International Journal of Psychiatry in Clinical Practice, 2010; 14: 212–219

ORIGINAL ARTICLE

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

Drug–drug interactions in general hospital and psychiatric hospital in-patients prescribed psychotropic medications

LINDSEY I. SINCLAIR1,2, SIMON J. C. DAVIES2, GRAHAM PARTON3 & JOHN P. POTOKAR1 1Psychopharmacology

Bristol, 3Avon

Unit, University of Bristol, 2Academic Unit of Psychiatry, Community Based Medecine, University of & Wiltshire Mental Health Partnership NHS Trust, Bristol, UK

Abstract Objectives. Drug–drug interactions (DDIs) present a serious, ever increasing clinical problem. Previous studies identified DDIs among psychiatric inpatients prescribed psychotropics, but none have focused on psychotropics prescribed to General Hospital inpatients. This study aimed to identify: putative drug–drug interactions; mechanisms; potential seriousness among patients prescribed psychotropes in both psychiatric and general hospital inpatients settings. We hypothesised that potential interactions per person would be greater in General Hospital inpatients on psychotropics, due to polypharmacy. Method. We surveyed psychotropic prescribing in hospital wards in a public sector mental health organisation and a 500-bed general hospital. Ward pharmacists collected drug prescription data. A computer based protocol evaluated DDIs. Results. A total of 7.4% of General Hospital inpatients and 100% of Psychiatric Unit inpatients surveyed were prescribed psychotropic medication. The General Hospital group had significantly more potential interactions per person (3.0) than Psychiatric inpatients (1.3) (P⬍0.05). There were significantly more potentially serious interactions in the general hospital group (P⬍0.025). Conclusions. DDIs affect those prescribed psychotropics in both General and Psychiatric Hospitals. The General Hospital patients had a higher number per person and more serious potential interactions, yet are often poorly served by psychiatric services, suggesting that liaison psychiatrists have a role in physician education and DDI assessment. Key Words: Drug interactions, psychotropic drugs, psychiatry, inpatients

Objectives Drug interactions are a major clinical problem and have been linked to emergency department visits [1], admissions and re-admissions to hospital [2], morbidity and mortality. Adverse interactions are defined as occurring “when two drugs are combined in a manner that alters either the effectiveness or the toxicity of one of the agents” [3]. Medications may interact in both a pharmacokinetic and pharmacodynamic manner. The reported incidence of drug–drug interactions (DDI) varies from 2.2 to 30% among in-patients [4] and 9.2 to 70.3% among out-patients [3,5–9]. Some of the variation is due to differing study methods, particularly those studies including interactions that were not considered clinically significant. There is also evidence that DDIs are becoming more common, with relevant factors including increased drug availability,

subspecialisation of doctors and polypharmacy [7,8]. Polypharmacy has been shown to increase the risk of drug–drug interactions [10,11]. The ability to prescribe appropriate drugs in a safe manner is a core skill for medical practitioners. Concerns have been raised recently regarding undergraduate education and the extent to which medical students are adequately prepared for prescribing after graduation [12]. In psychiatric practice the increase in polypharmacy may be due to the increased use of maintenance treatment and the ageing population [7–9] with other medical conditions requiring treatment [2,13–16]. Other reasons for using more than one drug per patient include: treatment of adverse effects from the first medication; augmentation of desired effect and acceleration of desired effect [17]. There have been several previous studies examining DDIs [1,4,7–9,11] although few amongst psychiatric

Correspondence: Dr L I Sinclair, Academic Unit of Psychiatry, Cotham House, Cotham Hill, Bristol BS6 6JL, UK. E-mail: [email protected] (Received 4 September 2009; accepted 8 April 2010) ISSN 1365-1501 print/ISSN 1471-1788 online © 2010 Informa Healthcare DOI: 10.3109/13651501.2010.486899

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

Drug interactions in inpatients on psychotropics in-patients [14,18]. In a retrospective study of 131 psychiatric in-patients in the USA, 141 potential pharmacokinetic interactions were identified, although no comment was made on clinical significance [18]. Previous work by our group found potentially clinically significant pharmacokinetic interactions in more than 20% of psychiatric in-patients on adult and functional elderly wards, although this study only considered interactions via CYP2D6 & 3A4 [14]. Patients in General Hospitals are often poorly served in terms of access to psychiatric expertise [19]. There is a high prevalence of mental disorders or symptoms amongst General Hospital in-patients, with up to 65% being affected [20] and this may affect the quality of medical care that they receive [21]. The psychiatric symptoms may be pre-existing or may arise as a consequence of the physical illness including its treatment [22]. The role of liaison psychiatry services includes education of general hospital colleagues about psychiatric symptoms and psychotropic medication [23]. If this service is not available, as is frequently the case [19,24] then the risk of inadvertent prescription of interacting drugs may be greater. In this study we aimed to identify putative drug– drug interactions, their mechanisms and potential seriousness using the online interaction prediction tool genemedrx.com among patients prescribed psychotropic agents in both psychiatric and general hospital inpatients settings. This tool is used locally in routine clinical practice. We hypothesised that in those patients on psychotropic drugs, the number of interactions per person would be greater in the General Hospital than in psychiatric inpatient settings, reflecting the potential for interaction with drugs for physical disorders.

