GPs' views on computerized drug interaction alerts: questionnaire survey

Ability of pharmacy clinical decision-support software to alert users about clinically important drug-drug interactions

OBJECTIVE

Pharmacy clinical decision-support (CDS) software that contains drug-drug interaction (DDI) information may augment pharmacists' ability to detect clinically significant interactions. However, studies indicate these systems may miss some important interactions. The purpose of this study was to assess the performance of pharmacy CDS programs to detect clinically important DDIs.

DESIGN

Researchers made on-site visits to 64 participating Arizona pharmacies between December 2008 and November 2009 to analyze the ability of pharmacy information systems and associated CDS to detect DDIs. Software evaluation was conducted to determine whether DDI alerts arose from prescription orders entered into the pharmacy computer systems for a standardized fictitious patient. The fictitious patient's orders consisted of 18 different medications including 19 drug pairs-13 of which were clinically significant DDIs, and six were non-interacting drug pairs.

MEASUREMENTS

The sensitivity, specificity, positive predictive value, negative predictive value, and percentage of correct responses were measured for each of the pharmacy CDS systems.

RESULTS

Only 18 (28%) of the 64 pharmacies correctly identified eligible interactions and non-interactions. The median percentage of correct DDI responses was 89% (range 47-100%) for participating pharmacies. The median sensitivity to detect well-established interactions was 0.85 (range 0.23-1.0); median specificity was 1.0 (range 0.83-1.0); median positive predictive value was 1.0 (range 0.88-1.0); and median negative predictive value was 0.75 (range 0.38-1.0).

CONCLUSIONS

These study results indicate that many pharmacy clinical decision-support systems perform less than optimally with respect to identifying well-known, clinically relevant interactions. Comprehensive system improvements regarding the manner in which pharmacy information systems identify potential DDIs are warranted.

[Drug-drug interactions in multicentre polypathological polymedicated patients]

OBJECTIVE

To determine the prevalence of relevant drug-drug interactions (DDIs) and associated predictor factors in a sample of patients with multiple complex chronic diseases (polypathological patients) receiving multiple drug therapy. Our secondary objective was to determine the acceptance of a drug interaction reporting program with recommendations addressed to the prescribing physicians.

SUBJECTS AND METHODS

A cross-sectional study performed in three primary care centres assigned to a teaching hospital. All patients with 2 or more chronic diseases and treated simultaneously with 5 or more drugs were recruited in the study. DDIs were detected by using Drug-Reax System((R)) (Micromedex) program, the Drug Data Base (Bot) Spanish General Council of Official Colleges of Pharmacists or literature search when needed. Those DDIs which, according to the opinion of the pharmacist investigators, required any intervention were considered relevant. Acceptance of the reported DDI recommendations was evaluated by means of a survey addressed by primary care physicians ("acceptable," pertinent recommendation to modify treatment).

RESULTS

A total of 283 polypathological polymedicated patients were included. Mean age was 74.5 years (range 43-100 years). Mean number of diseases per patient was 2.5 and prescriptions 9.7). Out of a total of 2748 drug prescriptions, 1053 DDIs in 250 patients (96.5%) were identified. Of these, 45% were filtered as relevant DDIs. The presence of ischemic heart disease, two or more hospital admissions and having received 7 or more prescriptions were associated with the presence of DDIs. 177 informs containing 473 recommendations about DDIs were sent to primary care physicians from our Pharmacy Department. 339 recommendations were answered by primary care physicians, and 84% were favourably accepted.

CONCLUSIONS

Almost every polypathological polymedicated patient is exposed to at least one DDI and about a 60% would require any intervention. Appropriate filtering and personalising recommendations in a collaborative way may represent an adequate manner to improve the risk-benefit ratio of the drug prescriptions.

Nationwide drug-dispensing data reveal important differences in adherence to drug label recommendations on CYP2D6-dependent drug interactions

AIMS

The study aimed to investigate the clinical adherence to drug label recommendations on important drug-drug interactions (DDIs). Dispensing data on drug combinations involving selective serotonin reuptake inhibitor (SSRI) antidepressants could help to identify areas for intensified medical education.

METHODS

This was a retrospective, cross-sectional analysis of individual dispensing data regarding all individuals > or =15 years old in Sweden. The study analysed the prescribing and dispensing of CYP2D6 drugs (metoprolol, donepezil, galantamine, codeine, tamoxifen) together with CYP2D6-blocking SSRIs (paroxetine/fluoxetine) or SSRIs without significant CYP2D6 inhibition (citalopram/escitalopram/sertraline), and the related prescribing of CYP2D6-independent comparator drugs (atenolol, rivastigmine, propoxyphene, anastrozole). Odds were calculated between each CYP2D6 drug and the corresponding comparator drug in patients on fluoxetine/paroxetine and citalopram/escitalopram/sertraline, respectively. The odds ratio (OR) was calculated by dividing the obtained odds in patients on fluoxetine/paroxetine by the corresponding odds in patients on citalopram/escitalopram/sertraline.

RESULTS

Compared with patients that were dispensed citalopram/escitalopram/sertraline, patients dispensed fluoxetine/paroxetine had lower prescribing rates of metoprolol (adjusted OR 0.80; 95% confidence interval 0.76, 0.85), donepezil (0.65; 0.49, 0.86) and galantamine (0.58; 0.41, 0.81). In contrast, the use of prodrugs codeine (compared woth propoxyphene) or tamoxifen (compared with anastrozole) was similar among patients on fluoxetine/paroxetine and citalopram/escitalopram/sertraline (adjusted OR 1.03; 0.94, 1.12 and 1.29; 0.96, 1.73, respectively).

CONCLUSIONS

Clinically important DDIs that are associated with impaired bioactivation of prodrugs might be more easily neglected in clinical practice compared with DDIs that cause drug accumulation and symptomatic adverse drug reactions.

The impact of traditional and smart pump infusion technology on nurse medication administration performance in a simulated inpatient unit

Objective

Assess the impact of infusion pump technologies (traditional pump vs smart pump vs smart pump with barcode) on nurses' ability to safely administer intravenous medications.

Design

Experimental study with a repeated measures design.

Setting

High-fidelity simulated inpatient unit.

Results

The nurses remedied 60% of “wrong drug” errors. This rate did not vary as a function of pump type. The nurses remedied “wrong patient” errors more often when using the barcode pump (88%) than when using the traditional pump (46%) or the smart pump (58%) (Cochran Q=14.36; p<0.05). The number of nurses who remedied “wrong dose hard limit” errors was higher when using the smart pump (75%) and the barcode pump (79%) than when using the traditional pump (38%) (Cochran Q=12.13; p<0.003). Conversely, there was no difference in remediation of “wrong dose soft limit” errors across pump types. The nurses' pump programming was less accurate when mathematical conversions were required. Success rates on secondary infusions were low (55.6%) and did not vary as a function of pump type.

Conclusions

These findings indicate that soft (changeable) limits in smart infusion pumps had no significant effect in preventing dosing errors. Provided that smart pumps are programmed with hard (unchangeable) limits, they can prevent dosing errors, thereby increasing patient safety. Until barcode pumps are integrated with other systems within the medication administration process, their role in enhancing patient safety will be limited. Further improvements to pump technologies are needed to mitigate risks associated with intravenous infusions, particularly secondary infusions.

Computerized clinical decision support for prescribing: provision does not guarantee uptake

There is wide variability in the use and adoption of recommendations generated by computerized clinical decision support systems (CDSSs) despite the benefits they may bring to clinical practice. We conducted a systematic review to explore the barriers to, and facilitators of, CDSS uptake by physicians to guide prescribing decisions. We identified 58 studies by searching electronic databases (1990-2007). Factors impacting on CDSS use included: the availability of hardware, technical support and training; integration of the system into workflows; and the relevance and timeliness of the clinical messages. Further, systems that were endorsed by colleagues, minimized perceived threats to professional autonomy, and did not compromise doctor-patient interactions were accepted by users. Despite advances in technology and CDSS sophistication, most factors were consistently reported over time and across ambulatory and institutional settings. Such factors must be addressed when deploying CDSSs so that improvements in uptake, practice and patient outcomes may be achieved.

