Multifaceted strategies for rational pharmacotherapy in the hospital: Optimizing pharmaceutical care from bench to bedside

Adverse drug events (ADEs) are an important cause of (potential) in-patient harm. ADE prevalence is estimated at 4.2 to 19.4 events per 100 hospital admissions, with 13-56% of events being avoidable. Additionally, patients are often admitted to the hospital because of an ADE. ADEs can be caused by adverse drug reactions or medications errors. Since the medication process is subjected to several vulnerabilities, the risk of errors significantly increases the awareness for implementing preventive strategies. Improving rational pharmacotherapy at the hospital-level can be achieved with the implementation of clinical decision support systems (CDSS) or complex interventions such as medication stewardship programs.

In this thesis, strategies focusing on rational pharmacotherapy of high-risk drugs or high-risk patients were evaluated in a tertiary care hospital setting.

First, a retrospective analysis of the hospital’s CDSS was performed to assess the CDSS performance, and clinical relevance of alerts for high-risk drug-drug interactions. Direct oral anticoagulants (DOACs) and CYP3A4 and/or P-gp modulating drugs were chosen as an exemplary high-risk combination. The latter are drugs known to influence DOAC pharmacokinetics, potentially increasing the risk of therapy failure (i.e., thromboembolic events) or side effects (i.e., bleeding). As physicians are known to experience alert fatigue to CDSS alerts, it is essential to know the clinical relevance of these alerts, which was explored in this study. We studied 166 individual patient records, in which 18 ADEs occurred that were potentially related to the concomitant use of these interacting drugs. Although the number of observed ADEs was relatively small, their importance should not be neglected due to their potential impact on patients. We showed that the physicians’ alert management was highly variable, indicating a need for more support in finding the most optimal management strategy for these drug-drug interactions. In a narrative review, we addressed this concern by exploring the current evidence base for the clinical relevance of DOAC and CYP3A4/P-gp inducer interactions. Present literature indicated a higher thromboembolic risk in patients on DOAC treatment combining CYP3A4- and/or P-gp inducers. DOAC management should be tailored to the individual patient through collaboration between expert healthcare professionals.

Secondly, CDSS can aid physicians to avoid inappropriate drugs in specific high-risk diseases. Little is known on the potential benefits and concerns of CDSS implementation for drug-disease interaction alerting in a hospital setting. Patients with liver cirrhosis were selected as a group of high-risk patients experiencing alterations in both pharmacokinetics and pharmacodynamics as a result of the disease. An exploratory study was performed to determine the prevalence of inappropriate drug use and potentially associated ADEs in hospitalized patients. A total of hundred randomly selected patient records were studied, in which ten percent of the prescribed drugs were considered potentially unsafe. In the patient cohort, three ADE cases were identified and discussed. Patients with liver cirrhosis admitted to the hospital are exposed to a significant number of potentially inappropriate drugs and face an increased risk of clinically relevant ADEs, possibly due to inadequate drug use. Supporting prescribers in drug-disease interaction identification is crucial by integrating CDSS within the electronic patient record.

In the second part of this PhD, we focused on interventions to improve rational pharmacotherapy through the introduction of a stewardship program. An intravenous (IV) fluid guideline was developed, given the identification of significant knowledge gaps on IV fluid management for adult patients amongst physicians and nurses in a hospital-wide survey. Subsequently, the IV fluid guideline was adapted to the survey results, implemented and evaluated in the UZ Brussel, as part of the program. To monitor guideline compliance, we decided to focus on the ED and determine the impact of the guideline on the IV maintenance fluid prescriptions’ appropriateness. This required the development and validation of a structured audit instrument, fit to the specific setting, followed by a retrospective analysis of the proportion of appropriate prescriptions pre- versus post guideline implementation. A substantial improvement of prescription appropriateness was determined, although documentation practice showed significant room for improvement. A second monitoring strategy included a
prospective, hospital-wide clinical audit using five KPIs. The feedback of the audit was reported to the ward’s fluid stewards. The impact of the feedback reports was studied, but was considered poor.

Although the most effective strategy for improving rational pharmacotherapy remains unknown, we identified several strengths and limitations of the approaches studied in this PhD thesis, demanding further study. Balancing a high alert burden imposed by CDSS while effectively ensuring the prevention of clinically relevant ADEs is challenging. We provided evidence on the potential safety threat to patients, acknowledging the importance of CDSS. On the other hand, we showed that literature does not yet provide an adequate knowledge base to recommend the most appropriate management strategy for several drug-related problems. Therefore, hospitals implementing interventions to enhance rational pharmacotherapy should be aware of these limitations and promote a management strategy that aligns closely with the proposed guidelines while considering the patient’s individual context. The impact on clinical outcomes, as well as cost-effectiveness of the described interventions is an additional aspect necessitating future study. While intervention or guideline noncompliance often stems from technical or organizational issues, the behavioral aspects of implementing changes in healthcare practice should also be examined to develop more performant strategies.