorder at a psychiatric care facility between 2004 and 2017 were retrieved from the Danish Psychiatric Central Research Register if they had been living in Denmark between 2004 and 2017, and had been treated for a mental illness at a psychiatric hospital or outpatient clinic in Denmark after January 1^st 1995.

Eighteen mental and substance use disorders were eligible for inclusion based on ICD-10 classifications, including alcohol use disorder, schizophrenia, anorexia nervosa, bulimia nervosa, personality disorders, ADHD, autism spectrum disorder, and conduct disorder. To establish mental disorder comorbidities, the authors counted the number of different types of mental disorders per individual per year. Using the ICD-10, 31 chronic somatic disorders were eligible for inclusion, including cancer, diabetes, hypertension, and allergies. The same approach was used to identify comorbid somatic disorders as for mental disorder comorbidity; the authors counted the number of different types of somatic disorders per individual per year among those with at least one mental disorder diagnosis. The authors therefore produced two count variables, one for the number of mental disorders, one for the number of somatic disorders within those with at least one mental disorder.

Four different health care costs were included; all were adjusted for inflation and presented in Euros. The health care costs measured were:

• Overall cost averages for the psychiatric service cost from in- and outpatient/emergency room visits
• National average operating expenses for the hospital service cost from somatic disorders
• Primary health care costs, including subsidies cost for visiting the primary health care provider
• Pharmaceutical cost from subsidised prescriptions

Annual assessments were conducted for the health care costs for every individual with at least one mental disorder. The nationwide annual cost and the annual cost per individual were calculated for each mental disorder and somatic disorder count. The outcomes were estimated as the average over the 14-year follow-up period. The nationwide total cost consists of both the cost per individual and the prevalence in each count category. All cost estimates were explored in additional sensitivity analyses where the results were standardised by age and sex.

Using data from almost 450,000 people, the authors estimated the annual health care cost, both for individuals and on the population level, by number of comorbid psychiatric and somatic diagnoses.

Results

The individual health care costs for each of the 447,209 individuals with any mental disorder was assessed annually. The average annual health care cost increased by number of mental disorders; across the 14-year study period, one mental disorder diagnosis was associated with an average annual health care cost of €4,471. For individuals with eight or more mental disorder diagnoses, this increased to €33,273.

For individuals with only one mental disorder diagnosis, the somatic service cost closely matched the psychiatric service cost (39% compared to 41% of the total annual health care cost), however this decreased as the number of mental disorder diagnoses increased. An increase in somatic disorder diagnoses resulted in increased annual healthcare costs per comorbid somatic disorder diagnoses (€4,613 for zero comorbid somatic disorder diagnoses, and up to €16,344 for fifteen comorbid somatic disorder diagnoses).

Nationwide healthcare cost was highest for individuals with only one mental disorder diagnosis (€786 million) and decreased with number of diagnoses. Within this, psychiatric service costs accounted for 70% of the total annual health care costs per individual with fewer than four somatic comorbid diagnoses. Primary health care and subsidised prescription costs was found to be increasingly higher the more comorbid somatic disorders an individual presented with.

Finally, when the results for each cost were standardised by age and sex, the patterns observed in the primary analyses persisted.

The average annual individual health care cost increased by number of mental disorders, while the population-level healthcare cost was highest for those with only one mental disorder, and decreased with number of diagnoses.

Conclusions

The yearly individual cost of healthcare increases with the number of comorbid mental and somatic diagnoses. However, as a small number of people fall into the higher frequency categories, the population-level healthcare cost decreases as the number of comorbidities increases.

Strengths and limitations

The data used for this analysis was high quality as it was collected from national registers, providing a fairly representative population-sample. The list of mental and chronic somatic disorders included in the selection criteria was thorough, although analyses could have been extended further to investigate costs associated with specific comorbid diagnoses.

Limitations include the focus on chronic somatic illnesses instead of a more complete list of somatic disorders, the exclusion of recovery rate per individual, and the exclusion of people with mental health problems who do not have a diagnosis from a hospital. Other important forms of economic costs beyond healthcare costs were not considered, such as national productivity loss, and this leaves a gap for future investigation. Furthermore, data from non-Danish nationals were excluded. This excluded immigrant populations, for example, who may present with specific mental health needs exacerbated by the process of cultural acclimation (Tinghög et al. 2010) which may have associated cost implications.

Future research could consider these research questions in alternative contexts, for example in the USA where healthcare is not free of charge (as it is in Denmark for Danish citizens). Finally, this data was collected pre-COVID-19. Future research could usefully investigate the implications of COVID-19 lockdowns and associated switches to online mental health and somatic services on the costs of healthcare among those with comorbid mental health and somatic diagnoses.

People who have mental health problems but do not have a diagnosis from a hospital were excluded, potentially resulting in an underestimation of costs.

Implications for practice

There are high levels of comorbidity between mental health and somatic diagnoses. Clinicians must avoid misattributing service users’ new symptoms to the original diagnosis, whilst also being aware of diagnostic overshadowing. These could be indicating a comorbid diagnoses that needs identifying, and without this identification the complexities of each case overall cannot be addressed. Clinicians and service providers should also be aware of the complexities comorbidities create for individuals and their carers, and should aim to provide (or signpost to) cross-sector support where possible.

Implications for policy

Policy makers should prioritise preventative public health strategies to promote good physical health among those with one or more mental disorder diagnoses, so as to reduce the risk of increased physical and somatic comorbidities. However, they should also be conscious not to perpetuate misconceptions and stigma that, at the population-level, individuals who have several comorbid diagnoses are costing the health service more. Large-scale policy changes building on these findings would need to consider changes to service use and service costs as a result of the COVID-19 pandemic.

Trends and patterns in service use and service costs are likely to have changed as a result of the COVID-19 pandemic; these would need to be investigated ahead of any large-scale policy changes based on the findings of this paper.

Statement of interests

None.

Links

Primary paper

Other references

Christensen, M.K., Lim, C.C.W., Saha, S., Plana-Ripoll, O., Cannon, D., Presley, F., Weye, N., Momen, N.C., Whiteford, H.A., Iburg, K.M. and McGrath, J.J., 2020. The cost of mental disorders: a systematic review. Epidemiology and psychiatric sciences, 29.

Christensen, M.K., McGrath, J.J., Momen, N.C., Whiteford, H.A., Weye, N., Agerbo, E., Pedersen, C.B., Mortensen, P.B., Plana-Ripoll, O. and Iburg, K.M., 2022. The cost of mental disorders in Denmark: a register-based study. npj Mental Health Research, 1(1), pp.1-7.

Firth, J., Siddiqi, N., Koyanagi, A. I., Siskind, D., Rosenbaum, S., Galletly, C., … & Stubbs, B. (2019). The Lancet Psychiatry Commission: a blueprint for protecting physical health in people with mental illness. The Lancet Psychiatry, 6(8), 675-712.

Plana-Ripoll, O., Pedersen, C.B., Agerbo, E., Holtz, Y., Erlangsen, A., Canudas-Romo, V., Andersen, P.K., Charlson, F.J., Christensen, M.K., Erskine, H.E. and Ferrari, A.J., 2019. A comprehensive analysis of mortality-related health metrics associated with mental disorders: a nationwide, register-based cohort study. The Lancet, 394(10211), pp.1827-1835.

Tinghög, Petter, Suad Al-Saffar, John Carstensen, and Lennart Nordenfelt. 2010. “The Association of Immigrant- and Non-Immigrant-Specific Factors with Mental Ill Health among Immigrants in Sweden.” The International Journal of Social Psychiatry 56 (1): 74–93.

Photo credits