A pharmacy led program to review anti-psychotic prescribing for people with dementia

This pharmacy-led program was conducted within General Practitioner (GP) surgeries in Medway Primary Care Trust (PCT). Medway PCT covers the Medway towns in Kent with a total population of 256,700 and a 60 plus population of 51,500 [17]. Medway is a relatively deprived area, ranked 132rd most deprived of 325 UK boroughs in the 2010 Index of Deprivation, with pockets of significant deprivation; eight geographical units called ‘Super Output Areas’ (SOAs) were ranked in the 10% most deprived nationally [18].

Pharmaceutical support within Medway PCT is provided by a medication management team containing pharmacists, pharmacy technicians and support staff; the team has a history of a close working relationship with GPs and undertaking clinical projects within individual practices. Following the publication of the NDS one of the key strategic priorities of the PCT was to limit the prescribing of anti-psychotics to people with dementia. This project, which was conducted from January to December 2011, had two aspects. First, in line with the NDS, data were collected on anti-psychotic usage to identify people on the dementia register within Medway PCT prescribed anti-psychotics. Second, a specialist pharmacist intervened in an appropriate cohort of people with dementia commenced on anti-psychotics by primary care identified by the data search.

Every practice information system, both the electronic and paper format, was searched, by an experienced pharmacy technician to identify people with dementia currently prescribed a low dose anti-psychotic, on acute or repeat prescription. Dementia registers were established within the UK as part of the Quality Outcome Framework (QOF) – an incentive scheme to reward good practice in primary care - to record people with dementia within the practice [19, 20]. The registers were established within Medway PCT in 2006/07 and were used to identify people with a confirmed diagnosis of dementia. The individual patient record including the medication history of every person on the dementia register was then searched to identify people prescribed low-dose anti-psychotics. Initially, this was restricted to the 6 most commonly prescribed anti-psychotics based on ePACT (electronic Prescribing Analysis and Cost Tabulation) data. However, to ensure completeness, patient’s notes were hand searched to identify other, less frequently prescribed, anti-psychotics. For the six most commonly prescribed anti-psychotics, low dose was defined a priori (olanzapine up to 10 mg daily, risperidone up to 2 mg daily, quetiapine up to 100 mg daily, amisulpiride up to 50 mg daily, sulpiride up to 200 mg daily and haloperidol up to 1.5 mg daily). For the other less frequently prescribed anti-psychotics the standard British National Formulary (BNF) dosing schedules were used to inform the decision on whether the dose was low [21].

Finally, the document storage system, where key documents such as referral letters are stored, was searched to determine where treatment was commenced (primary care, secondary care mental health trust, acute trust or Learning Disability Team). If the anti-psychotic had been initiated by a secondary care organisation, the patient was not included in the medication review, because the anti-psychotic may have been used for under-lying psychosis or a complex mental health problem. The residential status (living at home or in a care home) was also collected.

After the cohort of people with dementia on long-term anti-psychotics initiated by primary care was identified, the intervention was delivered. Practices were ranked according to the frequency of prescribing to prioritise where the intervention should be delivered. The pharmacist delivering the intervention sent a copy of a report to the GP, which identified the cohort of people with dementia where anti-psychotic withdrawal should be attempted and that the pharmacist was able to give support to the GP in the process by providing advice on a suitable withdrawal scheme, advising on local and national guidance and linking with the care home [14, 22‐24]. This initial communication was followed up either verbally or in writing.

The pharmacist collaborated with the GP in identifying, where withdrawal was potentially suitable and the level of pharmacy support including whether a pharmacist-led medication review, was required. Medication was withdrawn for clinical, not research reasons. Hence, consent was not required. All participants received usual care and any decision to alter medication was made by the GP based on individual clinical needs, following the pharmacist-led review. Withdrawal was generally considered if the patient was not under secondary care services, was receiving an anti-psychotic for non-acute behavioural problems, and the anti-psychotic had not been reviewed in the previous 12 months.

