CLINICAL OPINION ON UPPER DOSE LIMITS

SUMMARY

Questionnaires enquiring about antipsychotic dosages of 11 commonly used neuroleptic drugs were sent to samples of 400 consultant psychiatrists in 1985 and 1991. The total number of useable replies was 147">

Back

 

CLINICAL OPINION ON UPPER DOSE LIMITS

OF ANTIPSYCHOTIC DRUGS BETWEEN 1985 & 1991

 

P.E. Harrison-Read  PhD FRCPsych

D. Marchevsky  MRCPsych

and J. Steinert  MB FRCPsych

 

SUMMARY

Questionnaires enquiring about antipsychotic dosages of 11 commonly used neuroleptic drugs were sent to samples of 400 consultant psychiatrists in 1985 and 1991. The total number of useable replies was 147">

Back

 

CLINICAL OPINION ON UPPER DOSE LIMITS

OF ANTIPSYCHOTIC DRUGS BETWEEN 1985 & 1991

 

P.E. Harrison-Read  PhD FRCPsych

D. Marchevsky  MRCPsych

and J. Steinert  MB FRCPsych

 

SUMMARY

Questionnaires enquiring about antipsychotic dosages of 11 commonly used neuroleptic drugs were sent to samples of 400 consultant psychiatrists in 1985 and 1991. The total number of useable replies was 147">

Back

 

CLINICAL OPINION ON UPPER DOSE LIMITS

OF ANTIPSYCHOTIC DRUGS BETWEEN 1985 & 1991

 

P.E. Harrison-Read  PhD FRCPsych

D. Marchevsky  MRCPsych

and J. Steinert  MB FRCPsych

 

SUMMARY

Questionnaires enquiring about antipsychotic dosages of 11 commonly used neuroleptic drugs were sent to samples of 400 consultant psychiatrists in 1985 and 1991. The total number of useable replies was 147, representing return rates of 15 & 22%. There were no appreciable differences between the results of the two surveys. Psychiatrists were generally conservative and more consistent than the British National Formulary. The stated maximum doses (95% upper confidence limits of the median) exceeded those advised in the British National Formulary (BNF) only in the case of fluphenazine decanoate and haloperidol decanoate. Although psychiatrists did, up to 1991, appear to underestimate the potency of oral and depot haloperidol, and perhaps some other depot neuroleptics, this was not a general trend with all antipsychotic drugs, and may be accounted for by incomplete or misleading recommendations about dosages in the BNF.

KEY WORDS: antipsychotic drugs; high dose limits; clinical opinion; prescribing; questionnaire surveys.

 

INTRODUCTION

The pros and cons of high dose neuroleptics in the treatment of functional psychosis have been discussed and summarised by the Royal College of Psychiatrists' Consensus Panel (Thompson, 1994) and related articles (Hirsch & Barnes, 1994; Kane, 1994). Possibly because of the implication of conservatism in BNF guidelines on high dose neuroleptics, it is sometimes supposed that, before the publication of the College's recommendations, there was widespread use by psychiatrists of doses well above the suggested upper end of the BNF range (Thompson, 1994). Although an audit of antipsychotic drug prescribing in a regional secure unit reported that 50% of patients were receiving doses above the BNF advisory maximum limits (Stanley & Doyle, 1993), other surveys suggest that this is not a general state of affairs (Gill, 1993; Torkington et al, 1994).

In 1985 and 1991 we carried out surveys of the views held by consultant psychiatrists on doses of neuroleptics suitable for the treatment of functional psychoses. These surveys were intended to reflect the opinions of jobbing psychiatrists over a period when the disadvantages and problems associated with prescribing high doses of neuroleptics were becoming better known. This issue received more attention in the United States (Baldessarini et al, 1988), and was not reflected by changes in BNF guidelines during this period. Our hypothesis was that psychiatrists would recommend the use of neuroleptic doses in excess of the BNF maximum limits in 1985, but with a smaller discrepancy in 1991. We surveyed "biological psychiatrists" (those belonging to the Biological Psychiatry Special Interests Group of the Royal College of Psychiatrists) separately from other psychiatrists. Since consultant psychiatrists with a professed interest in biological psychiatry might be expected to have taken more notice of clinical research in this area, we predicted that the maximum dose limits recommended by biological psychiatrist would be substantially lower than those recommended by other psychiatrists in 1991, but not in 1985.

