Abstract
Medication use in the child and adolescent psychiatric population has increased in recent years, in the absence of robust research findings to support the use of psychotropic medication in this population [1–6]. Jensen et al. [1] report that 80% of medication used in children is used for other reasons than that which is written in the product information (‘off label’) and there is a significant mismatch between the level of prescribing of nonstimulant psychotropic medication and the empirical knowledge regarding its use. Gadow [4] has noted the absence of a recognized standard for determining medication overuse and raised the concern of potential misuse of medication in this population. In addition, children and adolescents may be more vulnerable to short- or longer-term adverse effects, given the unclear pharmacokinetics and pharmacodynamics of these medications in this population [7].
Little is known about how specific services are using medication, the rationale for this and the outcome. Identifying differences in the way services prescribe and the reasons for such differences facilitates the development of more uniform guidelines for psychotropic medication use in these services. Therefore, published audits of the use of psychotropic medication in the child and adolescent population are reviewed.
We identified six studies. Two of the studies were postal questionnaires to a random number of prescribers and did not address how specific services are prescribing medications [8], [9]. Bramble's study [9] inquired predominantly about antidepressant use, while Quinn et al. [10] and Jensen et al. [1] presented the prescribing data from large samples, which again, did not increase our knowledge of how medication is being prescribed in more discrete populations such as child and adolescent mental health services (CAMHS) in order to assist decisions at the level of service provision. In contrast, Kaplan et al. [11] investigated the appropriateness of prescribing practice in two outpatient settings and found a similar profile of prescribing practices. Dosetor et al.'s [12] 1-year audit dealt with three different patient populations with differing medication requirements, and highlighted the importance of prescribing practice being open to scrutiny by medical and non-medical colleagues to facilitate increased awareness and learning. Connor et al. [13] looked specifically at combination pharmacotherapy, noting that this is little researched, and emphasized the importance of documenting its prevalence in different treatment populations in order to increase empirical data on prescribing practices for child and adolescent psychiatrists. Wilens et al. [14] have further suggested tentative clinical indications for the practice of combined pharmacotherapy. Several authors stressed the importance of considering diagnostic and clinical information to determine the appropriateness of prescribing practices [10–12],[15].
In summary, key themes to emerge in the literature review were the lack of research of prescribing practice in this population compared to adult services, the current inadequate training of child and adolescent psychiatry fellows in psychotropic medication use and the need for this to be a clearly established part of specialty training [12].
An audit of a CAMHS is described below. The primary aims of this audit were to determine the prevalence and types of medications used in a regional CAMHS, including combination pharmacotherapy, and the symptoms and/or syndromes and/or diagnoses for which they were prescribed and the profession of the medication prescriber. Additional aims were to gain information on: (i) the rationale behind the prescribing of combination pharmacotherapy; and (ii) the practitioners perceptions of effectiveness.
Method
Subjects
Subjects were children and adolescents treated with psychotropic medication, aged 3–18 years, living in the catchment area of the participating CAMHS and suffering from psychiatric disorders. The diagnoses of these psychiatric disorders were confirmed by the prescribing consultant psychiatrist/paediatrician and the doses of medication given were adjusted to achieve an agreed (prescriber, case manager and child/parent) decrease in core symptom dimensions. The case manager and prescriber recorded the agreed decrease. All medications used were prescribed according to accepted practice procedures. All patients being actively treated in the service at the stated census time and their treating physicians were included in the study. All cases were in a shared care arrangement if private paediatricians and/or general practitioners were prescribers.
Study 1
Measures
Data regarding age, gender, diagnoses and comorbid diagnoses, psychotropic medication prescribed (single and combination) and occupation of prescriber were obtained.
Procedure
The Maroondah Hospital Ethics Committee approved the audit. A list of all cases active as of 1 December, 1999 was obtained and all case notes were reviewed and relevant information documented, including the psychotropic medication, if any, which was prescribed on this date. Reliability was tested for each of the categories described, by an independent rater, for a random selection of cases.