An evidence-based assessment of the clinical significance of drug-drug interactions between disease-modifying antirheumatic drugs and non-antirheumatic drugs according to rheumatologists and pharmacists

BACKGROUND

Clinically relevant drug-drug interactions (DDIs) must be recognized in a timely manner and managed appropriately to prevent adverse drug reactions or therapeutic failure. Because the evidence for most DDIs is based on case reports or poorly documented clinical information, there is a need for better assessment of their clinical relevance.

OBJECTIVE

This study evaluates the interdisciplinary agreement between rheumatologists and clinical (hospital) pharmacists in assessing the clinical relevance of DDIs with disease-modifying antirheumatic drugs (DMARDs) and non-DMARD medications.

METHODS

Potential DDIs were identified from the medical literature using MEDLINE and EMBASE for the years 1968-2009. The following search terms were used for the key word, title, and abstract sections of the publications: interaction(s), DMARD, disease-modifying antirheumatic drug(s), antirheumatic, rheumatology, rheumatoid arthritis, and the names of the individual DMARDs of interest (abatacept, adalimumab, anakinra, auranofin, aurothioglucose, aurothiomalate, d-penicillamine, etanercept, gold, [hydroxy]-chloroquine, interleukin-1 receptor antagonist, IL1-RA, infliximab, leflunomide, methotrexate, rituximab, and sulfasalazine/sulphasalazine). Reference lists of the retrieved publications were searched for further information on potential DDIs. All pharmacodynamic or pharmacokinetic DDIs between a DMARD and a non-DMARD identified were included in the study, with the exception of evidence regarding DMARD doses higher than used in the treatment of rheumatoid arthritis and interactions with phytotherapeutic or homeopathic preparations. Using a standard information set for each DDI (eg, from product labeling, textbooks, and the medical literature), a group of rheumatologists and a group of clinical pharmacists independently assessed whether the individual drug-DMARD combinations interacted and whether they required immediate intervention. Both groups consisted of 3 members (2 men and 1 woman), aged 40 to 60 years, who had >5 years of clinical experience and were currently involved in clinical practice in large, nonacademic teaching hospitals in the Netherlands.

RESULTS

Forty potential DDIs with DMARDs were retrieved and assessed by the 2 groups. For 30 (75%) of these, rheumatologists and clinical pharmacists agreed about the requirement for immediate intervention. Specifically, 17 drug combinations (43%) were judged to interact and to require immediate intervention, and 13 combinations (33%) were judged either not to interact or to interact but not to require immediate intervention. For 10 combinations (25%), rheumatologists and clinical pharmacists were not in agreement. Overall, agreement between the groups was good (kappa = 0.80) for judging whether the drug combinations were interactions, and agreement was fair (kappa = 0.39) for judging whether immediate intervention was required. Prospective analysis of the data showed that rheumatologists tended to recommend immediate intervention more often when the adverse reaction to the DDI involved an increased risk of toxicity of the DMARD. In contrast, clinical pharmacists more often advocated immediate intervention when the adverse reaction involved decreased effectiveness of the DMARD.

CONCLUSION

For a subset of DMARD-drug combinations, rheumatologists and clinical pharmacists differed in their assessments of clinical relevance.

Timely follow-up of abnormal diagnostic imaging test results in an outpatient setting: are electronic medical records achieving their potential?

BACKGROUND

Given the fragmentation of outpatient care, timely follow-up of abnormal diagnostic imaging results remains a challenge. We hypothesized that an electronic medical record (EMR) that facilitates the transmission and availability of critical imaging results through either automated notification (alerting) or direct access to the primary report would eliminate this problem.

METHODS

We studied critical imaging alert notifications in the outpatient setting of a tertiary care Department of Veterans Affairs facility from November 2007 to June 2008. Tracking software determined whether the alert was acknowledged (ie, health care practitioner/provider [HCP] opened the message for viewing) within 2 weeks of transmission; acknowledged alerts were considered read. We reviewed medical records and contacted HCPs to determine timely follow-up actions (eg, ordering a follow-up test or consultation) within 4 weeks of transmission. Multivariable logistic regression models accounting for clustering effect by HCPs analyzed predictors for 2 outcomes: lack of acknowledgment and lack of timely follow-up.

RESULTS

Of 123 638 studies (including radiographs, computed tomographic scans, ultrasonograms, magnetic resonance images, and mammograms), 1196 images (0.97%) generated alerts; 217 (18.1%) of these were unacknowledged. Alerts had a higher risk of being unacknowledged when the ordering HCPs were trainees (odds ratio [OR], 5.58; 95% confidence interval [CI], 2.86-10.89) and when dual-alert (>1 HCP alerted) as opposed to single-alert communication was used (OR, 2.02; 95% CI, 1.22-3.36). Timely follow-up was lacking in 92 (7.7% of all alerts) and was similar for acknowledged and unacknowledged alerts (7.3% vs 9.7%; P = .22). Risk for lack of timely follow-up was higher with dual-alert communication (OR, 1.99; 95% CI, 1.06-3.48) but lower when additional verbal communication was used by the radiologist (OR, 0.12; 95% CI, 0.04-0.38). Nearly all abnormal results lacking timely follow-up at 4 weeks were eventually found to have measurable clinical impact in terms of further diagnostic testing or treatment.

CONCLUSIONS

Critical imaging results may not receive timely follow-up actions even when HCPs receive and read results in an advanced, integrated electronic medical record system. A multidisciplinary approach is needed to improve patient safety in this area.

A drug interactions elective course

OBJECTIVES

To evaluate the impact of a drug interactions elective course on student knowledge and skills.

DESIGN

A drug interactions elective which focused on assessment and application of drug interaction information and identification and management of commonly encountered drug interactions by therapeutic category was offered to third-year PharmD students. Students were expected to (1) determine whether a given interaction was clinically significant or required pharmacist intervention, and (2) make rational, scientifically sound, practical recommendations for management of drug interactions. EVALUATION AND ASSESSMENT: Assessment included course evaluations, student self-assessments, and knowledge and skills assessments. Students who completed the course were more confident in their abilities relating to drug interactions than students who did not complete the course. Students who completed the course scored significantly better in all areas of the assessment compared to students who did not complete the course. Course evaluation results were also positive.

CONCLUSION

A course devoted to the identification and management of drug interactions improved PharmD students' knowledge and skills and could potentially improve the patient care they provide in the future.

What evidence supports the use of computerized alerts and prompts to improve clinicians' prescribing behavior?

Alerts and prompts represent promising types of decision support in electronic prescribing to tackle inadequacies in prescribing. A systematic review was conducted to evaluate the efficacy of computerized drug alerts and prompts searching EMBASE, CINHAL, MEDLINE, and PsychINFO up to May 2007. Studies assessing the impact of electronic alerts and prompts on clinicians' prescribing behavior were selected and categorized by decision support type. Most alerts and prompts (23 out of 27) demonstrated benefit in improving prescribing behavior and/or reducing error rates. The impact appeared to vary based on the type of decision support. Some of these alerts (n = 5) reported a positive impact on clinical and health service management outcomes. For many categories of reminders, the number of studies was very small and few data were available from the outpatient setting. None of the studies evaluated features that might make alerts and prompts more effective. Details of an updated search run in Jan 2009 are included in the supplement section of this review.

Assessing the value of electronic prescribing in ambulatory care: a focus group study

INTRODUCTION

Drug allergy and interaction alerts are a core function of most electronic prescribing (e-prescribing) systems. To characterize the value of e-prescribing and medication safety alerts, especially in small and medium-size practices, we undertook a hypothesis-generating focus group study of Massachusetts clinicians. We sought to understand the reasons for adoption and use of e-prescribing, as well as clinicians' complaints about and perceived benefits of drug allergy and interaction alerts.

