The number of patients with mental disorders is increasing across the world. A report from the World Health Organization (WHO) described that lifetime prevalence of mental problems is about 20% to 35%   . A survey conducted in Japan by the Ministry of Health, Labour and Welfare in 2014 estimated that 533 thousand patients (total of inpatients and outpatients) were diagnosed to have any of the mental and behavioral disorders (ICD 10 Code, F00 to 99)  . Kawakami et al surveyed 1663 Japanese persons in the several regions and found that 2.9% and 1.6% of them had major depression and alcohol abuse/dependence, respectively, in the previous 12 months  .
Mental problems often need long time treatment. For example, 40% to 60% of the patients with schizophrenia suffer from lifelong disorder  . Regarding health resource utilization, the rate of hospitalization in Japan is higher for mental and behavioral disorders than others, and the period of hospitalization exceeds 1 year for 12.8% of hospitalized patients. Among OECD countries, the number of psychiatric hospital beds is the highest and the hospitalization days are the longest in Japan  . Furthermore, the social scenario indicates a high prevalence of depression and schizophrenia in the person who committed suicide   .
Many studies in and outside of Japan have analyzed the medical and societal costs of mental disorders  -  . In these analyses, only the cost of mental disorders was analyzed. But clinically, many patients with mental disorders have physical disorders as well. It is reported that the co-existence of mental and physical disorders leads to poor control of mental conditions, insufficient management of physical disorders, and low QOL of the patients     . In order to improve the cost/benefit of treatment and quality of medical services for these patients, the conditions of concurrent physical disorders have to be addressed. Also, the necessity of preventing aggravation and acute phase treatment for physical disorders can be minimized in addition to discussions about the number of beds.
However, few reports have analyzed the physical comorbidities of the patients with mental disorders or the places where they have received treatment. This is a limitation of the patient surveys conducted by the Ministry of Health, Labour, and Welfare and the analyses of medical cost based on health insurance claim data.
In the present study, we analyzed the cost of mental disorders by analyzing and comparing the cost for mental disorders with the total medical costs (including mental and physical disorders) incurred by patients who were diagnosed with mental disorders. To analyze the medical cost of each disease, we used a medical cost grouping technique to link disorders listed on the health insurance claim sheets to the drugs and/or interventions administered according to their medical indications. Based on these analyses, we tried to clarify the details of the medical cost of patients with mental disorders.
The study population is limited to the residents who are covered by national health insurance in one city. However, it is expected that patients with mental disorders will continue to increase, and the improvement of QOL in these patients is an important issue. By analyzing health resource utilization and physical comorbidities of patients with mental disorders, our study may contribute to the improvement of medical service and clarify the cost-benefit of treatment provided to these patients whose cost is partly paid by the insurer.
The subjects analyzed in our study were the residents (age ≤ 74 years) of City A who were covered by national health insurance and suffered from one or more of 4 mental disorders that have the highest morbidity and lifetime prevalence among mental disorders: alcohol related disorders (ICD10 Code, F10), schizophrenia (ICD10, Code F20), bipolar disorder (ICD 10, Code F31), and depressive episode (ICD10 Code, F32). For the analysis of diseases of the hemodialysis patients, all persons covered by the national health insurance were included in the study.
Outline of persons covered in the national health insurance in City A (as of February 2014)
In total, 94,380 residents in City A were covered by the national health insurance. Among them, 47,150 (50.0%) were males and 47,230 (50.0%) were females. When grouped by age, the age group ranging 65 to 69 years accounted for the highest percentage (11,795 persons [12.5%]) in the residents covered by the national health insurance, followed by the age group 60 to 64 years (10,726 persons [11.4%]) and the 70 to 74 years age group (9,888 persons [10.5%]). Among all persons covered by the national health insurance, 23.2% were the elderly (age ≥ 65 years).
City A is a tertiary industry based local city. About 30% of the residents are covered by the national health insurance and 19.9% of the residents are the elderly in 2015. For 100 thousand residents in City A, there are 78.58 hospitals/clinics (national average [hereinafter same will be applied in this paragraph] 67.88), and among them, 12.84 (4.99) hospitals/clinics are specialized in psychiatry. Clinics have 127.73 beds (80.90), hospitals have 1012.73 beds (1215.00) and psychiatric hospitals/clinics have 230.09 beds (259.31). Psychiatric beds account for 22.7% of the total beds  .
2.2. Study Design
2.3. Data Collection and Analysis
We collected data from health insurance claim sheets issued over 12 months (March 2013 to February 2014) and extracted and analyzed data on age, sex, date of use of health resource, disease, medical institution, days of use, and medical cost. There are 3 types of health insurance claim sheets: sheets for hospitalization cost, sheets for outpatient cost, and sheets for prescription cost. Each month, a sheet that describes the medical service provided and its cost for the previous month is issued at each medical institution.