Data analysis
Number and the associated percentage of the total sample of: (i) patients for which there were active prescriptions in particular diagnostic groups; and (ii) patients prescribed for by different professions, are presented.
Study 2
Measures
A semistructured interview was devised, covering questions regarding rationale for prescribing (for example, target symptoms, syndromes and diagnoses), outcome, and adverse effects. Adverse effects were asked about explicitly and the prescriber response recorded. Part of the Clinical Global Impressions Scale (CGI) formed one of the items on the interview schedule, that is, Global Improvement and Efficacy Index. Global Improvement takes into account many factors, such as the effect of psychotropic medication and psychosocial factors that may have been crucial to a patient's altered clinical state. The Efficacy Index attempts to determine whether adverse effect burden outweighs the therapeutic effect of the medication. Efficacy Indices were calculated for each medication used in combination, and then for the combination itself. An average of these figures was then calculated. This gives an indication of the Efficacy Index of each medication in terms of how effective it was considered as part of the combination, as well as an indication of the effectiveness of the combination itself. Taking into consideration the adverse effect burden as well as the therapeutic score gives a more accurate picture of the usefulness of the medication combination. Efficacy Indices greater than 1 indicate that the perceived benefit of the medication outweighs the adverse effects and vice versa for efficacy indices less than one.
Procedure
Based on the information obtained from Study 1, a list of practitioners responsible for prescribing combination pharmacotherapy was obtained. The practitioners were sent a letter outlining the study and a copy of the interview schedule. To maximize the response rate all the prescribing practitioners were then contacted individually and the interview schedule discussed with them, in order to clarify any queries and increase the accuracy with which it was completed. Interviews were conducted with the prescribers for 69 out of 71 patients and each patient was discussed.
Data Analysis
Quantitative descriptive data are presented.
Results
Study 1
The total number of patients being treated by the service on 1 December, 1999 was 734. There were no missing files. Of this number, 235 (32%) were treated with medication. Of the 235 patients being prescribed medication, 164 (70%) were on one medication and the remaining 71 (30%) were being prescribed combination pharmacotherapy; 9.7% of the total number of patients in the service was therefore prescribed combination medication.
The age range of patients on pharmacotherapy was 3 years (one patient only) to 20 years. The mean (SD) age of patients on single pharmacotherapy was 12.8 years (3.7) compared to 11.5 years (4.1) for those on combination pharmacotherapy. This difference was not statistically significant (t(304) = 1.62, p = 0.11). The youngest child to receive combination pharmacotherapy was aged 5 years and 61.2% of patients treated by the service were male. Of the 235 patients receiving medication, 164 (70%) were male. Male patients comprised 68.3% of those patients on single medication and 81.7% of those on combination pharmacotherapy. These differences were again not statistically significant (χ2 (1) = 1.66, p = 0.10).
Table 1 outlines the single medications prescribed for particular diagnoses. The most common reason for prescribing medication was attention deficit hyperactivity disorder (ADHD), with stimulant medication the most frequently prescribed, followed by antidepressants prescribed for major depressive disorder.
Patients on single medication
Consultant psychiatrists working in the service were the most frequent prescribers of medication (χ2 (1) = 4.34, p = 0.02). Private paediatricians did a considerable amount of prescribing for the service, being responsible for 25% of those on single medications and for almost one-third of those on combination pharmacotherapy. Of note, the one child psychiatrist trainee and the three psychiatry registrars did comparatively little prescribing (Table 2).