METHODS

We recruited 25 Massachusetts clinicians to participate in three focus groups regarding the use and value of e-prescribing and medication safety alerts. The participants included high-volume (>100 electronic scripts per month) physicians, nurse practitioners, and practice assistants in family practice, internal medicine, pediatrics, and subspecialty practices who used a common commercial e-prescribing system.

RESULTS

Most clinicians were in small and medium-size group practices. Participants were, on average, 25 years post-medical school graduation (range 8-36), had used e-prescribing for 2.5 years (range 1.0-5.5), and wrote the majority (89%, range 15-100%) of prescriptions electronically. The participants' decision to adopt e-prescribing was driven largely by financial incentives offered by insurers, and was viewed as a step toward implementation of an electronic medical record. Although participants agreed that the system was easy to learn, few anticipated efficiencies were realized until clinicians configured the device to meet their needs. The participants were ambivalent about whether e-prescribing improved their own or overall office efficiency, and--in the absence of payer incentives--few were willing to pay for the systems out-of-pocket. The most valuable aspects of e-prescribing were the ease of changing doses, renewing prescriptions, ensuring legibility, and transmitting prescriptions to in- and out-of-state pharmacies. Participants were dissatisfied with the unreliability of transmitting prescriptions successfully to the pharmacy, and with their inability to merge duplicate patient entries, to create a comprehensive, allprescriber medication list, to write prescriptions for commonly ordered medications and supplies, and to enter allergy information into the system. Participants were critical of the volume of drug allergy and interaction alerts. Many alerts were of trivial clinical significance or were generated by interactions with out-of-date medications. As a result, many clinicians habitually ignored these alerts. Alerts were most helpful to clinicians who were unfamiliar with a particular drug or patient. Although alerts rarely led the clinicians to abort or alter a prescription, alerts did prompt clinicians to counsel patients about medication side effects, to educate themselves about potential interactions, to check physical examination findings, or to order laboratory tests. Despite problems, few clinicians were willing to forego receiving alerts for fear that they would miss a potentially dangerous drug interaction.

CONCLUSION

Electronic prescribing is a potential boon to ambulatory medical practice, although its value may be compromised by inappropriate and irrelevant medication safety alerts and by features of the e-prescribing system that prove burdensome to frontline clinicians. While alerts infrequently result in changed or aborted prescriptions, they may trigger a variety of other provider behaviors that help to ensure safe care.

Prescribers' knowledge of and sources of information for potential drug-drug interactions: a postal survey of US prescribers

BACKGROUND

Given the high prevalence of medication use in the US, the risk of drug-drug interactions (DDIs) and potential for patient harm is of concern. Despite the rise in technologies to identify potential DDIs, the ability of physicians and other prescribers to recognize potential DDIs is essential to reduce their occurrence. The objectives of this study were to assess prescribers' ability to recognize potential clinically significant DDIs and to examine the sources of information they use to identify potential DDIs and prescribers' opinions on the usefulness of various DDI information sources.

METHODS

A postal questionnaire was developed to assess prescriber knowledge of medications that may interact and prescribers' usual sources of DDI information. Recipients were asked to classify 14 drug pairs as 'contraindicated', 'may be used together but with monitoring' or 'no interaction'. A response option of 'not sure' was also provided. The questionnaires were sent to a national sample of 12 500 prescribers based on past history of prescribing drugs associated with known potential for DDI, who were identified using data from a pharmacy benefit manager covering over 50 million individuals.

RESULTS

Usable questionnaires were obtained from 950 prescribers. The percentage of prescribers who correctly classified specific drug pairs ranged from 18.2% for warfarin and cimetidine to 81.2% for paracetamol (acetaminophen) with codeine and amoxicillin, with 42.7% of all combinations classified correctly. The number of drug pairs correctly classified by the prescribers ranged from 0 to 13. For half of the drug pairs over one-third of the respondents answered 'not sure'; among those drug pairs, two were contraindicated. When asked what source was used to learn more about a potential DDI, a quarter of the prescribers reported using personal digital assistants and another quarter used printed material. The majority of the prescribers (68.4%) reported that they were usually informed by pharmacists about their patients' potential exposure to DDIs. Compared with the prescribers who used other sources, those who used computerized DDI alerts as their usual source of DDI information consistently gave a lower rating score to the five statements that assessed the usefulness of the information.

CONCLUSION

This study suggests that prescribers' knowledge of potential clinically significant DDIs is generally poor. These findings are supported by other research and emphasize the need to develop systems that alert prescribers about potential interactions that are clinically relevant. Physicians most commonly reported learning about potential DDIs from pharmacists, suggesting further work is needed to improve the drug-prescribing process to identify potential safety issues earlier in the medication use process.

Analysis of overridden alerts in a drug-drug interaction detection system

OBJECTIVE

The aim of this study was to evaluate the relevance of the signals generated by a computerized drug-drug interaction detection system and to design a classification of overridden drug-drug interaction alerts.

STUDY DESIGN

Prospective study over two months.

SETTING

Five hundred and ten-bed university paediatric hospital.

MAIN OUTCOME MEASURES

In Robert Debré Hospital physicians generate drug orders online using a computerized physician order entry system that also detects drug-drug interactions in real time. We analysed the relevance of a sample of alerts overridden by physicians.

RESULTS

We analysed a sample of 613 overridden alerts. We defined three categories of overridden alerts: informational errors (35); system errors (244) and accurate alerts (334). Two reasons account for 40% of false-positive alerts: an inability of the system to recognize real conflicts between drug treatments and guidelines stating that the two drugs can be used together, because the benefit outweighs the risk of side effects due to the drug-drug interaction.

CONCLUSIONS

We created a classification of overridden alerts, in the context of computerized physician order entry system coupled with a drug-drug interaction detection system. There is clearly room for improvement in the development of drug-drug interaction software. This classification should make it possible to break this work down into smaller tasks, making it possible to decrease the sensitivity to background noise of drug-drug interaction detection systems.

A clinical evaluation of the Janus Web Application, a software screening tool for drug-drug interactions

PURPOSE

To evaluate the clinical relevance of the Janus Web Application (JWA) in screening for potential drug-drug interactions (DDIs).

METHODS

One hundred and fifty patients taking two drugs or more were studied. Potential DDIs were identified by the JWA. Interviewing the patient and looking into his/her medical records provided complementing information. A clinical pharmacologist judged which potential DDIs were clinically relevant. Potentially relevant DDIs identified by the JWA were then correlated with clinically relevant DDIs.

RESULTS

A total of 150 significant potential DDIs were found. Sixteen percent (24/150) were judged to be clinically relevant.

CONCLUSIONS

A very small proportion of DDIs was considered clinically relevant in the specific clinical context. To optimise the software's user-friendliness, the following points need to be considered: the possibility of eliminating trivial potential DDIs, individualising drug alerts, and providing written information, accessible via a hyperlink.

A randomized trial of the effectiveness of on-demand versus computer-triggered drug decision support in primary care

OBJECTIVES

Prescribing alerts generated by computerized drug decision support (CDDS) may prevent drug-related morbidity. However, the vast majority of alerts are ignored because of clinical irrelevance. The ability to customize commercial alert systems should improve physician acceptance because the physician can select the circumstances and types of drug alerts that are viewed. We tested the effectiveness of two approaches to medication alert customization to reduce prevalence of prescribing problems: on-physician-demand versus computer-triggered decision support. Physicians in each study condition were able to preset levels that triggered alerts.

DESIGN

This was a cluster trial with 28 primary care physicians randomized to either automated or on-demand CDDS in the MOXXI drug management system for 3,449 of their patients seen over the next 6 months.

MEASUREMENTS

The CDDS generated alerts for prescribing problems that could be customized by severity level. Prescribing problems included dosing errors, drug-drug, age, allergy, and disease interactions. Physicians randomized to on-demand activated the drug review when they considered it clinically relevant, whereas physicians randomized to computer-triggered decision support viewed all alerts for electronic prescriptions in accordance with the severity level they selected for both prevalent and incident problems. Data from administrative claims and MOXXI were used to measure the difference in the prevalence of prescribing problems at the end of follow-up.

