A number of patients whom the admitting doctor considers to be acutely suicidal are denied admission to a psychiatric ward or discharged after a short time. The usual justification is that the receiving doctor believes the patient to be ‘chronically suicidal’ and that hospitalisation will exacerbate the condition.
Patients who are described as ‘chronically suicidal’ are often diagnosed with borderline personality disorder (1). This is the only diagnosis for which self-harm and suicidal behaviour are diagnostic criteria. Patients with this diagnosis have a fundamental difficulty in regulating emotions and impulses, and in forming and maintaining trusting relationships. Their reactions can be abrupt, intense and without nuance, and may entail drastic behaviour, conflicts and misunderstandings. However, in our opinion, the term ‘chronically suicidal’ should not be used, as the expression does not distinguish between thoughts and plans on the one hand and high-risk preparation and suicide attempts on the other (1).
Genuine risk of suicide
In a Swedish study of patients with personality disorder who had been admitted to psychiatric wards in the period 1987–2013, a total of 5 % of the women and 7 % of the men had taken their own lives during a twelve-year period (2, 3). The risk was highest in the group of personality disorders that encompassed the borderline type. It is estimated that in total at least 75 % of patients with borderline personality disorder have attempted to take their own lives, and that approximately 10 % die from suicide (4).
Hospitalisation of these patients in psychiatric wards has been fraught with controversy (5, 6). According to national guidelines for suicide prevention in mental health care, hospitalisation is inappropriate for this patient group where there is an ‘acute suicide risk’, but may nevertheless be imperative in an ‘acute suicide crisis’ (7). The supposed difference between acute ‘suicide risk’ and ‘suicide crisis’ is unclear.
Some of the patients have inflicted very severe self-harming or have attempted suicide. A potentially life-threatening acute situation characterised by mental imbalance and high suicide risk may involve acute psychiatric inpatient admission even though it is not recommended in the guidelines (7). In practice, a subgroup is hospitalised frequently and sometimes for long periods (8). A number of circumstances, including diagnostic unclarity, a disjointed treatment continuum, intense help-seeking behaviour, difficult risk assessments and fear of error, may result in hospitalisation (8). In some cases, it can be attributed to a dearth of outpatient or ambulant therapy options. Such situations represent significant challenges for patients, their next of kin and the health service.
Despite the fact that patients with personality disorder may have extensive mental health symptoms and during certain periods may need complex, coordinated services across different levels of the health service, as yet no national guidelines exist in Norway for the treatment and follow-up of personality disorder (9). There is evidence that targeted psychotherapy and early intervention result in a significant reduction in suicide attempts, self-harming behaviour and hospitalisations (10).
Patient care pathway
Oslo University Hospital has prepared a patient care pathway for personality disorders, from home via the health service and home again (11). In another article in this issue of the Journal of the Norwegian Medical Association, we write about acute treatment for this patient group (10). In cases of suicidal behaviour, the patient’s survival must first of all be safeguarded, before further treatment contact is then established or ensured.
If hospitalisation is necessary in a crisis situation, it should preferably be voluntary. Crises that may render hospitalisation necessary are generally characterised by symptoms of psychosis, serious self-harming and acute risk of suicide, especially following a suicide attempt. In cases of repeated, prolonged and serious self-harming and suicidality, the patient should not be admitted as an inpatient ‘because the fundamental principle of the treatment is that patients are encouraged to govern and control their own lives and have responsibility for their behaviour as far as they are capable of this, even though the behaviour may be destructive’ (11). However, this presupposes that the patient is not too unstable and has an outpatient therapist who has responsibility and the necessary frameworks to uphold the treatment agreement that is entered into (11). A possible exception is for patients who do not have an adequate outpatient treatment programme, in which case hospitalisation may be used to establish such a programme with the responsible therapist. The patient should not be discharged until agreements are in place. Planned, regular hospitalisations may be an option where such treatment pathways have been prepared.
Exacerbated by hospitalisation?
A Swedish study that compared mortality in patients with personality disorder who had been hospitalised, with patients who had only received outpatient treatment, found a somewhat higher prevalence of unnatural death among the hospitalised patients, but the mortality rate was also considerably higher for those who had only received outpatient treatment. The probable explanation is that the group that was admitted as inpatients was more ill (selection bias) and it cannot therefore be concluded that inpatient treatment increases the suicide risk (2, 3).
