Medical costs and related factors associated with mental disorders in Jilin Province, China, 2020–2022

Abstract

Background

Mental disorders are a major cause of the economic burden of disease worldwide, and in China, mental disorders consume a large amount of health care resources and impose a heavy economic burden on society, families, and individuals. The aim of this study was to analyze the medical costs associated with mental disorders in Jilin Province, China, and the related factors affecting medical costs.

Methods

Patients were selected by multistage stratified cluster random sampling. All patient records from 446 health care institutions in Jilin Province from 2020 to 2022 were analyzed to characterize their medical costs. T tests, ANOVA and hierarchical regression were used to explore the related factors influencing medical costs.

Results

The average medical cost in Jilin Province from 2020 to 2022 was 225.9 million dollars. Drugs accounted for the largest proportion of medical costs. Hierarchical regression analysis revealed that days of hospitalization, the medical payment method, and the institutional level were positively correlated with medical costs; gender, age, the disease type, comorbidities, the institutional type, the nature of institution, and treatment services were negatively correlated with medical costs(p < 0.05).

Conclusions

Mental disorders often present a substantial economic burden. Efforts must be made to help policy-makers develop effective strategies to reduce the medical burden of mental disorders.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-025-13259-7.

Background

Mental disorders are a complex and wide-ranging group of health problems that encompass a variety of disorders affecting an individual’s physical, mental, cognitive and behavioral functioning. Mental disorders can be defined as a collection of symptoms and signs that are associated with distress (e.g., anxiety, low mood) or impairment in one or more important areas of functioning (e.g., social relationships, work, self-care) and that result from a dysregulation of brain circuits and neurotransmitter systems, often in the context of genetic vulnerability and environmental stressors [1].

According to data released by the World Health Organization, nearly 1 billion people worldwide currently suffer from mental disorders [2]. There are more than 350 million people with depression and approximately 264 million with anxiety disorders globally. After the COVID-19 pandemic, the number of people with depression and anxiety disorders increased significantly. The number of depressed patients increased by 53 million, representing a substantial increase of 27.6%, whereas the number of patients with anxiety disorders increased by 62 million, for a rate of increase of 20.8% [3]. By analyzing data from the Global Burden of Disease Study in 2019, researchers estimated that mental disorders in 2019 led to 418 million disability-adjusted life-years (DALYs), accounting for 16% of total global DALYs, and the economic value associated with these disorders was estimated to be approximately $5 trillion. At the regional level, losses in sub-Saharan Africa may account for 4% of gross domestic product (GDP), whereas in high-income North America, this figure is 8% [4]. In China, rapid economic and social development over the past 30 years has led to a general increase in psychological stress [5], from 3.2% to 7.2% in the mid- to late 1970 s to the current 17.5%, with more than 2 million people suffering from mental illnesses today [6–8]. A study based on the China Mental Health Survey (CMHS) in 2013 revealed that [9, 10] the weighted prevalence of mental disorders (excluding dementia) was 9.3%, which was higher than that reported in 1982 (point prevalence of 1.1%, lifetime prevalence of 1.3%), 1993 (point prevalence of 1.1%, lifetime prevalence of 1.4%), and 2002 (12-month prevalence of 7.0%, lifetime prevalence of 13.2%) but lower than that reported in 2009 (1-month prevalence of 17.5%). In early 2009, statistics from Chinese health organizations indicated that the number of people with mental disorders in China exceeded 100 million and that the burden of disease for mental disorders accounted for approximately 20% of the total disease burden, making mental disorders the diseases with the highest burden [11]. Mental disorders are a large and diverse category of diseases that have a profound impact on the physical and mental health, social functioning and quality of life of individuals, and as a result of these problems, mental disorders consume a considerable amount of the medical and health care resources of society, resulting in a heavy economic burden of disease. Located in the center of Northeast China with a population of 23.477 million in 2022, Jilin Province is a province in China with relatively backward socio-economic development and a serious aging population [12]. Although it differs from some Chinese provinces in terms of economic structure and health resource allocation, it is representative to a certain extent in terms of the contradiction between supply and demand of mental health services and the characteristics of the disease spectrum [13].