RESULTS

During follow-up, 50% of the physicians receiving computer-triggered alerts modified the alert threshold (n = 7), and 21% of the physicians in the alert-on-demand group modified the alert level (n = 3). In the on-demand group 4,445 prescribing problems were identified, 41 (0.9%) were seen by requested drug review, and in 31 problems (75.6%) the prescription was revised. In comparison, 668 (10.3%) of the 6,505 prescribing problems in the computer-triggered group were seen, and 81 (12.1%) were revised. The majority of alerts were ignored because the benefit was judged greater than the risk, the interaction was known, or the interaction was considered clinically not important (computer-triggered: 75.8% of 585 ignored alerts; on-demand: 90% of 10 ignored alerts). At the end of follow-up, there was a significant reduction in therapeutic duplication problems in the computer-triggered group (odds ratio 0.55; p = 0.02) but no difference in the overall prevalence of prescribing problems.

CONCLUSION

Customization of computer-triggered alert systems is more useful in detecting and resolving prescribing problems than on-demand review, but neither approach was effective in reducing prescribing problems. New strategies are needed to maximize the use of drug decision support systems to reduce drug-related morbidity.

Evaluation of frequently used drug interaction screening programs

OBJECTIVE

Drug-drug interaction (DDI) screening programs are an important tool to check prescriptions of multiple drugs. The objective of the current study was to critically appraise several DDI screening programs.

METHODS

A DDI screening program had to fulfil minimal requirements (information on effect, severity rating, clinical management, mechanism and literature) to be included into the final evaluation. The 100 most frequently used drugs in the State Hospital of Baden, Switzerland, were used to test the comprehensiveness of the programs. Qualitative criteria were used for the assessment of the DDI monographs. In a precision analysis, 30 drugs with and 30 drugs without DDIs of clinical importance were tested. In addition, 16 medical patient profiles were checked for DDIs, using Stockley's Drug Interactions as a reference.

MAIN OUTCOME MEASURE

Suitability of DDI screening program (quality of monographs, comprehensiveness of drug list, statistical evaluation).

RESULTS

Out of nine programs included, the following four fulfilled the above mentioned criteria: Drug Interaction Facts, Drug-Reax, Lexi-Interact and Pharmavista. Drug Interaction Facts contained the smallest number of drugs and was therefore the least qualified program. Lexi-Interact condenses many DDIs into one group, resulting in less specific information. Pharmavista and Drug-Reax offer excellent DDI monographs. In the precision analysis, Lexi-Interact showed the best sensitivity (1.00), followed by Drug-Reax and Pharmavista (0.83 each) and Drug Interaction Facts (0.63). The analysis of patient profiles revealed that out of 157 DDIs found by all programs, only 18 (11%) were detected by all of them. No program found more than 50% of the total number of DDIs. A further evaluation using Stockley's Drug interactions as the gold standard revealed that Pharmavista achieved a sensitivity of 0.86 (vs Drug Interaction Facts, Lexi-Interact and Drug-Reax with a sensitivity of 0.71 each) and a positive predictive value of 0.67.

CONCLUSION

None of the four DDI screening programs tested is ideal, every program has its strengths and weaknesses, which are important to know. Pharmavista offers the highest sensitivity of the programs evaluated with a specificity and positive predictive value in an acceptable range.

Identification of drug interactions in hospitals--computerized screening vs. bedside recording

BACKGROUND AND OBJECTIVE

Managing drug interactions in hospitalized patients is important and challenging. The objective of the study was to compare two methods for identification of drug interactions (DDIs)--computerized screening and prospective bedside recording--with regard to capability of identifying DDIs.

METHODS

Patient characteristics were recorded for patients admitted to five hospitals. By bedside evaluation drug-related problems, including DDIs, were prospectively recorded by pharmacists and discussed in multidisciplinary teams. A computer screening programme was used to identify DDIs retrospectively--dividing DDIs into four classes: A, avoid; B, avoid/take precautions; C, take precautions; D, no action needed.

RESULTS

Among 827 patients, computer screening identified DDIs in 544 patients (66%); 351 had DDIs introduced in hospital. The 1513 computer-identified DDIs had the following distribution: type A 78; type B 915; type C 38; type D 482. By bedside evaluation, 99 DDIs were identified in 73 patients (9%). The proportions of computer recorded DDIs which were also identified at the bedside were: 5%, 8%, 8%, 2% DDIs of types A, B, C and D respectively. In 10 patients, DDIs not registered by computer screening were identified by bedside evaluation. The drugs most frequently involved in DDIs, identified by computerized screening were acetylsalicylic acid, warfarin, furosemide and digitoxin compared with warfarin, simvastatin, theophylline and carbamazepine, by bedside evaluation.

CONCLUSION

Despite an active prospective bedside search for DDIs, this approach identified less than one in 10 of the DDIs recorded by computer screening, including those regarded as hazardous. However, computer screening overestimates considerably when the objective is to identify clinically relevant DDIs.

Drug-drug interaction programs in clinical practice

There are multiple publications and electronic programs available to aid clinicians in the management of drug-drug interactions; however, clinicians should understand the advantages and limitations of these programs before using them in clinical practice. Because time constraints can make evaluation of the primary literature challenging, clinicians and health-care providers frequently rely on drug-interaction computer programs that are integrated in the prescribing process to flag drug interactions.

The impact of clinical pharmacists on drug-related problems and clinical outcomes

Drug-related problems are frequent and may result in reduced quality of life, and even morbidity and mortality. Many studies have shown that clinical pharmacists can effectively identify and prevent clinically significant drug-related problems and that physicians acknowledge and act on the clinical pharmacist's suggestions for interventions to the drug-related problems. A pro-active rather than a reactive approach on the part of the pharmacists seems prudent for obtaining most benefit. This includes participation of pharmacists in the multidisciplinary team discussions - at the stage of ordering and prescribing - where all types of drug-related problems, including also potential problems, should be discussed. In addition, counselling by pharmacists about medication on discharge and follow-up after discharge resulted in better outcomes. Furthermore, clinical pharmacists can positively influence other outcomes, such as improvement of levels of markers for drug use (e.g. optimization of lipid levels, anticoagulation levels and blood pressure). Some studies have reported positive effects on hard clinical outcomes, such as reduced length of stay, fewer re-admissions and fewer disease events (e.g. heart failure events and thromboembolism). However, more studies should be undertaken with larger patient populations, including patients from multiple sites. More knowledge about patient-specific factors that predict improved care is also needed. In conclusion, there is increasing evidence that participation and interventions of clinical pharmacists in health care positively influence clinical practice.

[Prospective assessment of drug interactions in hospitalized patients using a computer programme]

OBJECTIVE

To describe the implementation, functioning and results of a prospective automated system monitoring clinically relevant interactions in hospitalised patients in a 400-bed hospital for the period between 1 January 2005 and 31 March 2006.

METHOD

We created a computer programme in Access(R) 97 that checks, twice daily, the drug treatments of all of the patients admitted to the hospital in order to search for the 198 pairs of drugs previously selected from: a validated tertiary source (Hansten PD, Horn JR. Hansten and Horn's Drug Interactions Analysis and Management. St. Louis, MO: Facts and Comparisons; 2001 and updates), most relevant primary sources, expert opinions and alerts from the Spanish Agency of Medicines and Health Products. The clinical pharmacist will assess the drug-drug interaction (DDI) taking into account the timeline sequence, dose, administration route, management opportunities, patient diagnosis, clinical relevance, etc. If necessary, the doctor is contacted by phone and/or letter to inform him/her of the type of interaction, the mechanism and possible management. The programme files the following variables every day: interaction, sex, age, service, number of drugs, pharmaceutical intervention and doctor response.