Methods This was a cross-sectional survey of psychotropic prescribing within the Bristol Royal Infirmary, a 500bed general hospital hereafter referred to as the “General Hospital” and a large mental health service’s in-patient wards. Ward based pharmacists identified all inpatients prescribed one or more psychotropic drug. Data collected included dose and regimen of all prescribed medications and whether they had been administered in the preceding 24 h. Data were collected on a standardised form over a period of 3 months (December 2006 to February 2007), with 16 representative psychiatric wards and all 26 wards within the general hospital being visited once in this time. Types of psychiatric wards visited included acute adult, adult rehabilitation, old age, drug rehabilitation, high dependency, intensive care and forensic.

213

We were advised by a local ethical committee that no ethical approval was required due to the anonymised survey design. For each patient a full list of prescribed medications was entered into www.genemedrx.com and predicted interactions examined. Potentially serious interactions were defined as a decrease greater than 60% or increase of 75% of more in the plasma concentration of the affected medication. Due to the unexpectedly much greater numbers in the Psychiatric hospital group, non-parametric statistics were used for continuous measures. A significance level of P⬍0.05 was used throughout. Unless the use of another test is specified Mann–Whitney U-tests were used to test for significance. Because of the exploratory nature of this survey it was not possible to perform a formal power calculation.

Results Overall sample characteristics Thirty seven (7.4%) inpatients in the General Hospital and 216 (100%) inpatients in the Psychiatric Units surveyed were prescribed psychotropic medication and were thus included in the study. The General Hospital group was smaller than expected, which limits the conclusions that can be drawn. Of those patients prescribed psychotropic medications the general hospital group were: significantly older (69.9 vs. 46.9 years, P⬍0.01); on significantly more prescribed medications (mean per person [pp] 9.0 SD 4.5 vs. 5.4 SD 3.1, P⬍0.05); [significantly fewer psychotropics (mean pp 1.8 vs. 3.0, P⬍0.001); significantly fewer antipsychotics (mean pp 0.2 vs. 1.1, P⬍0.001); significantly more antidepressants (mean pp 0.8 vs. 0.4, P⬍0.001); significantly fewer hypnotics and benzodiazepines (mean pp 0.6 vs. 1.0, P ⫽0.004); significantly fewer mood stabilizers (mean pp 0.2 vs. 0.4, P⫽0.006)]; and significantly more were female (67.6 vs. 38.0%, Chi-square test, P⬍0.01). A total of 50 different psychotropic agents were being prescribed in these settings, with a grand total of 712 prescriptions. The majority of the psychotropics were prescribed orally, with only 57 prescriptions for depot medications (11.8% of all atypical antipsychotic prescriptions and 37.7% of all typical antipsychotic prescriptions).

Drug–drug interactions A total of 380 potential DDIs were identified: 111 in the General Hospital group and 269 in the Psychiatric Hospital group. Of these, the General Hospital group had significantly more potential

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

214

L. I. Sinclair et al.

Figure 1. The predicted percentage change in the plasma concentration of the affected medication for all DDIs. The majority of interactions resulted in only a minor change in plasma concentration. In the general hospital group there were significantly more potentially serious interactions, namely a more than a 60% decrease or a 75% increase in the plasma concentration of the affected medication (χ2 statistic 6.10, P⬍0.025).

interactions per person (mean 3.0 SD 2.4 vs. 1.3 SD 1.3, P⬍0.05) (see Figure 1). Only 21.6% of General Hospital patients included and 32.9% of Psychiatric patients had no potential interactions. A wide range of both psychotropic and non-psychotropic medications were involved in potential interactions in both settings (Tables I and II) with olanzapine and lamotrigine being most commonly affected and carbamazepine, diclofenac and ciprofloxacin being most commonly causative. Table I. Individual drugs and interaction causation potential. The number of instances of grug being affected in an interaction per prescription is given to allow direct comparison between the two groups. Only drugs causing ⱖ0.5 interactions per prescription in either group are shown.

