# 2003 Taylor & Francis
International Journal of Psychiatry in Clinical Practice 2003
Volume 7
Pages 231 /238 231
Prescribing pattern and indicators for performance in a psychiatric practice SANGEETA SHARMA1, RK CHADDA2, RK RISHI3, RAJ K GULATI1 AND JS BAPNA1
AIMS: To assess the prescribing pattern and to measure some specific aspects of the behaviour of the prescribers (psychiatrists) before and after educational interventions using core drug use indicators.
1
METHODS:
Department of Neuropsychopharmacology; Psychiatry Institute of Human Behaviour and Allied Sciences, Delhi, India; 3Department of Pharmaceutical Sciences, H.N.B. Garhwal University, Srinagar, Uttaranchal, India 2
Correspondence Address Dr. Sangeeta Sharma, (Assistant Professor Neuropsychopharmacology), Institute of Human Behaviour & Allied Sciences, PO Box 9520, Jhilmil, Dilshad Garden, Delhi 110095, India Tel.: /91 11 25790183/22583058/22114032 Fax: /91 11 25791840/22599227 E-mail: [email protected]
In the present randomized retrospective controlled pre-post intervention prescription survey of schizophrenia and depression, 100 prescriptions each for schizophrenia and depression were obtained before and after each intervention. The prescriptions were analyzed for the following prescriber-specific indicators: number of drugs prescribed, prescribing by generic names, prescriptions for essential drugs, antiparkinsonian and benzodiazepines, nature of drugs and number of combinations prescribed. Based on the results of pre-intervention data, two interactional workshops were conducted 1 and 6 months after preintervention data collection. The first workshop focused on the results of the prescription audit feedback and reasons thereof. The second workshop focused, in addition, on appropriate management of schizophrenia and depression using consensus treatment guidelines with the aim of evolving a consensus on the treatment in a given hospital setting.
Before intervention, the essential drugs accounted for 80.95 and 96.91% of the total number of drugs prescribed in depression and schizophrenia, respectively. Prescription for essential drugs improved further significantly in the post intervention period to 95.26% (P B 0.04) for depression; whereas, in schizophrenia, prescriptions for essential drugs declined to 80.95%. The average number of drugs prescribed per encounter marginally declined in both schizophrenia (2.4690.94 to 2.3490.65) and depression (2.0990.79 to 2.0090.65) after the second intervention. The patients receiving two or more drugs from the same group together declined from 12 to 9% in schizophrenia, but increased from 5 to 10% in depression after intervention. Trihexyphenidyl, an antiparkinsonian drug, was co-prescribed (90%) with antipsychotic agents (98%) in schizophrenia; however, use of benzodiazepines declined significantly after intervention to 28% compared to 48% in the preintervention period. Also, benzodiazepine use was high (68%) and remained so (70%) after interventions in depression cases.
RESULTS:
/
/
/
/
/
The present study demonstrates excessive use of antiparkinsonian agents in schizophrenia and benzodiazepines in depression. Monitoring for the use of antiparkinsonian and benzodiazepines can form an important component for measuring specific aspects of prescriber’s behaviour, which can be used as an indicator for comparisons at different time intervals and between health facilities. (Int J Psych Clin Pract 2003; 7: 231 238) Keywords prescription pattern schizophrenia depression rational drug use psychiatric practice prescriber behaviour
CONCLUSION:
/
Received 4 March 2003; accepted for publication 13 May 2003
DOI: 10.1080/13651500310002616
232
S Sharma et al
INTRODUCTION
T
he aim of any drug management system is to deliver the correct drug to the patient who needs that medicine in doses that meet their own requirements, for an adequate period of time, and at the lowest cost to them and their community. To conform to these criteria, prescribers should follow a standard process of prescribing. On the contrary, inappropriate prescribing in many forms is common and constitutes a major health problem in both developing and developed countries. These include prescription for drugs when not needed, prescription with a wrong drug or ineffective or unsafe drug; prescription of the right drug but at an inadequate dose and for an inadequate period. Rational prescribing is of especially greater relevance in psychiatric practice where patients usually require treatment for a longer duration. Lack of insight and denial of illness in psychiatric patients compounds the problem of non-compliance. The prevalence of non-compliance with medication among discharged patients who have schizophrenia has been reported to be approximately 50% after 1 year and 75% at 2 years.1,2 The practice of polypharmacy further promotes non-compliance, especially when regimes for two drugs are different and involve multiple daily drugs.3 Concerns over excessive and irrational prescribing of psychotropic drugs have been raised for many years.4 11 Wide variations have been reported in the prescribing patterns among different countries and among different hospitals within the same country. The process of diagnosis and pharmacological treatment is complex. Techniques do not yet exist for adequately assessing the quality of this process in a standardized objective way. WHO has developed objective measures/indicators that can describe the drug use situation in three general areas related to the rational use of drugs, which include prescribing practices of health providers and factors specific to the patient care and health facility which support rational use. Such measures or indicators serve as reproducible and simple supervisory tools to detect problems in the performance by individual providers or health facilities. These indicators allow health planners, managers and researchers to make basic comparisons between situations at different times or in different areas. Though patient care- and facilityspecific indicators remain the same for all types of health facilities, prescriber-specific indicators vary depending on the type of practice. There are no well-defined reproducible prescriber-specific indicators pertaining to psychiatric practice. Before addressing the issue of rationality, it is necessary to evaluate the prescribing pattern of psychotropic drugs using core drug use indicators described by WHO.12 For a general health facility these include average number of drugs per encounter, percentage of drugs prescribed by generic name and from the essential drugs list, percentage of encounters with an injection and percentage of encounters with antibiotics. For a specialty setting like psychiatry, percent encounters with antibiotics would not be an
appropriate indicator to bring out any relevant information on drug use practices, whereas encounters with antidepressants, psychotropic drugs, antiparkinsonians and benzodiazepines would reflect on key issues in the treatment of schizophrenia and depression. However, the indicators described above are not intended to measure all dimensions of the appropriateness of pharmaceutical care. These indicators are best understood as a first-line measure, intended to stimulate further questioning and to guide subsequent action. The broad objectives of the present study were to assess the prescribing practices of the group of psychiatrists and identify some objective indicators to measure some specific aspects of the behaviour of the psychiatrists, which can be used as a measure of performance. The specific objectives were (i) to study the prescribing pattern using core drug use indicators12 and identify priority problem areas in a given setting, (ii) to mount an effective educational intervention programme to improve some aspects of drug use and (iii) to measure the impact of any intervention undertaken to improve the quality of care.
MATERIALS AND METHODS STUDY SETTING The study was undertaken in a tertiary care multidisciplinary postgraduate institute in Delhi (India), having disciplines of psychiatry, neurology, and allied sciences. The institute has a wide catchment area and it is estimated to serve a population of about 30 /40 million. At the time of the study, five consultants and 10/12 senior residents (having post graduate degrees in psychiatry) and 20 /24 junior residents were working in the Department of Psychiatry. As a policy, only consultants and senior residents who are qualified psychiatrists write the prescriptions. On an average day about 250 patients visit the outpatient clinic for consultation.
DATA COLLECTION An audit of prescribing against standards, followed by an educational intervention and then a re-audit after 1 and 6 months in the year 1999 and 2000, were carried out. The pre- and post-intervention data on prescriptions for patients suffering from schizophrenia and depression, attending the outpatient department, were collected from the medical record section by a systematic random allocation method uniformly spread over all days. Diagnosis recorded on the front sheet of the case files was considered. Last prescription from the file was considered for analysis. Due care was taken to minimize bias during selection of prescriptions and to represent all the prescribers in the health facilities during data collection by spreading prescriptions on different days. Case records with more than one psychiatric diagnosis or other co-morbid conditions and illegible prescriptions were not considered for this study.