We analyzed the following particulars from the data: (1) relationship between the number of the patients and medical costs, (2) percentage of inpatients and days of hospitalization classified by hospital function classification, (3) details of physical comorbidities, and (4) diseases of patients on hemodialysis. The following 2 procedures were used for analysis. First, we used a medical cost grouping technique to analyze medical cost used for the target disorders only (4 mental disorders described above). For analysis of outpatient medical costs, medical charge of out-patient/prescription health insurance claim sheet was linked with diseases and summarized. If there were multiple diseases, costs were divided by the number of diseases and assigned equally to each disease. Medical charges that were not linked to any disease were assigned to the primary disease requiring outpatient clinic visits. The hospital costs listed on health insurance claim sheets were linked with the disease primarily responsible for hospitalization and summarized. If there were multiple diseases, costs were divided by the number of diseases and assigned equally to each disease. Second, all costs on the claim sheets of patients with target disorders were analyzed. Namely, the medical costs were analyzed regardless of the type of disease (mental and physical disease).
Data used in this analysis were obtained from the insurer through opt-out procedure.
The subjects included in our analysis were residents of City A who were covered by national health insurance and suffered from one or more of 4 mental disorders: schizophrenia (F20), alcohol related disorders (F10), bipolar disorder (F31), and depressive episode (F32).
3.1. Medical Cost Used for Treatment of Target Mental Disorders (Medical Cost Grouping)
First, we used a medical cost grouping technique to clarify the medical cost used for the 4 disorders as follows (Table 1). It should be noted that some patients were counted twice or more times because they had more than 1 disorder. The total number of patients per year was 7403 (6522 outpatients and 881 inpatients, with duplication). The medical cost was 1.02 billion yen for outpatients (approximately 80 thousand to 260 thousand yen per patient) and 2.24 billion yen for inpatients (approximately 1.05 million to 2.94 million yen per patient). Schizophrenia accounts for the highest percentage of this cost both for outpatients or inpatients.
Among the total medical costs used for the residents covered by national health insurance in City A, 7.2% was used for outpatients and 18.9% was used for inpatients for the above 4 mental disorders.
3.2. Total Outpatient Costs of Patients with Target Mental Disorders
Next, the total outpatient costs of the patients with one or more of the target
Table 1. Medical cost used for treatment of the target mental disorders.
In the period included in this analysis (1 year), 5,601 patients received treatment as outpatients (2650 males [47.3%] and 2,951 females [52.7%]). The age group that accounted for the highest percentage was the 60 to 69 years age group for both males and females (males 23.5% and females 22.3%). The age group that used the highest amount of medical cost per patient was the 40 to 49 years age group, which was approximately 390 thousand yen, followed by the 50 to 59 years age group and the 60 to 69 years age group. This amount was relatively higher than the outpatient cost per patient shown in Table 1, which indicates that the patients had multiple diseases for which they received treatment.
Regarding the medical cost used by each age group, the 40 to 49 years age group (20.9% of the total patients) used 33.1% of the total medical costs, followed by the 60 - 69 years age group (20.9% of the total patients) and the 50 - 59 years age group (20.9% of the total patients), which used 23.8% and 18.9%, respectively, of the total medical costs. The patients with target mental disorders used 9.8% of the total outpatient cost of the residents covered by national health insurance in City A. This was 2.6% higher than the percentage of medical cost obtained by the medical grouping technique for the 4 mental disorders.
The number of patients who were diagnosed with the target mental disorders (schizophrenia [F20], alcohol related disorders [F10], bipolar disorder [F31] or depressive episode [F32]) is summarized in Table 2. The most frequent disorder was depressive episode, followed by schizophrenia and bipolar disorder. Since the total number of outpatients was 5601, it was indicated that the patients with target mental disorders had more than one diagnosis of mental disorder.
Figure 1. The number of outpatients and medical cost per patient.
Figure 2. Medical cost by age of outpatient.
Table 2. Diagnosis of outpatients.
3.3. Total Inpatient Costs of Patients with Target Mental Disorders
Next, the health insurance claim data were analyzed to determine the cost for hospitalization due to one or more of the target mental disorders (4 ICD 10 code names specified above) (Figure 3, Figure 4). The total hospitalization costs were 2.3 billion yen. The breakdown of this total amount by sex and age group is as follows.
In the period included in this analysis (1 year), 1128 patients were hospitalized in total (637 males [56.5%] and 491 females [43.5%]). The age group that
Figure 3. The number of inpatients by age group and medical cost per patient.