RESULTS

Clinically relevant drug interactions were detected in 3% of patients during their stay in hospital. These patients were an average of 10 years older and received an average of four drugs more than other patients. A total of 538 interactions were detected in 15 months. Forty-three of 198 possible DDls appeared at some time. The pharmacist intervened on 126 (23%) occasions. The doctor accepted the recommendation on at least 66 (52%) occasions. Fourteen drugs were responsible for 91% of the interactions reported.

CONCLUSIONS

The patients with interactions are older and receive more drugs. The prior intervention of the pharmacist eliminated 77% of unnecessary alerts.

An electronic prompt in dispensing software to promote clinical interventions by community pharmacists: a randomized controlled trial

What is already known about this subject. Computerized prompts and reminders have been shown to be effective in changing the behaviour of health professionals in a variety of settings. There is little literature describing or evaluating electronic decision-support for pharmacists. What this study adds. An electronic prompt in dispensing software for a targeted clinical intervention has a significant effect on pharmacists' behaviour. A markedly increased rate of recording and performing the targeted clinical intervention was found. The effect of the prompt reduces markedly once the prompt is deactivated.

AIM

To evaluate the effect of an electronic prompt in dispensing software on the frequency of clinical interventions recorded by community pharmacists.

METHOD

An electronic decision-support prompt identifying patients for a targeted proactive clinical intervention was developed and implemented. Each time an oral antidiabetic agent was dispensed, a prompt was displayed reminding pharmacists to discuss the suitability of aspirin therapy in eligible patients with diabetes. The prompt was randomly assigned to 31 of 52 metropolitan pharmacies in Melbourne (Australia) for 6 weeks, with the remaining pharmacies as controls.

RESULTS

One hundred and fifty pharmacists in 52 pharmacies recorded a total of 2396 clinical interventions at an intervention rate of 0.92 interventions per 100 patients [95% confidence interval (CI) 0.58, 1.23]. Pharmacists recorded a total of 201 target interventions related to aspirin therapy in diabetes at an intervention rate of 2.55 interventions per 100 diabetic patients (95% CI 0.85, 4.24). All of the targeted clinical interventions were recorded in the prompt arm; no targeted interventions were recorded in the control group. The effect of the prompt decreased over the study period and was not maintained after prompt deactivation.

CONCLUSION

An electronic prompt significantly increased pharmacists' recording of the targeted clinical intervention in diabetic patients. An electronic decision-support prompt has significant potential to promote community pharmacists' contribution to the quality use of medicines.

Management of drug-interaction alerts in community pharmacies

BACKGROUND AND OBJECTIVE

Drug-interaction alert systems are commonly used in community pharmacies to identify potential drug-drug interactions. However, depending on the software default setting, pharmacists may override alerts because they are too numerous. We explored the handling of drug-interaction alerts by community pharmacies in Switzerland.

METHODS

Data were collected by 15 trained pharmacy students in 15 Swiss community pharmacies. The medication history and the drug-interaction alerts of 600 patients who had >or=2 drugs on prescription were assessed, and the pharmacists in charge were interviewed about their management of drug-interaction alerts.

RESULTS

In the 15 pharmacies studied, the computer systems were programmed to flag only 'severe' drug interactions in four, 'severe or moderate' in six or 'severe, moderate or minor' in five pharmacies. The median frequency of drug-interaction alerts increased with decreasing default severity level from 0.5 to 40, respectively, to 76 per 40 patient visits and pharmacy. Because of these default settings, 277 (35 x 2%) of 787 potential drug-interaction alerts on new or repeated prescriptions were overridden by the computer systems. Only 256 (32 x 5%) of 787 potential drug interactions emerged from new prescriptions. The alert systems produced 656 alerts of which 146 were irrelevant because of multiple alerting of the same interaction or of drug combinations currently no longer taken. Of the 510 remaining relevant drug-interaction alerts, 289 (56 x 7%) were overridden by community pharmacists without any action taken. If the pharmacist took care of a patient's prescription him- or herself (as opposed to just controlling a prescription after a technician took care of the patient), fewer drug-interaction alerts were overridden by the pharmacist [Odds ratio (OR) 0 x 6, 95% confidence interval (CI) 0 x 42-0 x 98; P=0 x 042). Technical overrides (by default settings) and pharmacists' overrides together accounted for 71 x 9% (566 of 787 potential drug interactions). Of the remaining 211 interactions alerts, 87 (41 x 2%) were checked more closely by consulting the literature, contacting the prescribing physician or discussion with the patient. This led to 55 (63 x 2%) interventions (close monitoring, adjustment of dose or ingestion time, therapy stop or switching to alternative therapy). Determinants associated with action taken after an interaction alert were potential high severity (severe or moderate) (OR 3 x 34, 95% CI 1 x 77-6 x 31; P<0 x 001) and alert flagged for the first time (OR 3 x 76, 95% CI 1 x 98-7 x 14; P<0 x 001). All severe potential drug interactions (n=10) generated an alert and all caused an intervention.

CONCLUSIONS

Pharmacists override a substantial proportion of drug-interaction alerts of minor or moderate potential severity by ignoring them or by programming the system to only flag drug interactions of potentially high severity. More sophisticated systems with improved sensitivity and specificity are required. Until these become available, it is important to ensure that at least potentially severe drug interactions are not missed; a goal that seems to be largely achieved.

Evaluation of outpatient computerized physician medication order entry systems: a systematic review

This paper provides a systematic literature review of CPOE evaluation studies in the outpatient setting on: safety; cost and efficiency; adherence to guideline; alerts; time; and satisfaction, usage, and usability. Thirty articles with original data (randomized clinical trial, non-randomized clinical trial, or observational study designs) met the inclusion criteria. Only four studies assessed the effect of CPOE on safety. The effect was not significant on the number of adverse drug events. Only one study showed a significant reduction of the number of medication errors. Three studies showed significant reductions in medication costs; five other studies could not support this. Most studies on adherence to guidelines showed a significant positive effect. The relatively small number of evaluation studies published to date do not provide adequate evidence that CPOE systems enhance safety and reduce cost in the outpatient settings. There is however evidence for (a) increasing adherence to guidelines, (b) increasing total prescribing time, and (c) high frequency of ignored alerts.

Patient safety features of clinical computer systems: questionnaire survey of GP views

AIM

To investigate general practitioners' (GPs') stated knowledge, use and training needs related to the patient safety features of computerised clinical systems in England.

DESIGN

Questionnaire survey.

SUBJECTS AND SETTING

GPs from six English primary care trusts.

OUTCOME MEASURES

GPs' views on the importance of specified patient safety features on their computer system; their knowledge of the presence of specified safety features; previous training and perceived future training needs.

RESULTS

Three hundred and eighty one GPs (64.0%) completed and returned the questionnaire. Although patient safety features were considered to be an important part of their computer system by the vast majority of GPs, many were unsure as to whether the system they were currently using possessed some of the specified features. Some respondents erroneously believed that their computers would warn them about potential contraindications or if an abnormal dose frequency had been prescribed. Only a minority had received formal training on the use of their system's patient safety features.

CONCLUSIONS

Patient safety was an issue high on the agenda of this GP sample. The importance of raising GPs' awareness of both the potential use and deficiencies of the patient safety features on their systems and ensuring that appropriate training is available should not be underestimated.

Clinical relevance of drug-drug interactions : a structured assessment procedure

INTRODUCTION

Computerised drug interaction surveillance systems (CIS) may be helpful in detecting clinically significant drug interactions. Experience with CIS reveals that they often yield alerts with questionable clinical significance, fail to provide relevant information on risk factors for the adverse reaction of the interaction and fail to detect all significant drug interactions. These problems highlight the importance of transparency and selectivity in choosing the drug interactions to be included in CIS. In The Netherlands, the Working Group on Pharmacotherapy and Drug Information is responsible for maintenance of the CIS of the Royal Dutch Association for the Advancement of Pharmacy (KNMP).

METHODS

The Working Group developed an evidence-based procedure for structured assessment of drug-drug interactions and revised all drug interactions in the CIS accordingly.

