Is it possible to predict hospitalisation during intensive home treatment? A retrospective cohort study

Intensive Home Treatment (IHT) is an alternative to acute inward treatment. The objective of this study was to assess which variables predict that a patient admitted to IHT required transfer to hospital for inward management.

We included the first 1000 episodes admitted to IHT and looked for crude associations between potential predictive factors and transfer to hospital. Then, we built a predictive model for this outcome.

The patients with a higher risk of transfer to hospital were those who had previous hospitalisations (OR = 2.6; 95% CI = 1.4-4.7), more admissions in the previous 5 years (median= 0, IQR = 0-1 vs. median = 0, IQR = 0-1.5; p = 0.0011) and a higher clinical severity at IHT admission (mean difference = 0.36; p50 = 0, IQR = 0-1.5 vs. p50 = 0, IQR = 0-1; p = 0.0011). The predictive model included age, previous admissions, clinical severity at IHT admission, and substance use at the beginning of the episode but had a low performance (R2 = 0.115; AUC = 0.752, 95% CI: 0.690-0.814).

Our results are consistent with those from previous studies in countries with different mental health systems. Far from cautioning us against using IHT in patients with severe symptoms or previous hospitalisations, these results should encourage us to find ways to offer them greater support at home.