Prescribing pattern and indicators for performance in a psychiatric practice
The prescribers were blind to the study methodology, as they were not aware of the intervention and the postintervention data collection. A total of 200 prescriptions were collected (100 each for schizophrenia and depression) by the systematic random method over a period of 2 months before intervention. Another two samples of 200 prescriptions (100 each for schizophrenia and depression) were collected after 1 month of the first intervention and then after the second intervention, spread over a period of 2 months using a similar methodology. For an inadequate performance level of 80% and adequate performance equal to 70%, 100 prescriptions each for schizophrenia and depression were studied to summarize values for facility-specific percentage indicators with a 95% confidence interval (CI) of 9/10%.14
STUDY TOOLS Prescriptions were examined for an average number of drugs per encounter, percentage of drugs prescribed by generic name and from the essential drugs list, percentage of encounters with an injection and percentage of encounters with antidepressants, psychotropic drugs, antiparkinsonians and benzodiazepines. The purpose and expected norms for optimum practice in a local health environment for all the above-mentioned indicators are shown in Table 1. Essential drugs list of Delhi State was used to define drugs as generics. The essential drugs list includes the following drugs: Trifluoperazine Chlorpromazine Haloperidol Thioridazine Fluphenazine decanoate
Imipramine Fluoxetine Amitriptyline *Trazodone
Lorazepam Diazepam Alprazolam Lithium
233
suitability and cost to make best use of limited resources. As part of the drug policy, only drugs on the EDL are procured for distribution in public health facilities in the state, and prescribers are expected to prescribe these drugs only. This is a dynamic list revised every 2 years. The second workshop focused on the appropriate management of schizophrenia and depression using available consensus treatment guidelines,13,14 in addition to prescription audit feedback after the first intervention. The aim was to evolve a consensus on the treatment of these conditions and develop departmental guidelines for the management of schizophrenia and depression. Misconceptions and misperceptions about precise management of schizophrenia and depression elicited in earlier discussions were addressed adequately during the second workshop. Written information in the form of consensus treatment guidelines was provided to all the prescribers.
DATA ANALYSIS The pre /post-intervention change in each of the study outcomes within the health facility was then computed, and Student’s t -test was used to test the difference between the average changes in the prescribing indicators.
RESULTS In the present retrospective pre /post-intervention study, a total of 300 prescriptions each for schizophrenia and depression were examined. The mean duration of treatment ranged from 18 to 29 months and the mean (9/SD) age of the patients was 35.539/12.29 years and 33.899/13.58 years in schizophrenia and depression, respectively.
*For selective use only
PRESCRIPTIONS INTERVENTION Based on the results of the pre-intervention data collection, two interactional workshops were conducted 1 and 6 months after pre-intervention data collection. These workshops consisted of interactive discussions involving 15 /20 prescribers each and included all prescribing doctors. A psychiatrist and a clinical pharmacologist who had been trained in appropriate management in how to lead a group discussion moderated the discussion. The participants were given an explanation of the purpose and implication of the indicators studied. The first workshop focused on the results of the prescription audit feedback and reasons thereof. The main objective of the discussion was to elicit the views of the participants on some of the key issues, like use of benzodiazepines and antiparkinson drugs in the initial phase and in the stabilization phase. The participants were also exposed to the essential drugs concept and the essential drugs list. The essential drugs list includes a limited list of drugs selected on the balanced criteria of efficacy, safety,
FOR
SCHIZOPHRENIA
Average (9/SD) number of drugs prescribed per encounter decreased from 2.469/0.94 to 2.349/0.65, though results were statistically insignificant. The maximum number of drugs prescribed per encounter decreased from 6 to 4 following intervention. However, mean number of doses per day remained same in the post-intervention (5.89/2.80 vs. 5.769/2.41) with a maximum of 14 tablets a day. Prescription by generic name improved significantly in the post-intervention period by 20% (P B/0.001). However, prescriptions for essential drugs declined significantly by 11% (P B/0.05) in the post-intervention prescription audit. The mainstay of the treatment was an antipsychotic agent as most patients (98%) received antipsychotics. The group of drugs prescribed and its extent is shown in Figure 1. The spectrum of preference for antipsychotic drugs remained similar after interventions. Haloperidol, risperidone, chlorpromazine and trifluoperazine were the most commonly prescribed antipsychotic drugs, and other drugs prescribed included carbamazepine and lithium. The number of patients
234
S Sharma et al
Table 1 The purpose and expected norms for optimum practice in a local health environment for the core drug use indicators Indicator
Purpose
*Expected norm
**Reflect specific aspect of behaviour
Average number of drugs per encounter
To measure the degree of polypharmacy
Preferred to be below 3
How secure are prescribers in their ability to diagnose and treat common illnesses? How strongly do prescribers feel that patient demand influences their practice?
Percentage of drugs prescribed by generic names
To measure the tendency to prescribe by generic name
As high as possible
Do prescribers know the correct generic names for most drugs? Does training of the prescribers affect their willingness to prescribe generically or to what extent the pharmaceutical representatives influence them? Are branded products being prescribed which are not available in the health facility
Percentage of drugs prescribed from essential drugs list (EDL)
Above 80% To measure the degree to which practices conform to a state drug policy, as indicated by prescribing from the EDL for the type of facility surveyed
What are the most common drugs being prescribed that are not on the EDL?
Is there an adequate supply of drugs on the EDL? Percentage of encounters with an To measure the overall level of use injection prescribed of important, but commonly overused and costly forms of drug therapy Percentage of encounters with antiparkinsonians agent in schizophrenia, benzodiazepines in affective disorder
Below 20%; high use may be justified for chronic non-complaint patients with schizophrenia
To measure the overall level of use Below 50% of important, but commonly coprescribed drugs
What are the beliefs and attitudes of patients and health providers about the relative efficacy of injections versus oral medications? Relative use of high potency and atypical antipsychotics?
What are the beliefs amongst prescribers about the role of these drugs in the acute phase and maintenance phase? Use of antiparkinsonians as prophylactic agent to prevent extrapyramidal reactions in schizophrenia? *There are no objective norms for either indicators. In a local health environment the above may be considered as the expected norm for optimum practice. **A level of performance on any of the indicators that would be considered unacceptable reflect some specific aspect of behaviour and should provoke specific kinds of follow up activities.
receiving two or more antipsychotic drugs together declined from 12 to 7% in the post-intervention period. An antiparkinsonian drug, trihexyphenidyl, was co-prescribed (90%) mostly with antipsychotics in the pre-intervention audit. After the first intervention, there was a decline in
concomitant use of antiparkinsonian agents from 90 to 78%. However, the pattern returned to near the pre-intervention level (94%) after the second intervention. A small proportion (12%) of the patients received concomitant antidepressant drugs, which declined to 7% in
Prescribing pattern and indicators for performance in a psychiatric practice
235
0.05). Prescribing by generic name also increased by 20%. Almost all prescriptions contained antidepressants, compared to 85% in the pre-intervention period, where 15% contained drugs other than antidepressants (Figure 2). Antipsychotics were co-prescribed in 16% of prescriptions in the pre-intervention audit, which showed a rise to 19% after the first intervention but declined significantly to 8% after the second intervention (P B/0.05). Accordingly, trihexyphenidyl, an antiparkinsonian drug, was co-prescribed with an antipsychotic in nearly 15% of prescriptions in the preand post-intervention audit. The spectrum of antidepressant drugs prescribed was similar after the intervention. Imipramine and fluoxetine were the two most commonly prescribed drugs. Other less commonly prescribed drugs included sertraline and clomipramine. In the pre-intervention period, 10% of the prescriptions contained two antidepressants and one contained three antidepressants, and this declined to 2% after first intervention but showed a rise to 5% in the second prescription audit. The percent benzodiazepine use (68%) was high and remained so even after the second intervention. The most commonly prescribed benzodiazepine was diazepam, followed by nitrazepam and lorazepam. Seven percent of prescriptions comprised drugs of other groups, which decreased after intervention.
Figure 1 Pattern of drugs prescribed in schizophrenia at baseline and 1 and 6 months after educational intervention. *PB/0.05.
the post-intervention period. Percent concomitant benzodiazepine use declined significantly from 48 to 28% after the second intervention (P B/0.05). The most commonly prescribed benzodiazepine was diazepam, followed by nitrazepam and lorazepam. Most of the patients were prescribed benzodiazepines in a once daily dose at bedtime.
