Psychotropicmedication in learning disabilities : audit as an alternative to legislation

Clinical standards were compiled for prescription of psychotropic medication in adults with learning disabilities, and used in a community population for comparisons of current practice in two audit studies 12 months apart. Completing the audit cycle led to improved clinical practice and a rationalisation of prescribing practices. Improvements were achieved on all standards and the methods used to obtain them are discussed.

The use of psychotropic medication in adults with learning disabilities (LD) is fraught with controversy because of the difficulties inherent in making psychiatric diagnoses in this population, where mental illness may present as challenging behaviour (CB), and where the use of psychotropic medication for CB in the absence of mental illness is controversial (Deb & Fraser, 1994).
In the US guidelines for the use of psychotropic medication in LD have been established as a result of litigation and legislation.The Civil Rights of Institutionalised Persons Act makes it illegal to prescribe antipsychotic medication to people with learning disability in institutional care.This has led to a search for alternative medication strategies and an emphasis on physical rather than chemical restraint.The current ethos is that there should be a move away from an outdated 'medical model', where drugs are seen as 'magic bullets', towards a more behaviourally orientated approach (Singh et al, 1992).
In Britain, where the rates for prescription of psychotropic drugs in adults with LD living in the community range from 10-19% (Clarke et al, 1990), an in-patient study fWressel et al, 1990) demonstrated a significant reduction in neuroleptic dosage and polypharmacy with the intro duction of a mandatory annual review of all prescriptions.This suggested that the process of clinical audit might be useful in limiting pre scriptions, and rationalise prescribing practices.Clinical audit, because it emphasises improving current practice, has advantages over litigation and legislation which, by restricting prescrip tions through fear of legal repercussions, will inevitably lead to some patients being deprived of medication which would benefit them.Gravestock (1996) described using a checklist to observe prescribing practices for depot neuroleptics in 79 patients, then developing a set of clinical standards for prescribing depot neuroleptics.Thirty-two patients were monitored over 24 months to audit changes in clinical practice, resulting in an increased frequency of six monthly psychiatric review and a reduction in depot dosages.Neither the outcome of the other 47 patients nor the steps taken to implement clinical standards are discussed.Harvey & Cooray (1993) developed a set of clinical standards for prescribing psychotropic medication in adults with LD.They described applying them to 32 in-patients, but did not complete the audit cycle by changing practice and re-auditing to assess whether changes in practice had resulted in achieving standards set.We expanded their indicators to include separate standards on information, consent, and on the use of psychotropic medication for behavioural disturbance in the absence of mental illness; and we carried out two audit studies 12 months apart, in a community-based population, com pleting the audit cycle.The improvements achieved and the methods used to obtain them are discussed.

The study
In central Manchester the LD mental health service is community based, liaising closely with the community LD teams (CLDT), and has five acute in-patient beds in a general ward.At the time of the study the psychiatric staff consisted of a consultant, a senior registrar and a registrar.The whole team and a representative from the CLDT took part in the study.
Following a literature review and discussion with colleagues at the regional LD audit group, a set of clinical standards was compiled (Table 1).For standard 1, we operationally defined infor mation sufficient for consent to treatment by four criteria (all had to be met) which were: doc umentation of explanations of the drug's target symptoms, efficacy, duration of treatment, and side-effects.These standards were incorporated into a data collection sheet, which was piloted and revised.The amount of information collected was limited to ensure the audit could be completed in an afternoon and was easily repeatable on future occasions.Drugs pre scribed for the treatment of epilepsy were excluded.All psychiatric diagnoses were accord ing to ICD-10 criteria (World Health Organiza tion, 1992).A data collection sheet was filled in for all current patients.The information was obtained from case notes and prescription sheets.
After the first audit we reviewed our clinical standards and decided that they were appro priate.We then took steps to improve our clinical practice and reviewed the success or failure of these by re-auditing a year later.