Previous studies have focused mostly on characterizing the economic burden of a particular type of mental disorder, and among the regression methods, ordinary least squares (OLS) and multiple linear regression have mainly been employed [14, 15]. In this work, based on describing the overall economic burden of mental disorders, we add univariate analysis and multilevel regression analysis. Through multilevel regression analysis, we can reduce the interference of confounding factors in the results, clarify the independent contribution of each independent variable to the dependent variable, and improve the accuracy of the study.

Methods

Data source and sample inclusion/exclusion criteria

The study extracted case data from the Jilin Hospital Information System (HIS) from January 1, 2020, to December 31, 2022, and coded diagnoses according to the International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10). The following criteria were used to determine data inclusion and exclusion: (1) The primary diagnosis is a mental disorder (identified by ICD-10 coding for the main diagnosis at discharge); (2) The discharge date falls between January 1, 2020, and December 31, 2022; (3) The actual length of hospital stay for inpatients is more than 1 day; (4) The medical costs less than or equal to 0 are excluded; (5) Data with missing information are excluded. A total of 564,181 patients with mental disorders were included in the sample pool, and 6,996 were excluded, resulting in a final total of 557,185 patients with mental disorders. Of these, 192,872 patients are included in 2020, 187,534 in 2021, and 176,779 in 2022. For two or more cases, each case was studied as a separate case, as our study focused on cases of mental disorders. Figure 1 shows the derivation of the final study sample.

Fig. 1.

Data screening process (statistics in 2020–2022)

Research methods

The System of Health Accounts 2011 (SHA2011) is a health cost accounting system for international classification and statistical reporting based on the system of national economic accounts and related principles to improve the international and domestic comparability of health cost accounting data.

On the basis of SHA2011, this study utilized the principle of “bottom-up” costing to obtain recurrent health costs, derived apportionment ratios, and then derived the medical costs of mental disorders in Jilin Province [16]. For the accounting method, see the New System of Health Accounts: An Introduction to SHA 2011 [17].

To eliminate the interference of the price factor in the medical costs of patients with mental disorders in different years, before the medical costs were analyzed, the costs were discounted by the consumer price index (CPI), with 2022 used as the baseline. All estimated costs were converted to US dollars (exchange rate: US $1 ≈ Chinese yuan ¥6.7261, Bank of China, January 2023).

Measure

The dependent variable for this study was medical costs, which included treatment, drug, bed, consultation, examination, surgery, laboratory, nursing and other costs. While the independent variables include the sociodemographic characteristics, medical characteristics, disease characteristics and institutional characteristics. Sociodemographic characteristics included gender and age. Medical characteristics included the number of days of hospitalization and the medical payment method. Disease characteristics included the disease type and comorbidities. Institutional characteristics included the institutional type, the institutional level, treatment services and the nature of the institution.

The classification of disease types is based on the World Health Organization (ICD-10) [18]. This classification method is a commonly used basis for judgment in the academic circle [19–22]. The specific classifications are as follows: (1) dementia and subtypes of dementia (F00-F03 in ICD-10); (2) mental and behavioral disorders due to psychoactive substance use (F10-F19 in ICD-10); (3) schizophrenia, schizotypal personality disorder, and delusional disorders (F20-F29 in ICD-10); (4) mood disorders, ICD-10 as F30-F39; (5) neurotic, stress-related and somatoform disorders (F40-F48 in ICD-10); (6) eating disorders (F50 in ICD-10); (7) sleep disorders (F51 in ICD-10); and (8) other mental disorders (F04-F09 and F52-F99 in ICD-10).