METHODS

Sample: For the 1985 survey, consultant psychiatrists working in the United Kingdom and the Republic of Ireland were selected at random from two lists derived from the 1985 Royal College of Psychiatrists Membership List. The first list comprised members of the Biological Psychiatry Section of the College, and the second list consisted of all other consultant members of the College. In 1985, questionnaires were sent to 200 biological psychiatrists and 200 other psychiatrists. In 1991, questionnaires were sent to the remaining 157 biological psychiatrists in the 1985 list and to 243 other psychiatrists selected at random from the same 1985 list who had not been surveyed previously.

Respondents were drawn from the same pool of psychiatrists on both occasions. Any change in the views expressed by psychiatrists surveyed on the two occasions would therefore be most likely to reflect changes in clinical opinion over time rather than the different opinions of a younger group of psychiatrists who might otherwise have been included in the later survey. Since the questionnaire was returned anonymously by many psychiatrists, it would not have been possible to ascertain the views of the same prescribers on the two occasions.

Questionnaire: The questionnaire was introduced by the statement : 'It seems to us there are discrepancies between recommended doses of some antipsychotic drugs, e.g.: given in the BNF, and doses actually required and used in practice.' Psychiatrists were requested to indicate for some commonly used antipsychotic drugs, the dose range which they had found to be effective in the majority of cases of functional psychosis (schizophrenia, manic-depressive illness, etc.), indicating the low and high end of the usual dose range. They were also asked to suggest minimum doses of each drug which they had found or would anticipate to be effective in unusually sensitive cases, as well as the maximum doses which they had found it necessary to use, or would be prepared to use, in severe resistant cases. It was emphasised that these doses should pertain to optimal treatment of active psychotic symptoms, rather than to prophylactic treatment, or to maintenance treatment of residual symptoms. A total of 11 antipsychotic drugs were included in the questionnaire, 5 oral drugs (chlorpromazine, haloperidol, trifluoperazine, thioridazine, and pimozide) and 6 drugs administered in the form of intramuscular depot injections (flupenthixol decanoate, fluphenazine decanoate, zuclopenthixol decanoate, haloperidol decanoate, pipothiazine palmitate, and fluspirilene). Clozapine was not included as it was not available in 1985, and sulpiride was omitted because its low potency militates against "high dose" regimes.

Statistics: The principle statistic used was the median value for each range of doses of each drug. The median was used in preference to the mean as the stated doses were not normally distributed. Ninety-five percent confidence limits of the median values were derived using tables published in 'Documenta Geigy'. Any reference value (e.g. BNF advisory maximum dose) falling outside these confidence limits of the median could be regarded as statistically significantly different, P<0.05.

RESULTS

Questionnaire Returns: Sixty-nine questionnaires out of 400 were returned in 1985 of which 61 (15.2%) were appropriately answered, (biological psychiatrist 32, other psychiatrists 29). The return in 1991 was 118, of which 86, (21.5%) were appropriately answered, (biological psychiatrists 39, others 47).

No psychiatrist who replied objected to being surveyed, but a few did not fill out the questionnaire on the grounds that they had retired or were no longer working in a field where antipsychotic drugs were used to any extent. Although the instructions to the questionnaire emphasised that we were concerned only with the treatment of active psychotic symptoms, some respondents complained that we had not explicitly excluded the control of excited and aggressive behaviour from our indications for neuroleptics. The implication was that they would use higher doses for behavioural control than for treatment of psychotic symptoms. One respondent in 1985 declined to state the high doses he sometimes felt obliged to use on the grounds that to do so might condone what he feared would later prove to be bad practice. Several respondents in both surveys indicated that their minimum doses were for prophylaxis against relapses. For this reason, the minimum dose estimates were considered to be unreliable and have been excluded from the analysis. Using the typical dose range for each drug cited by each respondent, a typical dose was calculated by taking the average of the high and low dose estimate.

