Abstract
‘Absent without leave’ (AWOL) and ‘absconding’ are terms commonly used to describe the departure of a patient without staff sanction from the physical boundaries of a hospital [1–3]. The behaviour usually disrupts the social climate of the inpatient unit [4] and staff frequently react with feelings of anger, embarrassment, guilt and concern [2]. They express a sense of failure and acknowledge that the AWOL event should have been foreseen and thereby prevented. Time and energy invested in conducting searches, writing incident reports, and informing the appropriate authorities detract from the care and treatment of other patients in the unit.
Absent-without-leave behaviour can also have medicolegal consequences for both the patient and the treating facility [5]. While absent from hospital, some patients commit offences such as assault and theft, while others engage in acts of self-injury and suicide [1–7]. Despite the risks associated with AWOLbehaviour for all involved, the available literature is sparse, rather dated and is of little assistance in identifying strategies to combat the problem.
Reported rates of absconding range from 2.5 to 34% of all admissions [6],[8–10]. Failure to distinguish between patients ‘absent without leave’ and those ‘discharged against medical advice’ (AMA) in US studies has contributed to much of the variation reported. In addition, a number of authors have classified patients as being absent without leave at the moment they could not be located on the unit, while others omitted this group and include only those patients who fail to return to hospital by midnight on the day of the incident.
Most studies have attempted to develop a profile of the typical AWOL patient. Despite conflicting results, AWOL behaviour appears to be associated with patients who are younger [3],[6], legally detained [3],[11], and who have a diagnosis of schizophrenia [3],[12]. However, it is possible that younger patients with a diagnosis of schizophrenia are more likely to be admitted to inpatient facilities and consequently lead to an over-representation of people with these characteristics in AWOL data. It is also probable that legally detained patients are more likely to be classified as being AWOL when they cannot be located.
Only two of the studies reviewed considered absconding behaviour from the patients' perspective. In interviews with 100 consecutive absconders, Folkowski et al. [1] found that patients were concerned with the stigma of being in a psychiatric hospital and disliked both the staff and food provided. However, almost one-quarter left hospital because of abnormal beliefs, including fears of being harmed while in hospital and in response to command hallucinations. McIndeo [2] carried out in-depth interviews with five patients within 72 h of returning from being AWOL and noted that most were unsure about the goals of their treatment, were dissatisfied with the programs offered, and questioned the need for their hospitalisation. All five patients perceived absconding as a positive event with positive consequences and given the same circumstances, would abscond again.
In the present study, we describe patient and environmental characteristics associated with absconding behaviour and present the patients' perspective on issues related to absconding. In addition, possible strategies to reduce the risk of absconding behaviour during hospitalisation are discussed.
Method
Setting
The study was conducted at an acute psychiatric unit (28 beds) located on the campus of a large, public general hospital (310 beds) in Queensland. The unit was opened in 1998 and was purposely built for the treatment of people with acute mental illness. It was designed to provide accommodation in a number of single and twin-share bedrooms. An activities area which includes a billiard room, crafts room and gymnasium is located at one end of the unit. There is also a small library and an outdoor courtyard area. The unit is unlocked from 06:00 h to 22:00 h each day. Restrictive measures such as appointing staff to monitor exit doors and placing patients in night attire to prevent absconding are not employed. During the 6-month study period, an average of 16 patients were admitted to the Unit each week with an average length of stay of 6.8 days.
Patients at risk of absconding can be transferred to the locked High Dependency Unit (HDU) located within the unit. The HDU has accommodation for four patients and has a high staff:patient ratio, usually 1:1, compared with a 1:5 ratio on the open ward. It provides short-term care (mean = 2.5 days) for acutely ill patients who cannot be managed in the open ward environment.
Procedure
In the event that a patient could not be located, nursing staff carried out a search of the area around the Unit. If this proved to be unsuccessful in locating the patient, the patient's case manager and/or police were notified. While all patients who could not be located were classified as being AWOL, only those patients who had not returned to hospital by midnight on the day in question were included in the study. At midnight, details of the AWOL incident were documented in the patient's medical record and the patient was marked ‘AWOL’ in the ward census register. A research nurse visited the Unit every other day to check the register. Data from the register and the case notes of each AWOLpatient were entered into a specially designed pro forma to ensure consistency in data collection.
In addition, in-depth interviews were conducted with a subsample of 14 patients (nine males and five females) to explore their perceptions of events related to their absconding behaviour. The patients ranged in age from 19 to 58 years (mean = 37), and were diagnosed with schizophrenia (n = 7), acute psychosis (n = 3), personality disorder (n = 2), and depression (n = 2). All patients were interviewed within 48 h of returning from being AWOL. While eight of the patients had absconded from the study wards during previous admissions, the current admission was the index admission for the remaining six patients.
