Early Intervention (EI) services aim to reduce progression to chronic illness for patients with schizophrenia. The Lambeth Early Onset study (2002) demonstrated reduced hospitalisation at 18 months for patients exposed to EI services. This study assesses the durability of these benefits at 5 years.

Hospital use in the LEO cohort was assessed by case note review.

There was no statistically significant difference in terms of ever being admitted at 5 years (OR 1.42; 95% CI 0.550 - 3.68; p=0.468)). Although the mean number of admissions was lower in the EI group: 1.65 (SD = 0.86) versus 1.83 (SD= 0.92), this difference was not significant (coefficient = 0.096; 95% CI -0.550 - 0.742; p=0.770). At 5 years patients assigned to EI used 42.25 days (SD 112.8 days) versus 51.41 days (SD 125 days); coefficient = 6.344; 95% CI -46 - 58.7; p= 0.810. The primary outcome was robust to potentially extreme admission rates for missing subjects on sensitivity analysis.

There is no evidence from these data that EI services statistically significantly reduce hospital use at 5 years follow-up (vis a vis care as usual) despite some interim benefits at 18 months. We only considered a limited number of outcomes and more work needs to be done in this area.