RESULTS

For every drug interaction four core parameters were assessed: (i) evidence on the interaction; (ii) clinical relevance of the potential adverse reaction resulting from the interaction; (iii) risk factors identifying patient, medication or disease characteristics for which the interaction is of special importance; and (iv) the incidence of the adverse reaction. On the basis of this assessment the drug-drug interactions for inclusion in the CIS were selected. After revision of the drug combinations in the KNMP-CIS, the Working Group judged 22% of the combinations to be not interacting and another 12% to be interacting but not requiring action. On the basis of this assessment the subset of drug combinations for which interaction alerts are generated and the information on management of a drug interaction alert for users of the CIS were adapted. When an alert is generated by the CIS, the user of the system is supplied with comprehensive information on the four core parameters, the mechanism of the interaction and critical information for management of the interaction for the individual patient.

DISCUSSION

This structured procedure offers the possibility for transparent and reproducible assessment of the clinical relevance of drug interactions.

CONCLUSION

A CIS selectively generating interaction alerts based on this assessment may help in realising the goal of good clinical practice and may offer a methodology to further increase drug safety.

Management of potential drug interactions in community pharmacies: a questionnaire-based survey in Switzerland

OBJECTIVE

To analyze the current drug-interaction management in Swiss community pharmacies, with a particular focus on electronic systems, and to compare the results with those expressed by German general practitioners in a recent survey.

METHODS

Data were collected with a postal questionnaire which was randomly sent to 500 out of 833 community pharmacies in the German part of Switzerland.

RESULTS

The response rate was 57.4%, and only 24.7% pharmacists reported that they were confronted less than daily with potential drug interactions. Use of computer software to identify potential drug interactions was widespread in community pharmacies (90.2%), and the software was the primary source of information (81.2 +/- 29.6%). The quality of the interaction software was judged sensitive (identifying all dangerous interactions) by 80.5 +/- 21.5%, but specific (identifying only relevant interactions) by only 38.3 +/- 32%. Pharmacists declared a low override rate (14%) of drug interaction alerts, although unjustified alerts were reported by 60.6 +/- 33.1%. In contrast to general practitioners, pharmacists opted less often for information on the mechanism of the interaction, and more frequently for details for dose adjustment. Both groups complained about deficient information on non-interacting alternatives.

CONCLUSION

The information needs of community pharmacists differed considerably from general practitioners, and pharmacists were overwhelmed by inappropriate alerts because of a lack of specificity of their drug-interaction systems. Substantial improvement of drug-interaction software systems is thus required at least in two important aspects, the suppression of inappropriate alerts and the tailoring to the needs of the user.

Practitioners' views on computerized drug-drug interaction alerts in the VA system

OBJECTIVES

To assess Veterans Affairs (VA) prescribers' and pharmacists' opinions about computer-generated drug-drug interaction (DDI) alerts and obtain suggestions for improving DDI alerts.

DESIGN

A mail survey of 725 prescribers and 142 pharmacists from seven VA medical centers across the United States.

MEASUREMENTS

A questionnaire asked respondents about their sources of drug and DDI information, satisfaction with the combined inpatient and outpatient computerized prescriber order entry (CPOE) system, attitude toward DDI alerts, and suggestions for improving DDI alerts.

RESULTS

The overall response rate was 40% (prescribers: 36%; pharmacists: 59%). Both prescribers and pharmacists indicated that the CPOE system had a neutral to positive impact on their jobs. DDI alerts were not viewed as a waste of time and the majority (61%) of prescribers felt that DDI alerts had increased their potential to prescribe safely. However, only 30% of prescribers felt DDI alerts provided them with what they needed most of the time. Both prescribers and pharmacists agreed that DDI alerts should be accompanied by management alternatives (73% and 82%, respectively) and more detailed information (65% and 89%, respectively). When asked about suggestions for improving DDI alerts, prescribers most preferred including management options whereas pharmacists most preferred making it more difficult to override lethal interactions. Prescribers and pharmacists reported primarily relying on electronic references for general drug information (62% and 55%, respectively) and DDI information (51% and 79%, respectively).

CONCLUSION

Respondents reported neutral to positive views regarding the effect of CPOE on their jobs. Their opinions suggest DDI alerts are useful but still require additional work to increase their clinical utility.

Personal digital assistants change management more often than paper texts and foster patient confidence

INTRODUCTION

Rapid retrieval of information, including drug treatment options, is critical to emergency department practice.

OBJECTIVES

To assess feasibility and patient acceptance of personal digital assistants and to determine the scope of management changes.

METHODS

Emergency medicine residents (EMRs, n = 18) and emergency medicine attending (EMAs, n = 12) used personal digital assistants with drug database and clinical references. Text versions were also available in the emergency department. We did a prospective, random, cross-over time-motion study, recording retrieval time, source, and changes to patient care for 16 and 8 h for EMRs and EMAs, respectively. We surveyed patients for confidence in EMRs and EMAs with personal digital assistants, and perceived efficiency.

RESULTS

EMRs accessed paper (n = 131) or personal digital assistant (n = 181) information on 92.3% of patients (n = 17, both). They accessed personal digital assistant on 61.4% of patients vs. 44.5% with texts (odds ratio 1.99, 95% confidence interval 1.4-2.80). Mean access times were 9.3 and 9.4 s, respectively, +1.4 for both. Personal digital assistant access was 75%/25% between pharmacopeia and clinical resource. Personal digital assistants changed drug choice in 39/181 patients (21.5%), and other management (diagnosis, treatment or disposition) in 15/181 patients (8.3%). Odds ratio for change in management for personal digital assistant vs. paper was 2.00 (95% confidence interval 1.11-3.60). We surveyed patient perception for 198 of 295 patients (67.1%). Fifty percent reported more confidence in their EMRs and EMAs with a personal digital assistant, while 5% reported less confidence. Sixty percent agreed strongly that there is too much medical information to remember.

CONCLUSIONS

Personal digital assistants are feasible in an academic emergency department and change management more often than texts. EMRs accessed personal digital assistants more often than paper texts. Patient perceptions of physicians who use personal digital assistants are neutral or favorable.

Frequency of high-risk use of QT-prolonging medications

PURPOSE

Prolongation of the QT interval has been associated with increased risk of torsades de pointes and death. Concurrent use of more than one QT-prolonging drug or a QT-prolonging drug with a drug that alters its pharmacokinetic profile is an important risk factor for adverse outcomes.

METHODS

Using a representative sample of 2 million health plan members from 10 health maintenance organizations with pharmacy benefits between January 1999 and July 2001, we identified potential drug interactions involving QT-prolonging medications. Prescription claims overlapping by at least 7 days for either 2 or more QT-prolonging drugs or a QT-prolonging drug with a drug that alters its clearance were considered potential drug interactions. We determined the number of drug interactions overall and the number of these interactions involving patients with other risk factors for torsades de pointes.

RESULTS

A total of 48 465 potential drug interactions were identified in 10 415 (4.6%) of the 228 550 patients with at least one prescription for a QT-prolonging drug. Amitriptyline was involved in 37 859 (78.1%) of the drug interactions. Of all potential drug interactions, 43 689 (90.1%) occurred in patients with at least one other risk factor for torsades de pointes, and 1053 (2.2%) were listed as a contraindicated combination in product labeling.

CONCLUSION

Potential drug interactions involving currently marketed QT-prolonging drugs occurred in 4.6% of patients who had a prescription for a QT-prolonging medication. The findings suggest several areas for targeted interventions to decrease the potential risk from QT-prolonging medications.

Janus computerised prescribing system provides pharmacological knowledge at point of care - design, development and proof of concept

OBJECTIVE

To develop and verify proof of concept for a user-defined prescribing system with decision support based on one single database consisting of several pharmacological sources.

METHODS

A multidisciplinary working group within the framework of a two-phase project developed the tool. A small-scale pilot study for proof of concept was carried out in an outpatient neurological polyclinic where four experienced physicians used the tool in patient care on a daily basis.