DISCUSSION PRESCRIPTIONS
FOR
DEPRESSION
The results of the present study show that the more than one specific drug is being used for the same indication in patients with schizophrenia and depression. The high concomitant use of antiparkinsonian agents and benzodiazepines in schizophrenia and depression, respectively, was common, although from each class of psychotropic drugs only a small number of compounds were used. Perhaps, being a public hospital, prescribing is restricted to drugs on the essential
Average (9/SD) number of drugs prescribed decreased from 2.099/0.79 to 2.009/0.65 after the second intervention. The mean (9/SD) number of tablets prescribed per day decreased from 4.909/2.73 to 4.469/2.76. Also, the maximum number of tablets prescribed decreased to 12 compared to 16 in the pre-intervention period. (Table 2). Prescriptions for essential drugs increased significantly by 15% after intervention (P B/
Table 2 Prescribing indicators for the treatment of schizophrenia and depression before and 1 and 6 months after educational intervention Schizophrenia Preintervention (n /100)
Postintervention I (n /100)
Postintervention II (n /100)
Preintervention (n /100)
Postintervention I (n /100)
Postintervention II (n /100)
2.46 (0.94) (max 6; min 1) 5.8 (2.56) (max 14; min 1) 12
2.41 (0.62) (max 4; min 1) 5.72 (2.27) (max 13; min 1.5) 7
2.34 (0.65) (max 4; min 1) 5.76 (2.45) (max 14; min 1) 5
2.09 (0.79) (max 3; min 1) 4.9 (2.73) (max 16; min 1) 10
2.18 (0.64) (max 3; min 1) 4.46 (2.25) (max 12; min 1) 7
2.00 (0.66) (max 4; min 1) 4.46 (2.76) (max 13; min 1) 5
3
0
4
1
2
/
7 96.91 76.06
7 87.60 90.08*
4 85.770 96.24
/ 80.95 62.85
/ 80.18 80.18
2 95.260 82.93
Indicators Number of drugs prescribed (mean (SD)) Mean number of doses per day (mean (SD)) More than two drugs from same group together (%) Combination prescribed (%) Injection use (%) EDL (%) Generic (%) 0
P B/0.05; *P B/0.001.
Depression
236
S Sharma et al
Figure 2 Pattern of drugs prescribed in depression at baseline and 1 and 6 months after educational intervention. *P B/0.05.
drugs list and also because of non-availability of some of the new generation drugs in this country, the newer atypical antipsychotics and antidepressants were seldom used. Diazepam was very commonly prescribed as a minor tranquilizer in both schizophrenia and depression. Nearly 90% patients suffering from schizophrenia received an antiparkinsonian agent, and about 12% were prescribed two or more neuroleptics, with one patient receiving four neuroleptics. Combinations of antipsychotics were prescribed to 13 /24% patients in some earlier studies.15 19 However, our data do not distinguish between temporary co-prescription of two neuroleptics, which occurs during a cross taper between two agents and results in full switch to the new agent, and true polypharmacy. The former is an accepted practice for switching medications. The latter ‘co-prescription of two antipsychotics as a means of ongoing treatment’ is more controversial because the evidence base supporting its use is weak. As regards extent of polypharmacy, over 27% of prescriptions for depression and over 55% of schizophrenia had at least three different drugs, respectively. The frequency of polypharmacy is lower than that reported in earlier studies.20,21 Though it further declined following intervention in the present study, 21 and 40% of patients still continued to receive more than two drugs in depression and schizophrenia, respectively. Polypharmacy is generally not recommended because of increased risk of adverse reactions and the lack of evidence for therapeutic advantage from using several psychotropic drugs instead of a properly chosen single drug. In particular, the sedative and anxiolytic properties of some antidepressants and neuroleptics make it hard to justify the common concurrent administration of minor transquilisers and/or hypnotics. Prescriptions for essential drugs accounted for 80.95 and 96.91% of the total number of drugs prescribed in depres-
sion and schizophrenia, respectively, in the pre-intervention analysis. Prescriptions from essential drugs improved further significantly after intervention in depression. However, prescription for essential drugs declined in schizophrenia following intervention, since risperidone had been recently introduced in the country and, because of fewer extrapyramidal adverse effects, became the first-line treatment compared to high-potency antipsychotics. The essential drugs list (EDL) is revised every 2 years and, at the time of the study, risperidone was not considered as essential and, therefore, was not available at the hospital pharmacy. One of the criteria for inclusion of new drugs in the list, in addition to definite efficacy, safety and cost advantage over older drugs, is a minimum post-marketing experience of 3 years. This could explain the fall in percent prescriptions for essential drugs in the post-intervention period in schizophrenia. In the present study, 12% of prescriptions for schizophrenia and 15 /16% of prescriptions for depression also carried antidepressants and antipsychotics, respectively, for possible accompanying depressive or psychotic symptoms. This is an another area riddled with controversies, but clinical data collected in the present study are not sufficient to allow us to comment any further on such and other similar aspects of drug prescribing. Thus, the specificity of drug choice cannot be discussed, although the prescription of antipsychotics in 15% of patients with depression seems to be acceptable. Percent encounters with antiparkinsonians and benzodiazepines were chosen as an indicator of performance since these are often used in a substantial proportion of patients, with the attitude that their use is insignificant.22 Use of such drugs can lead to a patient staying on another medication because of the failure to identify the critical therapeutic combination. In the present study, as well as earlier studies from India, antiparkinsonian agents were prescribed in almost all patients receiving neuroleptics.10,11 Regardless of the indication, adjunctive drugs used in prevention or treatment of side effects were co-prescribed in 47 /65% cases.19 23 25 Non-prescription of these drugs with antipsychotics in new patients in an out-patient setting has been reported to lead to non-compliance, but the available guidelines do not recommend prophylactic use of these agents at the time of commencing treatment or during maintenance therapy once the patient has stabilized.13,14 With the availability of atypical antipsychotics in recent years, prophylactic use of antiparkinsonian agents may not be required because of the lower potential for these drugs to cause extrapyramidal side effects.13,26 However, clinical experience in the Indian population shows a similar frequency of extrapyramidal side effects with risperidone compared to the conventional high-potency antipsychotics. In our setting, the prophylactic use of antiparkinsonian medication may be considered because of the high rate of extrapyramidal side effects among patients receiving antipsychotic medications, especially with the high-potency antipsychotic agents, and with risperidone being the most frequently prescribed
Prescribing pattern and indicators for performance in a psychiatric practice
antipsychotic in schizophrenia. However, the need for antiparkinsonian agents in the present study is not justified, particularly when the acute phase of treatment was over, since most prescriptions were for chronic patients. This question is of great clinical importance since, in a proportion of patients, antiparkinsonians can be safely withdrawn after a few months of treatment.27 There is a risk that some patients may remain unnecessarily on these medications,28 and there is evidence that these drugs promote development of tardive dyskinesia.29 Moreover, these drugs may be abused for their euphoriant effects.29 Similarly, high use of benzodiazepines as a co-prescription drug in affective disorders was reported in the present and other studies.23 Use of benzodiazepines along with antidepressants may be more justified in patients presenting with severe anxiety and agitation; however, use of these agents, particularly after the acute phase is over, needs attention. The next questionable pattern concerns prescription of drugs to be taken three or more times a day and the number of tablets taken per day, as was the case with nearly half of the oral neuroleptics.24 The pharmacokinetics of most of these drugs makes it possible, except in the initial phase of treatment, to administer the entire daily amount at bed time, or at most in two doses. Less frequent administration could improve compliance by reducing the number of tablets taken per day and making treatment more convenient. An educational intervention in the form of small group discussion was chosen to promote co-operation and consensus and disseminate key behavioural messages; these were addressed in a highly interactive discussion rather than in conventional didactic lectures. Moreover, interactive small group discussions were shown to be effective in bringing about behavioural change in some of the prescribing practices, but did not demonstrate significant impact in improving the appropriate use of antiparkinsonian agents and benzodiazepine use. Use of multiple drug therapy appears to be broad based, which is often driven by clinical need rather than evidence-based data. As a result, clinicians tend to rely on their clinical experience in such cases. Secondly, clinicians’ access to information about advances in drug therapies and pharmaceutical industry-sponsored information also may play a substantial role in physicians’ prescribing choices. Some of the factors associated with physicians’ prescribing behaviour elicited during discussions were fear of extra-pyramidal adverse effects and loss of follow-up. In the out-patient department, whenever a patient develops extra-pyramidal adverse effects, the chances of future non-compliance increases and such patients usually do not report back. Due to a general lack of awareness and the stigma associated with mental illnesses, a visit to a psychiatrist is often feared, and adverse effects further increase the chances of not seeking any medical help from qualified psychiatrists. In India, the stigma associated with psychiatric illnesses arises from a number of false beliefs, viz. mental illnesses are incurable, the mentally ill are aggressive, impulsive and unpredictable. Drugs used for mental illnesses further increase stigma because of troublesome adverse
237
effects. Patients on antipsychotics drugs may often feel isolated because of marked extra-pyramidal adverse effects like stooping gait, masked facies and poverty of movements.30 Other reasons suggested included belief among patients that the more drugs prescribed, the better the effect. These misperceptions can perhaps be targeted by periodic reinforcement so that prescribers are persuaded by strong behavioural messages from their peers about a relevant issue. It is important to mention here that objective norms do not exist for either indicator to make a comparison; however, if in a given setting an average number of drugs prescribed per prescription is four, with 90% patients receiving antiparkinsonians and 70% receiving benzodiazepines in schizophrenia and depression during maintenance phase, respectively, as shown in the present study and this practice may be considered inadequate. The health manager carrying out the study may feel that, in a local health environment, the expected norm should be two or three drugs per prescription, with 50% patients needing antiparkinsonians and benzodiazepines. In such a setting, priority should be given to finding out why antiparkinsonians and benzodiazepines are prescribed and to reduce their use if reason for such high use is inappropriate. Interpretation and comparison of findings from different countries is limited by differences in definitions used, methodologies and the availability of drugs, besides patient characteristics. The indicators used in the present study are simple, reproducible, less expensive and provide useful data to inform quality of care activities at the health facility compared to prospective descriptive prescription reviews. This study also suffers from the limitation that, being a crosssectional study, though it can provide information about point prevalence of polypharmacy and co-prescription of adjunctive drugs, it does not distinguish among prescriptions with cross tapers in progress or ongoing co-prescription of two or more drugs from the same therapeutic group and response to therapy. Determining the exact rate of true coprescription and tracking the course of the same would require much more extensive prescription reviews. Also, this type of analysis does not reveal practitioners’ rationale for particular medication choices, whether or not patients’ actually take the medication as prescribed, or patients’ clinical response to the medication.