The clinical medication review (based on modified National Prescribing Centre Level 3 medication review criteria) was delivered by an experienced senior clinical PCT pharmacist, who had a further qualification in dementia with full access to medical and care home notes [25]. The review included an individual withdrawal plan, which took into account the length of time the person had been on the anti-psychotic. In general the dose was gradually reduced usually by 50% every two to four weeks following discussions with the prescriber. During the withdrawal patients continued to be monitored, followed up, and treated by their own physicians according to their individual clinical needs. Care home staff were encouraged to log behaviours using ABC (Antecedents, Behaviour, Consequences) charts and monitor residents for any change in general health [23].

Care staff discussed the project with next-of-kin and whenever possible the person with dementia; if required the pharmacist supported these discussions. Any decisions from these discussions incorporated the views of the person with dementia, next–of-kin and professional carers. The pharmacist also provided support and training to the care home, which was designed to help staff manage BPSD, and was available for further advice and follow-up. The training utilised treatment guidelines for use in primary care to treat BPSD that were developed by a multi-disciplinary team of clinicians from the local mental health trust and local PCTs [23]. The overall concept of the training was based upon a person-centred approach including the need to understand the causes of any behaviour and accurately record any behaviour, non-pharmacological treatments options, and the appropriate pharmacological management of behavioural problems in dementia [22, 23].

This project was approved by Medway PCT. Ethics committee approval was not required because the project was a service evaluation and data were collected to evaluate this new service and not for the purposes of research. The decision to publish was made retrospectively and individual patient data was not identifiable to anyone in the team (IM, CF) conducting the secondary analysis of the results for publication.

Data were collected from 59 out of 60 (98.3%) practices (or groups of practices) within Medway PCT; one practice declined to take part. There were 1051 people on the dementia register in the 59 practices; of these 462 were in residential care and 589 lived in their own home. In total 161 people on the dementia register were receiving low-dose anti-psychotics; of these 118 were in residential care and 43 lived in their own home. People with dementia living in residential homes were nearly 3.5 times more likely to receive an anti-psychotic; 25.5% (118/462) of people resident in care homes compared with 7.3% (43/589) of people living in their own homes (p < 0.0001; Fisher’s exact test) [26].

On average (mean value) each practice treated 2.77 (range – 0 to 26; +/− SD 4.88) people with dementia with low-dose anti-psychotics. In 26 practices no-one with dementia was receiving low-dose anti-psychotics. Put another way, if one assumes that the dementia register was accurate, approaching half (44.1%) of all practices in Medway PCT were not treating anyone with dementia with a low-dose anti-psychotic. Five practices accounted for over 50% (81/161) of the prescribing; 3 of these 5 practices were significantly larger (16,575, 13,805 and 8,866) than the mean practice in Medway PCT of 4,683 patients (report generated from http://​www.​epact.​ppa.​nhs.​uk on the 13^th April 2012). The remaining two practices were responsible for large care homes.

The most commonly used anti-psychotic was amisulpride (52 out of 161; see Table 1 for full details of anti-psychotics used).

Of the 161 people on the dementia register prescribed low-dose anti-psychotics, 87 were receiving on-going treatment from local secondary care mental health services and 4 from the local Learning Disability Teams. The remaining 70 patients were included in the pharmacy led review. Different surgeries accessed different levels of pharmaceutical support and in total, following the pharmacy-led intervention, anti-psychotics were withdrawn, or the dosage of the anti-psychotic was reduced, in 43 instances (61.4%). The pharmacist directly liaised with 11 practices to withdraw anti-psychotics in 33 people (47.1%). In a further two practices, one which supported a specialist dementia home, six patients identified during data collection were withdrawn from anti-psychotics. In one further practice, which supported a large EMI (elderly mentally ill) home, the pharmacist developed and provided withdrawal schedules for four patients.