CORE study: different interpretation of the results

Lloyd-Evans et al ¹ published results from a cluster-randomised trial looking at the effect on patients of an improvement programme for mental health crisis resolution teams, in which the aim was to increase fidelity with the crisis resolution team model. In the intervention group, the authors found a reduction in admissions and in-patient bed days but no increase in average patient satisfaction. We have two comments about interpretation of their results.

First, the authors report that there was no difference in average patient satisfaction score between the intervention and the control group. They offer a ceiling effect as a possible explanation, given that average patient satisfaction was already high before the intervention. We wonder whether this ceiling effect can be at least partially explained by the timing of their assessment? The authors measured patient satisfaction around the time of discharge from the home treatment team. Patient satisfaction, however, tends to be lower if the time interval between intervention and measurement is larger.² The Mind report, Listening to Experience ³ – cited by the authors – suggests that patients are far more critical about crisis care, when questioned at a much later date following discharge. Studies reporting patient satisfaction 6 months or longer after the crisis episode are desperately needed.

Second, there remains the question of whether the observed reduction in admissions and in-patient bed days found in the intervention group is related to an increase in the fidelity scores. The crisis resolution teams in the intervention group received additional support to increase both their fidelity to the model and their scores on the fidelity scale. And yet despite this, the authors also mention in the article, and in the supplementary material (pp. 47–50), that there is no relationship between the fidelity scale scores and the reduction in admissions and in-patient bed days.

This makes us wonder about what are the causal factors in reducing admissions and in-patient bed days? It seems that an increase in scores on the fidelity scale is not necessarily essential to achieving this. This observation is important for us as practicing clinicians. The results here suggest that we ought to be aiming to secure the actual intervention itself, namely the access to a facilitator, the opportunity to discuss team improvement at a specially arranged day and the development of a service improvement plan and not be focusing on getting higher scores on the fidelity scale.