Biological psychiatrists versus others: Examining the typical and maximum doses cited by biological and other psychiatrists in the two surveys revealed almost identical returns, and so the results of the two groups of respondents were combined.

Survey dose limits compared to doses equivalent to 1000mg chlorpromazine and BNF maximum doses: Table 1 shows the median typical and maximum doses of neuroleptics uses by psychiatrists in the two surveys, along with the 95% confidence limits for the medians, and the percentage of the responding psychiatrists who made an estimate for each drug and dose level.

There was a striking similarity between the stated doses used in 1985 and 1991, with no trends towards higher or lower typical or maximum doses over this period. For fluphenazine decanoate and haloperidol decanoate, the lower 95% confidence limit of the median maximum dose was above the BNF advisory maximum dose. For oral haloperidol, flupenthixol decanoate and fluspirilene, the upper 95% confidence limit of the maximum dose was below the BNF maximum. The lower confidence limits of the maximum doses from each survey exceeded the doses estimated to be equivalent to 1000 mg per day of chlorpromazine in the case of oral and depot haloperidol, but for all other drugs, this high dose equivalent was either within or above the confidence limits of the median maximum doses from the surveys.

Survey & BNF dose limits expressed as equivalent doses of chlorpromazine: In Table 2 the upper confidence limits of the typical and maximum doses from the surveys, and the BNF advisory maximum doses for each drug, are expressed as equivalent daily doses of chlorpromazine. The upper dose limits expressed in this way were less than the equivalent of 2000 mg per day of chlorpromazine for all oral drugs except haloperidol. In the case of depots, fluphenazine, haloperidol and pipothiazine were used in maximum doses equivalent to 2000 mg per day or more of chlorpromazine. BNF advisory maximum doses also exceeded the equivalent of 2000 mg per day of chlorpromazine in the case of oral haloperidol and depot fluphenthixol. Considering all the neuroleptic medications from the two surveys, maximum upper dose limits expressed as chlorpromazine equivalent dose averaged out at slightly less than the average BNF advisory maximum doses expressed in the same way (1800mg & 1700mg versus 1900mg, not statistically significant). The spread of these values was less for the psychiatrists' upper dose limits than for the BNF maximum doses.

DISCUSSION

The response rate in these surveys was relatively low at 15 & 22%, so the views expressed by the 147 respondents may not be representative of the profession as a whole. However this is a relatively large sample including psychiatrists working in many different fields and locations throughout Great Britain and Eire. No doubt many respondents in the surveys consulted the current BNF before stating their opinions about neuroleptic doses. However the questionnaire invited deviations from the BNF in order to discourage a mere representation of BNF recommendations, instead of psychiatrists' actual views and practices, and in this it appears to have been successful. Nevertheless because of the low response rate, we cannot exclude the possibility that it was only psychiatrists whose opinions were well considered who were prepared to make their views known.

In both surveys there were no significant differences between biological and other psychiatrists with regard to minimum, typical and maximum neuroleptic doses. It could be supposed that in 1985, before much research had emphasised the low benefit-to-risk ratio of "mega-dose"neuroleptic regimes, biological psychiatrists might be more adventurous with high doses than other psychiatrists. In contrast, by 1991, it was expected that biological psychiatrists would be more aware of this research and would be more conservative with doses than other psychiatrists. Clearly these expectations were not fulfilled.

With most drugs, psychiatrists appeared to be no less conservative than the BNF with regard to maximum doses of neuroleptics. The exceptions were fluphenazine decanoate and haloperidol decanoate for which the BNF advisory maximum doses were, and remain, quite low at the equivalent of less than 1000 mg per day of chlorpromazine. There was a similar trend for psychiatrists to use pipothiazine palmitate at relatively high maximum dose limits, equivalent to 2000 mg per day of chlorpromazine. For these three high potency depot drugs, the respondents' typical dose upper limits correspond more closely to the BNF maximum doses. The remaining depot drugs were used at maximum dose limits approximately equivalent to 1000mg per day of chlorpromazine or less.