Methods
During the 6 months of data collection, a total of 390 patients were admitted to the Unit. Of these, 51 (13.1%) were responsible for 77 consecutive incidents of absconding behaviour. While 35 (68.6%) patients absconded only once, 16 others (31.4%) absconded on more than one occasion. In fact, one male patient with a diagnosis of schizophrenia absconded four times during a 4-week admission.
Forty-five (58.4%) of those who absconded were male and 74% were under the age of 40 years. The single largest group of absconders (41.6%) had a diagnosis of schizophrenia, followed by acute psychosis (25.2%), personality disorder (10.4%), depression (9.1%), bipolar disorder (6.5%), and other diagnoses (12.9%). The majority (77.9%) were admitted involuntarily under the Queensland Mental Health Act (1974) and were legally detained in hospital at the time they absconded. Thirty-nine percent of those who absconded had had previous admissions to hospital.
The length of time between admission to hospital and absconding ranged from a few minutes to 35 days. While 11 incidents (14.3%) occurred on the day patients were admitted, almost one-half of all AWOLincidents (49.4%) had occurred within 7 days of admission. Early morning (07:00–11:00 h) was a high-risk period with 34.1% of incidents occurring during this time.
While police were responsible for returning patients to hospital on 26 (33.8%) occasions, a smaller number returned of their own volition (22.1%) or with the assistance of a family member or friend (13.3%). Others (14.3%) were returned by community/unit staff or ambulance following presentation at the general hospital (1.3%). On 27 (35.1%) occasions, patients were not returned to hospital at all and were discharged whilst absent. However, 59.7% of these had been in hospital longer than the average length of stay (i.e. 7 days) when they absconded. The consequences of absconding for a small number of patients were relatively serious: two patients attempted suicide, one vandalised his mother's home, one had an injury while ‘train surfing’ in the local rail yard, and one patient suffered severe sunburn while wandering in a confused state in bushland near the hospital.
On returning to the unit, more than one-fifth (22.1%) were transferred to the High Dependency Unit (HDU) for close observation and containment. This was more likely if patients had absconded previously or if they had deteriorated mentally and or physically as a result of noncompliance with medication whilst AWOL. The mean time spent in the HDU was 2.5 days.
Patient perceptions
The interviews with patients were guided by two broad areas: (i) patients' perceptions of the treatment and care provided; and (ii) reasons for absconding. All interviews were audiotaped and transcribed in an attempt to arrive at an ‘authentic’ account of the patients' experience. It is not intended that the qualitative information provided by this sample should be generalised to a larger population, rather the primary aim was to gain an understanding of how the patients perceived the AWOLexperience [13]. The interviews helped to provide a clearer understanding of how absconding behaviour occurs within the complex social setting in which patients find themselves. Six issues emerged from the analysis of the interview data: boredom, lack of interesting activities, disturbed ward environment, perceptions of the need for hospitalisation, concerns about issues at home, and the perceived rewards from absconding. These are discussed in more detail below.
Most of the patients found that their stay in hospital was boring and it was difficult to ‘pass the time’, particularly on the weekends when there were less structured activities available. Patients who had been in hospital longer than 2 weeks were more likely to identify boredom as a reason for absconding. Most patients felt that the group activities provided were not helpful and in many cases contributed to their decision to leave hospital. They felt anxious about having to participate in group discussions and about having to have close contact with patients they feared or disliked.
Those who absconded suggested that ‘other’ patients were far more disturbed and in greater need of treatment than themselves. Five patients complained about the ‘madness’ and the tension in the Unit. They thought that the hospital would be a ‘quiet’ place where they could have some peace, instead they found it to be a stressful and volatile place with lack of freedom and rigid ward rules. One patient claimed that, ‘if I hadn't run away I would have gone mad’!
The need for hospitalisation was also raised. While six of the patients recognised that they were having difficulties, they believed that they were not sick enough to have been ‘scheduled’ and forced to stay in hospital. They failed to understand how staying in hospital was going to address the problems they were experiencing. For others, leaving hospital was an impulsive act, ‘I went out to get smokes and just kept going’. Another found a car in the visitors' car park with the keys in the ignition and decided to ‘go for a drive around town’. The police found him (and the car) at his parents' place the next morning.
Five of the patients stated that they left hospital to sort out problems at home. For example, one patient left hospital to make arrangements for the ongoing care of her cat. Another was concerned that he would lose his accommodation and left hospital to pay rent. When he got home he thought he would be able to cope, but after 2 days he decided to return to hospital. Finally, one patient claimed ‘I left to get glasses, I thought I was going blind’. His family doctor recognised that he was experiencing side effects from his medication and made arrangements for his return to hospital.