In the Chinese insurance system, patients are covered by a health insurance system based on China’s urban‒rural dichotomy and patients’ employment status. In this study, the types of health insurance were divided into three main categories: (1) resident basic health insurance, which included urban employees’ basic medical insurance, urban and rural residents’ basic medical insurance and new rural cooperative medical schemes; (2) out-of-pocket, which included patients without medical insurance or patients who sought medical care in nonlocal institutions; and (3) other types of health insurance, which included commercial insurance, injury insurance and maternity insurance.

Initially, there were 14 types of institutions, namely, general hospitals, traditional Chinese medical hospitals, various types of specialty hospitals and primary health care institutions. For the sake of simplicity, we consolidated them into four categories: “general hospitals” “traditional Chinese medical hospitals” “psychiatric hospitals” and “others”.

Statistical analysis

The data were analyzed using STATA/SE 16 and SPSS 25.0. The factors influencing the total medical costs associated with mental disorders were explored by t tests, ANOVA and hierarchical regression analysis. The test level was α = 0.05, and P < 0.05 was considered statistically significant.

Results

Status of treatment costs for patients with mental disorders in Jilin Province

Using 2022 as the baseline, the costs of treating mental disorders in 2020 and 2021 were discounted through the CPI (Table 1) and converted into US dollars.

Table 1.

Discounts on medical costs for people with mental disorders, 2020–2022

Discharged years	CPI (%)	Discount formula
2020 (Y1)	100.0	102.1%*100.6%*Y1
2021 (Y2)	100.6	102.1%*Y2
2022 (Y3)	102.1	Y3

The total costs of mental disorders were $217.3 million in 2020, $122.4 million in 2021 and $337.9 million in 2022. Total costs declined in 2021 and recovered rapidly in 2022. The three-year cumulative total costs was $677.7 million, of which $343.8 million was spent on drugs, which accounted for the largest proportion of the total costs, approximately 50.73%. The second highest were treatment costs and examination costs, accounting for approximately 15.59% and 8.93%, respectively. The costs of surgery and consultation were minimal (Table 2). In addition to distinguishing changes in medical costs over time and population size, this study measured the per capita medical costs of the sample population. Detailed information can be found in Appendix Tables 1 and 2.

Table 2.

Composition of medical costs for mental disorders, 2020–2022 (unit: billion USD)

Characteristic	2020	2021	2022	Cumulative value	percentage(%)
Treatment costs	0.415	0.067	0.575	1.056		15.59
Drug costs	1.099	0.854	1.485	3.438		50.73
Bed costs	0.184	0.011	0.320	0.515		7.60
Consultation costs	0.058	0.009	0.091	0.159		2.35
Examination costs	0.178	0.067	0.360	0.605		8.93
Surgery costs	0.005	0.007	0.004	0.017		0.25
Laboratory costs	0.118	0.049	0.249	0.417		6.16
Nursing costs	0.073	0.014	0.155	0.242		3.57
Others	0.042	0.146	0.140	0.327		4.83
Total Costs	2.173	1.224	3.379	6.777		100.00