In the case of flupenthixol decanoate, where the BNF maximum advisory dose is probably unjustifiably high at 400mg per week (equivalent to 2500mg chlorpromazine per day), psychiatrists in our surveys were decidedly more conservative than the BNF. This also appears to apply to the Maudsley psychiatrists surveyed by Mullen et al, (1994) whose average estimate for the dose of flupenthixol decanoate equivalent to 500mg per day of chlorpromazine was 30mg per week. This is comparable to approximately 40mg per week of flupenthixol cited by psychiatrists in our surveys as a typical antipsychotic dose, equivalent to about 500mg per day of chlorpromazine (Table 1).

Although the maximum doses of oral haloperidol cited by psychiatrists in our surveys are about half the then BNF maximum limit of 200mg per day, they are still equivalent to at least 5000mg per day of chlorpromazine (Table 2). The Psychopharmacology Subcommittee of the Royal College of Psychiatrists (Hirsch and Barnes, 1994) regards more than 80mg per day of haloperidol as 'very-high-dose neuroleptic medication', and its effectiveness in 'otherwise treatment-refractory patients to be an open issue'. Much the same can probably be said for all neuroleptics since there are no controlled studies to support the use of such high doses for an antipsychotic effect even in exceptional patients.

BNF 21, current at the time of our second survey (July 1991), suggests the dose of haloperidol should be 'initially 1.5-2.0mg daily in divided doses, gradually increased to 100mg (and occasionally 200mg) daily in severely disturbed patients.' By BNF 32 (September 1996), this wording has been amended to 'initially 1.5-3mg 2-3 times daily or 3-5mg 2-3 times daily in severely affected or resistant patients; in resistant schizophrenia up to 100mg (rarely 120mg) daily may be needed.' Whilst inviting more caution in prescribing higher doses, this change also actually confuses the issue further. In particular, the wording giving the indications for haloperidol (for 'schizophrenia and other psychoses, mania, short-term adjunctive management of psychomotor agitation, excitement, and violent or dangerously impulsive behaviour') continues to confuse the antipsychotic and major tranquillising actions of haloperidol which are unlikely to share identical pharmacological mechanisms or dose-response relationships. Furthermore, the suggestion from the antipsychotic dose equivalent table published in BNF 32 is that up to 100mg per day of haloperidol may be used 'in special psychiatric units where very high doses are required'. This seems confusing and unhelpful firstly because it is less than the 120mg per day maximum cited later in the same section of BNF 32 (200mg per day in earlier editions), and secondly because it makes no attempt to justify why 'very high doses' should be required at all.

Psychiatrists in our surveys tended to show greater consistency than the BNF in the maximum dose limits that they set for each drug, (Table 2). Since chlorpromazine equivalent doses have mostly been estimated from meta-analysis of controlled clinical trials (e.g.: Davis, 1976), psychiatrists' opinions appeared to be well informed by clinical experience, and in some respects better rationalised than the recommendations in the BNF. However, whilst the psychiatrists in our surveys were asked for their opinion on maximum dose limits of neuroleptics for an antipsychotic action, it would appear that this is more closely reflected by the typical doses cited, whereas the maximum dose limits given are more appropriate for controlling disturbed behaviour.

In the introduction to the section on antipsychotic drugs, BNF 32 (1996), as with earlier editions, makes no distinction between maximum doses based on considerations of toxicity and safety, and those considerably lower doses above which no further increase in antipsychotic efficacy can be expected. Whilst the BNF is dependent on manufacturers' data sheets for detailed guidance on doses of individual drugs, general guidelines which also draw attention to this distinction would be welcome. For example, it might be reasonable for the BNF to publish a table of relative clinical potency values and to suggest that maximum antipsychotic efficacy in the majority of patients can be achieved at doses equivalent to 1000 mg per day of chlorpromazine or less. For short-term control of disturbed behaviour, the advisory upper dose limit might be extended to the equivalent of 1500 mg per day of chlorpromazine for low potency neuroleptics or 2000 mg per day for high potency drugs. This would translate to a relatively modest maximum of 40 to 60 mg per day of haloperidol, although lower doses supplemented by other drugs such as benzodiazepines to help anxiety and behavioural disturbance might still be preferable.

