Prescripción en pacientes con alergia a medicamentosAnálisis crítico de un sistema de alertas integrado en la historia clínica

Résumé

The prevalence of drug allergies is increasing, with as many as 25% of the general adult population affected in USA. In hospitals, drug-allergy interactions are one important cause of adverse drug events (ADEs). These events are preventable when the patient had a previous documented allergy to the drug that is prescribed. Computerized physician order entry (CPOE) with clinical decision support (CDS) has been implemented in most Health Care Systems to improve patient safety and the quality of care. These systems generate alert warnings regarding potential ADEs. CPOE with a drug allergy alert system (DAAS) is an important tool to prevent the prescription of medications to which the patient has a documented allergy and allergy-related medication events. However, these systems are not free of drawbacks (e.g. alert fatigue, unnecessary alerts) and their evaluation is necessary to improve their effectiveness. In our hospital, a homemade DAAS was implemented in 2007 and its evaluation was warranted. The objectives of this study were to review the evidence regarding DAAS, and to evaluate the DAAS of our information system (Clínica Universidad de Navarra Information System (SI-CUN)) analysing drug allergy alerts triggered by our DAAS, and the clinical consequences of drug allergy alert overrides. A PubMed and Cochrane databases search was conducted looking for articles describing electronic DAAS. Seventeen different DAAS were identified. Systems differ in recorded drug allergy information, rule bases that generate the alerts, type of matches or the information shown in the alerts. We found no comparative studies between them to identify the most effective system. However, we can extract some general recommendations. For example, to codify drug allergies information and override reasons and to register reactions suffered by patients. Alerts information has to be easy to understand. To evaluate SI-CUN drug allergy system, we conducted a retrospective observational study of DAA generated from 2014 to 2016. A total of 6,123 drug allergy alerts were triggered and after duplicated alerts were eliminated, 4,564 alerts remained for analysis. DAA were more frequently triggered by a ‘non-exact’ match and when ordering antiinfectives and nervous system drugs. The override rate was 44.8% and the main reasons given by prescribers were that the patient had tolerated the ordered drug or had not allergy or the allergy was uncertain, and that the patient had a prior administration of the drug or was taking it already. Clinical consequences of alert overrides were analysed in a retrospective observational study of a random sample of 194 overridden DAA generated between 2014 and 2016. A chart review was conducted to identify allergic ADEs. Six (3.09%) allergic ADEs were identified. Most ADEs were cutaneous and were classified as mild. Most alerts were appropriately overridden. The literature analysis together with the SI-CUN drug allergy alerts analysis and the clinical consequences of overrides evaluation allow making recommendations for systems improvement. Drug allergy information has to be to be accurate, codified and updated. It is necessary to identify important and clinical value alerts to reduce alert override and alert fatigue; some alerts that are frequently overridden with no clinical consequences, such as non-antibiotic sulfonamide alerts, can be eliminated or changed to non-interruptive information. Every medication has to be prescribed through the standard procedure to generate alerts when needed. Intelligent systems, that automatically update the information and change to non-interruptive information those alerts that are continuously overridden, would make the process easier. A committee must be created to regularly analyse the system, implement improvements and evaluate their impact as part of a quality control process to improve patients safety.