CONCLUSIONS The results of the present study, as well as other similar studies, have shown that such audits have proved helpful in improving the performance and prescribing habits from as early as 1976.31 Monitoring the use of antiparkinsonian agents and benzodiazepines using drug use indicators described in the present study can form an important component for measuring specific aspects of prescriber’s behaviour. and can be used as an indicator for comparison at different time intervals and between health facilities in a
238
S Sharma et al
reproducible manner. These indicators can also serve as simple and reproducible supervisory tools to detect problems in performance by individual providers or health facilities. Once the broad outlines of drug use practices and behaviour are known, these indicators can be used to identify facilities whose performance falls below a specific standard of quality, so that they can be targeted for intensive supervision.
ACKNOWLEDGEMENTS We gratefully acknowledge the technical support from the Delhi Society for Promotion of Rational Drug Use under the INDIA /WHO Essential Drugs Programme.
REFERENCES 1.
2. 3. 4.
5. 6.
7. 8.
9. 10.
11.
12.
13.
14. 15.
16.
Weiden P, Rapkin B, Mott T et al (1994) Rating of medication influences (ROMI) scale in schizophrenia. Schizophr Bull 20: 297 / 310. Weiden PJ, Oljson M (1995) Cost of relapse in schizophrenia. Schizophr Bull 21: 419 /29. Cramer JA, Mattson RH (1991) Monitoring Compliance in Medical Practice and Clinical Trials . New York: Raven Press: 123 /38. Hemminki E (1988) In: Ghodse, Khan (eds.) Factors Influencing Prescribing in Psychoactive Drugs: Improving Prescribing Practices . Geneva: World Health Organization: 22 /35. Morgan R, Gopalaswamy AK (1984) Psychotropic drugs: another survey of prescribing patterns. Br J Psychiatry 144: 298 /302. Edwards S, Kumar V (1984) A survey of prescribing of psychotropic drugs in a Birmingham psychiatric hospital. Br J Psychiatry 145: 502 /7. Muizen M, Silverstone T (1987) A comparative hospital survey of psychotropic drug prescribing. Br J Psychiatry 150: 501 /4. Clark AF, Holden NL (1987) A survey and follow-up of prescribing to chronic hospital in-patients. Br J Psychiatry 150: 88 /91. Holloway F (1988) Prescribing for the long-term mentally ill: A study of treatment practices. Br J Psychiatry 152: 511 /5. Amarjeet, Chadda RK, Bapna JS (1997) Pattern of use of psychotropic drugs in an inpatient unit in a mental hospital. J Ment Health Hum Behav 1: 29 /34. Aggarwal AK, Natu MV, Kalra R, Deswal RS (2001) Prescribing patterns of psychotropic drugs in a tertiary level hospital. Indian J Pharmacol 33: 243 /4. How to Investigate Drug Use in Health Facilities. Action Programme on Essential Drugs. World Health Organization. WHO/DAP/93.1. Expert Consensus Working Group (1999) Expert Consensus Guideline Series: Treatment of Schizophrenia 1999. J Clin Psychiatry 60 (Suppl II): 1 /80. Practice guidelines for the treatment of patients with major depressive disorder (2000) Am J Psychiatry (Suppl) 157: 1 /45. Keks NA, Altson K, Hope J, Krapivensky N, Culhane C, Tanaghow A, Doherty P, Bootle A (1999) Use of antipsychosis and adjunctive medications by an inner urban community psychiatric service. Aust N Z J Psychiatry 33: 896 /901. Clark RE, Bartels SJ, Mellman TA, Peacock WJ (2002) Recent trends in antipsychotic combination therapy of schizophrenia and schizoaffective disorder: implication for state mental health policy. Schizophr Bull 28: 75 /84.
17.
18.
19.
20. 21.
22. 23.
24.
25.
26.
27.
28. 29.
30. 31.
Weissman EM (2002) Antipsychotic prescribing practices in the veterans health care administration / New York metropolitan region. Schizophr Bull 28: 31 /42. Procyshyn RM, Kennedy NB, Tse G, Thompson B (2001) Antipsychotic polypharmacy: a survey of discharge prescriptions from a tertiary care psychiatric institution. Can J Psychiatry 46: 334 /49. Chong SA, Sachdev P, Mahendran R, Chua HC (2000) Neuroleptic and anticholinergic drug use in chinese patients with schizophrenia resident in a state psychiatric hospital in Singapore. Aust N Z J Psychiatry 34: 988 /91. Nunley M (1996) Why Indian psychiatrists prescribe so many drugs. Culture Med Psychiatry 20: 165 /97. Konard M, Kolakowska T (1981) A survey of prescribing psychotropic drugs in two psychiatric hospitals. Br J Psychiatry 138: 217 /21. Dufresne RL (1996) Issues in polypharmacotherapy focus on major depression. Psychopharm Bull 32: 547 /53. Fourrier A, Gasquet I, Allicar MP, Bouhssira M, Lepine JP, Begaud B (2000) Patterns of neuroleptic drug prescription a national cross-sectional survey of a random sample of French psychiatrists. Br J Clin Pharmacol 49: 80 /6. Ungvari GS, Pang AH, Chiu HF, Wong CK, Lum FC (1996) Psychotropic drug prescription in rehabilitation. A survey in Hong Kong. Soc Psychiatr Psychiatr Epidemiol 31: 288 /91. Chiu HF, Shum PS, Lam CW (1991) Psychotropic drug prescribing to chronic schizophrenics in a Hong Kong Hospital. Int J Soc Psychiatry 37: 187 /91. Luo RD, Belleti DA, Tran D, Arcona S, Salen PN (2002) National prescribing patterns in the management of extra pyramidal symptoms among patients with schizophrenia. Int J Psychiatr Med 32: 261 /9. Hsin-Tung, Simpson GM (2000) Medication induced movement disorders. In: Sadock BJ, Sadock VA (eds.) Comprehensive Textbook of Psychiatry, VII Edn. 2265 /71. Putten GP, Best NR, McGiure J (1984) A anticholinergic drug abuse. A common problem? Br Med J 289: 612 /4. Meyer JM, Simpson GM. Anticholinergics and amantadine. In: Sadock BJ, Sadock VA (eds.) Comprehensive Textbook of Psychiatry, VII Edn. 2276 /82. Wig NN (1997) Stigma against mental illness editorial. Indian J Psychiatry 39: 187 /9. Diamond H, Tislow R, Snyder T, Rickels K (1976) Peer review of prescribing patterns in a CMHC. Am J Psychiatry 133: 697 /9.