Fluoxetine Citalopram Sertraline Paroxetine Olanzapine Risperidone Clozapine Haloperidol Diazepam Lamotrigine Clomipramine Impramine Trazodone Amitryptilline Omeprazole Aspirin Diclofenac Ciproflozacin

General Hospital

Psychiatric Hospital

Total

0.5 0.6 0.5 0.0 N/A 1.0 N/A 0.3 0.6 1.0 N/A N/A 0.5 1.0 0.6 0.2 1.0 0.5

0.3 0.4 0.0 0.7 0.7 0.1 0.6 0.5 0.2 0.6 1.0 0.5 0.0 0.5 1.5 0.1 0.0 0.0

0.4 0.4 0.2 0.3 0.7 0.2 0.6 0.5 0.3 0.7 1.0 0.5 0.1 0.6 0.3 0.2 0.2 0.3

The algorithm, “genemedrx” identified some interactions as “major”, which either reflected a greater than 150% change in plasma concentration of the affected drug, or major potential clinical consequences, e.g., death. By far the most frequent “major” interactions (Table III) were those; between clozapine and lorazepam where there have been case reports of delirium and death [25,26] despite the relatively widespread use of this combination; and the co-prescription of tramadol with a serotonergic antidepressant, which may precipitate the serotonin syndrome [27–29]. In addition the use of tramadol

Table II. Individual drugs and interaction causation potential. The number of causations per prescription is given to allow direct comparison between the two groups. Only drugs causing ⱖ0.5 interactions per prescription in either group are shown.

Fluoxetine Sertraline Escitalopram Paroxetine Quetiapine Clozapine Diazepam Valproate Carbamazepine Venlafaxine Trazodone Amitryptilline Omeprazole Aspirin Diclofenac Ciprofloxacin

General Hospital

Psychiatric Hospital

Total

2.2 1.0 N/A 1.0 1.0 N/A 0.6 1.3 4.0 1.0 0.5 1.0 2.6 2.6 0.0 2.5

0.8 1 0.5 1 0.8 0.5 0.3 0.9 2.3 0.4 0 2 1.3 0.1 2.5 1.3

1.5 1.0 0.5 1.0 0.8 0.5 0.3 0.9 2.5 0.5 0.1 1.8 1.8 1.7 2.4 2.0

Drug interactions in inpatients on psychotropics

215

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

Table III. Drug pairs resulting in potentially serious predicted drug–drug interactions on three or more occasions in the entire sample and their putative mechanisms.

Drug pair of interest

No. of instances

Clozapine/Lorazepam

9

Major

Unknown

Tramadol/TCA or SSRI or venlafaxine Haloperidol/Clozapine

9

Major

4

Codeine/SSRI

4

Risperidone or olanzapine/ Carbamazepine Citalopram/Omeprazole

3

Clozapine/Fluoxetine

3

Quetiapine/Valproate

5

Haloperidol/Diazepam

4

Valproate/Citalopram

3

Major decrease in [haloperidol] 75–150% increase in [codeine] 60–90% decrease in [antipsychotic] 25–75% increase in [citalopram] 25–75% increase in [clozapine] 25–75% increase in [quetiapine] 25–75% increase in [haloperidol] 25–60% decrease in [valproate]

2D6, 3A4 plus pharmacodynamic 3A4

4

Predicted effect of interaction

with a 2D6 inhibitor such as fluoxetine may cause failure of tramadol’s analgesic effect [29]. In the General Hospital group there were significantly more potentially serious interactions namely a more than a 60% decrease or a 75% increase in the plasma concentration of the affected medication (χ2 statistic 6.10, P⬍0.025). Causative agents of interactions with potentially major consequences included fluoxetine (2D6 & 3A4 inhibitor), ciprofloxacin, (1A2 inhibitor) amiodarone (1A2, 2C9, 2D6 & 3A4 inhibitor) and diltiazem (3A4 inhibitor). Forty-nine percent of all interactions were predicted to cause a less than 25% change in the concentration of the affected medication and were classed as minor. In the General Hospital group significantly more (92.8%) of the potentially interacting combinations had been administered in the preceding 24 h versus 73.6% of those at the Psychiatric Hospitals (χ2 statistic 19.7, P⬍0.001). Although significantly fewer of the interactions identified in the General Hospital group (50.0%) involved a psychotropic than in the Psychiatric wards (84.0%) (χ2 statistic 46.1, P⬍0.001), they still account for half of all potential interactions in this patient group. Six different CYP enzymes were predicted to be involved in the potential interaction (Table IV) the most frequent being 3A4 (34.4%), 2D6 (10.8%) and 2C19 (8.9%). Some potential interactions mediated through phase II conjugation pathways, e.g., glucuronidation (14.9% of interactions) were also identified.

