Psychotropic drug prescribing in a Nigerian psychiatric hospital.

Psychopharmacotherapy dominates the therapeutic arsenal of psychiatrists and, not surprisingly, psychotropic drugs are widely consumed in psychiatric practice. The pattern of prescribing of these drugs needs to be appraised in terms of 'rational drug use', which may be defined as 'the use of the least number of drugs to obtain the best possible effects in the shortest possible time and at a reasonable cost' (Gross, 1981).

Observations on prescribing practice can be the basis of useful biomedical research (Barbul & Tansella, 2005) and are a form of self-audit for hospitals (Edwards & Kumar, 1984). Several studies have noted variations in prescribing patterns in different settings (Xiang et al, 2007). These variations (which may be related to poverty, lack of community resources and treatment adherence) and conformity with rational drug use in Nigeria were investigated in this study.

Methods

We enquired into the pattern of prescribing psychotropic drugs at the Uselu psychiatric hospital, Benin City, Nigeria, a 250-bed institution run by the federal government which serves an estimated population of 13 million.

We randomly chose a census day in March 2007 and extracted prescribing data on all 143 in-patients, none of whom had a physical disorder that might have influenced the psychotropic prescription. Five case files were excluded because the record was inadequate. The data colleced were key demographic factors, type of medication, frequency of use and administration schedules of prescribed drugs, their mean daily doses, combinations, use of anticholinergics, ‘when required’ (p.r.n.) orders and monthly costs of the drugs to the patients.

The data were processed using SPSS version 10 and a 5.0% significance level was used for comparison of frequencies using chi-square statistics.

Results

Prescriptions of antipsychotics, anxiolytics, anticholinergics, antidepressants and mood stabilisers varied significantly across some demographic variables: gender, Kruskal–Wallis value (KW) = 13.76, d.f. = 4, P < 0.001; marital status, KW = 20.23, d.f. = 16, P < 0.005; age (10-year groupings), KW = 19.73, d.f. = 16, P < 0.05; and employment, KW = 21.51, d.f. = 8, P < 0.05. These distributions reflect the characteristics of patients seen at the hospital rather than the specifics of prescribing practice there.

The survey revealed a high frequency of prescribing of conventional antipsychotics to all patients with functional psychotic disorders: 51.1% of these patients were prescribed haloperidol, 45.5% chlorpromazine, 38.5% trifluoperazine, 10.5% fluphenazine decanoate. There was less frequent prescribing of the atypical (second-generation) antipsychotics: 4.9% risperidone and 1.4% olanzapine.

Of the 15 patients with a depressive disorder, 10 were prescribed amitriptyline (a tricyclic antidepressant), three sertraline, one paroxetine and one fluoxetine. Regarding the mood stabilisers, for bipolar affective disorder, 15 patients were prescribed carbamazepine and 4 sodium valproate. One patient received a high dose (450 mg) of thioridazine.

The mean daily doses of all drugs prescribed were within acceptable therapeutic ranges for the respective diagnoses; however, there was a remarkable consistency in the p.r.n. duo of intramuscular chlorpromazine 150 mg and intravenous diazepam 30 mg for rapid tranquillisation of severely agitated patients. Generally we found no anomalies in the matching of drug types with diagnostic groups.

Polypharmacy was noted in 92.3% of all prescriptions, with 62.2% of patients being on antipsychotic/anticholinergic combinations (Table 1). The commonest adjunctive drug was an anticholinergic agent, benzhexol, which was noted in the prescriptions of 82.6% of patients placed on regular antipsychotics and was administered once or twice daily. The frequency of p.r.n. prescribing was 57.7% and was for patients with non-affective psychotic disorders, either for sedation or to control extrapyramidal reactions. The other major finding was that 36.4% of prescriptions were to be administered twice and 6.8% thrice daily.