Findings
On the first audit of 95 patients, 59 (62%) were receiving psychotropic medication (30 men and 29 women, average age 39 years, s.d.= l 1), and of these, 58 were out-patients and one was an inpatient.On the second audit of 108 patients, 75 (69%) were receiving psychotropic medication (41 men and 34 women, average age 39.3 years, s.d.= 12).and 71 were out-patients and four were in-patients.
In the first study, 48 (81%) patients were receiving medication for a psychiatric illness (28 for an affective disorder, 18 for schizophre nia, schizotypal and delusional disorders, and two for generalised anxiety disorder), and 11 (19%) for challenging behaviour in the absence of mental illness (CB) (one had autism and one Asperger's syndrome).In the second study, 68 (91%) of the 75 patients were receiving medica tion for a psychiatric illness (37 for an affective disorder, 27 for schizophrenia, schizotypal and delusional disorders, four for neurotic and stress-related disorders), and seven (9%) for CB (three had autism).
Table 1 shows that performance improved on all standards between audits.Table 2 shows that overall medication prescribed was broadly appropriate to diagnostic group, and that in the second study the increased number of prescriptions is accounted for by the increased frequency of mental illness, with significantly fewer patients receiving psychotropic medication for CB, and of those who were, none were receiving depot or more than one antipsychotic drug.In the first study four patients were receiving benzodiazepines, all of which had been withdrawn by the second study.
In the second study, significantly fewer pa tients were receiving two or more drugs com pared with the first study (44% v. 51%; P<0.05, X2=4.46).Eight patients were receiving more than one antipsychotic drug in the second audit, compared with 10 in the first audit (P<0.05.

Table 1. Audit indicators
Clinical standards for prescribing (information which should be recorded incase notes)1 .The patient (or carer) should be given information sufficient for consentregarding the drug when it was started2.
The patient (or carer) should consent to the drugNumber meeting standards 1 and 23.
All prescriptions should be within the BritishNational Formulary guidelines4.
The reason for the prescription should be recorded at the start of theprescription and at least once in the last 12 months5.
There should be 6-monthly review of medication by a psychiatrist6.
There should be 6-monthly review of side-effects by a psychiatrist7.
There should be 6-monthly review of target symptoms by a psychiatrist8.
There should be yearly consultant reviewNumber meeting standards 3 to 8Number of people in each audit with CB in the absence of mental illness9.

Discussion
The first audit highlighted a number of areas in which we needed to improve our clinical practice.We were pleased to discover that on both occasions all patients were receiving medication within the BJVF(1995) guidelines.We performed particularly poorly on the items of consent and information-giving (standards 1 and 2), partly because we had recently had a number of people resettled in our area who had been in long-term institutional care and who had been discharged on medication with very little information provided.In addition some patients probably had been given information, but this was either not documented or insufficient for consent to treatment.When we re-audited we found although our performance had improved significantly it was still well below 100%.In view of this we have changed our clinical practice from sending all correspondence to the patient's GP to sending clinic letters directly to patients with a copy to their GP.We sent an initial letter of explanation to all GPs involved, asking whether they had any objections, but none did.Patients and carers have found this beneficial and this compares with the advantages found for client held records in the non-LD population (Laughame & Stafford, 1996).Future audits might cover topics such as how well patients and carers have understood the information pro vided, and documentation of the effects of a drug from their point of view.
We managed to improve our recording of the reason for prescriptions, and of 6-monthly re views by a psychiatrist of medication, side-effects and target symptoms (standards 4 to 7) by circulating our list of clinical standards, with a copy of the results of the first audit including recommendations on improved recording of information, and by improving the system for ensuring the handover of patients from junior doctors.We have since incorporated standards 1 to 7 into stickers which will be placed in patients' notes every six months to act as an aide-mÃ©moire to meet these standards.
In the first study we identified a particular problem in recording all types of information about people with challenging behaviour.The problem was partly because our case note system was not designed to cope with recording this type of information.The implementation of the care programme approach helped to ensure better recording of this information.The second audit revealed a significant drop in those on medication with CB, and an improvement in those achieving standards 9 and 10, although still below an ideal of 100%.
Completing the audit cycle enabled us to target our prescribing of psychotropic medication to patients with mental illness, and with regard to those patients receiving psychotropic medication in the absence of mental illness, to ensure medication was used together with, and not as a substitute for, alternative strategies.We successfully discontinued all prescriptions of benzodiazepines, significantly reduced polypharmacy and number on more than one antipsychotic drug, and improved monitoring of mood stabilisers.
There is still room for improvement and we plan to re-audit in a further six months to assess the efficacy of the changes we have made in our clinical practice.
The psychiatrist who works with adults with LD has to pilot a course between the Scylla of failing to treat mental illness manifesting as a behaviour problem, and the Charybdis of treat ing a behaviour problem inappropriately with psychotropic medication.We have found the process of clinical audit to be a useful naviga tional device for steering a safe passage.
case conference or documented programme of behavioural therapy or environmental manipulation 10.There should be a 6-monthly riskbenefit assessment(such as side-effects 4 (36) 5 (71) v. quality of life) and/or evaluation of alternative approaches Number meeting standards 9 and 10 *P<0.1, '*P<0.001(two-tailed).

Table 2 .
Breakdown of prescribing patterns in audit one (Al) and audit two (A2) by diagnosis (ICD-10 codes in brackets)