Impact of social-psychiatric services and psychiatric clinics on involuntary admissions

Abstract

Background:

Germany provides a wide range of highly developed mental health care to its citizens. The aim of this study was to identify factors influencing the voluntariness of admissions to psychiatric hospitals. Especially the impact of demographic factors of the region, characteristics of the psychiatric hospitals and characteristics of the psychosocial services was analyzed.

Method:

A retrospective analysis of hospital admission registers from 13 German adult psychiatric hospitals in 2009 was conducted. Public data on the regional psychiatric accommodation and demographic situation were added. Hospitals were dichotomously divided according to their index of involuntary admissions. Group comparisons were performed between the clinics with low and high involuntary admission indices. Analysis was conducted with clinical, psychiatric provision and demographic data related to inpatients in the Landschaftsverbands Westfalen-Lippe (LWL)-PsychiatryNetwork.

Results:

Especially the range of services provided by the social-psychiatric services in the region such as number of supervised patients and home visits had an influence on the proportion of involuntary admissions to a psychiatric hospital. Some demographic characteristics of the region such as discretionary income showed further influence. Contrary to our expectations, the characteristics of the individual hospital seem to have no influence on the admission rate.

Conclusion:

Social-psychiatric services show a preventive impact on involuntary acute psychiatry interventions. Sociodemographic factors and patient variables play a role with regard to the number of involuntary hospitalizations, whereas characteristics of hospitals seemed to play no role.

Keywords

Psychiatric epidemiology social-psychiatric services psychiatric clinics involuntary admission risk factors

Introduction

Since the Psychiatry Enquête Report (1975) on the status of the Psychiatry in Germany was published by the lower house of the German parliament in 1975, a fundamental improvement of the mental health care in Germany has been observed. An integral part of the improvement was the creation of a nationwide network of day clinics and outpatient centers. Emphasis was on leveraging the range of ambulatory health care, in particular social-psychiatric services, flat-sharing communities, day care centers and labor rehabilitative measures. One central aim of the changes in the psychiatric health-care system was the reduction of involuntary admissions to psychiatric hospitals.

Various surveys indicate that the number of involuntary hospitalizations of mentally ill people in Germany is increasing: in the late 1990s, Crefeld (2002) reported a doubling of the numbers since 1986 according to the administration of justice. Müller (2004) reported a drastic increase of involuntary admissions based on a study by Darsow-Schütte and Müller (2001) as well as Müller and Josipovic (2003). The data collection of these studies was done in a small section of Germany. Furthermore, Marschner and Volckart (2001) documented an increase of involuntary hospitalizations using interpretations of statistical data provided by judicial authorities on civil law or public law-based hospitalizations. In a pan-European study, Salize and Dreßing (2004) likewise stated an increase of the total figures for involuntary hospitalizations in Germany. However, there are studies with contrary findings. Spengler (2007) did not find evidence for a strong rise of compulsory admissions in Germany, but suspected the number to stabilize on a high level. Related to the statistics of the German district clinic of Regensburg between 1990 and 2003, Cording and Binder (2004) could not confirm increasing numbers of involuntary hospitalizations.

Legal foundations of the involuntary admission are the laws for psychiatric illnesses of the respective federal states of Germany called PsychKG and BtG. The PsychKG implies that a forced hospitalization of adults is only permitted if and as long a current considerable risk of self-endangerment or endangerment of others exists, which is due to the persons behavior caused by illness, and if the danger cannot be averted otherwise (Dodegge & Zimmermann, 2003; Hoffmann, 2009). The involuntary hospitalization is a temporary measure, ending either after a certain period which has been set by a higher authority that has judged the comparativeness or based on the patients consent to treatment.

When compared internationally, Germany operates a widely developed psychiatric provision. This relates to the stationary sector (hospitals, clinics) as well as to the outpatient (day clinics) and ambulatory sector (doctor’s surgeries, outpatient clinics, outreach clinics; Statistisches Bundesamt, 2011). The clinical stationary sector of the mental health care in Germany is regionalized. Each region is obliged to facilitate a sufficient provision. Throughout Germany, the structure, size and scope of the social-psychiatric services differ considerably. Being the base service, this level is responsible to accommodate patients according to PsychKG. It covers crisis intervention, and the personalized coordination of required assistance, in particular for people suffering from chronic mental illness or chronic addicts (Psychiatrie in Deutschland, 2003).

The structural alignment of the different psychiatric provision sectors and region-specific demographic parameters referring to the number of involuntary hospitalizations could provide new findings on the increase of the involuntary hospitalizations. The reduction of inpatient stays from an average of 65 days in 1991 to 23 days in 2008 (Spengler, Dreßing, Koller, & Salize, 2008) and changing clinical structures (flow of work in the daily clinic routine, personnel, specialist divisions etc.) as well as demographic factors (population figures, rate of unemployment, disposable income etc.) could influence the number of involuntary hospitalizations.