Prescribers of medication
Study 2
Sixty-seven of the 69 respondents reported prescribing the medication sequentially, after an inadequate response with a single medication. There was a large range of symptoms and diagnoses for which medications were prescribed (Tables 3,4). Reasons for prescribing combinations rather than monotherapy included combating the adverse effects of the initial medication, to treat a comorbid diagnosis, or because the initial medication was failing to treat symptoms adequately (97%). All made it clear that their preference was for monotherapy, low doses to reduce the likelihood of adverse effects, a long enough period of use at an appropriate dose for prescribers to be clear about their effectiveness, and an optimal frequency of review. Only this last point varied between prescribers. The stimulants were most frequently used in combination for ADHD symptoms, while clonidine was most frequently used for initial insomnia related to stimulant use, along with motor overactivity, impulsiveness and aggression in this patient population. Risperidone was used in combination with (i) stimulants and (ii) clonidine for aggression and to decrease the frequency and intensity of affect change (that is, mood lability), again in the ADHD population primarily, and for the treatment of psychotic disorders (Tables 3,4,5). The rate of comorbidity for those treated with combination pharmacotherapy was 83%. Only nine of the 60 cases of ADHD treated with combination pharmacotherapy did not have a comorbid diagnosis. Oppositional defiant disorder was the most frequently comorbid condition with ADHD, occurring in 20 of the 60 cases of ADHD.
Prescribing rationale
Prescribing rationale
Medication combinations used by prescribing practitioners
Fifty-three of the 69 (77%) patients were rated as being ‘Very Much’ or ‘Much Improved’ using the Clinical Global Impressions Scale (CGI). Ten percent of patients were rated as being ‘Unchanged’ or ‘Minimally Worse’ and no patients were rated as ‘Much Worse’ or ‘Very Much Worse’. The main adverse effect with individual medications was sedation, which occurred in 20 out of 29 cases in which risperidone was used. Insomnia was the second most common adverse effect, occurring 16 out of 56 times in which stimulant medication was prescribed. No new adverse effects were described when combination pharmacotherapy was used. In fact, there were 72 adverse effects mentioned using single pharmacotherapy compared to 43 with combination pharmacotherapy. This reduction is partly due to the addition of clonidine to alleviate sleep disturbance, with only three of the 16 patients experiencing sleep disturbance on stimulants continuing to do so. In addition, 25 rather than 35 patients continued to be described as having sedation once combination pharmacotherapy was introduced. All efficacy indices were greater than one, indicating that therapeutic benefit outweighed adverse effects for the various medication combinations (Table 5).
Discussion
The rate of medication prescription found in this audit is consistent with that reported by Dosetor et al. [12] and illustrates the importance of pharmacotherapy as a treatment modality available to clinicians. While the crosssectional data presented cannot suggest that the rate of medication use is increasing, several recent authors have documented the increasing use of pharmacotherapy, both singly and in combination [1–4],[10], although the reasons for this increase are many and varied. For example, a greater appreciation of the biological basis for many disorders [4], a more favourable adverse effect profile of newer medications and increased safety in overdose, as evidenced by the higher rates of selective serotonin re-uptake inhibitors (SSRI) and atypical antipsychotics used in this audit, and family expectations and those of the education system may influence the decision to use medication. Rey et al. [3] have emphasized that the awareness of effective non-pharmacological interventions as viable alternatives to the unproven use of medication needs to be fostered in the community, albeit also emphasizing the current limited knowledge base for many forms of current individual, family and group psychological treatment, aside from cognitive behaviour therapy (CBT). They also sympathetically note that clinicians may feel urged to prescribe without an evidence base to do so. The use of potentially unsafe practices was still felt unjustified, however, and the authors urged that the NHMRC make funding of research in this area a priority and look at ways of overcoming the complex ethical issues involved. Kaplan et al.'s findings [11] were replicated in the current study with many prescribers using medication not tested in or indicated for the child and adolescent population. Gadow [4] laments that this is not a new situation. In 1975 he documented over a dozen different psychotropic medications that lacked clinical indications for preschoolers, but in 1997 none of these medications have yet been labelled for use by children under 6 years.
Differences in prescribing practices between practitioners were noted. The amount of debate, teaching and review of practices within a service and the provision of more guidelines for prescribing would possibly enable a more consistent pattern of prescribing to occur. It is arguable that, although practitioners’ professionalism and prescribing autonomy should be respected, general consensus should also be reached regarding what is considered appropriate treatment for specific mental illnesses.