The status of dimensions of medical costs for mental disorders

In terms of gender, the cumulative three-year costs for female mental disorder patients was $364.0 million, accounting for 53.71% of the total, with medical costs for females being higher than those for males. In terms of age, the cumulative three-year costs for patients with mental disorders under 45 years of age was $287.0 million, accounting for the largest proportion (42.34%). This group was followed by the 45–64 age group, with $277.0 million (40.88%). The least costly group was the 65 years and older age group, with a proportion of 16.77%. In terms of the medical payment method, the cumulative three-year costs of resident basic health insurance patients with mental disorders was $448.1 million, accounting for the highest proportion of 66.11%. The costs of out-of-pocket patients was $139.9 million, accounting for 20.65% of all patients. In terms of the disease type, among the eight types of mental disorders, the cumulative three-year costs for mood disorders was $199.1 million, accounting for 29.37%. The three types of disorders, namely, mood disorders, schizophrenia, schizotypal personality disorder, and delusional disorders and neurotic, stress-related and somatoform disorders, accounted for more than 75.00% of the total. In terms of comorbidities, the cumulative three-year costs for patients without comorbidities was $527.2 million (77.79%). In terms of the institutional type, the three-year cumulative costs incurred in psychiatric hospitals was $379.1 million, constituting the highest percentage (55.94%). This is followed by general hospitals, with $201.9 million, (29.79%). In terms of the institutional level, the highest three-year cumulative costs of $395.9 million (58.41%) was incurred in secondary hospitals. The three-year cumulative costs of $238.1 million (35.13%) was incurred in tertiary hospitals. In terms of treatment services, the three-year cumulative costs for inpatient services were $375.8 million, which was slightly higher than that for outpatient services (55.44%). In terms of the nature of the institution, most medical costs were incurred in public hospitals, with a three-year cumulative cost of $674.1 million (99.46%) (Table 3). Since drug costs account for a relatively large portion of medical costs, we conducted a specific flow analysis, as shown in Appendix Table 3.

Table 3.

Flow of medical costs associated with mental disorders (unit: billion USD)

Characteristic	2020	2021	2022	cumulative value	percentage(%)
Gender
Female	1.148	0.713	1.779	3.640	53.71
Male	1.025	0.512	1.601	3.137	46.29
Age
< 45	0.942	0.527	1.401	2.870	42.34
45–64	0.925	0.437	1.408	2.770	40.88
≥ 65	0.304	0.261	0.572	1.137	16.77
Medical payment method
Resident basic health insurance	1.332	0.485	2.664	4.481	66.11
Out-of-pocket	0.001	0.691	0.708	1.399	20.65
Others	0.074	0.050	0.007	0.131	1.94
Disease type
Dementia and subtypes of dementia	0.029	0.008	0.027	0.063	0.94
Mental and behavioral disorders due to psychoactive substance use	0.044	0.009	0.054	0.107	1.58
Schizophrenia, schizotypal personality disorder, and delusional disorders	0.611	0.115	0.968	1.694	24.99
Mood disorders	0.542	0.449	0.999	1.991	29.37
Neurotic, stress-related and somatoform disorders	0.322	0.370	0.722	1.415	20.88
Eating disorders	0.006	0.008	0.022	0.035	0.52
Sleep disorders	0.023	0.015	0.025	0.064	0.94
Other mental disorders	0.597	0.250	0.562	1.410	20.80
Comorbidities Yes	0.133	0.187	1.186	1.505	22.21
No	2.040	1.038	2.194	5.272	77.79
Institutional type
General hospitals	0.579	0.407	1.033	2.019	29.79
Traditional Chinese medical hospitals	0.171	0.264	0.219	0.654	9.65
Psychiatric hospitals	1.296	0.534	1.960	3.791	55.94
Other hospitals	0.127	0.020	0.167	0.313	4.62
Institutional level
Class-I	0.002	0.006	0.004	0.011	0.17
Secondary	1.064	0.665	2.230	3.959	58.41
Tertiary	0.683	0.553	1.146	2.381	35.13
Unrated	0.423	0.002	0.000	0.425	6.27
Treatment services
Inpatient	1.075	0.267	2.415	3.758	55.44
Outpatient	1.098	0.958	0.964	3.020	44.56
Nature of the institution
Public hospitals	2.164	1.202	3.375	6.741	99.46
Private hospitals	0.009	0.023	0.005	0.036	0.54