Demographic aspects are an important factor for the social environment of humans. The influence of the environment, the economic situation and also the educational background of the people has changed over time (Statistisches Bundesamt, 2011). The dependency of the psychiatric care from sociodemographic characteristics was already detected by Creed, Tomenson, Anthony, and Tramner (1997) and Kisely, Preston, and Rooney (2000). They analyzed the severity of psychiatric illness with regard to various variables. Hudson (2005) confirmed these results; the study showed the direct influence of the socioeconomic status on the level of psychiatric illnesses.

Purpose of this study is to determine the influence of institutional characteristics of the clinics and the social-psychiatric services on the legal status of the hospitalization of a patient. Furthermore, the influence of demographic characteristics of the region on admissions was taken into consideration.

Material and methods

The Landschaftsverbands Westfalen-Lippe (LWL)-PsychiatryNetwork is the biggest provider of psychiatric services in Germany with a catchment area of 8.5 million people in North Rhine-Westphalia. It consists of 27 administrative districts with 13 psychiatric hospitals for adults covering both rural and urban areas. Quantitative data from hospitals’ admission registers about all inpatients from 2004 to 2009 were provided by the LWL.IT. In all, 230,678 cases were registered in this period. Collected variables were date of admission and discharge, length of stay, department of admission, diagnoses, legal basis, mechanical restraints (amount and duration), number of previous admissions, clinic, year, gender, date of birth, age at date of admission and nationality.

Questionnaires for medical and therapeutic treatment in clinics and for the general psychiatric accommodation by social-psychiatric services were developed in order to acquire data from the clinics and the social-psychiatric services.

The interpretation of the data of voluntary and involuntary hospitalization in psychiatric clinics requires background information on accommodation and treatment and the structure of the clinic. Therefore, questionnaires for medical and therapeutic treatment in clinics and for the general psychiatric accommodation by social-psychiatric services were developed. The information concerning the clinics was collected in 13 clinics of the LWL-PsychiatryNetwork. The response rate of the clinics was 12 out of 13. To assess the local supplementary structure, the social-psychiatric services were also interviewed regarding their structure and financial and human resources. The response rate of the psychosocial services was 8 out of 14. The collection of the demographic data was carried out from different databases like ‘Landesdatenbank Nordrhein-Westfalens’ and ‘Wegweiser Kommune der Bertelsmann Stiftung’ and information from publications of the corresponding cities and communities. The data were collected in 2009 (Bertelsmann Stiftung, 2011; Stadt Bochum, 2010).

The evaluation of the data of treated patients in the LWL-PsychiatryNetwork was anonymized by Hash-MD5 Code. The retrospective analysis of hospital admission registered from 11 adult psychiatry hospitals over 6 years (2004–2009) was conducted. The single data record of the adult psychiatry was adopted on a core matrix. The total sample (N = 230,678) was divided in two sub-samples, voluntary admission group (N = 196,389) and involuntary admission group (N = 17,206). Admissions based on guardianship and other reasons were not relevant for the analysis and were excluded from the sample (N = 17,083).

An involuntary admission index was calculated by relating the total number of admissions to the number of PsychKG-related admissions per clinic in 2009. Based on this quotient, a classification of the psychiatric clinics was made in clinics with low quotient of involuntary admission (0.00–0.05) and high quotient of involuntary admission (0.06–0.15). Therefore, the thresholds shown are specific to the analyzed scope of the study and may differ from similar setups in other regions or different time periods. In order to investigate underlying factors for the involuntary admission index, the initial classification (low/high index figure) has been established.

Characteristics of clinics

The clinic questionnaire inquires the organizing institution, the catchment area of the clinic, the structure of the clinic and the existence and practice of involuntary admission. Also human resources, patient treatment and cooperation with other psychiatric accommodations were investigated.

Characteristics of social-psychiatric services

For the data inquiry of the psychosocial services, a questionnaire was created, which included items on the existence of psychosocial accommodations, cooperation of the different psychosocial accommodations and psychiatric clinics, exposure to involuntary admission, human resources and their structure and accommodation and care of the patients.

Demographic characteristics of the regions data

It contained data of population, population density, rate of unemployment and social benefit, percentage of people with migration background, index of education, discretionary income and living circumstances.