International Journal of Psychiatry in Clinical Practice 2003
Volume 7
Pages 231 /238 231
Prescribing pattern and indicators for performance in a psychiatric practice SANGEETA SHARMA1, RK CHADDA2, RK RISHI3, RAJ K GULATI1 AND JS BAPNA1
AIMS: To assess the prescribing pattern and to measure some specific aspects of the behaviour of the prescribers (psychiatrists) before and after educational interventions using core drug use indicators.
1
METHODS:
Department of Neuropsychopharmacology; Psychiatry Institute of Human Behaviour and Allied Sciences, Delhi, India; 3Department of Pharmaceutical Sciences, H.N.B. Garhwal University, Srinagar, Uttaranchal, India 2
Correspondence Address Dr. Sangeeta Sharma, (Assistant Professor Neuropsychopharmacology), Institute of Human Behaviour & Allied Sciences, PO Box 9520, Jhilmil, Dilshad Garden, Delhi 110095, India Tel.: /91 11 25790183/22583058/22114032 Fax: /91 11 25791840/22599227 E-mail: [email protected]
In the present randomized retrospective controlled pre-post intervention prescription survey of schizophrenia and depression, 100 prescriptions each for schizophrenia and depression were obtained before and after each intervention. The prescriptions were analyzed for the following prescriber-specific indicators: number of drugs prescribed, prescribing by generic names, prescriptions for essential drugs, antiparkinsonian and benzodiazepines, nature of drugs and number of combinations prescribed. Based on the results of pre-intervention data, two interactional workshops were conducted 1 and 6 months after preintervention data collection. The first workshop focused on the results of the prescription audit feedback and reasons thereof. The second workshop focused, in addition, on appropriate management of schizophrenia and depression using consensus treatment guidelines with the aim of evolving a consensus on the treatment in a given hospital setting.
Before intervention, the essential drugs accounted for 80.95 and 96.91% of the total number of drugs prescribed in depression and schizophrenia, respectively. Prescription for essential drugs improved further significantly in the post intervention period to 95.26% (P B 0.04) for depression; whereas, in schizophrenia, prescriptions for essential drugs declined to 80.95%. The average number of drugs prescribed per encounter marginally declined in both schizophrenia (2.4690.94 to 2.3490.65) and depression (2.0990.79 to 2.0090.65) after the second intervention. The patients receiving two or more drugs from the same group together declined from 12 to 9% in schizophrenia, but increased from 5 to 10% in depression after intervention. Trihexyphenidyl, an antiparkinsonian drug, was co-prescribed (90%) with antipsychotic agents (98%) in schizophrenia; however, use of benzodiazepines declined significantly after intervention to 28% compared to 48% in the preintervention period. Also, benzodiazepine use was high (68%) and remained so (70%) after interventions in depression cases.
RESULTS:
/
/
/
/
/
The present study demonstrates excessive use of antiparkinsonian agents in schizophrenia and benzodiazepines in depression. Monitoring for the use of antiparkinsonian and benzodiazepines can form an important component for measuring specific aspects of prescriber’s behaviour, which can be used as an indicator for comparisons at different time intervals and between health facilities. (Int J Psych Clin Pract 2003; 7: 231 238) Keywords prescription pattern schizophrenia depression rational drug use psychiatric practice prescriber behaviour
CONCLUSION:
/
Received 4 March 2003; accepted for publication 13 May 2003
DOI: 10.1080/13651500310002616
232
S Sharma et al
INTRODUCTION
T
he aim of any drug management system is to deliver the correct drug to the patient who needs that medicine in doses that meet their own requirements, for an adequate period of time, and at the lowest cost to them and their community. To conform to these criteria, prescribers should follow a standard process of prescribing. On the contrary, inappropriate prescribing in many forms is common and constitutes a major health problem in both developing and developed countries. These include prescription for drugs when not needed, prescription with a wrong drug or ineffective or unsafe drug; prescription of the right drug but at an inadequate dose and for an inadequate period. Rational prescribing is of especially greater relevance in psychiatric practice where patients usually require treatment for a longer duration. Lack of insight and denial of illness in psychiatric patients compounds the problem of non-compliance. The prevalence of non-compliance with medication among discharged patients who have schizophrenia has been reported to be approximately 50% after 1 year and 75% at 2 years.1,2 The practice of polypharmacy further promotes non-compliance, especially when regimes for two drugs are different and involve multiple daily drugs.3 Concerns over excessive and irrational prescribing of psychotropic drugs have been raised for many years.4 11 Wide variations have been reported in the prescribing patterns among different countries and among different hospitals within the same country. The process of diagnosis and pharmacological treatment is complex. Techniques do not yet exist for adequately assessing the quality of this process in a standardized objective way. WHO has developed objective measures/indicators that can describe the drug use situation in three general areas related to the rational use of drugs, which include prescribing practices of health providers and factors specific to the patient care and health facility which support rational use. Such measures or indicators serve as reproducible and simple supervisory tools to detect problems in the performance by individual providers or health facilities. These indicators allow health planners, managers and researchers to make basic comparisons between situations at different times or in different areas. Though patient care- and facilityspecific indicators remain the same for all types of health facilities, prescriber-specific indicators vary depending on the type of practice. There are no well-defined reproducible prescriber-specific indicators pertaining to psychiatric practice. Before addressing the issue of rationality, it is necessary to evaluate the prescribing pattern of psychotropic drugs using core drug use indicators described by WHO.12 For a general health facility these include average number of drugs per encounter, percentage of drugs prescribed by generic name and from the essential drugs list, percentage of encounters with an injection and percentage of encounters with antibiotics. For a specialty setting like psychiatry, percent encounters with antibiotics would not be an
appropriate indicator to bring out any relevant information on drug use practices, whereas encounters with antidepressants, psychotropic drugs, antiparkinsonians and benzodiazepines would reflect on key issues in the treatment of schizophrenia and depression. However, the indicators described above are not intended to measure all dimensions of the appropriateness of pharmaceutical care. These indicators are best understood as a first-line measure, intended to stimulate further questioning and to guide subsequent action. The broad objectives of the present study were to assess the prescribing practices of the group of psychiatrists and identify some objective indicators to measure some specific aspects of the behaviour of the psychiatrists, which can be used as a measure of performance. The specific objectives were (i) to study the prescribing pattern using core drug use indicators12 and identify priority problem areas in a given setting, (ii) to mount an effective educational intervention programme to improve some aspects of drug use and (iii) to measure the impact of any intervention undertaken to improve the quality of care.
MATERIALS AND METHODS STUDY SETTING The study was undertaken in a tertiary care multidisciplinary postgraduate institute in Delhi (India), having disciplines of psychiatry, neurology, and allied sciences. The institute has a wide catchment area and it is estimated to serve a population of about 30 /40 million. At the time of the study, five consultants and 10/12 senior residents (having post graduate degrees in psychiatry) and 20 /24 junior residents were working in the Department of Psychiatry. As a policy, only consultants and senior residents who are qualified psychiatrists write the prescriptions. On an average day about 250 patients visit the outpatient clinic for consultation.
DATA COLLECTION An audit of prescribing against standards, followed by an educational intervention and then a re-audit after 1 and 6 months in the year 1999 and 2000, were carried out. The pre- and post-intervention data on prescriptions for patients suffering from schizophrenia and depression, attending the outpatient department, were collected from the medical record section by a systematic random allocation method uniformly spread over all days. Diagnosis recorded on the front sheet of the case files was considered. Last prescription from the file was considered for analysis. Due care was taken to minimize bias during selection of prescriptions and to represent all the prescribers in the health facilities during data collection by spreading prescriptions on different days. Case records with more than one psychiatric diagnosis or other co-morbid conditions and illegible prescriptions were not considered for this study.