Analysis of influencing factors

Characteristics of the study sample

Table 4 shows an overview of the sociodemographic, medical, disease and institutional characteristics of patients with mental disorders from 2020 2022. The number of female patients with mental disorders included in the study was 296,745, accounting for 53.26% of the sample. The number of patients in the 45–64 years age group was 222,473, accounting for the largest proportion (39.93%). The average number of days of hospitalization was 2.33 ± 12.618. The number of patients who were out-of-pocket was 356,759, accounting for 64.03% of all patients. The number of patients with neurotic, stress-related and somatoform disorders was the greatest, at 196,359 patients, accounting for 35.23%. This group was followed by patients with mood disorders, 175,249, accounting for 31.44% and with both groups accounting for more than 65%. The majority of patients had no comorbidities, accounting for 98.94%. The largest number of patients, 362,144 (65.00%), were treated in psychiatric hospitals. Most patients went to secondary and tertiary hospitals, accounting for 75.48% and 22.47%, respectively. The majority of patients received services on an outpatient basis, accounting for 97.01%. A total of 553,415 patients were treated in public hospitals, accounting for 99.32%. The total costs were log-transformed to conform to a normal distribution. A univariate analysis of the variables above and the total costs was performed for gender (t = 18.42, p < 0.001), age (F = 8771.06, p < 0.001), the number of days of hospitalization (t = 219.43, p < 0.001), medical payment method (F = 6127.14, p < 0.001), the disease type (F = 16897.42, p < 0.001), comorbidities (t = 194.00, p < 0.001), the institutional type (F = 2043.44, p < 0.001), the institutional level (F = 8937.59, p < 0.001), treatment services (t = 336.38, p < 0.001), the nature of the institution (t = 67.68, p < 0.001) and total costs, and the results were statistically significant. In addition, we conducted a univariate analysis on the drug cost as the dependent variable. The results showed that the above variables also had specific statistical significance for the drug cost. Detailed information can be found in Appendix Table 4.

Table 4.

Univariate analysis of patients with mental disorders

Characteristic	N	percentage(%)	t/F	P
Gender			18.4193	< 0.001
Female	296,745	53.26
Male	260,440	46.74
Age			8771.06	< 0.001
< 45	209,992	37.69
45–64	222,473	39.93
≥ 65	124,720	22.38
Days of hospitalization	557,185	100.00	219.43	< 0.001
Medical payment method			6127.14	< 0.001
Resident basic health insurance	188,482	33.82
Out-of-pocket	356,759	64.03
Other	11,994	2.15
Disease type			16897.42	< 0.001
Dementia and subtypes of dementia	1061	0.19
Mental and behavioral disorders due to psychoactive substance use	5023	0.9
Schizophrenia, schizotypal personality disorder, and delusional disorders	57,894	10.39
Mood disorders	175,247	31.44
Neurotic, stress-related and somatoform disorders	196,345	35.23
Eating disorders	2262	0.41
Sleep disorders	47,501	8.53
Other mental disorders	71,852	12.91
Comorbidities Yes	5899	1.06	193.997	< 0.001
No	551,286	98.94
Institutional type			2043.44	< 0.001
General hospitals	111,082	19.94
Traditional Chinese medicine hospitals	67,584	12.13
Psychiatric hospital	362,144	65.00
Other hospitals	16,375	2.94
Institutional level			8937.59	< 0.001
Class-I	4011	0.72
Secondary	420,549	75.48
Tertiary	125,189	22.47
Unrated	7436	1.33
Treatment services			336.3792	< 0.001
Inpatient	16,665	2.99
Outpatient	540,520	97.01
Nature of the institution			67.6789	< 0.001
Public hospitals	553,415	99.32
Private hospitals	3770	0.68