Statistics

The statistical analysis was progressed with IBM SPSS Statistics 20.0^®. Group comparisons between regions with low and high quotient of involuntary admissions were performed by t-test. A regression analysis was performed to predict the influence of demographic data on the involuntary admission index. Correlations between the average lengths of stay in the clinic with demographic data were assessed. Furthermore, bed capacity on the locked acute psychiatric ward and demographic data of the region were correlated.

Results

Involuntary admission index

The number of involuntary admissions in the 13 clinics in 2009 was between 53 and 1,006 admissions. This results in a median of 277.2 involuntary admissions on average (standard deviation (SD) = 266.4). Related to the total amount of admissions of the clinics, the rate of involuntary admissions turned out between 2.4% and 14.9%. The median of the share of involuntary admission was 6.9% in 2008 (SD = 4.3).

The analysis of the data of involuntary admission and admission resulted in a quotient of involuntary admission. The involuntary admission index was calculated by relating the total number of admissions to the number of involuntary admissions per clinic in 2009. According to the mean of the involuntary admission index, two types of clinics were defined. Based on this definition, eight clinics had a low quotient of involuntary admission (0.00–0.05) and six had a high quotient of involuntary admission (0.06–0.15).

Characteristics of clinics

The influence of the clinical characteristics on quotient of involuntary admission was examined. Due to the low sample size (N = 10), a statistical comparison was not reliable. Therefore, only descriptive analysis of the data was realized. Table 1 displays the characteristics of the clinic sub-divided in low and high quotient of involuntary admission.

Table 1.

Descriptive variables of characteristics of the clinic sub-divided in low and high quotient of involuntary admission.

	Clinics with involuntary admission index low (n = 8)	Clinics with involuntary admission index high (n = 6)
Number of inhabitants a regional mandatory care	M = 24,2687.5	M = 28,8550.0
	SD = 112,464.9	SD = 93,188.0
	t(10) = −0.700; p = .500
Admission ward
Yes, n (%)	6 (75)	3 (60)
No, n (%)	2 (25)	2 (40)
Proportion of closed stations to total ward	M = 0.168	M = 0.1892
	SD = 0.086	SD = 0.136
	t(12) = −0.363; p = .723
Total number of accommodations	M = 3797.6	M = 3877.8
	SD = 1911.6	SD = 1784.5
	t(9) = −0.072; p = .944
Total number of wards	M = 219	M = 333.5
	SD = 101.4	SD = 174.2
	t(12) = −1.544; p = .149
Length of stay in general psychiatry (days)	M = 28.7	M = 31.9
	SD = 4.2	SD = 5.4
	t(11) = −1.230; p = .244
Residence time in addiction medicine (days)	M = 12.0	M = 11.9
	SD = 2.5	SD = 1.3
	t(10) = 0.081; p = .937
Residence time in geriatric psychiatry (days)	M = 27.8	M = 29.1
	SD = 3.6	SD = 4.3
	t(10) = −0.586; p = .571

SD: standard deviation; M: mean.

The data show that the existence of an admission ward did not show an influence on involuntary admission. Furthermore, other characteristics of the clinics like length of admission, number of admission, admission capacity in involuntary wards and number of inhabitants in a regional mandatory care were considered with regard to low and high quotient of involuntary admission. Based on the multiple response option, a comparison of means with t-test was realized. No significant differences were detected (Table 1).

Also, in matters of the ambulant crisis intervention and cooperation with psychosocial facilities, there was no significant difference between the clinics with low and high quotient of involuntary admission considered (data not shown).

Characteristics of social-psychiatric services

The structure of social-psychiatric services was examined in terms of low and high quotient of involuntary admission. The analysis was carried out for 10 clinics and 8 social-psychiatric services. Due to the small sample size, a statistical comparison was not reliable. Therefore, a description of the evaluated data was realized. Table 2 displays the characteristics of social-psychiatric services. It turns out that some social-psychiatric service characteristics show an impact on the involuntary admission index. The number of supervised patients and the number of home visits executed by the social-psychiatric services indicated that in regions with a lower involuntary admission index, a lower amount of clients needed to be covered than in regions with a higher involuntary admission index (Table 2). The client–staff ratio of the social-psychiatric services showed similar values in both regions with low and high involuntary admission indices (Table 2). In regions with a low involuntary admission index, on average, a higher number of admissions were issued by the social-psychiatric services or the general practitioner than in regions with a high involuntary admission index. Regarding the admissions initiated by registered psychiatrists or psychiatrists in clinics, a contrary interdependency was detected (Table 2).

Table 2.

Characteristics of the social-psychiatric services and characteristics term of low and high quotient of involuntary admission.

