Understanding the Length of Stay (LOS) in Mental Health Acute Wards for Patients of the Joint Homelessness Team vs Other Westminster Patients

Abstract

Aims

The Joint Homelessness Team (JHT) is a specialist mental health service for rough sleepers in Westminster. Patients have many years of declining mental health, leading to a downward social decline, resulting in homelessness.There is a concern that mental health patients are discharged prematurely due to bed pressures. This means treatment may not be optimised, with social issues and risks remaining. Serious mental illness (SMI) is associated with reduced life expectancy and high morbidity. Thus, homelessness and SMI are a double-hit of risks for people. For JHT patients, there are many barriers to admission, including coordinating complex street or S135(1) assessments and police availability; some cases have taken months or years to plan. If patients remain sub-optimally treated at discharge, there's a high risk they abandon accommodation and return to rough-sleeping; leading to further lost years of illness and homelessness.

Methods

This was a retrospective analysis. Via SystmOne, we identified all JHT admissions in 2021 and their LOS. Data were collected, including demographics, mode of admission, discharge destination and whereabouts at three months (as a secondary endpoint). LOS figures were gathered for other Westminster patients from general adult wards for 2021.

Results

There were 57 JHT admissions in 2021. 1 patient was excluded as still admitted. 22 patients were already care-coordinated by JHT and admitted for relapses; for the remainder it was a first admission.

For Westminster patients, there was a low variation throughout the year in number of monthly admissions and LOS; monthly average LOS range was 30-38 days. For JHT, there was higher variability for number of admissions and LOS with no seasonal pattern; monthly average LOS ranged from 4-95 days.

At three months, 22 (39%) patients were not housed in the community. Of these: 3 were discharged to the streets; 9 became street homeless; 3 were discharged to ‘Stepdown’ and went AWOL soon after; 5 patients were readmitted; 2 patients their location was unknown. 28 (50%) were housed in homeless hostels.

The data were presented at a borough-wide academic meeting.

Conclusion

While the dataset is small, the LOS for JHT was inconsistent, reflecting the variability of the cohort. This may need further exploration.

With 39% of admissions having unsatisfactory endpoints, this suggests that many patients were not well enough to work collaboratively out in the community. There was a consistent pattern of shorter admissions leading to poorer endpoints at three months.