Table 1

Drug combinations

Drug combinations	Number of prescriptions	Percentage of total
Antipsychotic alone	11	7.7
Antipsychotic + anticholinergic	89	62.2
Antipsychotic + anxiolytic	3	2.1
Antipsychotic + anxiolytic + anticholinergic	5	3.5
Antipsychotic + antidepressant	5	3.5
Anxiolytic + antidepressant	1	0.7
Antipsychotic + mood stabiliser	5	3.5
Antipsychotic + mood stabiliser + anticholinergic	13	9.1
Antipsychotic + antidepressant + anticholinergic	9	6.3
Antipsychotic + mood stabiliser + anticholinergic + anxiolytic	1	0.7
Anxiolytic alone	1	0.7
Total	143	100.0

The monthly costs of the typical antipsychotics prescribed varied between US$2.44 for fluphenazine decanoate to US$12.0 for thioridazine, while the generally preferred atypical antipsychotics cost much more: risperidone US$17.00 and olanzapine US$36.70. For the treatment of depression, the monthly cost of the most frequently prescribed antidepressant, amitriptyline, was US$5.00, which is much cheaper than the selective serotonin reuptake inhibitors: sertraline US$18.0, paroxetine US$60.90 and fluoxetine US$65.00.

Discussion

The most discernible practice was polypharmacy, as has previously been reported in Nigeria (Famuyiwa, 1988) and in numerous other studies (see Cookson et al, 2002) but its rationale has been questioned, because of its high risk of dangerous interactions and the burden it imposes on both nurses and the patient, as well as carers. However, polypharmacy may be justified for some patients who are unresponsive to single-drug treatments, provided close monitoring for undesirable effects or toxicity is ensured (Broekema et al, 2007). Combinations of two or more drugs should generally be avoided, but if a combination is judged to be necessary then the drugs should belong to different chemical groups (Cookson et al, 2002).

The liberal use of anticholinergic drugs was probably due to the high rate of prescribing typical antipsychotics, as also reported in several other studies (Xiang et al, 2007). Although anticholinergic drugs are useful in controlling extrapyramidal reactions, they may be misused by some patients, as they can give the experience of euphoria, and they can cause cognitive impairment (e.g. memory loss, particularly in the elderly), aggravate tardive dyskinesia and precipitate psychiatric symptoms (Cookson et al, 2002). The current consensus is that anticholinergic drugs should be administered only to patients who are experiencing extrapyramidal symptoms. However, the majority of patients on therapeutic doses of typical antipsychotics do not experience gross neurological side-effects.

Another significant finding was the high rate of p.r.n. prescribing, which was associated with high doses and polypharmacy, and should be discouraged. A Cochrane systematic review of randomised trials comparing ‘as required’ medication regimens with regular, fixed, non-discretionary regimens for schizophrenia found no evidence to support the common practice of p.r.n. prescription (Chakrabrati et al, 2007). The other objection is that the attending nurse may not have the requisite expertise to decide on the appropriateness and safety of extra doses. Thus, p.r.n. prescribing may be appropriate for unpredictable patients but, in terms of safety, it is good practice for the prescriber to specify indications for each p.r.n. dose, for example ‘one hour delay in falling asleep’, or to restrict the maximum number of doses within a given period, for example ‘not more than 3 p.r.n. doses in 24 hours’.

Tranquillisation with parenteral antipsychotics should be avoided, although this is rather impracticable in a low-income country, due to staff shortages and high case-loads. Relating multiple administration to daily schedules (e.g. three times daily) is pharmacologically not justified because most psychotropic drugs have a long biological half-life; furthermore, it consumes considerable nursing time.

A notable feature was the low prescription rate for atypical antipsychotics, despite their efficacy and satisfactory side-effect profile. The clinicians were apparently aware that the large majority of patients had a low income (the minimum wage in Nigeria is US$54 per month), hence they were unwilling to prescribe them, but curiously they are comparable in efficacy to some typical antipsychotics (Gureje et al, 2007).

Overall, the pattern of prescribing psychotropic drugs observed was similar to that in several other low- and medium-income countries. Increasing sensitivity to patients’ needs had led to a number of ameliorative measures. For instance, education of prescribers by the use of a manual (Baker et al, 2008), routine checks by and feedbacks from pharmacists on adherence to rational drug use and a combination of education and reminders on medication charts (Thompson et al, 2008) have been found to be useful. However, presently in Nigeria, as perhaps elsewhere in sub-Saharan Africa, such feedback seems impracticable, in that most pharmacists are in private practice and hence have no statutory links with psychiatric hospitals. Alternatively, measures to stress the importance of rational drug use should be in place in psychiatric training curricula and prescribing guides for clinicians.