	Involuntary admission index low	Involuntary admission index high
Number of supervised patients by social-psychiatric services per year	M = 2250	M = 1790
	SD = 1455.1; N = 3	SD = 575.4; N = 4
Client–staff ratio of the social-psychiatric services per 100,000 inhabitants	M = 2.23	M = 2.47
	SD = 1.75; N = 3	SD = 0.839; N = 3
Number of home visits by social-psychiatric services per year	M = 4,950	M = 1,682.3
	SD = 5,303.3; N = 2	SD = 1,201; N = 4
Percentage of medical certificates for involuntary admission by social-psychiatric services	M = 10.1	M = 4.4
	SD = 2.8; N = 3	SD = .15; N = 3
Percentage of medical certificates for involuntary admission by psychiatrists	M = 4.8	M = 15.1
	SD = 2.75; N = 3	SD = 22.9; N = 3
Percentage of medical certificates for involuntary admission by general practitioner	M = 19.4	M = 10.7
	SD = 17.7; N = 3	SD = 12.6; N = 3
Percentage of medical certificates for involuntary hospitalization by clinics	M = 50.4	M = 60.4
	SD = 26.6; N = 3	SD = 29.9; N = 3

SD: standard deviation; M: mean.

Demographic characteristics of the regions

The analysis of the sociodemographic data in the investigated area between 2004 and 2008 showed a reduction with regard to the total number of inhabitants, the population density, the share of immigrants, the unemployment rate and the homogeneity of incomes. The share of single-person households and the income of the private households increased between 2004 and 2008. With regard to education, a trend to higher school degrees was observed. While the share of special school degrees increased only slightly, the number of people without graduation remained stable. The percentage of secondary modern school degrees decreased from 37.8% to 18.2%. The share of general certificate of secondary education degrees grew from 29.6% to 38.6%, and the proportion of advanced technical college entrance qualification or university-entrance diploma increased from 28.6% to 38.7%.

The characterization of demographic parameters with regard to the quotient of involuntary admission occurred by t-test. Demographic parameters like population, population density, rate of unemployment and social benefit, percentage of foreigners, index of education, discretionary income and living circumstances were reviewed. The surrounding area of the psychiatric hospitals with high or low PsychKG quotient differed concerning the discretionary income only (t(6.758) = −3.071; p = .019; Table 3).

Table 3.

Demographic parameters of Germany and the surrounding area of the clinics with regard to the quotient of involuntary admission.

		Surrounding areas				Germany
		Involuntary admission index low		Involuntary admission index high
		M	SD	M	SD
Percentage of unemployment rate		9.163	3.4977	11.767	41951	7.8
		t_(9.691) = -1.233; p = .247
Population		272,180	105,815.393	422,875.33	156858,652	82,120,000
		t_(8.305) = -2,032; p= ,075
Percentage of foreigners		8.296	3.0199	9.458	3.6622	8.76
		t_(9.614)= -.633; p = .542
Discretionary income (Euro)		17,718.375	955.7121	20,405.333	1,976.8964	18,974
		t_(6.758) = -3.071; p = .019
Index of education (Percent)	low	4.533	.8733	4.750	1.5897	N/A^a
	low	t_(7.766) = -.293; p = .777
	middle	56.483	4.4960	54.267	8.4602	N/A^a
		t_{(7.616 )}= -.567; p = .587
	high	38.983	3.9458	40.983	9.5552	N/A^a
		t_(6.657) = -.474; p = .651

SD: standard deviation; M: mean.

Comparable data of the overall population are not available (N/A) for the specific scope.

Correlations

To verify the relation between the frequency of admissions based on PsychKG and demographic specifications of the coverage area, the data were examined using Pearson’s correlation analysis. For this purpose, the data of the LWL.IT from 2008 were used. The proportion of involuntary admissions was set into context to the demographic data and some characteristics of the hospitals such as average length of stay and the number of beds in locked wards.

Correlation of demographic data to the involuntary admission index

According to the correlation analysis (Table 4), a slightly negative correlation between age at the date of admission and the involuntary admission rate can be observed. This means, a higher proportion of younger patients are admitted in clinics with a higher involuntary admission index. By trend, patients with multiple previous admissions tend to be admitted in clinics with a higher involuntary admission index, and these clinics show on average a higher length of stay per patient.

Table 4.

Correlation of demographic data to the involuntary admission index (2008), length of stay and the share of secured treatment places. Not significant (NS) correlations are not displayed.

