During the one-year study period, 492 calls related to mental health crises were identified in the police database. With a population of 198,000 inhabitants in the police district, this constitutes a rate of 2.5 calls per 1000 inhabitants per year. Table
1 presents characteristics of the mental health calls received by the police. Importantly, in one-fifth of the calls it was the individual experiencing the crisis who contacted the police. In addition, in approximately half of the cases, the crisis was judged to constitute a problem to others (46%: nuisance or danger to others), while only a quarter of the cases were deemed to constitute a problem for the individual experiencing the crisis (23%: self-neglect or danger to self).
Table 1
Characteristics of mental health calls to the police (N=492)
Caller | | |
Neighbour or bystander | 167 | 34 |
Agency (nuisance complaints desk; pension; etc.) | 95 | 19 |
Person self | 93 | 19 |
Family or friend | 75 | 15 |
Victim or police surveillance | 30 | 6 |
Unclear or missing | 32 | 7 |
Time of call
| | |
Working day; outside office hours | 191 | 39 |
Working day; during office hours | 176 | 36 |
Weekend or holiday | 125 | 25 |
Location where person was found
| | |
Public space | 220 | 45 |
At home | 184 | 37 |
Unclear or missing | 88 | 18 |
Police incident code
| | |
Nuisance by disturbed person | 250 | 51 |
Assistance of citizen | 161 | 33 |
Domestic or neighbours’ quarrel | 34 | 7 |
Suicide attempt | 18 | 4 |
Violence | 13 | 3 |
Missing | 16 | 3 |
Nature of mental health crisis
| | |
Nuisance to others | 178 | 36 |
Danger to self | 108 | 22 |
Danger to others | 48 | 10 |
Self-neglect | 5 | 1 |
Other or unclear | 153 | 31 |
Individuals experiencing public mental health crises
The 492 mental health calls concerned 336 individuals (1.7 per 1000 inhabitants per year). Their mean age was 40.1 years (SD=16.3), ranging from 13 to 93 years. Sixty percent of these individuals were male, and 18% were homeless. For 74 individuals there were multiple calls (range 2–11) within the study year.
Fifteen individuals came from outside the region covered by the Psychiatric Case Register, meaning there was no information on their contact with mental health services. These individuals were excluded from further analyses. Of the remaining 321 individuals, 162 (50%) did not have regular mental health care contact in the year before the crisis, 52 (16%) had some care contacts (up to two inpatient days and up to nine outpatient contacts), and 107 (33%) had frequent care contacts (three or more inpatient days or ten or more outpatient contacts).
Table
2 shows the nature of the mental health problems experienced by individuals with and without care contact in the preceding year. Due to the lack of contact with mental health services, no formal diagnosis was available for 36% of individuals without regular care contact in the year before the crisis. Nevertheless, the diagnoses that were available (from crisis interventions, or care contact prior to a year before the crisis, or contact after the crisis) revealed that a substantial portion of individuals seen by the police for mental health crises experience serious mental health problems that warrant referral to mental health services. For example, at least 26% of individuals without care contact had a psychotic or bipolar disorder, which are generally regarded as serious mental illnesses, and another 30% (controlling for comorbidity with psychotic or bipolar disorders) had a potentially serious cognitive, substance use or depressive disorder. For individuals with care contact, these figures were 45% and 36%, respectively.
Table 2
Mental health problems of individuals with public mental health crises
Psychotic disorder | 59 | 37 | 33 | 20 |
Bipolar disorder | 13 | 8 | 10 | 6 |
Other mood disorder | 15 | 9 | 5 | 3 |
Dementia or other cognitive disorder | 3 | 2 | 11 | 7 |
Substance use disorder | 75 | 47 | 52 | 32 |
Personality disorder | 49 | 31 | 14 | 9 |
Other mental disorder | 44 | 28 | 22 | 14 |
Unclear mental health problem#
| 9 | 6 | 59 | 36 |
Police response to mental health crises
The frequencies of different types of police response to a mental health crisis are presented in the left panel of Table
3. Only the responses to the first crisis per individual in the study period are shown, but an almost identical pattern of results was observed for all crises. In approximately half of the cases the police dealt with the crisis themselves, without contacting any care services (e.g. by calming the individual or contacting a relative to look after them). In 27% of crises the police contacted mental health services (e.g. the client’s case manager or a psychiatric emergency service), and in 10% the police contacted another care service (e.g. the person’s general practitioner or social services). Finally, 14% of individuals were taken to the crisis drop-off centre at the police station, for a mental health evaluation by a public health physician of the municipality or a professional of the mental health services. This option was often chosen outside office hours and for clients who did not live in the area or were homeless [
10]. The 24/7 crisis drop-off centre therefore appears to be used by police officers as an additional entry point to mental health services, at times when access to these services is difficult [
10].
