Evaluation of psychiatry assessment and planning units

To the Editor

Mitchell et al. (2020) have produced an interesting article reporting a comprehensive evaluation of three emergency department (ED)-embedded psychiatric assessment and planning units (PAPUs). EDs are normally not the best place to offer therapeutic interventions for psychiatric crises; hence, we welcome the initiative of the authors. However, we also have a number of remarks about their evaluation.

Table 4 lists the diagnostic-related groups (DRGs) of patients who attended the PAPUs. DRG is a controversial concept (Hunter and McFarlane, 1994). Some of the categories in Table 4 are very specific such as borderline personality disorder (BPD), and some are not such as ‘suicidal ideation/attempt’, ‘unspecified’ or ‘medication changes’. We noticed that in Eastern 3.3% of the patients had been assigned to DRG suicidal ideation/suicide attempt while in Austin it was 22.7%. Furthermore, in Eastern 34.1% of the patients received DRG BPD yet conversely in Austin it was 9.9%. It could well be that clinicians in Austin did not want to label patients with BPD.

The overall evaluations of the units were quite positive and the authors also rightly investigated the effect on the overall system. One of their findings was that the average waiting time before definite care did not always decrease. We think that their analysis could be enhanced by also investigating the patients who were deemed not suitable for the PAPU, especially those deemed to be too high risk, as this was also a theme in the qualitative interviews with staff. PAPUs may not be the way forward, if patients who need to remain in the EDs end up having a worse experience (e.g. because staff have been moved from the ED to staff the PAPU instead).

The last point we would like to offer is to ask whether the authors considered long-term evaluation before concluding that ED-embedded PAPUs are safe and effective? The rate for 28-day re-admissions was approximately 10% in all units and we are wondering about the re-admission rate over a longer period of time (1 or 2 years) and whether it is related to diagnosis? For example, there is some evidence that admission to hospital or ED as such is associated with less favourable outcome in patients with BPD (Coyle et al., 2018).