Very few studies show that patients who are in fact acutely suicidal become worse by being hospitalised until the situation has stabilised
Some mental healthcare professionals believe that hospitalisation may increase the risk of suicidal behaviour. Removing patients from the usual demands associated with coping with life, particularly dealing with situations that elicit strong emotions, might result in patients failing to learn how to deal with these emotions, and second, might cause suicidal behaviour to be reinforced (12). A study of female patients with borderline personality disorder who received outpatient treatment found that those who had previously been hospitalised attended the acute psychiatric service more frequently than those who had not been hospitalised. Those who attended the acute psychiatric service during treatment were also those who had the most suicide attempts in the first year after treatment (12).
One reason why hospitalisation of patients with borderline personality disorder has generally been discouraged is based on claims that after only a few days’ stay in an inpatient institution, their symptoms may be exacerbated or they may become involved in negative interaction with staff and fellow patients (8, 13). An increase in threats of suicide, self-harming or suicidal behaviour is feared, which in turn may result in prolonged hospitalisation or provoke a premature discharge or other negative reactions from the hospital staff (2, 3).
We are critical of these claims in relation to acute suicide crises. Very few studies show that patients who are in fact acutely suicidal become worse by being hospitalised until the situation has stabilised. For the majority of patients with borderline personality disorder, the acute risk of suicide is short-lived, even though the risk may be significantly heightened in the long term. If the patient has active plans and has no control over his/her suicidal impulses, they are in such great need and danger that therapists must implement measures to prevent suicide. If this can happen outside the inpatient institution, that is well and good, but in many cases hospitalisation is necessary until the patient has regained sufficient control.
Any potential hospitalisation should have the objective of supporting long-term, ongoing psychotherapy where this is established, or of establishing such a service. It is unlikely that patients in a state of chaotic despair who are in imminent danger of taking their own lives are in a position to learn better life-coping skills.
If it is thought that patients at risk of suicide will be made worse by being admitted to a psychiatric ward, it ought to be investigated whether this is actually the case and what needs to be done – in other words, what factors are harmful for the patients.
Length of hospitalisation
There is broad agreement that long stays of several weeks are not indicated. Hospitalisations should be short, but not too short. Many patients have been hospitalised overnight and discharged the next day. Within that time it will not have been possible to establish sufficient help outside the hospital, or to manage the suicide crisis to an adequate extent.
When assessing whether a patient should be hospitalised or rapidly discharged, symptoms, diagnosis and the patient’s current life situation must be considered, including their social network. If the patient has no network, there is a greater risk of self-destructive behaviour, and it should then be ensured that the patient is more stable before he or she is discharged than if the patient has a supportive network.
Assessment of suicide risk
Although the patient also has responsibility for destructive actions, clinicians have both a professional and legal responsibility to prevent suicide and the most serious forms of self-harming, particularly in patients whose mental functioning is poor.
It may be difficult to assess the patient’s mental state and ability to take responsibility in a crisis situation where the patient is emotionally distraught, which may affect their ability to think rationally about their own situation. Understanding and validating the patient’s subjective experience is always the first step, and a prerequisite for cooperation in identifying the situation, what has triggered the crisis and whether it is possible to find alternatives to hospitalisation.
There must be room for discretion in the assessment of whether there is an acute risk of suicide
There must be room for discretion in the assessment of whether there is an acute risk of suicide, not least because ambivalent feelings are involved that may fluctuate rapidly. Naturally, the better one knows the patient and the more a mutual therapeutic alliance has been established, the safer the assessment. On the other hand, when one does not know the patient, hospitalisation may be necessary to avoid running too high a risk (14). A patient who does not know the therapists will likely tend to feel rejected more easily than one who is more familiar with the therapists and the clinical justification for a possible discharge.
Nor should suicidal intention be underestimated in a patient who has taken an overdose and then sought help, because intention can change after intake and while under the influence of a medicinal drug or an intoxicant. Emotions may be more erratic, ambivalence greater and coping ability reduced. Some patients are suicidal when left to themselves or living in a stressful situation, but not in a secure hospital ward.
It is difficult to assess suicidal behaviour in patients with borderline personality disorder because external conditions and patients’ internal state can change rapidly. Some patients experience rejection both as outpatients and while in inpatient wards. This increases the risk of self-destructive behaviour.
Although the national guidelines for suicide prevention in mental health care state that these patients should not be treated in psychiatric inpatient wards, clinical practice ought to have scope for nuances. A reasonable objective for hospitalisation is primarily to assist the patient through an acute suicidal crisis, and then to identify whether she/he has an adequate outpatient treatment option, and motivate the patient to undertake further therapy. Acute hospitalisation then becomes one element of a larger care pathway.