RESULTS

Automatically generated functions, such as recommended drugs, alerts for interactions, alerts for drug therapy during pregnancy and breast-feeding and a search tool for adverse drug effects, were quickly adopted into the daily outpatient working regime. Functions such as treatment strategies and a link to a producer independent website were less frequently used but still rated as useful and educational.

CONCLUSION

Searches for information that can be concentrated in one system saves time. Alerts inevitably draw physicians' attention to the information. Instant availability to drug recommendations in a computerised prescribing system such as Janus should increase adherence to recommendations, but this needs to be evaluated systematically. Small-scale pilot studies such as the one reported here have been shown to be invaluable in providing the theoretical basis for implementation of the system and for gaining an understanding of the complex change processes involved. Small-scale projects can therefore provided a base for further development and broader implementation of pharmacological tools and services.

Overriding of drug safety alerts in computerized physician order entry

Many computerized physician order entry (CPOE) systems have integrated drug safety alerts. The authors reviewed the literature on physician response to drug safety alerts and interpreted the results using Reason's framework of accident causation. In total, 17 papers met the inclusion criteria. Drug safety alerts are overridden by clinicians in 49% to 96% of cases. Alert overriding may often be justified and adverse drug events due to overridden alerts are not always preventable. A distinction between appropriate and useful alerts should be made. The alerting system may contain error-producing conditions like low specificity, low sensitivity, unclear information content, unnecessary workflow disruptions, and unsafe and inefficient handling. These may result in active failures of the physician, like ignoring alerts, misinterpretation, and incorrect handling. Efforts to improve patient safety by increasing correct handling of drug safety alerts should focus on the error-producing conditions in software and organization. Studies on cognitive processes playing a role in overriding drug safety alerts are lacking.

Community Pharmacy Managers' Perception of Computerized Drug—Drug Interaction Alerts

OBJECTIVE

To examine community pharmacists' attitudes toward computerized drug-drug interaction (DDI) alerts and identify factors associated with more favorable perceptions of these alerts.

DESIGN

Cross-sectional postal survey.

SETTING

18 metropolitan statistical areas (MSAs) in the United States.

PARTICIPANTS

3000 community pharmacy managers.

INTERVENTIONS

34-item survey instrument designed to collect data about the pharmacy including demographics, workload issues, handling of DDIs, and pharmacists' attitudes toward computerized DDI alerts.

MAIN OUTCOME MEASURE

Responses to items concerning community pharmacy managers' attitudes toward DDI alerts and factors associated with more favorable attitudes toward these alerts.

RESULTS

A total of 736 usable surveys were returned (25.3% response rate). Pharmacy managers generally disagreed that DDI alerts were a waste of time (mean +/- SD, 2.1 +/- 1.1 on a scale of 1, strongly disagree, to 6, strongly agree). However, they were not completely confident that their computer systems provided them with meaningful DDI alerts (mean +/- SD, 4.5 +/- 1.2). They were confident in their ability to identify DDIs (mean +/- SD, 4.9 +/- 0.9) and discuss DDIs with physicians (mean +/- SD, 5.2 +/- 0.7). Pharmacy software that provided detailed DDI information as well as the ability to customize DDI alerts were associated with more favorable perceptions of DDI alerts.

CONCLUSION

Despite being presented with a large proportion of clinically unimportant alerts, community pharmacy managers did not believe DDI alerts were meaningless or a waste of time. Incorporation of features that streamline DDI alerts may improve their effectiveness in community pharmacy practice.

A similarity measure for case based reasoning modeling with temporal abstraction based on cross-correlation

Adverse drug events (ADEs) are a major limitation of drug safety. They are often caused by inappropriate selection of dose and the concurrent use of drugs modulating each other (drug interaction). Risk assessment and prevention strategies must therefore consider co-administered drugs, individual doses, and their timing. In a new approach we evaluated the performance of cross correlation, commonly used in signal processing, to determine similarities in patient treatments. To achieve this, patient treatments were modeled as groups of vectors representing discrete time intervals. These vectors were cross-correlated and the results evaluated to find clusters in time courses indicating similarity in treatment of different patients. To evaluate our algorithm, we then created a number of test cases. The focus of this article is on each treatment, and its pattern in time and dosage. The algorithm successfully produces a relatively low similarity score for cases that are completely different with respect to their pattern of time and dosage but high scores when they are equal (score of 0.699) or similar (score of 0.528) in their therapies, and thus succeeds in having a relatively high specificity (27/30). Such an approach might help to considerably reduce the problem of false alarms which hampers most existing alerting systems for medication errors or impending ADEs.

Incidence and possible causes of prescribing potentially hazardous/contraindicated drug combinations in general practice

BACKGROUND

Preventing the use of medications where there is the potential for serious drug-drug interactions or drug-disease interactions (contraindications) is essential to ensure patient safety. Previous studies have looked at the incidence of prescribing contraindicated drug combinations, but little is known about the underlying reasons for the co-prescribing events. The objectives of this study were to estimate the incidence of prescribing contraindicated drug combinations in general practice and to explore the clinical context, possible causes and potential systems failures leading to their occurrence.

METHODS

A list of contraindicated drug combinations was compiled according to established references. A search of computerised patient medication records was performed, followed by detailed chart review and assessment. The patient records from four general practices in an area of England were searched for a period of 1 year (1 June 1999-31 May 2000) to identify contraindicated drug combinations. All patients registered with the four participating practices during the study period were included (estimated n = 37 940). Medical records of the cases identified by the computer search were reviewed in detail and relevant information was extracted. Each case was then independently assessed by a pharmacist and a physician who judged whether the co-prescribing was justified and whether it was associated with an adverse drug event. Proximal causes and potential systems failures were suggested for each co-prescribing event.

MAIN OUTCOME MEASURES AND RESULTS

Fourteen patients with potential drug-drug interactions and 50 patients with potential drug-disease interactions were identified. Overall, these represent an incidence of 1.9 per 1000 patient-years (95% CI 1.5, 2.3) or 4.3 per 1000 patients being concurrently prescribed > or =2 drugs per year (95% CI 3.2, 5.4). 62 cases involving 63 co-prescribing events were reviewed. Two-thirds of these events involved medications that were initiated by hospital doctors. Awareness of the potential drug-drug or drug-disease interactions was documented in one-third of the events at the time of initial co-prescribing. Within the study period, the co-prescribing was judged to be not justified in 44 events (70%). Potential drug-drug interactions possibly resulted in two adverse drug events. The majority of contraindicated co-prescribing related to drug-disease interactions involved the use of propranolol or timolol eye drops for patients receiving bronchodilators and the use of amiodarone for patients receiving levothyroxine sodium.

CONCLUSION

The prescribing of contraindicated drug combinations was relatively rare in this study. Multiple possible causes and systems failures were identified and could be used to develop strategies for the prevention of prescribing errors involving contraindicated drug combinations in primary care.

Evaluation of a Drug-Drug Interaction Alert Structure through the Retrospective Analysis of Statins-Macrolides Co-Prescriptions

The aim of this work was the evaluation of the existent drug interaction alert structure in Spain, which is based on yellow cards notifications and circulation of drug alert letters, through the retrospective analysis of CYP3A-metabolized statins and macrolides co-prescriptions in the Spanish province of Badajoz between May and September 2001. The period of study was planned to include the release of 2 drug alert letters released by the Spanish Drug Agency in June and July, addressed to all healthcare professionals, which warned against the concomitant prescription of statins and inhibitors of their metabolism, e.g. macrolides antibacterials. 4,600,764 prescriptions were examined, 664 of which corresponded to combinations of statins and macrolides. Although a decrease was detected in the number of these co-prescriptions throughout the study, 80 of these corresponding to 67 patients were still being prescribed in September, after the warnings by the Spanish Drug Agency had been released. 431 physicians prescribed these drugs simultaneously, with 22.9% of them having more than one patient at potential risk. Doctors working at rural healthcare centres or not directly attached to any healthcare facility were more prone to prescribe unsafe coprescriptions than those working at urban health centre. This study shows that the present drug alert system is not fully efficient when facing a situation like the one retrospectively reviewed in this study, in which a prompt action, in this case termination of potentially hazardous coprescriptions, was required. New systems developed to improve prescribing, including a new method based on personal contact between Drug Surveillance Centres and general practitioners, are discussed.