Mechanism

Comments Case reports of delirium & death despite widespread use Risk of serotonin syndrome

2D6, 3A4 3A4 3A4, 2C19 1A2 Unknown Glucuronidation Unknown

Discussion In this study we hypothesised that among in-patients prescribed psychotropic drugs, the number of interactions per person would be greater in the General Hospital than on Psychiatric wards. We report that the general hospital patients prescribed psychotropic medication did indeed have a higher number of predicted DDIs per person, which were potentially more serious. The General Hospital patients prescribed psychotropic medication had a different pattern of psychotropic use to the psychiatric hospital patients. It is likely that they represent a group of patients with

Table IV. Frequency of individual cytochrome P450 enzyme involvement in interactions among general hospital and psychiatric hospital inpatients prescribed psychotropic medication. Figures given are the number of involvements in interactions per 100 patients prescribed psychotropic drugs in each setting and in total across the two settings. Cytochrome P450 2D6, 3A4 and 2C19 were the most frequently involved at both sites.

CYP Enzymes 1A2 2D6 3A4 2C9 2C19 2C8

General hospital interactions/ 100 patients 0.0 51.4 124.3 16.2 48.6 2.7

Psychiatric hospital interactions/ 100 patients

Total interactions/ 100 patients

3.2 9.3 36.1 2.3 6.5 0.0

2.8 15.4 49.0 4.3 12.6 0.4

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

216

L. I. Sinclair et al.

a different profile of psychiatric illnesses, with psychotic disorders in particular being less prevalent. The psychiatric illness profile among General Hospital in-patients is perhaps more similar to psychiatric out-patients than to psychiatric in-patients. The percentage of patients prescribed interacting combinations is higher than previous estimates for General Hospital in-patients and closer to estimations for all out-patients [3]. Antidepressant use, which was more common in the General Hospital group may account for some of this increase in interactions as many antidepressants inhibit CYP450 enzymes. Alternative explanations include: age; polypharmacy; methodological differences between our study and others such as the computer interaction software used in this study predicting minor interactions that were not included in previous studies; or recent alterations in prescribing patterns [10]. Comparison with previous research In this study a higher proportion of interacting combinations had been administered in the preceding 24 h compared with our earlier study (which only considered psychiatric in-patients [14]) in which less than 60% of interacting combinations had been administered in the preceding 24 h. Previous studies have reported that the risk of DDIs increases with age [2,5,15,16]. In one study the odds ratio for a potential DDI was 2.11 for those aged between 65 and 74 years compared to those aged 50–64 years [16]. Here the General Hospital inpatients were significantly older than the psychiatric inpatients and this may be one reason for the higher number of interactions. Previous studies have also shown that the risk of DDIs increases with polypharmacy [5,10]. In our study the General Hospital in-patients were predicted to have more drug–drug interactions. The mean number of drugs prescribed was significantly higher in the General Hospital group and we considered that this may account for some of the betweengroup differences. There was however no significant relation between the number of drugs prescribed per patient and the number of interactions per patient (Spearman ρ 0.026, P⫽0.676].

Mechanisms of drug–drug interactions Cytochrome P450-based interactions are more likely to be clinically significant for drugs with narrow therapeutic windows and if breakdown of an active drug is inhibited [30]. Prior to excretion the vast majority of pharmaceutical agents require metabolism to