Prescribing pattern and indicators for performance in a psychiatric practice
The prescribers were blind to the study methodology, as they were not aware of the intervention and the postintervention data collection. A total of 200 prescriptions were collected (100 each for schizophrenia and depression) by the systematic random method over a period of 2 months before intervention. Another two samples of 200 prescriptions (100 each for schizophrenia and depression) were collected after 1 month of the first intervention and then after the second intervention, spread over a period of 2 months using a similar methodology. For an inadequate performance level of 80% and adequate performance equal to 70%, 100 prescriptions each for schizophrenia and depression were studied to summarize values for facility-specific percentage indicators with a 95% confidence interval (CI) of 9/10%.14
STUDY TOOLS Prescriptions were examined for an average number of drugs per encounter, percentage of drugs prescribed by generic name and from the essential drugs list, percentage of encounters with an injection and percentage of encounters with antidepressants, psychotropic drugs, antiparkinsonians and benzodiazepines. The purpose and expected norms for optimum practice in a local health environment for all the above-mentioned indicators are shown in Table 1. Essential drugs list of Delhi State was used to define drugs as generics. The essential drugs list includes the following drugs: Trifluoperazine Chlorpromazine Haloperidol Thioridazine Fluphenazine decanoate
Imipramine Fluoxetine Amitriptyline *Trazodone
Lorazepam Diazepam Alprazolam Lithium
233
suitability and cost to make best use of limited resources. As part of the drug policy, only drugs on the EDL are procured for distribution in public health facilities in the state, and prescribers are expected to prescribe these drugs only. This is a dynamic list revised every 2 years. The second workshop focused on the appropriate management of schizophrenia and depression using available consensus treatment guidelines,13,14 in addition to prescription audit feedback after the first intervention. The aim was to evolve a consensus on the treatment of these conditions and develop departmental guidelines for the management of schizophrenia and depression. Misconceptions and misperceptions about precise management of schizophrenia and depression elicited in earlier discussions were addressed adequately during the second workshop. Written information in the form of consensus treatment guidelines was provided to all the prescribers.
DATA ANALYSIS The pre /post-intervention change in each of the study outcomes within the health facility was then computed, and Student’s t -test was used to test the difference between the average changes in the prescribing indicators.
RESULTS In the present retrospective pre /post-intervention study, a total of 300 prescriptions each for schizophrenia and depression were examined. The mean duration of treatment ranged from 18 to 29 months and the mean (9/SD) age of the patients was 35.539/12.29 years and 33.899/13.58 years in schizophrenia and depression, respectively.
*For selective use only
PRESCRIPTIONS INTERVENTION Based on the results of the pre-intervention data collection, two interactional workshops were conducted 1 and 6 months after pre-intervention data collection. These workshops consisted of interactive discussions involving 15 /20 prescribers each and included all prescribing doctors. A psychiatrist and a clinical pharmacologist who had been trained in appropriate management in how to lead a group discussion moderated the discussion. The participants were given an explanation of the purpose and implication of the indicators studied. The first workshop focused on the results of the prescription audit feedback and reasons thereof. The main objective of the discussion was to elicit the views of the participants on some of the key issues, like use of benzodiazepines and antiparkinson drugs in the initial phase and in the stabilization phase. The participants were also exposed to the essential drugs concept and the essential drugs list. The essential drugs list includes a limited list of drugs selected on the balanced criteria of efficacy, safety,
FOR
SCHIZOPHRENIA
Average (9/SD) number of drugs prescribed per encounter decreased from 2.469/0.94 to 2.349/0.65, though results were statistically insignificant. The maximum number of drugs prescribed per encounter decreased from 6 to 4 following intervention. However, mean number of doses per day remained same in the post-intervention (5.89/2.80 vs. 5.769/2.41) with a maximum of 14 tablets a day. Prescription by generic name improved significantly in the post-intervention period by 20% (P B/0.001). However, prescriptions for essential drugs declined significantly by 11% (P B/0.05) in the post-intervention prescription audit. The mainstay of the treatment was an antipsychotic agent as most patients (98%) received antipsychotics. The group of drugs prescribed and its extent is shown in Figure 1. The spectrum of preference for antipsychotic drugs remained similar after interventions. Haloperidol, risperidone, chlorpromazine and trifluoperazine were the most commonly prescribed antipsychotic drugs, and other drugs prescribed included carbamazepine and lithium. The number of patients
234
S Sharma et al
Table 1 The purpose and expected norms for optimum practice in a local health environment for the core drug use indicators Indicator
Purpose
*Expected norm
**Reflect specific aspect of behaviour
Average number of drugs per encounter
To measure the degree of polypharmacy
Preferred to be below 3
How secure are prescribers in their ability to diagnose and treat common illnesses? How strongly do prescribers feel that patient demand influences their practice?
Percentage of drugs prescribed by generic names
To measure the tendency to prescribe by generic name
As high as possible
Do prescribers know the correct generic names for most drugs? Does training of the prescribers affect their willingness to prescribe generically or to what extent the pharmaceutical representatives influence them? Are branded products being prescribed which are not available in the health facility
Percentage of drugs prescribed from essential drugs list (EDL)
Above 80% To measure the degree to which practices conform to a state drug policy, as indicated by prescribing from the EDL for the type of facility surveyed
What are the most common drugs being prescribed that are not on the EDL?
Is there an adequate supply of drugs on the EDL? Percentage of encounters with an To measure the overall level of use injection prescribed of important, but commonly overused and costly forms of drug therapy Percentage of encounters with antiparkinsonians agent in schizophrenia, benzodiazepines in affective disorder
Below 20%; high use may be justified for chronic non-complaint patients with schizophrenia
To measure the overall level of use Below 50% of important, but commonly coprescribed drugs
What are the beliefs and attitudes of patients and health providers about the relative efficacy of injections versus oral medications? Relative use of high potency and atypical antipsychotics?
What are the beliefs amongst prescribers about the role of these drugs in the acute phase and maintenance phase? Use of antiparkinsonians as prophylactic agent to prevent extrapyramidal reactions in schizophrenia? *There are no objective norms for either indicators. In a local health environment the above may be considered as the expected norm for optimum practice. **A level of performance on any of the indicators that would be considered unacceptable reflect some specific aspect of behaviour and should provoke specific kinds of follow up activities.
receiving two or more antipsychotic drugs together declined from 12 to 7% in the post-intervention period. An antiparkinsonian drug, trihexyphenidyl, was co-prescribed (90%) mostly with antipsychotics in the pre-intervention audit. After the first intervention, there was a decline in
concomitant use of antiparkinsonian agents from 90 to 78%. However, the pattern returned to near the pre-intervention level (94%) after the second intervention. A small proportion (12%) of the patients received concomitant antidepressant drugs, which declined to 7% in
Prescribing pattern and indicators for performance in a psychiatric practice
235
0.05). Prescribing by generic name also increased by 20%. Almost all prescriptions contained antidepressants, compared to 85% in the pre-intervention period, where 15% contained drugs other than antidepressants (Figure 2). Antipsychotics were co-prescribed in 16% of prescriptions in the pre-intervention audit, which showed a rise to 19% after the first intervention but declined significantly to 8% after the second intervention (P B/0.05). Accordingly, trihexyphenidyl, an antiparkinsonian drug, was co-prescribed with an antipsychotic in nearly 15% of prescriptions in the preand post-intervention audit. The spectrum of antidepressant drugs prescribed was similar after the intervention. Imipramine and fluoxetine were the two most commonly prescribed drugs. Other less commonly prescribed drugs included sertraline and clomipramine. In the pre-intervention period, 10% of the prescriptions contained two antidepressants and one contained three antidepressants, and this declined to 2% after first intervention but showed a rise to 5% in the second prescription audit. The percent benzodiazepine use (68%) was high and remained so even after the second intervention. The most commonly prescribed benzodiazepine was diazepam, followed by nitrazepam and lorazepam. Seven percent of prescriptions comprised drugs of other groups, which decreased after intervention.
Figure 1 Pattern of drugs prescribed in schizophrenia at baseline and 1 and 6 months after educational intervention. *PB/0.05.
the post-intervention period. Percent concomitant benzodiazepine use declined significantly from 48 to 28% after the second intervention (P B/0.05). The most commonly prescribed benzodiazepine was diazepam, followed by nitrazepam and lorazepam. Most of the patients were prescribed benzodiazepines in a once daily dose at bedtime.