Evaluation of drug interaction software to identify alerts for transplant medications

BACKGROUND

The ability of computerized physician order entry (CPOE) systems to identify clinically significant drug interactions is dependent upon the integrity of the drug information populating the software. A CPOE system with incomplete or inaccurate drug information will fail to identify clinically important drug interactions and, therefore, fail to reduce preventable adverse drug events (pADEs).

OBJECTIVE

To evaluate, from the prescribers' perspective, the ability of a common drug interaction database to identify clinically important drug interactions involving drugs used in transplantation.

METHODS

The clinical significance of drug interactions involving 5 transplant drugs was evaluated by an expert panel to determine whether alerts should be generated for physicians not involved in the transplant at the time of order entry. Drug interactions included in the analysis were generated from the expert panel, a common drug interaction database, and 2 standard drug interaction references. Responses on the clinical significance were used to calculate the sensitivity, specificity, and positive and negative predictive values for each severity setting of a common electronic drug interaction database.

RESULTS

Overall, the database failed to identify approximately 70% of interactions considered significant by the expert panel. Of the alerts that were generated, >85% were considered clinically significant. The database was most deficient in identifying interactions resulting from additive toxicity.

CONCLUSIONS

To expect a decrease in pADEs caused by drug interactions, the information used to populate CPOE systems must be validated. Establishing consistency and integrity of this information may be a future role for pharmacists.

How to Design Computerized Alerts to Ensure Safe Prescribing Practices

BACKGROUND

Medication errors and preventable adverse drug events are common, and about half of medication errors occur during medication ordering. This study was designed to develop and evaluate medication safety alerts and processes for educating prescribers about the alerts.

METHODS

At Kaiser Permanente Northwest, a group-model health maintenance organization where prescribers have used computerized order entry since 1996, qualitative interviews were conducted with 20 primary care prescribers.

RESULTS

Prescribers considered alerts helpful for providing prescribing and preventive health information. More than half the interviewees stated that it would be unwise to let clinicians control or avoid safety alerts. Common frustrations were (1) being delayed by the alert, (2) having difficulty interpreting the alert, and (3) receiving the same alert repeatedly. Most prescribers preferred small-group educational sessions tied to existing meetings and having local physicians conduct education sessions.

DISCUSSION

The findings were used to design a strategy for introducing and promoting the interventions, modifying the alert text and tools, and focusing the education on how clinicians could use the alerts effectively.

Requirements for a successful implementation of drug interaction information systems in general practice: results of a questionnaire survey in Germany

OBJECTIVE

To determine drug interaction information requirements in general practice with respect to both content and mode of presentation.

METHODS

In a mail survey among 2,000 general practitioners in south-west Germany, we collected information on risk evaluation of drug interactions and combinations of concern, usage of and satisfaction with the current information sources, desirable content and mode of future presentation and demographic variables. Categorical variables were compared using chi2 test. Trends were analysed with Cochran-Armitage test and determinants of literature usage with logistic regression.

RESULTS

Response rate was 60.8%. The majority of general practitioners considered drug interactions a risk factor in prescribing (88.6%). For 18.2% of the drug combinations most frequently indicated as interacting, there was no published evidence of a clinically relevant interaction. More than half of the participants were dissatisfied with the information on severity, mechanism, and dose adjustment currently available in their sources. In particular, non-interacting alternatives were thought to be lacking (86.9%). Users of drug interaction software more frequently retrieved drug interaction information than non-users [odds ratio (OR) 1.95; 95% confidence intervals (CI) 1.50, 2.52], but only 28.6% of general practitioners had access to such systems. There was a significant trend towards electronic sources among younger physicians, but at present, 41.7% of general practitioners favour printed sources, and 8.8% would refuse to use electronic sources.

CONCLUSION

General practitioners wish for more informative support on drug interactions, especially concerning management. Despite a trend towards electronic information sources, printed documents are presently still required to reach all prescribers.

Using information technology and community-based research to improve the dental health-care system

It is commonly acknowledged that the United States' health-care system produces some of the finest care in the world for some people but fails to meet the needs of others. The Institute of Medicine (IOM) issued six aims for a redesigned health-care system, that it be: safe, effective, patient-centered, timely, efficient, and equitable. The purpose of this paper is to use an ongoing community-based study to illustrate current problems in the provision of oral health services that could be addressed through information technology. Appropriate use of information technology can assist dental schools and clinics in community-based clinical outcomes research needed to assemble the evidence base for improving oral health care. This conference serves as an important steppingstone to establish a means for information technology to improve the community's oral health.

Improving prescribing using a rule based prescribing system

OBJECTIVE

To test the hypothesis that the prescribing behaviour of doctors would improve after having experience with a computerised rule based prescribing system.

DESIGN

A prospective observational study of changes in prescribing habits resulting from the use of a computerised prescribing system in (1) a cohort of experienced users compared with a new cohort, and (2) a single cohort at the beginning and after 3 weeks of computer aided prescribing.

SETTING

64 bed renal unit in a teaching hospital.

INTERVENTION

Routine use of a computerised prescribing system by doctors and nurses on a renal unit from 1 July to 31 August 2001.

MAIN OUTCOME MEASURES

Number of warning messages generated by the system; proportion of warning messages overridden; comparison between doctors of different grades; comparison by doctors' familiarity with the system.

RESULTS

A total of 51,612 records relating to 5995 prescriptions made by 103 users, of whom 42 were doctors, were analysed. The prescriptions generated 15,853 messages, of which 6592 were warning messages indicating prescribing errors or problems. Doctors new to the system generated fewer warning messages after using the system for 3 weeks (0.81 warning messages per prescription v 0.42 after 3 weeks, p = 0.03). Doctors with more experience of the system were less likely to generate a warning message (Spearman's rho = -0.90, p = 0.04) but were more likely to disregard one (Spearman's rho = -1, p<0.01). Senior doctors were more likely than junior doctors to ignore a warning message.

CONCLUSIONS

Doctors are influenced by the experience of using a computerised prescribing system. When judged by the number of warning messages generated per prescription, their prescribing improves with time and number of prescriptions written. Consultants and registrars are more likely to use their clinical judgement to override warning messages regarding prescribed drugs.

Improving prescribing using a rule based prescribing system

OBJECTIVE

To test the hypothesis that the prescribing behaviour of doctors would improve after having experience with a computerised rule based prescribing system.

DESIGN

A prospective observational study of changes in prescribing habits resulting from the use of a computerised prescribing system in (1) a cohort of experienced users compared with a new cohort, and (2) a single cohort at the beginning and after 3 weeks of computer aided prescribing.

SETTING

64 bed renal unit in a teaching hospital.

INTERVENTION

Routine use of a computerised prescribing system by doctors and nurses on a renal unit from 1 July to 31 August 2001.

MAIN OUTCOME MEASURES

Number of warning messages generated by the system; proportion of warning messages overridden; comparison between doctors of different grades; comparison by doctors' familiarity with the system.

RESULTS

A total of 51,612 records relating to 5995 prescriptions made by 103 users, of whom 42 were doctors, were analysed. The prescriptions generated 15,853 messages, of which 6592 were warning messages indicating prescribing errors or problems. Doctors new to the system generated fewer warning messages after using the system for 3 weeks (0.81 warning messages per prescription v 0.42 after 3 weeks, p = 0.03). Doctors with more experience of the system were less likely to generate a warning message (Spearman's rho = -0.90, p = 0.04) but were more likely to disregard one (Spearman's rho = -1, p<0.01). Senior doctors were more likely than junior doctors to ignore a warning message.

CONCLUSIONS

Doctors are influenced by the experience of using a computerised prescribing system. When judged by the number of warning messages generated per prescription, their prescribing improves with time and number of prescriptions written. Consultants and registrars are more likely to use their clinical judgement to override warning messages regarding prescribed drugs.