	Involuntary admission index	Average length of stay of the patients in the clinic	Share of secured treatment places
Admission age	r = −.037 (p < .001); N = 40,596	NS	NS
Number of prior admissions	r = .024 (p < .001); N = 40,599	NS	NS
Population	r = .790 (p < .001); N = 38,297	r = −.278 (p < .001); N = 177.920	r = .320 (p < .001); N = 177.920
Residence time	r = .033 (p < .001); N = 40,599	NS	NS
Population density	r = .240 (p < .001); N = 38,297	r = −.294 (p < .001); N = 177.920	r = .502 (p < .001); N = 177.920
Percentage of foreigners	r = .478 (p < .001); N = 38,297	r = −.760 (p < .001); N = 177.920	r = .070 (p < .001); N = 177.920
Rate of unemployment	r = .675 (p < .001), N = 38,297	r = −.538 (p < .001); N = 177.920	r = −.375 (p < .001); N = 177.920
Homogeneity of incomes	r = .364 (p < .001) N = 38297	NS	NS
Average income	NS	r = −.043 (p < .001); N = 139.623	r = −.273 (p < .001); N = 139.623

Furthermore, a tendency between a higher involuntary admission index and a higher population density and homogeneity of incomes can be observed. Moreover, a distinct positive correlation between population figure and involuntary admission index was detected (r = .790; p < .001; N = 38,297). The higher the population figure in a certain catchment area, the higher was the involuntary admission index of the respective hospital. There is also a positive correlation between involuntary admission index and share of immigrants as well as unemployment rate (r = .675; p < .001; N = 38,297). This implies the higher the unemployment rate and the share of immigrants, the higher the involuntary admission index. Stepwise multiple regression was used in order to assess the ability of several variables to predict the involuntary admission index. Preliminary analyses were conducted to ensure no violation of the assumptions of normality, linearity, multicollinearity and homoscedasticity. Variables entered in the model were population, population density, percentage of foreigners, homogeneity of incomes, number of prior admissions, rate of unemployment and residence time. The total variance explained by the model as a whole was 81%, F(7, 111,323) = 31,218, p < .001. For the explanation of the variance of the involuntary admission index, the demographic data play an important role.

Correlation of length of stay to demographic data

The analysis of the average length of stay of the patients in the clinic with regard to the different demographic variables in the specific region, as shown in Table 4, resulted in a slightly negative correlation between the average length of stay and the population figure (r = −.278; p < .001; N = 177,920) as well as between the average length of stay and the population density (r = −.294; p < .001; N = 177,920). A very low correlation between average length of stay and average income (r = −.043; p < .001; N = 139,623) was realized. In contrast, a distinct negative correlation between average length of stay and share of immigrants (r = −.760, p < .001; N = 177,920) as well as between average length of stay and unemployment rate (r = −.538; p < .001; N = 177,920) was detected. This shows that a higher share of immigrants and a higher unemployment rate are accompanied by a shorter length of stay of the patients in the clinic.

Correlation of number of places in a residential treatment in a secure environment and demographic data

Based on the figures in Table 4, the number of treatment places in a secure environment was set in relation to the total number of beds of a clinic. This proportion was correlated with different demographic variables and showed a slightly positive correlation to the population figure(r = .320; p < .001; N = 177,920). With regard to the share of immigrants, there is a positive tendency between the share of beds in relation to the total number of beds of a clinic.

A slightly negative correlation between the share of secured treatment places in relation to the total number of beds of a clinic and the unemployment rate (r = −.375; p < .001; N = 177,920) as well as the average income (r = −.273; p < .001; N = 139.623) was detected. A considerable positive correlation is shown between the share of beds in closed wards in relation to the total number of beds of a clinic and the population density (r = .502; p < .001; N = 177.920). In summary, the analysis confirmed that a higher share of secured treatment places in relation to the total number of beds of a clinic goes along with a higher population density and a higher population figure as well as a lower unemployment rate and a lower average income.

Discussion

The increase of involuntary admission based on PsychKG over the last years since the reformation of the psychiatric health provision care in Germany (Cording & Binder, 2004; Crefeld, 2002; Darsow-Schütte & Müller, 2001; Dreßing & Salize, 2004; Marschner & Volckart, 2001; Müller, 2004; Müller & Josipovic, 2003; Salize & Dressing, 2004) raises the question, whether the structure of the regional mental health care partly accounts for the increased number of involuntary admissions. Therefore, the present study closely investigated the structures of the stationary and ambulant psychiatric provision units. The examination of the number of admissions based on PsychKG in the 11 analyzed clinics shows a strong variance in the number of involuntary admissions. In 2008, the rate in the examined clinics differed from 53 to 1,006 admissions by PsychKG. However, these numbers were not put into context with the size of the clinic and the total number of admissions per clinic.