Hierarchical regression analysis

Prior to hierarchical regression analysis, the independent variables were measured for collinearity, with a VIF value of < 2, and there were no collinearity. Hierarchical regression analysis were performed for variables that were significant in the univariate analysis, and the dependent variable was log-transformed total costs. The variables were divided into four layers and incorporated into the model. The first stratum consisted of sociodemographic characteristics, the second stratum consisted of medical characteristics, the third stratum consisted of disease characteristics, and the fourth stratum consisted of institutional characteristics. The results show that gender (β=−0.106, p < 0.001) and age (β=−0.407, p < 0.001) affected medical costs. Female were associated with higher medical costs, and younger patients bear higher medical costs. Although gender has statistical significance for medical costs, its effect size was small and may lack practical clinical significance. After controlling for sociodemographic variables, the F value of Model 2 was 29685.555, days of hospitalization (β = 0.037, p < 0.001) and medical payment method (β=−0.432, p < 0.001) affected total medical costs. The longer the length of hospital stay, the more medical costs were incurred. Days of hospitalization, while statistically significant, its effect size was small, and the practical significance might be very weak. Patients who pay with resident basic health insurance bear higher medical costs. After controlling for sociodemographic and medical characteristic variables, the F value of Model 3 was 11107.715, disease type (β=−33.141, p < 0.001) and comorbidities (β=−145.898, p < 0.001) affected medical costs. The patients with comorbidities spend more on medical expenses. After controlling for sociodemographic variables, medical characteristics and disease characteristics, the F value for Model 4 was 13316.064, institutional type (β=−64.514, p < 0.001), treatment services (β=−178.765, p < 0.001), nature of the institution (β=−87.548, p < 0.001) and institutional level (β = 67.549, p < 0.001) effected medical costs. Patients visiting general hospitals spend more on medical costs. Inpatients have higher medical costs compared to outpatients. Patients visiting tertiary hospitals and public hospitals were associated with higher medical costs (Table 5).

Table 5.

Hierarchical regression analysis of patients with mental disorders

Characteristic	Model 1			Model 2		Model 3		Model 4
	β		t	β	t	β	t	β	t
Gender	−0.106	−22.905***		−0.117	−26.574***	−0.107	−24.720***	−0.075	−18.011***
Age	−0.407	−133.165***		−0.443	−150.528***	−0.454	−157.157***	−0.415	−149.555***
Days of hospitalization				0.037	208.352***	0.030	164.602***	0.007	34.893***
Medical payment method				−0.432	−97.716***	−0.372	−84.424***	−0.299	−70.319***
Disease type						−0.048	−33.141***	−0.104	−68.657***
Comorbidities						−3.209	−145.898***	−0.571	−22.577***
Institutional type								−0.185	−64.514***
Institutional level								0.168	67.549***
Treatment services								−3.295	−178.765***
Nature of the institution								−2.236	−87.548***
constant	5.916	637.290		6.639	546.174	13.184	289.149	16.845	323.538
R²		0.031			0.125		0.158		0.232
△R²		0.031			0.093		0.034		0.073
F		9041.556***			29685.555***		11107.715***		13316.064***

Note: ***P < 0.001, **P < 0.01, *P < 0.05

Robustness check

To verify the robustness of the results, we performed K-value cluster analysis on medical costs and used ordinal logistic regression as a robustness test. The results showed that the direction and significance of the main variables were largely consistent, confirming the stability of the findings. Full regression results were provided in Appendix Table 5.

