Medical and psychiatric comorbidity in psychiatric beds in general hospitals: A cross-sectional study in Tokyo

Abstract

Aims:  Although somatic diseases in psychiatric patients are increasing with the increase of the aged population, psychiatric wards in general hospitals in Japan have progressively been decreasing. The purpose of this cross-sectional study was to clarify whether psychiatric beds in general hospitals play sufficient roles in medical comorbidities of psychiatric patients or not.

Methods:  This was a cross-sectional study performed all over Tokyo during the 2-month period from April to May 2007. The total number of patients who require admission due to both somatic and psychiatric diseases was investigated with their demographic and clinical characteristics.

Results:  The total number of patients admitted to psychiatric beds in general hospitals for the above-mentioned reason was 326, while the number of patients who could not be admitted to them despite the same reason was 88. The rate of surgical diseases in the latter group was higher than that in the former group. In the latter group, diseases requiring orthopedic surgery (22%) and abdominal surgery (22%) were the most frequent, followed by gastrointestinal and hepatic diseases (8%), and gynecological diseases (7%). Patients who had attempted suicide were included more in the latter group than in the former group. Even in the former group, general hospitals could not respond to 34% of requests for emergency admission.

Conclusion:  Psychiatric beds in general hospitals do not necessarily function for medical comorbidities in psychiatric patients, especially in severe and emergency cases. Not only the quantity but also the quality of psychiatric wards in general hospitals should be reconsidered.

ALTHOUGH SOMATIC DISEASES in psychiatric patients are increasing with the increase of the aged population, psychiatric wards in general hospitals in Japan have progressively been decreasing. For example, 9.3% of psychiatric beds in general hospitals in Japan disappeared during the 3 years from 2002 to 2005.¹ In such a situation, the limitation of access to medical care in psychiatric patients may increase.

Although some studies have examined medical comorbidities among psychiatric inpatients, results have not been from population-based designs, but from hospitals only.^2,3 Another study was population-based, but focused on mortality among psychiatric outpatients.⁴ Few cohort studies have thus been performed in which the severities of both somatic and psychiatric diseases are equivalent to the level of admission required.

The purpose of this cross-sectional study in Tokyo was to clarify whether psychiatric beds in general hospitals play sufficient roles in medical comorbidities of psychiatric patients or not. Such epidemiological data in Tokyo, where there are approximately 12 million inhabitants, might contribute to public policy not only in Japan but also in other countries.

METHODS

1. Design

This was a cross-sectional study performed all over Tokyo during the 2-month period from April 1 to May 31 2007.

In general, when medical or surgical diseases occur in patients with severe psychiatric symptoms, patients are treated in psychiatric beds in general hospitals. However, it is difficult to arrange such medical care for psychiatric emergency patients during the night and holidays, or for inpatients of psychiatric hospitals.

In Tokyo, the medical comorbidity service for inpatients of psychiatric hospitals is provided by the Tokyo Metropolitan Government. Five hospitals (204 beds) are responsible for this system (P-GHP). Moreover, the psychiatric emergency service provided by the Tokyo Metropolitan Government practically covers medical or surgical diseases occurring in psychiatric emergency patients during the night and holidays. Three hospitals (12 beds) are responsible for this system (N-GHP). Other situations, i.e. during the daytime on weekdays, patients who require admission due to both somatic and psychiatric diseases are ordinarily admitted to a psychiatric ward in one of 28 general hospitals (1135 beds, D-GHP). There were no beds overlapping with each other in these three systems. The total number of patients who require admission due to both somatic and psychiatric diseases in these three systems means the total number of those in Tokyo.

Twenty-one of 28 D-GHP, which correspond to 75.2% of psychiatric beds of all the 28 D-GHP, participated in the study. Moreover, all of the three N-GHP (100%) and all of the five P-GHP (100%) participated in the study.

The study protocol was approved by the institutional review board of Juntendo University School of Medicine. The approved protocol did not require informed consent from patients because the data in this observational study remained anonymous and were analyzed in the aggregate.