For a closer analysis, the involuntary admission index was calculated which provides insight into the ratio of involuntary admissions to the total number of admissions per clinic. Clinics with a high or a low involuntary admission index were compared with regard to various characteristics of the hospital. Overall, the comparison provided no significant differences. The involuntary admission index seems to be independent of the respective institutional form. There are only a few studies, which analyze the number of involuntary admissions with regard to the clinic structures.

This study examined the structures of social-psychiatric services, community-psychiatric cooperation and complementary facilities with regard to the involuntary admission index. Basically, it turned out that a high involuntary admission index goes along with a low usage of social-psychiatric services. In regions with a low involuntary admission index, application for involuntary treatment were more frequently issued by the social-psychiatric service or the general practitioner and only rarely by resident psychiatrists and clinics than in regions with a high involuntary admission index. The client to staff ratio of the social-psychiatric services was comparable with regard to a low versus high involuntary admission index. Differences are indicated with regard to the number of clients in assisted accommodation, where regions with a low involuntary admission index show lower proportion of clients per staff than regions with a high involuntary admission index. By trend, in regions with a high involuntary admission index, a low of social-psychiatric services can be stated. This indicates a preventive impact of the social-psychiatric service on the acute psychiatry interventions. Therefore, a clear correlation between involuntary admission index and the activities of social-psychiatric services, community-psychiatric cooperations and complementary facilities exists. Mantas and Mavreas (2012) showed similar correlations with regard to early intervention services, which enable cooperation with the families and the mobile mental health units. Out of these result short lengths of stay and a close connection to the mobile mental health units, which ensure an after-treatment and minimize the relapse rate.

A further aspect for the interpretation of the present result is the accessibility of the clinic/the social-psychiatric services to the clients in the gathering grounds with a high involuntary admission index. Potentially, no compatible program is offered there. Regarding this set of problems, Kluge, Becker, Kallert, Matschinger, and Angermeyer (2007) described the consideration of structural factors when analyzing the individual usage of psychosocial salvage as an important help for the interpretation.

Another aspect of the present study is the consideration of sociodemographic data with regard to the gathering grounds of clinics with a low or high involuntary admission index, which did not result in significant differences. Consequently, sociodemographic data cannot explain the difference between clinics with a low or high involuntary admission index. There is a positive relation between demographic characteristics of the catchment area of the hospital (number of inhabitants, population density, homogeneity of incomes, share of immigrants, unemployment rate) and the proportion of involuntary admissions to hospital in 2008. Furthermore, the length of stay in hospital showed a positive correlation to the average income. Solely, the correlation between the age at admission and the involuntary admission index showed a negative relation. This suggests that there is an influence of demographic parameters of the region and patient variables on the level of the involuntary admission index in a region. The execution of multivariate analysis (regression analysis) confirmed this assumption. By considering population figure, population density, homogeneity of incomes, share of immigrants, and unemployment rate, a relatively high variance could be clarified. By additionally involving the patient variables, the share of the explained variance increases slightly. This means that the demographic environment variables play a more important role for the explanation of the involuntary admission index variance than the patient variables. The dependency of the psychiatric care from sociodemographic characteristics was already described by Creed et al. (1997) and Kisely et al. (2000). They analyzed the severity of psychiatric illness with regard to various variables. The present results are confirmed by a study from Hudson (2005) that showed the direct influence of the socioeconomic status on the level of psychiatric illnesses.

By considering the relation between length of stay and socioeconomic variables, a short length of stay goes along with a higher unemployment rate, a higher number of inhabitants and a higher population density. A shorter length of stay in a clinic located in a region with a higher number of inhabitants and a higher population density could indicate the hospital’s need for a quicker turnover, because of limited stationary treatment places. On the one hand, a higher rate of unemployment and share of immigrants could relate to a higher population figure, reflecting the demographic differences between urban and rural areas. Due to language barriers, immigrants could, for example, have difficulties to access the medical provision system and as a result illnesses remain treated insufficiently for a longer period. In case of a culmination of the illness, they would need to be hospitalized for a crises intervention in a clinic. Furthermore, this could reflect the limitations of a clinic treating patients with language barriers that eventually requires more time. An alternative explanation of this fact could be the potentially better familial support and familial integration of immigrants. Various reasons for a shorter length of stay at higher unemployment rates could be discussed accordingly. For instance, an easier ambulant provision based on more flexible scheduling opportunities or accumulation of specific diagnoses could play a role. Compton, Craw, and Rudisch (2006) also saw the variance in the length of stay of patients in dependency to demographic factors. Further studies also showed the influence of various demographic factors on the length of stay (Ballesteros, Martinez, Martin, Ibarra, & Bulbena, 2002; Creed et al., 1997; Lerner & Zilber, 2010; Munley et al., 1977; Rocca et al., 2010; Zhang, Harvey, & Andrew, 2011).