Discussion

With rapid socioeconomic development, increased social rhythm and competitive pressure, people’s mental health problems have increased, and an increasing number of families are suffering from the psychological and economic pressures caused by mental disorders. The results of this study show that the economic burden caused by mental disorders has increased. When the effects of inflation factors are removed, the cost of medicines accounts for more than 50% of medical costs in Jilin Province in 2020–2022, which reflects the high reliance on drugs in the treatment of mental disorders in China, especially in the treatment of chronic diseases such as schizophrenia and mood disorders, where long-term medication management is the norm [13]. This cost structure suggests that we should strengthen the optimization of rational drug use, basic drug protection and health insurance payment policies to reduce the economic burden of drug expenditures. The total medical costs in Jilin Province increased from $217.3 million in 2020 to $337.9 million in 2022, with an annual growth rate of 55.45%, accounting for 0.06%−0.17% of the GDP of Jilin Province. Notably, medical costs were significantly lower in 2021 compared to 2020 and 2022, a change that may be partially attributed to the disruption of healthcare services during the COVID-19 pandemic. At that time, there were short-term fluctuations in the supply of and demand for healthcare services due to factors such as the redeployment of healthcare resources and fewer patient visits [23, 24]. The downward trend may be a result of limited accessibility of services. Therefore, the medical costs recorded in 2020 and 2021 may underestimate the true economic burden of mental disorders. In European countries, the economic burden of mental illness is approximately 3.50% of GDP [25]. The population of Jilin Province was 23,992,000 in 2020 and 23,477,000 in 2022, and the per capita cost of treatment increased from $9.057 in 2020 to $14.394 in 2022. The total medical costs of mental disorders and the per capita medical costs are increasing rapidly, which is in line with the results of Wang Li. There is also a large gap between the utilization rates of mental health services in China and abroad. It is estimated that fewer than 10% of people with mental disorders in China receive specialized mental health services, whereas the treatment rates for people with mental disorders in Germany, Belgium and Canada are 67%, 66% and 52%, respectively [26, 27]. This difference may be because the treatment and prevention of mental health disorders in China started relatively late and because patients with mental disorders usually delay treatment and visit large hospitals only when their illnesses are abnormally serious, which, to a certain extent, increases the costs of treatment [28, 29]. Therefore, mental disorders should be prioritized for disease prevention and control from the perspective of medical costs control.

Although the economic burden of mental disorders is high, the impact is not uniform across all segments of society. At the sociodemographic level, medical costs are greater for women than for men, which is consistent with Bandelow’s findings. Possibly due to innate factors, women are more sensitive and insecure than men are [30], especially when experiencing adverse life events [31]. This result may also be associated with gender discrimination, violence, sexual abuse, prenatal and postnatal depression and poor sociocultural norms [32, 33]. There was a negative effect of age and medical costs, which decreased with increasing age. Medical costs were significantly greater for people under 45 years old and those aged 45–64 years than for people over 65 years old. This result is consistent with the findings of Junfang Xu [34]. The reasons may be that young patients are the hope for the future or the backbone of their families; once they become ill, the rate of consultation is higher; patients and their families will seek rehabilitation at all costs; and the loss of productivity due to illness is often greater in young patients than in older patients [35, 36].

At the level of medical characteristics, medical costs increase with the days of hospitalization; that is, the longer the hospital stay is, the more health services are utilized, thus generating more medical costs, which is consistent with the findings of Life and Arquivos de Neuro-Psiquiatria et al. [37–39]. Although the number of days of hospitalization positively predicted higher total costs, the relatively small regression coefficient suggests that daily hospitalization costs may not be increasing substantially. This implies that while longer stays accumulate costs, the marginal daily cost may be stable or modest. Studies have shown that the implementation of clinical pathways can reduce the number of ineffective hospitalization days, effectively reduce hospitalization costs, and in turn reduce medical costs [40]. The medical payment method has a negative impact on medical costs. In the 2020–2022 period, the cumulative total costs for fully self-funded patients was $139.9 million, accounting for 20.65%, with a larger proportion of individual out-of-pocket payments for patients with psychiatric disorders. Xiong et al. also suggested that the setting of different health insurance policies has an impact on medical costs [41]. The medical costs were greater for resident health insurance patients than for fully self-financed patients or other patients, and the difference was statistically significant. This result may be because the inclusion of health insurance in recent years has reduced people’s sensitivity to medical costs, and relatively high reimbursement rates have led to an increase in total medical costs [42].

At the level of disease characterization, medical costs vary among the different types of mental disorders. The three-year cumulative costs for mood disorders, schizophrenia and anxiety disorders amounted to $509.9 million, which accounted for 75.24% of the total costs for all types of mental disorders, which is consistent with the findings of Yang X, Vos T et al. [43–46]. Notably, the number of patients with schizophrenia was 10.39% of all mental disorder patients, and their share of medical costs was 24.99%, making the burden of treatment costs heavier for this group of patients. Comorbidities had a negative impact on treatment costs; that is, patients with comorbidities had higher medical costs than those without comorbidities did, which is consistent with the findings of Farhane-Medina NZ, Ress S et al. [47–49]. It is possible that medical costs have increased because of increased attention to and examination of the body due to comorbidities [50].