2. Data collection

The subjects of this study were patients who required admission due to both somatic and psychiatric diseases. Psychiatric patients comorbid with somatic diseases, patients with somatic diseases who had recently contracted psychiatric diseases, and patients whose acute psychiatric symptoms were caused by somatic diseases, were all included, although patients who did not require admission were not included. Data collection was consecutive.

The information collected included the following: (i) demographic characteristics such as age and gender; (ii) medical or surgical diagnoses; (iii) psychiatric discharge diagnoses according to ICD-10; (iv) the level of emergency, e.g. need to be admitted within the day of request, within 2 days, within a week, or no need for haste; (v) the duration of waiting for admission; (vi) kinds of facilities where patients are introduced; (vii) suicidal behavior; (viii) restraint or seclusion; (ix) the Excited Component for the Positive and Negative Syndrome Scale (PANSS-EC: Excitement, Hostility, Tension, Uncooperativeness, Poor impulse control) and Lack of Judgment and Insight;^5,6 and (x) the length of hospital-stay for the somatic diseases.

At the same time, we surveyed cases that could not be admitted to psychiatric beds in general hospitals despite the existence of somatic diseases requiring admission. The information collected includes the following: (i) demographic characteristics such as age and gender; (ii) suspected medical or surgical diagnoses; (iii) suspected psychiatric diagnoses according to ICD-10; (iv) suicidal behavior; and (v) the reason that the patient could not be admitted.

3. Analyses

Differences between categorical variables were calculated using Fisher's exact test. Differences between sequential variables were calculated using the Student's t-test. The statistical test was two-tailed. A P-value of less than 0.05 was regarded as statistically significant.

RESULTS

1. Demographic and clinical characteristics of patients who were admitted to psychiatric beds in general hospitals due to both somatic and psychiatric diseases requiring admission

The number of patients who were admitted to D-GHP during the study period was 997. Of the 997 patients, 174 patients (17.5%) were due to both somatic and psychiatric diseases. The number of patients who were admitted to N-GHP during the study period was 242. Of the 242 patients, 10 patients (4.1%) had accompanying somatic diseases requiring admission. The number of patients who were admitted to P-GHP due to both somatic and psychiatric diseases during the study period was 142. Accordingly, the total number of patients who were admitted to psychiatric beds in general hospitals due to both somatic and psychiatric diseases during the study period was 326 (Table 1). The mean age of the 326 patients was 61.7 years (SD 16.2, Table 1). Forty-eight percent of these patients were 65 years or older and 46% were male.

Table 1.  Characteristics of psychiatric patients with medical comorbidities who were admitted to psychiatric beds in general hospitals and of those who could not be admitted

Group	n	Age*** (year)	Gender (%, men)	Diagnosis (%)			Suicidal attempt*** (%)	Diseases requiring surgery** (%)
				F0	F2	F3
Data are presented as percentage except age (mean [SD]). Diagnoses were made with ICD-10. **P < 0.001, ***P < 0.0001. Admitted, psychiatric patients with medical comorbidities who were admitted to psychiatric beds in general hospitals; F0, organic, including symptomatic, mental disorders; F2, schizophrenia, schizotypal and delusional disorders; F3, mood disorders; Not admitted, psychiatric patients with medical comorbidities who could not be admitted to psychiatric beds in general hospitals.
Admitted	326	61.7 [16.2]	46	28	40	15	10	40
Not admitted	88	48.0 [20.2]	48	22	41	16	35	58

Medical or surgical diagnoses of the 326 patients at discharge are shown in Table 2. Of these, 194 patients (60%) were medical, and the remaining 132 patients (40%) were surgical. Respiratory diseases were the most frequent (19%), followed by diseases requiring orthopedic surgery (13%), diseases requiring abdominal surgery (10%), and gastrointestinal and hepatic diseases (10%). Detailed descriptions of frequent somatic diseases were in the following order: pneumonia (14%), femoral neck fracture (6%), diabetes mellitus (4%), cerebral infarction (3%), and colon cancer (3%).