Figure 4. Inpatient medical cost by age group.
accounted for the highest percentage of inpatients was the 60 to 69 years age group (males 33.1% and females 27.3%). The age group that used the highest amount of medical cost per patient was the 60 to 69 years age group, followed by the 50 to 59 years age group and the 40 to 49 years age group.
Regarding the medical cost used by each age group, the 60 to 69 years age group (30.6% of the total patients) used 36.9% of the total hospitalization costs, followed by the 50 to 59 years age group (22.8% of the total patients) and the 40 - 49 years age group (18.7% of the total patients), which used 24.8% and 15.8%, respectively, of the total hospitalization costs. Of the total hospitalization costs used by the residents covered by national health insurance in City A, 19.4% was used for the patients with target mental disorders.
Analysis of the number of patients by each diagnosis of the 4 ICD 10 code names specified above is summarized in Table 3. The most frequent disorder was schizophrenia, followed by depressive episode and alcohol related disorders.
The hospitalization period was analyzed in patients with schizophrenia, a disorder which used highest hospitalization cost. As a result, 462 patients (61.2% of the patients with schizophrenia) were hospitalized for 91 days (3 months) or longer per year, and of them, 401 patients (53.1% of the patients with schizophrenia) were hospitalized for 180 days (6 months) or longer. The total
Table 3. Diagnosis of inpatients.
hospitalization costs of the patients who were hospitalized for schizophrenia for 91 days or longer was approximately 1.72 billion yen, which accounted for 74.9% of the total hospitalization costs of the target mental disorders*.
*This cost is the total hospitalization costs of patients who were hospitalized for schizophrenia and includes the cost used for treatment of mental disorders other than schizophrenia and/or physical diseases.
3.4. Hospitalization Cost by Hospital Functional Classification
In Japan, each prefecture decides hospital functional classification of medical institutions matched by its medical/operational plans  . There are the following bed categories of hospital functional classification: 1) beds for the advanced acute phase = beds for highly intensive treatment for acute conditions, e.g., high- care unit and stroke care unit; 2) beds for the acute phase = beds for treatment of conditions that are in the acute phase but do not need highly intensive treatment; 3) beds for the convalescent phase = beds for rehabilitation or preparation for discharge after the acute phase; and 4) beds for the chronic phase = beds for long-term treatment. One medical institution may have more than one category of beds. In the present study, hospitalization was analyzed by hospital functional classification (Table 4). When a medical institution has more than one category of beds, it was grouped according to the highest category.
As a result, the percentage of the hospitalization cost for target disorders was the highest in psychiatric beds that do not belong in any of the above categories (59.1%), followed by the convalescent phase (24.2%) and the chronic phase (7.0%) categories. For these 3-bed categories that used most of the hospitalization cost as above, the hospitalization period per month was 26.2 days or longer, which was markedly longer than the hospitalization period for beds in other categories. Among the inpatients with the target disorders, 283 inpatients (19.6% of the total inpatients) were elderly (≥ 70 years), most of whom were hospitalized in one of these 3-bed categories. Of note, 8 patients whose deaths were recorded in the health insurance claim sheet were hospitalized in one of the above 3 bed categories having long hospitalization periods and had stayed in the hospital until their deaths.
For the medical cost per month, the cost was the high for the advanced acute phase and acute phase categories, for which highly intensive treatments are administered.
Table 4. Hospitalization by hospital functional classification.
3.5. Concurrent Disorders of Patients with Mental Disorders
The 30 medical conditions that are most frequently reported in the health insurance claim sheet of the patients with target mental disorders are summarized in Table 5. To specify the condition, we used the minor classification of ICD10 disorder name. Commonly observed conditions were insomnia, a condition often observed as a complication of mental disorders, and constipation, chronic gastroenteritis, diabetes/glucose tolerance abnormal, headache, deformation/ fracture of joint/spine, obesity, and drug-induced hepatic/renal disorders, which are conditions often observed as adverse reactions to antipsychotics.
Frequently reported disorders included lifestyle-related diseases such as hypertension, dyslipidemia, diabetes, hyperuricemia and obesity.
3.6. Diseases of Patients on Hemodialysis
Finally, Japan has the second highest density of patients requiring hemodialysis therapy in the world  . As renal failure is possible to manage, we analyzed patients requiring hemodialysis.
