Background and importance Drug–drug interactions leading to QT prolongation and potentially fatal torsades de pointes arrhythmias (QT-DDIs) are very common. Clinical decision support (CDS) triggers alerts for such QT-DDIs to warn physicians while prescribing. An additional safeguard mechanism is real-time follow-up of all alerts by clinical pharmacists who intervene by telephone when necessary.

Aim and objectives The first objective was the evaluation of the alert acceptance of QT-DDIs and QT interventions over a period of 5 years (2016–2020). The second objective was to evaluate the risk profile of patients with a QT intervention.

Material and methods In a tertiary hospital, QT-DDIs and pharmacist interventions were prospectively registered in a Microsoft Excel database. Possible interventions were electrocardiogram (ECG) or electrolyte monitoring, therapy change (eg, drug switch), the choice between monitoring and therapy change, and drug (re)initiation. Measured risk factors were female sex, age >65 years, impaired renal function (creatinine clearance <60 mL/min), electrolyte disturbances (potassium, calcium or magnesium), structural heart disease, number of QT prolonging drugs, bradycardia, no recent ECG, recent prolonged QT interval. Three types of acceptance were evaluated: CDS alert acceptance, telephone acceptance (ie, oral confirmation by physician) and intervention acceptance. Chi-square tests were used to compare frequencies.

Results In total, the CDS triggered 11 084 QT-DDIs, of which 2679 (24.2%) alerts were accepted. Pharmacists intervened for 192 QT-DDIs (1.7% of all QT-DDIs) with a telephone acceptance of 177 (92.2%). When verified in the patient records, the true intervention acceptance was significantly lower (145, 75.5%; p=0.037). Of 192 interventions, monitoring was advised for 85 (44.3%), therapy change for 51 (26.6%), and re(initation) for 31 (16.1%). There was no significant difference in intervention acceptance between the intervention types (p=0.087). On average, patients with a QT intervention had five risk factors. The most prevalent risk factors were age >65 years (121, 63.0%), structural heart disease (120, 62.5%), female sex (88, 45.8%) and prolonged QT interval (88, 45.8%).

Conclusion and relevance Telephone acceptance was very high, which can be interpreted as the pharmacist interventions being highly appropriate and complementary to CDS alerts. However, reasons for the difference between telephone acceptance and intervention acceptance need to be explored.

References and/or acknowledgements K Muylle and S Wuyts contributed equally.

Conflict of interest No conflict of interest