Table 2.  The list of medical and surgical diseases of psychiatric patients who were admitted to psychiatric beds in general hospitals

Disease	n (%)	Detailed descriptions (n)
Medical
Respiratory	61 (19%)	Pneumonia (44), Pulmonary tuberculosis (4), Pneumothorax (3) Acute exacerbation of chronic respiratory failure (3) Acute interstitial pneumonia (2), Others (5)
Gastrointestinal and hepatic	32 (10%)	Ileus (9), Hematemesis and melena (7), Liver cirrhosis (3) Alcoholic liver disease (2), Acute pancreatitis (2), Others (9)
Neurological	25 (8%)	Cerebral infarction (11), Parkinson's disease (3), Herpes encephalitis (2) Status epileptics (2), Others (7)
Endocrinological	22 (7%)	Diabetes mellitus (14), Others (8)
Nephrological	16 (5%)	Acute renal failure (7), Chronic renal failure (2), Others (7)
Hematological	15 (5%)	Anemia (5), Malignant lymphoma (3), Others (7)
Cardiological	10 (3%)	Acute heart failure or acute exacerbation of chronic heart failure (6) Others (4)
Collagen disease	3 (1%)	Systemic Lupus erythematodes (1), Others (2)
Infectious disease	3 (1%)	HIV (2), Neurosyphilus (1)
Others	7 (2%)
Surgical
Orthopedic	42 (13%)	Femoral neck fracture (19), Pelvic fracture (5), Vertebral fracture (4) Finger tendon injury (3), Other fractures (9), Others (2)
Abdominal	32 (10%)	Colon cancer (10), Gastric cancer (5), Acute cholecystitis (5) Liver cancer (2), Abdominal stab wound (2) Perforation of gastrointestinal tract (2), Others (6)
Gynecological	15 (5%)	Breast cancer (8), Cesarean section (3) Carcinoma of the uterine cervix (2), Ovarian cancer (2)
Neurosurgery	14 (4%)	Subdural hematoma (7), Brain contusion (3), Brain tumor (3) Others (1)
Chest	10 (3%)	Lung cancer (8), Others (2)
Dermatological	7 (2%)	Pressure ulcer (2), Others (5)
Urological	6 (2%)	Bladder cancer (1), Prostatic cancer (1), Others (4)
Ophthalmological	5 (2%)	Cataract (4), Other (1)
Oral surgery	1 (0%)	Osteomyelitis (1)
Total		326

Psychiatric diagnoses (ICD-10) of the 326 patients at discharge were distributed as follows (Table 1): organic, including symptomatic, mental disorders (F0), 28%; mental and behavioral disorders due to psychoactive substance use (F1), 7%; schizophrenia, schizotypal, and delusional disorders (F2), 40%; mood disorders (F3), 15%; neurotic, stress-related and somatoform disorders (F4), 2%; behavioral syndromes associated with physiological disturbances and physical factors (F5), 1%; disorders of adult personality and behavior (F6), 2%; mental retardation (F7), 4%; epilepsy (G40), 1%. Detailed descriptions of the category F0 were as follows: dementia and organic amnesic syndrome (F00-F04), 18%; delirium and others (F05-F07), 10%.

Discrepancy between the level of emergency and the duration of waiting for admission was as follows. The total number of patients who need to be admitted immediately due to severe somatic diseases was 88. Of these, 30 patients could not be admitted within the day (34%). Similarly, the total number of patients who needed to be admitted within 2 days of request due to the severity of somatic disease was 66. Of these, 14 patients could not be admitted within 2 days (21%). The total number of patients who need to be admitted within a week of request due to the severity of somatic diseases was 84. Of these, 20 patients could not be admitted within a week (24%).