At the level of institutional characteristics, the type of institution has a negative effect on medical costs, whereas the institution level has a positive effect on medical costs. The highest medical costs are incurred in psychiatric hospitals, followed by general hospitals. Given the specialization and severity of mental disorders, some new technologies and medical specialists cannot be used in primary or other specialized hospitals, and patients usually visit developed central hospitals and higher-level hospitals, resulting in higher medical costs [51]. Treatment services have a negative effect on the costs of treatment; that is, medical costs are greater for inpatients than for outpatients. The cumulative costs was $375.8 million for inpatient services and $302.0 million for outpatient services in the 2020–2022 period. Compared with inpatients, outpatients are significantly more common, mainly because the number of patients with mild to moderate mental disorders can be effectively managed through outpatient treatment, such as psychological counseling, psychotherapy, and medication. Patients with severe mental disorders, on the other hand, need to be hospitalized, their conditions are more complex, and their treatments are more varied than those of outpatients, resulting in higher costs. The nature of the institution has a negative impact on the costs of medical care, with most costs being incurred in public hospitals, which is inconsistent with the findings of Akhigbe et al. [52]. This result may be because private hospitals are more developed in the United States, whereas public hospitals are predominant in China, and public hospitals enjoy favorable reimbursement policies [53].

Although our large sample size may have led to statistical significance, a number of findings, notably treatment services, nature of the institution, comorbidities, age, medical payment method, institutional type, and institutional level, demonstrated a relationship with increased medical costs.

This study has several limitations. First, only direct medical costs were included, not indirect costs. Second, while the data in this study were collected from multiple medical institutions and has a larger sample size and greater representativeness compared to data from a single institution, it still lacks national representativeness. Third, for two or more cases, each case was studied separately, which may lead to underestimation. Fourth, although this study maximized the inclusion of information and variables contained on the first page of the medical record, it did not include clinically relevant variables or information about patients after discharge. Therefore, this study has relatively limited guidance for clinical practice. Future studies will consider additional factors to more accurately assess medical costs.

Conclusions

This study describes the economic burden of mental disorders in Jilin Province from 2020 to 2022 in terms of four dimensions—patient demographic characteristics, medical characteristics, disease characteristics and institutional characteristics—and analyzes the magnitude and direction of the impact on medical costs through multilevel regression. Mental disorders impose great economic burdens on patients and society in Jilin Province, China. The economic burdens borne by patients with different types of mental disorders vary to some extent, and it is more conducive to alleviating diseases and reducing economic burdens when the demand for health care services of all patients with mental disorders is satisfied.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

DALYs

Disability-adjusted Life-years

ICD-10

International Statistical Classification of Diseases and Related Health Problems 10th Revision

CCMD-3

Chinese Classification of Mental Disorders Version 3

WHO

World Health Organization

GDP

Gross domestic product

OLS

Ordinary least squares

SHA2011

System of Health Accounts 2011

CPI

Consumer price index

Author contributions

XY participated in the conception and design of this study. JY participated in the data collection, data analysis, and interpretation of data and drafted and revised the manuscript. YY participated in the data collection and drafted and revised the manuscript. YY, HZ, HL and HZ participated in the data analysis. XY, LN, SG and YX reviewed and revised the manuscript.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Data availability

The datasets supporting the conclusions are not publicly available owing to data confidentiality agreements but are available from the corresponding author (Yu X, xhyu@jlu.edu.cn) upon reasonable request.

Declarations

Ethics approval and consent to participate

Our study adheres to the ethical principles set forth in the Declaration of Helsinki and has received approval from the Medical Ethics Committee of the School of Public Health, Jilin University (IRB code: 20230707). Prior to commencing the investigation, informed consent was obtained from all participants in the study. All participants have the autonomy to withdraw from the study at any point.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.