DISCUSSION PRESCRIPTIONS
FOR
DEPRESSION
The results of the present study show that the more than one specific drug is being used for the same indication in patients with schizophrenia and depression. The high concomitant use of antiparkinsonian agents and benzodiazepines in schizophrenia and depression, respectively, was common, although from each class of psychotropic drugs only a small number of compounds were used. Perhaps, being a public hospital, prescribing is restricted to drugs on the essential
Average (9/SD) number of drugs prescribed decreased from 2.099/0.79 to 2.009/0.65 after the second intervention. The mean (9/SD) number of tablets prescribed per day decreased from 4.909/2.73 to 4.469/2.76. Also, the maximum number of tablets prescribed decreased to 12 compared to 16 in the pre-intervention period. (Table 2). Prescriptions for essential drugs increased significantly by 15% after intervention (P B/
Table 2 Prescribing indicators for the treatment of schizophrenia and depression before and 1 and 6 months after educational intervention Schizophrenia Preintervention (n /100)
Postintervention I (n /100)
Postintervention II (n /100)
Preintervention (n /100)
Postintervention I (n /100)
Postintervention II (n /100)
2.46 (0.94) (max 6; min 1) 5.8 (2.56) (max 14; min 1) 12
2.41 (0.62) (max 4; min 1) 5.72 (2.27) (max 13; min 1.5) 7
2.34 (0.65) (max 4; min 1) 5.76 (2.45) (max 14; min 1) 5
2.09 (0.79) (max 3; min 1) 4.9 (2.73) (max 16; min 1) 10
2.18 (0.64) (max 3; min 1) 4.46 (2.25) (max 12; min 1) 7
2.00 (0.66) (max 4; min 1) 4.46 (2.76) (max 13; min 1) 5
3
0
4
1
2
/
7 96.91 76.06
7 87.60 90.08*
4 85.770 96.24
/ 80.95 62.85
/ 80.18 80.18
2 95.260 82.93
Indicators Number of drugs prescribed (mean (SD)) Mean number of doses per day (mean (SD)) More than two drugs from same group together (%) Combination prescribed (%) Injection use (%) EDL (%) Generic (%) 0
P B/0.05; *P B/0.001.
Depression
236
S Sharma et al
Figure 2 Pattern of drugs prescribed in depression at baseline and 1 and 6 months after educational intervention. *P B/0.05.
drugs list and also because of non-availability of some of the new generation drugs in this country, the newer atypical antipsychotics and antidepressants were seldom used. Diazepam was very commonly prescribed as a minor tranquilizer in both schizophrenia and depression. Nearly 90% patients suffering from schizophrenia received an antiparkinsonian agent, and about 12% were prescribed two or more neuroleptics, with one patient receiving four neuroleptics. Combinations of antipsychotics were prescribed to 13 /24% patients in some earlier studies.15 19 However, our data do not distinguish between temporary co-prescription of two neuroleptics, which occurs during a cross taper between two agents and results in full switch to the new agent, and true polypharmacy. The former is an accepted practice for switching medications. The latter ‘co-prescription of two antipsychotics as a means of ongoing treatment’ is more controversial because the evidence base supporting its use is weak. As regards extent of polypharmacy, over 27% of prescriptions for depression and over 55% of schizophrenia had at least three different drugs, respectively. The frequency of polypharmacy is lower than that reported in earlier studies.20,21 Though it further declined following intervention in the present study, 21 and 40% of patients still continued to receive more than two drugs in depression and schizophrenia, respectively. Polypharmacy is generally not recommended because of increased risk of adverse reactions and the lack of evidence for therapeutic advantage from using several psychotropic drugs instead of a properly chosen single drug. In particular, the sedative and anxiolytic properties of some antidepressants and neuroleptics make it hard to justify the common concurrent administration of minor transquilisers and/or hypnotics. Prescriptions for essential drugs accounted for 80.95 and 96.91% of the total number of drugs prescribed in depres-
sion and schizophrenia, respectively, in the pre-intervention analysis. Prescriptions from essential drugs improved further significantly after intervention in depression. However, prescription for essential drugs declined in schizophrenia following intervention, since risperidone had been recently introduced in the country and, because of fewer extrapyramidal adverse effects, became the first-line treatment compared to high-potency antipsychotics. The essential drugs list (EDL) is revised every 2 years and, at the time of the study, risperidone was not considered as essential and, therefore, was not available at the hospital pharmacy. One of the criteria for inclusion of new drugs in the list, in addition to definite efficacy, safety and cost advantage over older drugs, is a minimum post-marketing experience of 3 years. This could explain the fall in percent prescriptions for essential drugs in the post-intervention period in schizophrenia. In the present study, 12% of prescriptions for schizophrenia and 15 /16% of prescriptions for depression also carried antidepressants and antipsychotics, respectively, for possible accompanying depressive or psychotic symptoms. This is an another area riddled with controversies, but clinical data collected in the present study are not sufficient to allow us to comment any further on such and other similar aspects of drug prescribing. Thus, the specificity of drug choice cannot be discussed, although the prescription of antipsychotics in 15% of patients with depression seems to be acceptable. Percent encounters with antiparkinsonians and benzodiazepines were chosen as an indicator of performance since these are often used in a substantial proportion of patients, with the attitude that their use is insignificant.22 Use of such drugs can lead to a patient staying on another medication because of the failure to identify the critical therapeutic combination. In the present study, as well as earlier studies from India, antiparkinsonian agents were prescribed in almost all patients receiving neuroleptics.10,11 Regardless of the indication, adjunctive drugs used in prevention or treatment of side effects were co-prescribed in 47 /65% cases.19 23 25 Non-prescription of these drugs with antipsychotics in new patients in an out-patient setting has been reported to lead to non-compliance, but the available guidelines do not recommend prophylactic use of these agents at the time of commencing treatment or during maintenance therapy once the patient has stabilized.13,14 With the availability of atypical antipsychotics in recent years, prophylactic use of antiparkinsonian agents may not be required because of the lower potential for these drugs to cause extrapyramidal side effects.13,26 However, clinical experience in the Indian population shows a similar frequency of extrapyramidal side effects with risperidone compared to the conventional high-potency antipsychotics. In our setting, the prophylactic use of antiparkinsonian medication may be considered because of the high rate of extrapyramidal side effects among patients receiving antipsychotic medications, especially with the high-potency antipsychotic agents, and with risperidone being the most frequently prescribed
Prescribing pattern and indicators for performance in a psychiatric practice
antipsychotic in schizophrenia. However, the need for antiparkinsonian agents in the present study is not justified, particularly when the acute phase of treatment was over, since most prescriptions were for chronic patients. This question is of great clinical importance since, in a proportion of patients, antiparkinsonians can be safely withdrawn after a few months of treatment.27 There is a risk that some patients may remain unnecessarily on these medications,28 and there is evidence that these drugs promote development of tardive dyskinesia.29 Moreover, these drugs may be abused for their euphoriant effects.29 Similarly, high use of benzodiazepines as a co-prescription drug in affective disorders was reported in the present and other studies.23 Use of benzodiazepines along with antidepressants may be more justified in patients presenting with severe anxiety and agitation; however, use of these agents, particularly after the acute phase is over, needs attention. The next questionable pattern concerns prescription of drugs to be taken three or more times a day and the number of tablets taken per day, as was the case with nearly half of the oral neuroleptics.24 The pharmacokinetics of most of these drugs makes it possible, except in the initial phase of treatment, to administer the entire daily amount at bed time, or at most in two doses. Less frequent administration could improve compliance by reducing the number of tablets taken per day and making treatment more convenient. An educational intervention in the form of small group discussion was chosen to promote co-operation and consensus and disseminate key behavioural messages; these were addressed in a highly interactive discussion rather than in conventional didactic lectures. Moreover, interactive small group discussions were shown to be effective in bringing about behavioural change in some of the prescribing practices, but did not demonstrate significant impact in improving the appropriate use of antiparkinsonian agents and benzodiazepine use. Use of multiple drug therapy appears to be broad based, which is often driven by clinical need rather than evidence-based data. As a result, clinicians tend to rely on their clinical experience in such cases. Secondly, clinicians’ access to information about advances in drug therapies and pharmaceutical industry-sponsored information also may play a substantial role in physicians’ prescribing choices. Some of the factors associated with physicians’ prescribing behaviour elicited during discussions were fear of extra-pyramidal adverse effects and loss of follow-up. In the out-patient department, whenever a patient develops extra-pyramidal adverse effects, the chances of future non-compliance increases and such patients usually do not report back. Due to a general lack of awareness and the stigma associated with mental illnesses, a visit to a psychiatrist is often feared, and adverse effects further increase the chances of not seeking any medical help from qualified psychiatrists. In India, the stigma associated with psychiatric illnesses arises from a number of false beliefs, viz. mental illnesses are incurable, the mentally ill are aggressive, impulsive and unpredictable. Drugs used for mental illnesses further increase stigma because of troublesome adverse
237
effects. Patients on antipsychotics drugs may often feel isolated because of marked extra-pyramidal adverse effects like stooping gait, masked facies and poverty of movements.30 Other reasons suggested included belief among patients that the more drugs prescribed, the better the effect. These misperceptions can perhaps be targeted by periodic reinforcement so that prescribers are persuaded by strong behavioural messages from their peers about a relevant issue. It is important to mention here that objective norms do not exist for either indicator to make a comparison; however, if in a given setting an average number of drugs prescribed per prescription is four, with 90% patients receiving antiparkinsonians and 70% receiving benzodiazepines in schizophrenia and depression during maintenance phase, respectively, as shown in the present study and this practice may be considered inadequate. The health manager carrying out the study may feel that, in a local health environment, the expected norm should be two or three drugs per prescription, with 50% patients needing antiparkinsonians and benzodiazepines. In such a setting, priority should be given to finding out why antiparkinsonians and benzodiazepines are prescribed and to reduce their use if reason for such high use is inappropriate. Interpretation and comparison of findings from different countries is limited by differences in definitions used, methodologies and the availability of drugs, besides patient characteristics. The indicators used in the present study are simple, reproducible, less expensive and provide useful data to inform quality of care activities at the health facility compared to prospective descriptive prescription reviews. This study also suffers from the limitation that, being a crosssectional study, though it can provide information about point prevalence of polypharmacy and co-prescription of adjunctive drugs, it does not distinguish among prescriptions with cross tapers in progress or ongoing co-prescription of two or more drugs from the same therapeutic group and response to therapy. Determining the exact rate of true coprescription and tracking the course of the same would require much more extensive prescription reviews. Also, this type of analysis does not reveal practitioners’ rationale for particular medication choices, whether or not patients’ actually take the medication as prescribed, or patients’ clinical response to the medication.