Among the 326 patients, 42 patients (13%) were associated with suicide. Thirty-two patients (10%) had attempted suicide (Table 1), and 10 patients (3%) had had obvious suicidal ideation.

Physical restraints were used in 114 patients (35%) among the 326 patients, and seclusion rooms were used in eight patients (3%).

The median scores of PANSS-EC and Lack of Judgment and Insight in the 326 patients were 13 (range 5–35) and 4 (range 1–7), respectively.

The median length of hospital-stay was 28 days.

2. Demographic and clinical characteristics of patients who could not be admitted to psychiatric beds in general hospitals despite the existence of both somatic and psychiatric diseases requiring admission

The number of patients who could not be admitted to D-GHP despite the existence of both somatic and psychiatric diseases requiring admission during the study period was 88 (Table 1). The rate of rejection of admission to D-GHP despite the existence of both somatic and psychiatric diseases requiring admission was 34%. Meanwhile, there were no patients who could not be admitted to the N-GHP or P-GHP. The mean age of the 88 patients was 48.0 years (SD 20.2, Table 1). Twenty-nine percent of these patients were 65 years or older and 48% were male.

Medical or surgical diagnoses of the 88 patients are shown in Table 3. Of these, 37 patients (42%) were medical, and the remaining 51 patients (58%) were surgical. Remarkably, the rate of surgical diseases in patients who could not be admitted to psychiatric beds in general hospitals was higher than that in patients who were admitted (Relative Risk = 0.70, P = 0.0038, Table 1). Diseases requiring orthopedic surgery (22%) and abdominal surgery (22%) were the most frequent, followed by gastrointestinal and hepatic diseases (8%), and gynecological diseases (7%). Detailed descriptions of frequent somatic diseases were in the following order: multiple fractures (7%), acute heart failure (6%), and femoral neck fracture (5%).

Table 3.  The list of medical and surgical diseases of psychiatric patients who could not be admitted to psychiatric beds in general hospitals

Disease	n (%)	Detailed descriptions (n)
Medical
Gastrointestinal and hepatic	7 (8%)	Ileus (2), Hematemesis and melena (2), Hepatic encephalopathy (2) Others (1)
Neurological	5 (6%)	Encephalitis (3), Consciousness disturbance (2)
Nephrological	5 (6%)	Chronic renal failure (3), Malignant syndrome (2)
Cardiological	5 (6%)	Acute heart failure or acute exacerbation of chronic heart failure (5)
Others	15 (17%)
Surgical
Orthopedic	19 (22%)	Multiple fractures (6), Femoral neck fracture (4), Other fractures (9)
Abdominal	19 (22%)	Abdominal stab wound (3), Inguinal hernia (3), Colon cancer (3) Esophageal cancer (2), Acute cholecystitis (2), Acute appendicitis (2) Others (4)
Gynecological	6 (7%)	Cesarean section (3), Ovarian cancer (2), Other (1)
Neurosurgery	3 (3%)	Head injury (2), Other (1)
Chest	2 (2%)	Lung cancer (2)
Dermatological	2 (2%)	Pressure ulcer (2)
Total	88

Psychiatric diagnoses (ICD-10) of the 88 patients were distributed as follows: F0, 22%; F1, 3%; F2, 41%; F3, 16%; F4, 10%; F6, 6%; and F7, 2%. Detail descriptions of the category F0 were as follows: F00-F04, 3%; F05-F09, 18%.

Among the 88 patients, 33 patients (38%) were associated with suicide. Thirty-one patients (35%) had attempted suicide, and two patients (2%) had had obvious suicidal ideation. Remarkably, the rate of having attempted suicide in patients who could not be admitted to psychiatric beds in general hospitals was higher than that in patients who were admitted (Relative Risk = 0.28, P < 0.0001, Table 1).

The reasons for refusal of admission despite the existence of both somatic and psychiatric diseases requiring admission were as follows: ‘psychiatric beds were full’, 68%; ‘impossible to manage such a patient because of an open ward’, 7%; ‘no specialist corresponding to the somatic disease’, 5%; ‘impossible to perform an operation without delay’, 3%; ‘others’, 17%.