Table 3
Police response to mental health crises and relationship with care contacts
*
Dealt with by police | 156 | 49 | 1.3¶
| 5.9 | 94 | 58 | 5 | 5§
| 19 | 20¶
|
Mental healthcare contacted | 88 | 27 | 7.4¶
| 12.6 | 29 | 18 | 12 | 41§
| 17 | 59¶
|
Evaluation at police station | 44 | 14 | 7.0¶
| 10.8 | 20 | 12 | 12 | 60§$
| 12 | 60¶
|
Other care contacted | 33 | 10 | 4.0 | 8.9 | 19 | 12 | 5 | 26§$
| 7 | 37 |
Relationship between police response and number of care contacts
The mean number of regular care contacts increased from 3.4 (SD=8.4) in the month before the crisis to 7.4 (SD=11.6) in the month after the crisis (paired t=7.53, df=320, p<.01), when individuals were seen for the first time in the study period. No interaction effect was found between the type of police response and the number of care contacts in the year before the crisis on the increase in number of contacts after the crisis (F=0.61, df=6;309, p=.72). In addition, there was no main effect of the number of care contacts on the increase in contacts (F=0.82, df=2;309, p=.44). However, the increase in the number of care contacts from the month before to the month after the crisis was related to the type of police response to the crisis (F=7.23, df=3;309, p=<.01), as illustrated in the left panel of Table
3. A greater increase was observed when police had contacted mental health services during the crisis or had taken the person to the crisis drop-off centre, compared with situations where crises were dealt with by the police themselves (t=5.13, df=242, p<.01 and t=4.66, df=198, p<.01, respectively).
Linking individuals to services
Of special interest is the effect of the police response for individuals with no regular care contact with mental health services prior to their mental health crisis. We tested whether the way in which the police handled the crisis affected the likelihood of regular care contact (the individual’s linkage with mental health services) after the crisis.
Of the 162 individuals without contact with mental health services in the year before the crisis, 34 (21%) had regular care contact (other than for crisis intervention) with these services in the month after the crisis. In addition, the likelihood of care contact was related to the police response to the crisis (χ
2=39.87, df=3, p<.01), as shown in the middle panel of Table
3. This likelihood was greater for individuals who were taken to the crisis drop-off centre or for whom the police directly contacted the mental health or another care service during the crisis, compared with those for whom the police did not contact a service to deal with the crisis (χ
2=38.86, 24.20 and 8.64, respectively, all df=1, p<.01). Taking the individual to the crisis drop-off centre was also more effective than contacting another care service for linking the person to mental health services (χ
2=4.50, df=1, p=.03). Importantly, the least effective response for linking individuals to mental health services was chosen most frequently by the police. In 58% of mental health crises of persons who were out of contact with services, the police dealt with the crisis themselves and did not contact a service provider.
Of the 34 individuals who were newly linked to the services, 25 were admitted (admission was compulsory for 12 individuals) to a psychiatric hospital in the month after the crisis, for 13.2 days on average (excluding the crisis day; SD=12.3; range 1–30), and the remaining nine had 1.4 outpatient contacts on average (SD=0.7; range 1–3).
Finally, we examined whether individuals who were newly linked to mental health services in the month after the crisis developed a lasting care relationship with the services that continued into the subsequent year. The right panel of Table
3 shows the number of individuals who had regular care contact with mental health services in the period from 1 to 13 months after the crisis. For all persons who were disengaged from the services before the crisis, this number increased from 34 (21%) in the first month after the crisis to 55 (34%) in the subsequent year. Again, this number was related to the police response to the crisis (χ
2=21.91, df=3, p<.01). When police contacted mental health services at the time of crisis, either directly or by taking the person to the crisis drop-off centre, individuals were more likely to develop a lasting care relationship with services, compared with individuals whose crisis was dealt with by the police themselves without contacting service providers (χ
2=15.79 and 13.19, respectively, both df=1, p<.01). Furthermore, this relationship more often consisted of frequent care contacts (i.e. three or more inpatient days or ten or more outpatient contacts) for individuals referred to mental health services directly or indirectly, compared with individuals dealt with by the police only (39% versus 12%; χ
2=14.24, df=1, p<.01, data not shown). For five of the 34 individuals who were newly linked to mental health services in the first month after the crisis, the care relationship did not persist into the subsequent year, and this was not related to the police response to the crisis (χ
2=0.65, df=3, p=.89).