Like the BNF, our questionnaire failed to remind psychiatrists that if more than one neuroleptic is prescribed at the same time (not recommended), usual maximum dose limits of individual drugs are no longer valid. In hindsight, we would have liked to ask prescribers' views on the maximum total neuroleptic dose in chlorpromazine equivalents. This will vary depending on the actual drugs prescribed, but should probably not exceed the equivalent of 2000 mg per day chlorpromazine under any circumstances.

It will be of interest to carry out a further, more representative survey of clinical opinion on antipsychotic drug dosage in order to check if the recent discussion of these issues lead by the Royal College of Psychiatrists (Thompson, 1994) has had a useful impact. Of course information about clinical opinion needs to be checked and validated against up to date surveys of actual clinical practice in this important area of psychopharmacology.

REFERENCES

Association of the British Pharmaceutical Industry (1996-97). ABPI Data Sheet Compendium, Haldol Decanoate (Janssen-Cilag Ltd.), p.451.

Baldessarini R.J., Cohen, B.M., & Teicher, M.H., (1988). Significance of neuroleptic dose & plasma level in the pharmacological treatment of psychoses. Archives of General Psychiatry 45, 79-91.

Davis, J.M. (1976). Comparative doses & costs of antipsychotic medication. Archives of General Psychiatry, 33, 858-861.

Gill, D. (1993). Audit of antipsychotic use in relation to BNF guidelines on dose, route & polypharmacy. Psychiatric Bulletin, 17, 773-774.

Hirsch, S. R., & Barnes, T.R.E. (1994). Clinical use of high-dose neuroleptics. British Journal of Psychiatry 164, 94-96.

Kane, J.M. (1994). The use of higher-dose antipsychotic medication. Comment on the Royal College of Psychiatrists' consensus statement. British Journal of Psychiatry 164, 431-432.

Mullen, R., Caan, A.W., and Smith, S. (1994). Perception of equivalent doses of neuroleptic drugs. Psychiatric Bulletin, 18, 335-337.

Ortiz, A., & Gershon, S. (1986). The future of neuroleptic psychopharmacology. Journal of Clinical Psychiatry 47, 5 (suppl.), 3-11.

Stanley, A. & Doyle, M. (1993). Audit of above BNF dosage medication. Psychiatric Bulletin, 17, 299-300.

Thompson, C. (1994). Consensus Statement. The use of high-dose antipsychotic medication. British Journal of Psychiatry 164, 448-458.

Torkington, B., Hogg, S., Powell, G., Main, C. & Barker, A. (1994). Antipsychotic medication use in relation to BNF guidelines. Psychiatric Bulletin, 18, 375-376.

 

Correspondence to:

Dr P. E. Harrison-Read

Consultant Psychiatrist

Park Royal Centre for Mental Health

Central Middlesex Hospital

Acton Lane, London NW10 7NS, U.K.

Tel: 0181 453 2710

Fax: 0181 961 6339

Acknowledgements: We thank Professor R.G. Priest for invaluable help and advice, and Lundbeck Limited for a grant towards the 1991 survey.

 

Tables 1a & 1b: 'Typical' and 'maximum' doses of neuroleptic drugs used by British psychiatrists for functional psychoses in 1985 (n = 61) and 1991 (n = 86). (Medians with 95% confidence limits and percentage response) in comparison with BNF advisory maximum doses and doses equivalent to 1000mg per day chlorpromazine.