DISCUSSION

1. Characteristics of patients who require admission to psychiatric beds in general hospitals due to both somatic and psychiatric diseases

The mean age of 61.7 years (SD 16.2) in the 326 patients who were admitted to psychiatric beds in general hospitals due to both somatic and psychiatric diseases may reflect the increase of the aged population. Unfortunately, there have been no similar reports with population-based design, so it is impossible to compare this result with others.

In the present study, the most frequent medical and surgical diseases were pneumonia and femoral neck fracture, respectively. To our knowledge, this appears to be a new finding with respect to the needs for admission to psychiatric beds in general hospitals. So far, a wide range of comorbidity has been described, with chronic medical illnesses such as hypertension, heart disease, pulmonary disease, and diabetes.^7,8 However, patients with such comorbidity do not necessarily require admission because of the inclusion of chronic state. Meanwhile, in a mortality survey, heart diseases and suicide have been reported as the leading causes of death.³ However, patients under such imminent conditions may be brought in not to a general hospital psychiatric unit but to a cardiac care unit or a critical care center. Therefore, the survey of mortality cannot estimate the needs for psychiatric beds in general hospitals. Thus, there are obvious differences in kinds of diseases between reports focusing on comorbidity and those focusing on mortality, both of which do not seem to reflect the needs for admission to psychiatric beds in general hospitals.

With respect to psychiatric diagnoses, schizophrenia and associate disorders (ICD-10: F2) was most frequent as previously reported.³ The second highest frequency was organic mental disorders (F1). In particular, the high frequency of dementia (F00-F04) was remarkable, indicating the increased aged population.

Although the median score of PANSS-EC was not so high, that of Lack of Judgment and Insight was moderate. Lack of judgment and insight may have disturbed the treatment of somatic diseases, which may explain the high frequency of physical restraints.

The reason for the finding of a significantly higher mean age of patients who could be admitted compared with patients who could not be admitted is unclear. In general, the average age of patients admitted to psychiatric hospitals is high. In the present study, 142 patients were admitted to P-GHP (to which patients from psychiatric hospitals were admitted), while there was no patient who could not be admitted to P-GHP. Inclusion of such aged population in the group of patients who could be admitted may have caused this difference in age.

2. Do psychiatric beds in general hospitals function in quality?

It is remarkable that the rate of having attempted suicide in patients who could not be admitted to psychiatric beds in general hospitals was higher than that in patients who were admitted (Table 1). The finding suggests that psychiatric beds in general hospitals did not necessarily accept psychopathologically severe cases with medical comorbidities. The findings that the median score of PANSS-EC was not so high, and that the median score of Lack of Judgment and Insight was moderate, may support this.

It is also remarkable that the rate of surgical diseases in patients who could not be admitted to psychiatric beds in general hospitals was higher than that in patients who were admitted (Table 1). The finding suggests that psychiatric beds in general hospitals did not necessarily accept medically severe cases with psychiatric diseases.

Furthermore, it is serious that 34% of patients who needed to be admitted immediately due to severe somatic diseases could not be admitted within the day. The finding suggests that psychiatric beds in general hospitals did not necessarily accept emergency cases with medical comorbidities.

Thus, psychiatric beds in general hospitals do not necessarily function for medical comorbidities in psychiatric patients, especially in severe and emergency cases.

3. Do psychiatric beds in general hospitals function in quantity?

The incidence of medical comorbidity for which psychiatric patients should be hospitalized appears to be at least 25 per 100 000 inhabitants in Tokyo, as described elsewhere.⁹ As there are approximately 12 million inhabitants in Tokyo, 3000 patients may require admission to psychiatric beds in general hospitals due to both somatic and psychiatric diseases. As the median length of hospital-stay was 28 days, the number of patients who can utilize one bed is calculated on 13.0 per year. As the estimated number of patients is 3000 per year in Tokyo, the number of beds needed is calculated as 231.

