Prevalence and types of drug interactions in hospitalized pediatric patients in Mexico City.

Objective: Drug-drug interactions may modify the therapeutic effect or the safety profile of the medicines used in pediatric populations.  Although interest on potential drug interactions in these age groups has  increased, information on clinically relevant drug-drug interactions is still  scarce. The aim of this study was to explore the prevalence and  characteristics of potential and clinically relevant drug-drug interactions  among pediatric patients hospitalized in two pediatric hospitals of Mexico  City.

Method: A cross-sectional study was conducted on patient records in critical, oncological, burns and other non-critical services by a pediatric resident physician at both hospitals. Micromedex® was used as a source of potential drug-drug interactions data. Subsequently, each  interaction's prevalence, severity and evidence level were estimated.  Additionally, drug-drug interaction causality with regard to diverse clinical  outcomes of hospitalized patients was determined through the Drug  Interaction Probability Scale. The clinical consequences of each interaction  were classified by severity.

Results: The observed prevalence of one or more potential drug-drug interactions in hospitalized patients was 61.3% (52.2-70.4%), whilst  the prevalence of real drug-drug interactions was 3.6% (0.1-7.1%). Of  potential drug-drug interactions, 60.5% were considered major and only  5.1% contraindicated. These were generally more common in intensive  care and burn units. The main pharmacological agents involved in  potential drug-drug interactions were opioids analgesics and anti-infective  and neurologic agents. Four clinically relevant drug-drug interactions  required a regimen change and another prompted an extension of the  patient's hospital stay.

Conclusions: Potential drug-drug interactions were common in the pediatric patients studied, whereas the frequency of real drug-drug  interactions was low. However, some drug-drug interactions required  medical actions in addition to routine monitoring. More information is  needed on real drug-drug interactions as those related to failed efficacy  might be underestimated.