Discussion

In our sample of adults with intellectual disability who were under the care of mental health services but were receiving their prescriptions for continuing antipsychotic medication from primary care, the quality of the clinical correspondence sent by mental health teams to primary care was associated with the quality of clinical review: when there was a clear statement about which service was responsible for reviewing the side-effects and continuing need for antipsychotic treatment, it was much more likely that such reviews had been documented in the previous 6 months.

Plans for reviewing the clinical benefits of antipsychotic medication were communicated to primary care in two-thirds of cases, and such reviews were documented in the mental health clinical records in around three-quarters of cases, with little difference between the subgroup for whom plans had been communicated and the subgroup for whom they had not. The most likely explanation for this finding is that there was an implicit understanding by both mental health services and primary care that the mental health team will be responsible for the continued monitoring of mental state and behaviour. The finding that no clinical plans relating to future reviews of the effectiveness of antipsychotic medication had been communicated to primary care for one patient in four overall is consistent with this interpretation.

Plans for monitoring the side-effects of antipsychotic medication were communicated to primary care in three out of every five cases, only a slightly lower proportion than for plans for monitoring mental state and behaviour. However, where such plans had been clearly communicated, a review of side-effects was >50% more likely to be documented in the previous 6 months than cases where plans had not been communicated. A possible explanation for this finding is that clinicians within mental health teams considered the monitoring of side-effects and the associated physical health checks to be evidently the responsibility of primary care, as an integral part of providing continuing prescriptions for antipsychotic medication, and therefore failed to make this explicit in communications. But where a clear plan for monitoring side-effects is not communicated, primary care may assume that their role is to provide access to antipsychotic medication, with monitoring remaining the responsibility of the mental health team, as part of shared care.

Plans for reviewing the continuing need for antipsychotic medication were communicated to primary care in only a quarter of cases; for cases where such a plan had been communicated, a review of this aspect of care was more than two and a half times more likely to have been documented compared with those cases where there was no such plan. The finding was the same in the subgroup of people with a diagnosed psychotic or affective disorder, for whom continuing antipsychotic medication is likely to be indicated, as well as the subgroup prescribed antipsychotic medication for other indications (including behaviour that challenges), where the rationale for continuing treatment with antipsychotic medication may be far less convincing. This suggests that correspondence with primary care may not consistently and sufficiently address an individual's clinical circumstances and the criteria on which to base a decision to continue or withdraw antipsychotic medication, thus reinforcing any assumption by primary care that determining the need for continuing antipsychotic medication remains the remit of the mental health team. But such an assumption increases the risk that for individuals who are clinically stable and discharged to the sole care of their general practitioner, prescribing may become long-term by default, which may at least explain some of the findings from surveys of practice in primary care.^{Reference Sheehan, Hassiotis, Walters, Osborn, Strydom and Horsfall1,Reference Henderson, Mcskimming, Kinnear, McCowan, McIntosh and Allan6}

Suboptimal communication across the interface between services is not a new issue nor is it specific to intellectual disability services or medication-related issues.^{Reference Lester16} For people prescribed antipsychotic medication for the treatment of schizophrenia, ‘shared-care’ agreements that clearly outline the responsibilities of primary and secondary care are commonly ratified at a systems level, but we are not aware of any such documents for the use of antipsychotic medication in people with intellectual disability who have behaviour that challenges. Although it would be possible to specify requirements for the monitoring of side-effects within such frameworks, there is insufficient evidence on which to base universal instructions for de-prescribing.^{Reference Shankar, Wilcock, Oak, McGowan and Sheehan7} Thus, the criteria on which to base decisions to reduce or stop antipsychotic medication in people with intellectual disability need to be individualised and clearly communicated.

In conclusion, when the responsibility for prescribing continuing antipsychotic medication for people with intellectual disability has been transferred to primary care, the monitoring and review of this medication may fall short of recommended best practice. Our findings suggest that this may partly reflect the failure of mental health services to communicate reliably with primary care about the allocation of responsibilities for reviewing antipsychotic medication. This is particularly the case for the monitoring of side-effects and consideration of whether to reduce or discontinue the medication. The use of a formatted letter that prompts inclusion of a clear message about which service is accountable for which aspects of monitoring may be an effective quality improvement intervention.