CONCLUSIONS The results of the present study, as well as other similar studies, have shown that such audits have proved helpful in improving the performance and prescribing habits from as early as 1976.31 Monitoring the use of antiparkinsonian agents and benzodiazepines using drug use indicators described in the present study can form an important component for measuring specific aspects of prescriber’s behaviour. and can be used as an indicator for comparison at different time intervals and between health facilities in a
238
S Sharma et al
reproducible manner. These indicators can also serve as simple and reproducible supervisory tools to detect problems in performance by individual providers or health facilities. Once the broad outlines of drug use practices and behaviour are known, these indicators can be used to identify facilities whose performance falls below a specific standard of quality, so that they can be targeted for intensive supervision.
ACKNOWLEDGEMENTS We gratefully acknowledge the technical support from the Delhi Society for Promotion of Rational Drug Use under the INDIA /WHO Essential Drugs Programme.
REFERENCES 1.
2. 3. 4.
5. 6.
7. 8.
9. 10.
11.
12.
13.
14. 15.
16.
Weiden P, Rapkin B, Mott T et al (1994) Rating of medication influences (ROMI) scale in schizophrenia. Schizophr Bull 20: 297 / 310. Weiden PJ, Oljson M (1995) Cost of relapse in schizophrenia. Schizophr Bull 21: 419 /29. Cramer JA, Mattson RH (1991) Monitoring Compliance in Medical Practice and Clinical Trials . New York: Raven Press: 123 /38. Hemminki E (1988) In: Ghodse, Khan (eds.) Factors Influencing Prescribing in Psychoactive Drugs: Improving Prescribing Practices . Geneva: World Health Organization: 22 /35. Morgan R, Gopalaswamy AK (1984) Psychotropic drugs: another survey of prescribing patterns. Br J Psychiatry 144: 298 /302. Edwards S, Kumar V (1984) A survey of prescribing of psychotropic drugs in a Birmingham psychiatric hospital. Br J Psychiatry 145: 502 /7. Muizen M, Silverstone T (1987) A comparative hospital survey of psychotropic drug prescribing. Br J Psychiatry 150: 501 /4. Clark AF, Holden NL (1987) A survey and follow-up of prescribing to chronic hospital in-patients. Br J Psychiatry 150: 88 /91. Holloway F (1988) Prescribing for the long-term mentally ill: A study of treatment practices. Br J Psychiatry 152: 511 /5. Amarjeet, Chadda RK, Bapna JS (1997) Pattern of use of psychotropic drugs in an inpatient unit in a mental hospital. J Ment Health Hum Behav 1: 29 /34. Aggarwal AK, Natu MV, Kalra R, Deswal RS (2001) Prescribing patterns of psychotropic drugs in a tertiary level hospital. Indian J Pharmacol 33: 243 /4. How to Investigate Drug Use in Health Facilities. Action Programme on Essential Drugs. World Health Organization. WHO/DAP/93.1. Expert Consensus Working Group (1999) Expert Consensus Guideline Series: Treatment of Schizophrenia 1999. J Clin Psychiatry 60 (Suppl II): 1 /80. Practice guidelines for the treatment of patients with major depressive disorder (2000) Am J Psychiatry (Suppl) 157: 1 /45. Keks NA, Altson K, Hope J, Krapivensky N, Culhane C, Tanaghow A, Doherty P, Bootle A (1999) Use of antipsychosis and adjunctive medications by an inner urban community psychiatric service. Aust N Z J Psychiatry 33: 896 /901. Clark RE, Bartels SJ, Mellman TA, Peacock WJ (2002) Recent trends in antipsychotic combination therapy of schizophrenia and schizoaffective disorder: implication for state mental health policy. Schizophr Bull 28: 75 /84.
17.
18.
19.
20. 21.
22. 23.
24.
25.
26.
27.
28. 29.
30. 31.
Weissman EM (2002) Antipsychotic prescribing practices in the veterans health care administration / New York metropolitan region. Schizophr Bull 28: 31 /42. Procyshyn RM, Kennedy NB, Tse G, Thompson B (2001) Antipsychotic polypharmacy: a survey of discharge prescriptions from a tertiary care psychiatric institution. Can J Psychiatry 46: 334 /49. Chong SA, Sachdev P, Mahendran R, Chua HC (2000) Neuroleptic and anticholinergic drug use in chinese patients with schizophrenia resident in a state psychiatric hospital in Singapore. Aust N Z J Psychiatry 34: 988 /91. Nunley M (1996) Why Indian psychiatrists prescribe so many drugs. Culture Med Psychiatry 20: 165 /97. Konard M, Kolakowska T (1981) A survey of prescribing psychotropic drugs in two psychiatric hospitals. Br J Psychiatry 138: 217 /21. Dufresne RL (1996) Issues in polypharmacotherapy focus on major depression. Psychopharm Bull 32: 547 /53. Fourrier A, Gasquet I, Allicar MP, Bouhssira M, Lepine JP, Begaud B (2000) Patterns of neuroleptic drug prescription a national cross-sectional survey of a random sample of French psychiatrists. Br J Clin Pharmacol 49: 80 /6. Ungvari GS, Pang AH, Chiu HF, Wong CK, Lum FC (1996) Psychotropic drug prescription in rehabilitation. A survey in Hong Kong. Soc Psychiatr Psychiatr Epidemiol 31: 288 /91. Chiu HF, Shum PS, Lam CW (1991) Psychotropic drug prescribing to chronic schizophrenics in a Hong Kong Hospital. Int J Soc Psychiatry 37: 187 /91. Luo RD, Belleti DA, Tran D, Arcona S, Salen PN (2002) National prescribing patterns in the management of extra pyramidal symptoms among patients with schizophrenia. Int J Psychiatr Med 32: 261 /9. Hsin-Tung, Simpson GM (2000) Medication induced movement disorders. In: Sadock BJ, Sadock VA (eds.) Comprehensive Textbook of Psychiatry, VII Edn. 2265 /71. Putten GP, Best NR, McGiure J (1984) A anticholinergic drug abuse. A common problem? Br Med J 289: 612 /4. Meyer JM, Simpson GM. Anticholinergics and amantadine. In: Sadock BJ, Sadock VA (eds.) Comprehensive Textbook of Psychiatry, VII Edn. 2276 /82. Wig NN (1997) Stigma against mental illness editorial. Indian J Psychiatry 39: 187 /9. Diamond H, Tislow R, Snyder T, Rickels K (1976) Peer review of prescribing patterns in a CMHC. Am J Psychiatry 133: 697 /9.