The correlation between the share of closed wards within the total number of clinic beds and sociodemographic parameters showed that a higher share of involuntary wards is associated to a higher population density and number of inhabitants, as well as to a lower rate of unemployment and a lower average income. The increased need for psychiatric inpatient care in highly populated areas might result in longer waiting periods and relatively progressed diseases. The corresponding severity of the symptoms could make a closed accommodation necessary. A similar explanation could play a role for the correlation of low unemployment rate and a higher share of involuntary wards, as employees increasingly wait longer to report their incapability for work and their willingness to get treatment in a clinic lacks until a culmination of the set of problems. Growing tendencies of the individuals endangering themselves and others could be related to the above-mentioned acute need for treatment and might result in higher rates of involuntary admission. Likewise, a low sociodemographic status and a low income could lead to a higher psychotic stress, which could go along with endangerment of themselves and others and could require involuntary admission or acute crises intervention.

For the substantiation of the obtained results, further standardized data should be collected to create a leveraged sample with reference to the social-psychiatric services and therefore to enable more detailed statements.

The present study has certain limitations that need to be taken into account when considering its results. Unfortunately, the servicing areas of psychiatric hospitals and social-psychiatric services in one region are not identical. In a region served by one social-psychiatric service, more than one psychiatric hospital can exist, which complicates the comparability. Further complicating matters, no exact demographic data for the areas of responsibility of the clinics could be obtained. The responsibility of a clinic may cover only some districts of one city or cover districts of different cities, but the demographic data are related to one city or rural district.

The data of the hospitals’ admission register allow only limited conclusions because there is only limited information concerning the history of a patient’s admission to the psychiatric hospital or his illness. Moreover, the data are not referred to single patients, but to patient cases, which means that one patient might appear several times in the data if he or she had been re-admitted during the study period. A further limitation of the study is that the investigated region covers only a part with 8.3 million inhabitants of North Rhine-Westphalia (46.6%).

Because of this and the fact that federal state law justifies admission, the generalizability of the results is limited. Further research on this topic is evidently needed, ideally in form of a prospective study.

Conclusion

In summary, the number of admissions based on PsychKG is determined from an interaction of multiple factors. On the one hand, the integration and orientation of the social-psychiatric service is essential for the popularity with the potential patients. On the other hand, sociodemographic factors and patients variables play a role with regard to the number of admissions based on PsychKG.

Footnotes

Acknowledgements

The authors gratefully acknowledge the collaborative partners of the social-psychiatric services and in the LWL-PsychiatryNetwork.

Declaration of conflicting interest

G.J. received fees for consulting from AstraZeneca, Bristol-Myers-Squibb/Otsuka, Janssen, Lilly, Lundbeck and Pfizer. He was advisory board member of AstraZeneca, Bristol-Myers-Squibb/Otsuka and Janssen-Cilag. He received research funding from Jansen, AstraZeneca, Lundbeck and BMS. All other authors declare that there is no conflict of interest.

Funding

This study was supported by the NRW Center for Health LZG.NRW (SZ-01/2010).

References

Ballesteros

R. J.

Martinez

P. L.

Martin

C. M.

Ibarra

G. N.

Bulbena

V. A.

(2002). Assessment of the Life Skills Profile and the Brief Psychiatric Rating Scale as predictors of the length of psychiatric hospitalization. Actas Españolas de Psiquiatría, 30, 225–232.

Bericht der Psychiatrie Enquete. (1975). Bericht über die Lage der Psychiatrie in der Bundesrepublik Deutschland. Retrieved from http://www.dgppn.de/schwerpunkte/versorgung/enquete.html

Bertelsmann Stiftung. (2011). Wegweiser Kommune der Bertelsmann Stiftung. Retrieved from http://www.wegweiserkommune.de

Compton

M. T.

Craw

Rudisch

B. E.

(2006). Determinants of inpatient psychiatric length of stay in an urban county hospital. Psychiatric Quarterly, 77, 173–188.

Cording

Binder

(2004). Unfreiwillige Behandlungen in der psychiatrisch-psychotherapeutischen, Versorgungsklinik (Vortrag). Berlin, Germany: DGPPN Kongress.

Creed

Tomenson

Anthony

Tramner

(1997). Predicting length of stay in psychiatry. Psychological Medicine, 27, 961–966.

Crefeld

(2002). Hilfe und Zwang – oder Zwang aus Hilfslosigkeit: Einführung in das Thema. In Brill

K. E.

(Ed.), Zehn Jahre Betreuungsrecht: Qualifizierung der Umsetzung oder erneute Rechtsreform? (pp. 219f). Recklinghausen, Germany: Eigenverlag.