increase their water solubility. This occurs in the liver in two phases. The cytochrome P450 enzymes in the liver perform phase I metabolism for 90% of commercially available drugs [31], with six isoenzymes performing the bulk of the metabolism: CYP3A4, 2D6, 2C9, 2C19, 2E1 and 1A2 [3]. These six isoenzymes are also the most important in terms of drug– drug interactions. Some substances are metabolised by a single P450 enzyme, others by multiple P450 enzymes in a variable manner. These enzymes are liable to induction and/or inhibition and this forms the basis of many drug–drug interactions. Previous work by our group looking at drug–drug interactions in psychiatric hospital in-patients found that more than a third of patients were prescribed combinations that interacted via CYP2D6 and a quarter combinations that interacted via 3A4 [14]. The current study was not restricted to 2D6 and 3A4 and found that six different CYP enzymes were predicted to be involved in the potential interaction: the most frequent being 2D6, 3A4 and 2C19. Interestingly in this study, unlike the previous study, more patients were prescribed combinations that interacted via 3A4 than 2D6. This may be because 3A4 has been more fully elucidated in the 7 years since the previous study and thus more potential interactions have been characterised. Detecting drug–drug interactions Given the number of different pharmacological agents now available it is difficult to keep all potential interactions in mind. Several previous studies have examined the use of computerised drug interaction databases [32]. Some have also compared such databases to active pharmacist interaction detection [32–35], but were restricted to General or Psychiatric Hospital or outpatient settings. Our study is the first to compare General Hospital and Psychiatric Hospital inpatients, all prescribed psychotropic medications. In the present study almost half of the interactions detected were thought to produce only a minor change in the plasma concentration of the affected drug. Identifying numerous interactions that are relatively unimportant clinically could have the undesirable effect of trivialising drug interaction software generated warnings, so that warnings of dangerous interactions are missed [33,36,37]. Studies comparing interaction prediction software to pharmacist detection of interactions found that such software detects many more interactions than pharmacists [33,35], In one study the sensitivity and specificity of such a database were acceptable (over 80%) but the positive predictive value was low [35]. This may be because computer programmes may not take

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

Drug interactions in inpatients on psychotropics dosage, disease severity grading or the fact that some interacting combinations are clinically useful into account [33]. It should, however, be remembered that with such a vast array of drugs available to prescribe and polypharmacy increasing in recent times [15] it is impossible for any one doctor or pharmacist to memorise all possible drug–drug interactions. Our previous work [14] used a panel of experts to detect drug–drug interactions. This has advantages in that flexibility, common sense and clinical knowledge can all be applied. Only clinically important or potentially clinically important interactions were considered. There are however a number of disadvantages to this approach as it is time-consuming and subject to human fallibility.

217

prevent pro-active trials of drug pairs known to interact and thus information on symptomatic patients comes only from naturalistic trials and case reports. Despite drug-interactions being identified as an important health problem it is important to remember that not all drug combinations interact in a deleterious manner and some are clinically useful (e.g. Li augmentation of antidepressant). In a large German primary care study [36] only 12.4% of all drug pairs were classified as interacting and the vast majority (⬎90%) of these were manageable, e.g. by dose reduction. Whether this reflects clinical practice is another question. Strengths and limitations of this study

Difficulties of studying drug interactions The majority of available drug information comes from clinical trials either pre or post marketing. Many patients in clinical trials are highly selected, making it difficult to generalise the results. Trials often do not reflect the “real world” where co-morbidity is the norm and multiple agents are often required to treat patients adequately [17,38]. Response in a clinical trial may not equate to remission and in practice combination treatment may be required for a full recovery. Combinations may also occur in the short term during drug switching with transient potential to cause side effects or toxicity [13]. The lack of generalisability of clinical trials undermines computer and paper systems for identifying drug interactions, resulting in a dearth of evidence for many drug pairs [36]. In-vitro models can help to predict interactions but it is unusual to study more than two drugs simultaneously. This clearly does not reflect the clinical norm of multiple medication use. It is also difficult to predict in-vivo effects from in-vitro models, possibly due to poor compliance [39]. Many healthy volunteer studies into interactions have used single doses, rather than steady state (which would be the clinical norm) and the pharmodynamic effects of these interactions have rarely been studied [30,39]. It is also questionable whether data from healthy young male volunteers is generalisable to the elderly who are most commonly those experiencing polypharmacy. There is confusion about when a theoretical drug–drug interaction becomes clinically significant. Another difficulty in studying interactions is that they can present as vague, non-specific symptoms or mimic physical disorders [38]. Conversely symptoms attributed to interactions may be common and be derived from unrelated psychological or physical causes such as generalised anxiety disorder or panic attacks [40]. Ethical considerations

Strengths of this study include the coverage of both general and psychiatric hospital in-patients, complete information on all drugs prescribed and whether they had actually been administered in the preceding 24 h. Limitations include the low number of general hospital in-patients prescribed psychotropic medication (only 7.4% of total), which was less than expected and raises the possibility of ascertainment bias. The unbalanced numbers in the two groups dictated the use of non-parametric statistics, limiting the study’s power. The cross-sectional design meant that we were unable to assess whether interactions whether any efforts were being made to manage them, e.g., dose reduction or indeed whether prescribers had taken pharmacokinetic interactions into account in determining prescribed doses. As our study considered a 24-h timeframe it is possible that some interactions involving drugs administered less than once daily or intermittently were missed. Further this design cannot distinguish combinations prescribed without reference to CYP interactions with scenarios where drugs had been co-prescribed in full knowledge of the potential interaction with appropriate dose adjustment, or where a “problem” drug was in the process of being progressively withdrawn with an alternative compound introduced at a starting dose. Finally we did not collect data on diet or smoking status, both of which can potentially make DDIs more likely.