'TYPICAL' DOSE

'MAXIMUM' DOSE

BNF MAXIMUM

'EQUIVALENT DOSE' (Eq.D)¹

Table 1a:

oral (mg/d)                        1985               1991                        1985                        1991            1985-91

chlorpromazine	500 425 - 550 97%	450 412 - 525 96%		1000 1000 - 1500 100%	1000 1000 - 1200 96%	1000	1000
haloperidol	32 23 - 42 98%	32 30 - 38 93%		100* 80 - 120 100%	80* 80 - 120 94%	200	20 (20-100)²
trifluoperazine	20 18 - 24 92%	22 18 - 25 89%		45 30 - 60 94%	50 40 - 60 88%	not given	40 (50)²
thioridazine	312 225 - 420 89%	312 225 - 375 80%		800 600 - 800 87%	600 500 - 800 80%	800	1000 (1000)²
pimozide	8 6 - 10 62%	8 7 - 10 65%		16 12 - 20 62%	20 12 - 20 65%	20	15 (20)²

 

Table 1b: depot (mg/wk)	1985	1991		1991	1991	1985-91
flupenthixol decanoate	42 30 - 55 92%	40 30 - 55 97%		100* 100 - 200 90%	100* 80 - 160 97%	400	160 (40)³
fluphenazine decanoate	31 19 - 53 90%	35 28 - 38 91%		100+ 100 - 150 93%	100+ 75 - 200 91%	50	100 (25)³
zuclopenthixol decanoate	200 150 - 250 57%	200 150 - 250 77%		600 40 - 600 59%	500 400 - 600 77%	600	1000
haloperidol decanoate	50 38 - 62 44%	62 44 - 75 67%		100+ 100 - 200 41%	100+ 100 - 200 66%	75	80 (50)³
pipothiazine palmitate	31 8 - 88 18%	50 36 - 75 27%		100 50 - 200 20%	100 50 - 200 24%	50	100
fluspirilene	6 4 - 10 21%	4 2 - 5 24%		12 6 - 20 18%	10* 4 - 10 23%	20	20

Footnote 1: Doses equivalent to 1000mg/d chlorpromazine are mainly derived from Davis (1976) and from Ortiz and Gershon (1986), based on efficacy data from controlled clinical trials. Depot doses per week are taken as approximately 3.5 times the daily oral equivalent (Davis, 1976). Chlorpromazine equivalents are uncertain for flupenthixol decanoate, pipothiazine palmitate and fluspirilene, but the potencies of these drugs are taken to relate to that of fluphenazine decanoate in the ratio of 0.625 : 1.0 : 6.25 to one.

Footnote 2: Recent (e.g. No.32, 1996) editions of the BNF suggest the equivalents given in brackets. For haloperidol, BNF 32 suggests that 'in specialist psychiatric units where very high doses are required, the (100mg per day chlorpromazine) equivalent dose of haloperidol might be up to 10mg' (i.e.: 100mg haloperidol is equivalent to 1000mg chlorpromazine).

Footnote 3: Equivalent doses suggested in ABPI (1996-97, p. 451).

* Significantly lower (P _ 0.05) than BNF advisory maximum dose.

+ Significantly higher (P _ 0.05) than BNF advisory maximum dose.

Table 2 Upper 95% confidence limits of 'typical' & 'maximum' neuroleptic doses from 1985 and 1991 surveys expressed as equivalent daily doses (mg) of chlorpromazine.

	'Typical Dose' upper limit 1985                   1991	'Maximum Dose' upper limit 1985                   1991	BNF Maximum   1985-91
oral (mg/d) chlorpromazine haloperidol trifluoperazine thioridazine pimozide	600                       500 2100                   1900 600                      600 400                      400 700                       700	1500                   1200 6000                   5000 1500                   1500 800                       800 1300                   1300	1000 10000 --- 800 1300
depot (mg/wk) flupenthixol decanoate fluphenazine decanoate zuclopenthixol decanoate haloperidol decanoate pipothiazine palmitate fluspirilene	300                        300 500                        400 200                        200 800                        900 900                        800 500                         200	1200                     1000 1500                     2000 600                         600 2500                     2500 2000                     2000 1000                     1000	2500 500 600 900 500 1000

 

 

Top of Page             back