Darsow-Schütte

K. I.

Müller

(2001). Number of hospitalizations according to German ‘PsychKG’ legislation has doubled in 10 years. Psychiatrische Praxis, 28, 226–229.

Dreßing

Salize

H. J.

(2004). Zwangsunterbringung und Zwangsbehandlung psychisch Kranker in den Mitgliedsländern der ′Europäischen Union. Psychiat Prax, 31, 34–39.

10.

Dodegge

Zimmermann

(2003). PsychKG NRW: Gesetz über Hilfen und Schutzmaßnahmenbei psychischen Krankheiten (Praxiskommentar). Stuttgart, Germany: Boorberg Verlag.

11.

Hoffmann

(2009). Freiheitsentziehende Unterbringung von Kindern und Jugendlichen – Rechtslage nach Neufassung des § 1631 b BGB und Inkrafttreten des FamFG. Recht & Psychiatrie, 27, 121–129.

12.

Hudson

C. G.

(2005). Socioeconomic status and mental illness: Tests of the social causation and selection hypotheses. American Journal of Orthopsychiatry, 75, 3–18.

13.

Kisely

Preston

Rooney

(2000). Pathways and outcomes of psychiatric care: Does it depend on who you are, or what you’ve got? Australian and New Zealand Journal of Psychiatry, 34, 1009–1014.

14.

Kluge

Becker

Kallert

T. W.

Matschinger

Angermeyer

M. C.

(2007). Impact of structure and context on utilization of socio-psychiatric services in Saxony – A multi-level-analysis. Psychiatrische Praxis, 34, 20–25.

15.

Lerner

Zilber

(2010). Predictors of cumulative length of psychiatric inpatient stay over one year: A national case register study. Israel Journal of Psychiatry & Related Sciences, 47, 304–307.

16.

Mantas

Mavreas

(2012). Establishing and operating an early intervention service for psychosis in a defined catchment area of northwestern Greece within the context of the local mental health network. Early Intervention in Psychiatry, 6(2), 212–217.

17.

Marschner

Volckart

(2001). Freiheitsentziehung und Unterbringung. München, Germany: Verlag C.H. Beck.

18.

Müller

(2004). Zwangseinweisungen nehmen zu. Deutsches Ärzteblatt, 101(42), A2794–A2798.

19.

Müller

Josipovic

(2003). Unfreiwillige Einweisungen nach Betreuungsrecht in acht Jahren verdreifacht. Psychiatrische Praxis, 30, 108–113.

20.

Munley

P. H.

Devone

Einhorn

C. M.

Gash

I. A.

Hyer

Kuhn

K. C.

(1977). Demographic and clinical characteristics as predictors of length of hospitalization and readmission. Journal of Clinical Psychology, 33, 1093–1099.

21.

Psychiatrie in Deutschland. (2003). Strukturen, Leistungen, Perspektiven erarbeitet von der Arbeitsgruppe Psychiatrie der Obersten Landesgesundheitsbehörden im Auftrag der Gesundheitsministerkonferenz. 76. Sitzung, Beschluss vom 02./03.07.2003.

22.

Rocca

Mingrone

Mongini

Montemagni

Pulvirenti

Rocca

Bogetto

(2010). Outcome and length of stay in psychiatric hospitalization, the experience of the University Clinic of Turin. Social Psychiatry and Psychiatric Epidemiology, 45, 603–610.

23.

Salize

H.J.

Dreßing

(2004). Epidemiology of involuntary placement of mentally ill people across the European Union. British Journal of Psychiatry, 184, 163–168.

24.

Salize

H. J.

Dressing

(2004). Epidemiology of involuntary placement of mentally ill people across the European Union. The British Journal of Psychiatry, 184, 163–168.

25.

Spengler

(2007). Compulsory admission in Germany – Epidemiological data 1992-2003. Psychiatrische Praxis, 34(Suppl. 2), S191–S195.

26.

Spengler

Dressing

Koller

Salize

H.J.

(2005). Zwangseinweisungen –bundesweite Basisdaten und Trends. Nervenarzt (Mitteilungen DGPPN), 76, 363–370.

27.

Stadt Bochum. (2010). Statistische Jahrbücher 2005-2009. Available from http://www.bochum.de

28.

Statistisches Bundesamt. (2011). Statistisches Jahrbuch 2010 (Fachserie 12, Reihe 7.2. 2011). Wiesbaden, Germany: Author.

29.

Zhang

Harvey

Andrew

(2011). Factors associated with length of stay and the risk of readmission in an acute psychiatric inpatient facility: A retrospective study. Australian and New Zealand Journal of Psychiatry, 45, 578–585.