Suggestions for future work A variety of alleles have been identified for each of the six major CYP450 enzymes and many of these affect enzymatic function [3]. Loss of functional alleles resulting in decreased enzymatic activity (poor metaboliser phenotype) have been identified for CYP 2D6 and 2C19 [30,41]. Two SNPs resulting in decreased function have been identified for CYP 2C9 [41]. It is now possible to identify genotypes for

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

218

L. I. Sinclair et al.

a number of these CYP450 enzymes and, from this information, predict phenotype. Research has shown that CYP2D6 genotype is a good predictor of metaboliser phenotype in a range of different populations [42–45], which may form the basis for dose adjustment according to phenotype. Many psychotropic medications are substrates for the CYP450 enzyme system, as well as acting as inducers (carbamazepine, 3A4) or inhibitors (paroxetine, 2D6). Genetic metaboliser status with respect to CYP2D6 and CYP2C19 is thought to be particularly relevant in psychiatry as so many of the medications that we use are metabolised by this system, as well as inducing or inhibiting the individual enzymes [46]. For example, SSRIs potentiate adverse effects of antipsychotics by inhibiting their metabolism via CYP2D6. Some drugs such as paroxetine may cause “phenocopying”, i.e. inhibition of a CYP450 enzyme producing a similar slowing of metabolism to that seen in those with a genetic poor metaboliser status. It is possible [38,41], although as yet unproven [47], that drug–drug interactions combined with poor metaboliser status could represent an increased risk rendering the patient far more likely to suffer with symptoms from their drug–drug interaction. This important area of medicine deserves further research. Drug companies are now reluctant to develop compounds found to be inhibitors of CYP2D6/2C19 for their metabolism and will seek alternative compounds where possible [17].

Conclusion In conclusion we report that the General Hospital patients prescribed psychotropic medication (when compared to psychiatric in-patients) were prescribed more drugs per person and were subject to more potential interactions per person and the interactions identified in the General Hospital group were theoretically more serious. This reinforces the relevance and importance of an understanding of clinical psychopharmacology for all clinicians if iatrogenic effects are to be avoided. Liaison Psychiatrists can play an important role in educating General Hospital clinicians as well as assessment of individual patients.

Key points • Drug interactions are a potential problem with any prescription of a drug, but particularly with psychotropic drugs • In this study interactions were more common in patients in the general hospital who were prescribed psychotropic medication

• Psychiatric expertise is often not available in the general hospital setting , thus we advocate education of general hospital physicians on CYP interactions to ensure they are recognised

Acknowledgements We gratefully acknowledge the assistance provided by the pharmacy staff of the Bristol Royal Infirmary and the Avon & Wiltshire Mental Health Partnership NHS Trust.

Statement of interest There are no direct conflicts of interest and this work was not commissioned nor funded by any external agency. Drs Sinclair and Potokar have no conflicts of interest to disclose. Mr Parton has been on an advisory board for AstraZeneca and was funded to attend a conference by Lilly. He has also been on advisory boards for Janssen-Cilag, Lilly, Servier and Lundbeck and has received funding as a conference delegate. Dr Davies has performed speaking engagements unconnected to this study for two pharmaceutical companies with fees being paid to the University of Bristol. References [1] Raschetti R, Morgutti M, Menniti-Ippolito F, Belisari A, Rossignoli A, Longhini P, et al. Suspected adverse drug events requiring emergency department visits or hospital admissions. Eur J Clin Pharmacol 1999;54(12):959–63. [2] Becker M, Kallewaard M, Caspers P, Visser L, Leufkens H, Stricker B. Hospitalisations and emergency department visits due to drug–drug interactions: a literature review. Pharmacoepidemiology and drug safety 2007;16:641–51. [3] Sikka R, Magauran B, Ulrich A, Shannon M. Bench to bedside: Pharmacogenomics, adverse drug interactions, and the cytochrome P450 system. Acad Emerg Med 2005;12(12): 1227–35. [4] Leape L, Bates D, Cullen D, Cooper J, Demonaco H, Gallivan T, et al. Systems analysis of adverse drug events. ADE Prevention Study Group. J Am Med Assoc 1995; 274(1):35–43. [5] Aparasu R, Baer R, Aparasu A. Clinically important potential drug–drug interactions in outpatient settings. Res Social Adm Pharm 2007;3(4):426–37. [6] Tulner LR, Frankfort SV, Gijsen GJP, van Campen JPC, Koks CHW, Beijnen JH. Drug–drug interactions in a geriatric outpatient cohort: prevalence and relevance. Drugs Aging 2008;25(4):343–55. [7] Becker, Visser L, van Gelder T, Hofman A, Stricker B. Increasing exposure to drug–drug interactions between 1992 and 2005 in people aged ⬎ or ⫽ 55 years. Drugs Aging 2008; 25(2):145–52. [8] Haider S, Johnell K, Thorslund K, Fastbom J. Trends in polypharmacy and potential drug–drug interactions across educational groups in elderly patients in Sweden for the period 1992–2002. Int J Clin Pharmacol Ther 2007;45(12): 643–53.