Five of the 14 patients were transferred to the High Dependency Unit (HDU) when they returned from being AWOL. This was more likely to occur if patients had been AWOL previously during their admission or if they had a history of absconding during previous admissions. While they resented the lack of freedom that the HDU offered, they claimed to have received more attention and had greater access to nursing staff while in the HDU. Interestingly, patients who were returned to the open ward area also claimed to have received more attention from the staff.
Discussion
In the present study, 13% of patients absconded from an ‘open’ acute psychiatric unit. While the finding is higher than the range of 2.5–9.3% reported in UK studies [3],[10], it is considerably lower than the 34% reported in a recent US study [8]. Although our study suffers from lack of a control or comparison group, the finding that those who absconded tended to be involuntary young males with a diagnosis of schizophrenia is in keeping with the profile of absconders developed in previous studies [1],[3],[6]. Unfortunately, the profile is rather broad and provides little guidance for staff wishing to address the problem of absconding.
One-third of all AWOL incidents resulted from repeated absconding by the same individuals. The finding is supported by previous research and suggests that a history of absconding increases the risk of this behaviour in the future [1–3]. Perhaps absconding provides this group with a sense of control and power over a system which normally assumes complete control. Indeed, absconding is understandable from the patients' perspective and could be perceived as a positive response in some patients in that they retain the initiative and motivation to abscond.
Although absconding had serious consequences for a small number of patients, to the best of our knowledge, the majority of incidents were uneventful. However, the low rate of serious incidents reported should not give rise to complacency. Absconding prolongs treatment and places the patient and members of the public at risk [6]. Indeed, the ‘danger’ that some patients pose to the public while AWOLis beginning to receive greater attention from the media in Australia [14]. Pressure from the media and the possible medicolegal consequences of absconding will place greater responsibility on mental health staff to prevent absconding.
Issues raised in the interviews with patients provide some additional pointers for managing the thorny issue of absconding more effectively at this and other hospitals. First, it is clear that staff involved in the facilitation of patient groups or other therapeutic activities must possess experience and skills that are commensurate with the needs of mentally ill people who require a period of hospitalisation. Patients are unlikely to participate in activities (and treatment) if these are perceived as being uninteresting and unhelpful. Clearly, the challenge is to provide a range of programs for patients who function at different levels. This highlights the need for specialist training for staff and indeed the provision of adequate numbers of staff to meet patient needs.
Second, in order to deal effectively with patients' concerns about personal issues such as accommodation or rent payments, there is a clear need for improved communication between staff and patients. While it is important to ensure that problem issues are not only identified and acted on, the results of these actions must be related back to the individuals involved to ease their concerns. Third, the provision of locked areas (two to four beds) within acute inpatient units enable staff to monitor patients at risk of absconding without the need to lock the entire unit and limit the freedom of other patients. Moreover, the segregation of disturbed patients in a locked area may help to alleviate fear and anxiety in the general ward population and thus reduce their risk of absconding.
Fourth, the extra attention that patients receive from staff as a consequences of absconding may be reinforcing for some patients and may increase the probability of this group absconding in the future. An alternative set of management strategies may need to be identified for these patients. For example, establishing a routine (e.g. regular meeting times with patients) which encourages patients to remain in the unit may be useful.
Finally, the influence of the social environment on behaviour needs to be acknowledged and examined when attempting to reduce the levels of AWOL in hospitals [15]. Morrison [16] noted how the attitudes of staff were closely linked with levels of assaults on staff by patients, a very controlling attitude among the staff was correlated with an increase in the rate of assaults. Similarly, the attitudes displayed by staff on acute admission wards with increasing numbers of very ‘difficult’ patients may contribute to a social climate of unpredictability and perceived threat. In such a setting, AWOL behaviour is more likely to occur [7].
Conclusion
The absconder profile to emerge from this and other studies is unlikely to be specific enough to identify those at risk of absconding. Issues raised in the interviews with patients suggest that situational and environmental factors are also likely to be associated with absconding behaviour and future research should begin to address these. The high rates of repeated absconding in our sample suggest that a small subgroup of patients who absconded repeatedly accounted for a large number of incidents. In the absence of all other indicators, a history of absconding is probably the best predictor of future absconding. Careful assessment of absconding risk in patients at the point of admission and the provision of a more ‘caring’ treatment environment, especially during the early days following admission, may reduce the risk of absconding. However, as Meyer et al. [17, p.304] concluded from their investigation of absconding over 30 years ago, ‘appropriate preventive intervention remains a real possibility as well as a challenge’.