Background and Objective: 

Earlier research at our institution revealed a high incidence of drug discrepancies (DD) at the intensive care unit (ICU). Whether some of these changes were intentional (however not documented) or not was not yet investigated. The objective of the study was to determine the number of unintentional changes. Only unintentional changes were considered as discrepancies.

Design: 

Observational, prospective cohort study. At hospital admission, the medication history was documented by a pharmacist and was compared with the physician's medication history with special focus on chronic drugs. The administered medication on the ICU (last 24h), on the hospital ward subsequently following the ICU stay and the medication scheme in the discharge letter were all compared with the pharmacist-acquired medication history. If there was a non-documented modification of the chronic medication, the responsible physician was contacted to determine whether the change was intentional or not.

Setting: 

A cardiosurgical intensive care unit of a Belgian university hospital (UZ Brussel).

Main outcome measures:

The percentage of patients with DD, as well as the incidence and the type of DD in chronic medication. Also the number of intentional however not documented changes was investigated.

Results: 

The study group consisted of 36 patients. At hospital admission 30 patients (83%) had one or more DD in the medication history. There were a total of 99 DD for 179 detected chronic preadmission drugs (median of 2 DD per patient, range 0-10). For the comparison at hospital admission, every difference between the physician's and pharmacist's medication history was considered as unintentional and thus DD. The two most frequent types of DD were erroneous listed drugs in the physician's medication history (25% of DD) and omission of drugs (21% of DD). Comparison of the administered chronic medication on the ICU (last 24h) with the preadmission drugs revealed 26 patients (72%) with DD. A total number of 65 DD were detected (median of 1 DD per patient, range 0-7). The most common type of DD was omission or unintentional stop of a chronic drug (58% of DD). Of these 38 DD, 19 were drugs that were already missing in the physician's medication history and the other 19 were unintentionally stopped at the ICU. There were also 48 intentional but not documented changes. On the hospital ward subsequently following the ICU stay, 28 patients (78%) had one or more DD. In total, 68 DD were detected (median of 2 DD per patient, range 0-7). The most common type of DD was an unintentional stop of a chronic drug (32% of DD) that was listed by the physician at hospital admission. There were also 45 intentional but not documented changes. Comparison of the discharge letter revealed 116 DD for 165 preadmission drugs and 30 patients (91%) had one or more DD (3 patients without discharge letter). The most common type of DD was again the unintentional stop of a chronic drug (37% of DD) that was listed by the physician at hospital admission.

Conclusions:

Acquiring complete and correct medication histories at hospital admission is a prerequisite for adequate pharmacotherapy during hospitalization. Intentional medication changes during hospitalization should be well documented to avoid discrepancies during further transition moments. At transition moments, structural medication reconciliation with respect towards the original preadmission medication is crucial to reduce the number of potential adverse drug events that can arise from drug discrepancies.