Unexpectedly, this number is much smaller than the total number of psychiatric beds in general hospitals in Tokyo (1135 beds). However, the rate of rejection of admission to D-GHP despite the existence of both somatic and psychiatric diseases requiring admission was not low (34%). An explanation is that the major roles of psychiatric beds in general hospitals include not only medical comorbidities but also electroconvulsive therapy, differential diagnosis using neuroimaging methods, and emergency cases with abnormal physiological conditions.¹⁰ Therefore, it is not conclusive that the total number of psychiatric beds in general hospitals in Tokyo may be enough for medical comorbidities.

In order to provide a medical care service for psychiatric patients efficiently, research concerning the roles of psychiatric units in general hospitals, i.e. not only medical comorbidities but also other functions mentioned above, will be needed. Thereafter, the priority among various roles of psychiatric units in general hospitals will be determined, which will lead to the provision of an efficient medical care service for psychiatric patients.

The strength and weaknesses of this study bear discussion. The strength of our study is that it included all psychiatric patients who lived in a defined area during the study period. One limitation is that our finding may represent only a metropolis. Therefore, a similar study in provincial areas is needed. Another limitation is that our findings may represent only the mental health system in Japan. However, such findings may have significance for comparison of differences in mental health systems among countries. Such a population-based study should be conducted regularly in order to improve the treatment of somatic diseases in psychiatric patients and to clarify the function of a general hospital psychiatric unit.

ACKNOWLEDGMENTS

The authors thank Dr Hirotsugu Kikumoto, Dr Ichiro Masudomi, Mr Ichiro Yoshida, and Ms Yukiko Masuda (Tokyo Metropolitan Government), Dr Kunitoshi Hato (Japanese Association of Mental Health Services), Dr Kunihiro Isse and Dr Mitsuru Nakamura (Tokyo Metropolitan Toshima General Hospital), Dr Takao Nishimura and Dr Hikaru Furuta (Tokyo Metropolitan Fuchu General Hospital), Dr Hiroshi Suwa (Tokyo Metropolitan Ebara General Hospital), Dr Mitsuo Suyama (Tokyo Metropolitan Bokuto General Hospital), Dr Hiroshi Umezu (Tokyo Metropolitan Matsuzawa Hospital), Dr Yoshihiro Yahiro (Tama Saisei Hospital), Dr Naohiro Fujimura (Tokyo Musashino Hospital), Dr Hiroshi Mitsushio (Ohme-City General Hospital), Dr Hirokatsu Kono (Kyosai Tachikawa General Hospital), Dr Masahiro Shintani (Tokyo Metropolitan Hiroo General Hospital), Dr Takashi Takeuchi (Tokyo Medical and Dental University Hospital), Dr Joichiro Shirahase (Keio University Hospital), Dr Ken Inada (Tokyo Women's University Hospital), Dr Amane Tateno (Nippon Medical School Hospital), Dr Shuichi Katsuragawa (Toho University Omori Medical Center), Dr Tsuyoshi Akiyama (Kanto Medical Center NTT EC), Dr Hiroki Kocha (National Hospital Organization Tokyo Medical Center), Dr Tomomichi Kameyama (Tokyo Teishin Hospital), Dr Jin Habu (JSDF Central Hospital), Dr Yukako Seki and Dr Kobun Imai (International Medical Center of Japan), Dr Kazunori Nakajima (Sanraku Hospital), Dr Takeo Muraki (JR Tokyo General Hospital), Dr Yosuke Ichimiya, Dr Yoichiro Matsubara, and Dr Ryo Kumagai (Juntendo Tokyo Koto Geriatric Medical Center), for collecting the data.

This work was supported by a grant from the Ministry of Health, Welfare, and Labor of the Japanese Government (Research on Psychiatric and Neurological Diseases and Mental Health, H19-009).