For the patients on hemodialysis among the residents covered by national health insurance in City A, risks of deterioration in renal function and complications were analyzed. Although temporal or causal relations between the hemodialysis and onset of the disease could not be analyzed in our study, we performed an analysis to identify the characteristics of this group (Table 6). Among these residents, 411 patients received hemodialysis, and common diseases observed among these patients were primary renal impairment, and secondary renal impairment associated with lifestyle diseases, such as hypertension, diabetes, and dyslipidemia. Patients with mental disorders frequently received hemodialysis as well.
4.1. Long Hospitalization and Changes in Treatment Environments for Mental Disorders
We analyzed several aspects of targeted mental disorders. It was revealed that
Table 5. Top 30 disorders that the patients with target mental disorders have.
patients in their 40 s accounted for the highest percentage of outpatients, while those in their 60 s accounted for the highest percentage of inpatients and were in the hospital for a long period. Also, it was revealed that many patients with mental disorders had complications with more than one mental or physical disorders and used medical cost for these comorbidities in addition to the underly-
Table 6. Top 20 diseases observed in patients on dialysis.
ing mental problems. Moreover, quite a few of them received treatment for the highly acute phase.
These findings depict that some patients live with several mental and physical problems. From their 40s and onwards, patients tend to need hospitalization because they require inpatient management for their mental and physical disorders. Sometimes, they shift from outpatient visits to hospitalization as they have difficulty in visiting hospitals and receiving care after being discharged, leading to prolongation of hospitalization. In our analysis, follow-up of the outcome was limited because it was based only on health insurance claim data and we found few records of death. However, we estimate that actual data will show that more end-of-life care has been administered in hospitals.
According to a report that describes the medical and societal costs for Japanese patients (age ≥ 20 years) with depression or schizophrenia, treatment cost of mental disorders accounts for 9% to 28% of the total costs, while societal cost associated with absenteeism (absent from work) or presenteeism (lowered productivity) accounts for 47% to 67% and mortality cost accounts for 45% to 47%   . Regarding societal cost, discussions are also taking place in abroad   . The survey by the Ministry of Health, Labour and Welfare (2014) reported that among patients who are hospitalized for an extended period, those who need urgent support for their daily or social life in addition to disease management and care often lacks access to these supports  . On the other hand, the patients who need relatively less support have to spend a long time in training for jobs and self-management and can obtain few medical services in the community such as home-visiting nursing care, visiting medical care, day care, or care for their physical conditions. Such lacks of medical management and care services are leading to longer hospitalizations.
4.2. Comprehensive Assessment of the Mental/Physical Health of Patients with Mental Disorders
Our study revealed that many patients with mental disorders have physical problems as well. Although temporal/causal relation between diseases was not clarified in our analysis, interaction between mental and physical conditions hinders proper control of the systemic condition and leads to lowered QOL and risks of death. The diseases commonly observed in the patients included in our analysis were symptoms associated with the underlying mental disorders, common adverse reactions to antipsychotics, and lifestyle-related diseases. Since we have universal coverage and a free access healthcare system in Japan, not all patients with mental disorders are followed up by specialists. It is important for medical professionals to comprehensively assess the mental/physical and living conditions of the patients as well as the available social resources to help them have access to better self-management education and medical care services. Patients with mental disorders often use several drugs. Medical professionals should recommend them to consult specialists when they show symptoms associated with their mental disorders or common adverse reactions to antipsychotics so that they can switch to a more appropriate treatment, dose, or non-phar- maceutical therapy.
Furthermore, Japan has the highest aging rate in the world (WHO, 2010), and the aging rate in 2015 is 26.7% (Ministry of Health, Labor and Welfare, 2015)   . The elderly population rate is expected to rise in the future, and accordingly, an increase in the number of elderly patients with dementia is estimated  . It is reported that patients with dementia have high prevalence of physical diseases   , and the home treatment becomes more difficult as the severity of the physical diseases increases    . In order to realize the dementia patient’s familiar living place and to improve patient's QOL, it is necessary to analyze the cost for mental disorders further and to improve service and infrastructure in the future.
In this study, there is a limit to generalization because the subject is covered by national health insurance in one area. However, the results of clarifying the health resource utilization and comorbidities in patients with mental disorders can be implemented to improve the health services and systems. It is necessary to expand the number of areas and sample size for further analysis.
Our study revealed that patients with mental disorders are shifting from outpatient treatment to hospitalization as they get older. Patients in general are often hospitalized for an extended period, and they have many concurrent mental/physical problems. Therefore, comprehensive assessment of their mental/physical health, self-management education suitable for each patient, organization of services, and support for decision making regarding treatment are essential for the improvement of their QOL.
We thank City A for providing us the data and Data Horizon Corporation for helping us with processing health insurance claim sheets. We obtained funding from City A to conduct this study as a joint research with the city, but analysis was approved by the ethics committee of Hiroshima University.