Int J Psych Clin Pract Downloaded from informahealthcare.com by University of Waterloo on 10/27/14 For personal use only.

Drug interactions in inpatients on psychotropics [9] Bjerrum L, Andersen M, Petersen G, Kragstrup J. Exposure to potential drug interactions in primary health care. Scand J Prim Health Care 2003;21(3):153–8. [10] Johnell K, Klarin I. The relationship between number of drugs and potential drug–drug interactions in the elderly: a study of over 600,000 patients from the Swedish Prescribed Drug Register. Drug Saf 2007;30(10):911–8. [11] Johnell K, Klarin I. The relationship between number of drugs and potential drug–drug interactions in the elderly: a study of over 600,000 elderly patients from the Swedish Prescribed Drug Register. Drug Saf 2007;30(10):911–8. [12] Aronson JK, Henderson G, Webb DJ, Rawlins MD. A prescription for better prescribing. Br Med J 2006;333(7566): 459–60. [13] Rosenbaum J. Managing selective serotonin reuptaje inhibitordrug interactions in clinical practice. Clin Pharmacokinet 1995;29(Suppl 1):53–9. [14] Davies S, Eayrs S, Pratt P, Lennard M. Potential for drug interactions involving cytochrome P450 2D6 and 3A4 on general adult psychiatric and functional elderly psychiatric wards. Br J Clin Pharmacol 2004;57(4):464–72. [15] Astrand E, Astrand B, Antonov K, Petersson G. Potential drug interactions during a three-decade study period: a cross sectional study of a prescription register. Eur J Clin Pharmacol 2007;63:851–9. [16] Gagne J, Maio V, Rabinowitz C. Prevalence and predictors of potential drug–drug interactions in Regione EmiliaRomagna, Italy. J Clin Pharmacol Ther 2008;33:141–51. [17] Preskorn SH. Pharmacogenomics, informatics, and individual drug therapy in psychiatry: past, present and future. J Psychopharmacol 2006 Jul 1;20(Suppl 4):85–94. [18] Markowitz J, Devine C. Drug interaction potential of fluoxetine, sertraline and paroxetine in four state psychiatric hospital populations. Ther Drug Monit 1997;19(2):244–5. [19] Sakhuja D, Bisson J. Liaison psychiatry services in Wales. Psychiatr Bull 2008;32:134–6. [20] Gomez J. Liaison psychiatry. Mental health problems in the general hospital. London: Croom-Helm; 1987. [21] Peveler R, House A. Developing services in liaison psychiatry; making the case of need. Liaison psychiatry: Planning services for specialist settings. London: Gaskell; 2000. p. 1–13. [22] Torem M, Saravay S, Steinberg H. Psychiatric liaison: benefits of an “active” approach. Psychosomatics 1979;20(9):598–611. [23] Kornfeld D. Consultation-liaison psychiatry: Contributions to medical practice. Am J Psychiatry 2002;159(1964):1972. [24] Howe A, Hendry J, Potokar JPP. A survey of liaison psychiatry services in the south-west of England. Psychiatr Bull 2003;27:90–2. [25] Klimke A, Klieser R. Sudden death after intravenous application of lorazepam in a patient treated with clozapine. Am J Psychiatry 1994;151(5):780. [26] Jackson C, Markowitz J, Brewerton T. Delirium associated with clozapine and benzodiazepine combinations. Ann Clin Psychiatry 1995;7(3):139–41. [27] Lange-Asschenfeldt C, Weigmann H, Hiemke C, Mann K. Serotonin syndrome as a result of fluoxetine in a patient with tramadol abuse: Plasma level-correlated symptomatology. J Clin Psychopharmacol 2002;22(4):440–1. [28] Kesevan S, Sobala G. Serotonin syndrome with fluoxetine plus tramadol. J R Soc Med 1999;92(9):474–5. [29] Hersh EV, Pinto A, Moore PA. Adverse drug interactions involving common prescription and over-the-counter analgesic agents. Clin Ther 2007;29(11 Suppl 1):2477–97. [30] Hemeryck A, Belpaire F. Selective serotonin reuptake inhibitors and cytochrome P450 mediated drug–drug interactions: An update. Curr Drug Metab 2002;3(1):13–37.

