How to interpret different results for CRHTT data

Jacobs & Barrenho ¹ used the same data as Glover et al ² when they were comparing admissions in primary care trusts with and without crisis resolution and home treatment teams (CRHTTs). However, they employed different methods for their analysis and reached conflicting conclusions. According to Jacobs & Barrenho, the introduction of CRHTTs did not have a statistically significant influence on the number of admissions, while Glover et al found a significant reduction especially for CRHTTs which offered a 24-hour service.

In their article, Jacobs & Barrenho ¹ do report a reduction in admissions (e.g. Fig. 4) but state that it was not statistically significant. They do not mention power calculations. There were usable data available from 229 primary care trusts (PCTs) and the authors conducted various complex analyses by using a number of control factors and by studying trends over time. It could be that their lack of statistically significant findings is because of a lack of power. If this is the case, there is no fundamental difference between their findings and the previous analysis. ²

At the end of their article, the authors make the suggestion that perhaps data should be analysed at the level of CRHTTs and not at the level of PCTs, given that there is huge variation between CRHTTs. We concur with that suggestion and we would like to go even further and suggest that future studies look at the service actually provided to individual patients in terms of how many visits are undertaken over a specified number of days. This information is readily available from most electronic notes systems. Further study is needed to investigate the types of interventions provided, such as whether medication was prescribed and administered, whether specific psychological treatments were offered, and so on. The availability of such data will allow an informed decision to be made about what is required to avoid admission to hospital and whether a CRHTT is the best organisational format to deliver that care.