Bramble's [9] study found that many child psychiatrists were concerned about the poor quality of training regarding psychopharmacotherapy. Other authors [12], [16] reiterated the need for guidelines for junior psychiatric registrars and the importance of peer review and ongoing education. The findings of the current study may point towards inadequate experience and training in the use of pharmacotherapy by general psychiatry trainees. At the time of the study, none of the three trainees completing their 6-month rotation in child and adolescent psychiatry were prescribing medication for a patient.
Combination pharmacotherapy is currently controversial in the child and adolescent population and authors have stressed the need to gather data regarding the use of combination pharmacotherapy to increase awareness of how medications are being prescribed [13]. The current study addresses this need. Further, Morgan et al. [15] found that discussion with prescribers to obtain a greater knowledge of the rationale behind their prescribing led to reduced rates of inappropriate prescribing. Unfortunately, we did not address this further question through the longitudinal assessment of prescribers’ habits.
All prescribers preferred monotherapy, with subsequent medication added to address an unsatisfactory response to a single medication, to treat an adverse effect arising from a single medication, or to treat another target comorbid disorder. Low doses were used to ensure that patients were not receiving higher doses than were necessary and to reduce the likelihood of adverse effects, and medications were used for a long enough period at an appropriate dose for prescribers to be clear about their effectiveness. Discussion with private practitioners revealed they sought further advice when patients were failing to improve on monotherapy and were unclear about the optimal frequency of review. Future guidelines for the use of combination pharmacotherapy should include the frequency of review for these patients.
Unlike the study by Connor et al. [13], this study did not find an association with disruptive behavioural disorders and no combination pharmacotherapy. Prescribers interviewed often commented that the patients on combination pharmacotherapy were their most difficult cases. Attention deficit hyperactivity disorder was the most common diagnosis for those on combined pharmacotherapy (85%), which reflects the high rates of a number of comorbid disorders in this group of patients [17]. In addition, the service studied is a tertiary referral centre that manages the most treatment non-responsive patients whom general practitioners and/or paediatricians have difficulty treating in a primary care setting. Attention deficit hyperactivity disorder and oppositional defiant disorder or conduct disorder are the most common clinical group referred.
The Global Improvement Scale results suggest a clear improvement in the clinical state of those children and adolescents on combined pharmacotherapy, although psychosocial variables are included in this overall assessment. In future studies, parent and child views should also be ascertained along with prescriber views.
No Efficacy Index scores for any of the medication combinations indicated that adverse-effects outweighed therapeutic effect, thus implying that the risk/benefit ratio of such combinations was low. However, the adverse effect data obtained was only from prescribers who are less reliable than patients at reporting and recording such data. Also, there are no studies reporting safety in the short- or long-term for any combination pharmacotherapy used in this population [1], [5]. While the receptor specificity of the newer medications (e.g. the SSRIs) does mean their adverse effect profile will be more favourable than older medications such as the tricyclic antidepressants, practitioners need to be aware of the possibility of drug interactions, particularly given the absence of research on the use of combinations in this population. However, in this study there was no evidence that these occurred for any patient using combination pharmcotherapy. Possible longer-term sideeffects are unknown.
Since prescribers feel strongly that combination pharmacotherapy is benefiting their patients, they are prescribing it. However faced with inadequate supportive research or guidelines it is arguable they should not. This raises the issue of the perhaps equally unethical situation of denying medication to children and adolescents which could assist their level of functioning and achievement. Further research is needed in clinical populations that are not responding to conventional psychological and single medication treatments to ascertain what symptoms, behaviours and neuropsychological and/or neurophysiological markers best respond to particular combined pharmacotherapy approaches. Also, training of CAMHS staff and trainee child psychiatrists in particular, is a priority in the recognition of such treatment non-responsive cases, and the incorporation of combined pharmacotherapy approaches, if indicated, and the frequent monitoring of such interventions are required if the needs of these patients are to be best met.
Footnotes
Acknowledgements
We wish to acknowledge Kwong Yeang for his assistance with data collection, as well as Maroondah CAMHS.