219

[31] Dalma-Weiszhausz D, Murphy GJ. Single nucleotide polymorphisms and their characterisation with oligonucleotide microarrays. Psychiatr Genet 2002;12(2):97–107. [32] Humphries T, Carroll N, Chester E, Magid D, Rocho B. Evaluation of an electronic critical drug interactions program compared with active pharmacist intervention. Ann Pharmacother 2007;41(12):1979–85. [33] Blix H, Viktil K, Moger T, Reikvam A. Identification of drug interactions in hospitals-computerised screening vs. bedside recording. J Clin Pharmacol Ther 2008;33:131–9. [34] Egger T, Dormann H, Runge U, Neubert A, Criegge-Rieck M, Gassmann K, et al. Identification of adverse drug reactions in geriatric inpatients using a computerised drug database. Drugs Aging 2003;20(10):769–76. [35] Dallenbach M, Bovier P, Desmeules J. Detecting drug interactions using personal digital assistants in an out-patient clinic. Q J Med 2007;100:691–7. [36] Bergk V, Gasse C, Rothenbacher D, Loew M, Brenner H, Haefeli WE. Drug interactions in primary care: Impact of a new algorithm on risk determination. Clin Pharmacol Ther 2004;76(1):85–96. [37] Juurlink DN, Mamdani M, Kopp A, Laupacis A, Redelmeier DA. Drug–drug interactions among elderly patients hospitalized for drug toxicity. J Am Med Assoc 2003;289(13): 1652–8. [38] Preskorn S. Drug–drug Interactions: Proof of relevance (Part I). J Psychiatr Pract 2005;11(2):116–23. [39] DeVane C. Antidepressant-drug interactions are potentially but rarely clinically significant. Neuropsychopharmacology 2006;31(8):1594–604. [40] Davies SJC, Jackson PR, Ramsay LE, Ghahramani P. Drug intolerance due to nonspecific adverse effects related to psychiatric morbidity in hypertensive patients. Arch Intern Med 2003;163(5):592–600. [41] Brosen K, Naranjo C. Review of pharmacokinetic and pharmacodynamic interaction studies with citalopram. Eur Neuropsychopharmacol 2001;11:275–83. [42] Evans W, Relling M. Concordance of P450 2D6 (debrisoquine hydroxylase) phenotype and genotype: inability of dextromethorpan metabolic ratio to discriminate reliably heterozygous and homozygous extensive metabolisers. Pharmacogenetics 1991;1(3):143–8. [43] Masimirembwa C, Hasler J, Bertilssons L, Johansson I, Ekberg O, Ingelam-Sundberg M. Phenotype and genotype analysis of debrisoquine hydroxylase (CYP2D6) in a black Zimbabwean population. Reduced enzyme activity and evaluation of metabolic correlation of CYP2D6 probe drugs. Eur J Clin Pharmacol 1996;51(2):117–22. [44] Chou WH, Yan FX, Robbins-Weilert DK, Ryder TB, Liu WW, Perbost C, et al. Comparison of two CYP2D6 genotyping methods and assessment of genotype-phenotype relationships. Clin Chem 2003;49(4):542–51. [45] Broly F, Gaedigk A, Heim M, Eichelbaum M, Morike K, Meyer U. Debrisoquine/sparteine hydroxylation genotype and phenotype: analysis of common mutations and alleles of CYP2D6 in a European population. DNA Cell Biol 1991; 10(8):545–58. [46] Chou W, Yan F, de Leon J, Barnhill J, Rogers T, Cronin M, et al. Extension of a pilot study: impact from the cytochrome P450 2D6 polymorphism on outcome and costs associated with severe mental illness. Clin Psychopharmacol 2000;20(2): 246–51. [47] de Leon J, Barnhill J, Rogers T, Boyle J, Chou WH, Wedlund PJ. Pilot Study of the cytochrome P450-2D6 genotype in a psychiatric state hospital. Am J Psychiatry 1998;155(9): 1278–80.

Drug-drug interactions in general hospital and psychiatric hospital in-patients prescribed psychotropic medications.

Abstract Objectives. Drug-drug interactions (DDIs) present a serious, ever increasing clinical problem. Previous studies identified DDIs among psychia...
142KB Sizes 0 Downloads 4 Views