Quality of antimicrobial use in hospitals

Methods

Based on A-teams’ preferences and feasibility, three stewardship objectives were selected for inclusion in the antimicrobial stewardship monitor: 1) prescribe restricted antimicrobials according to the local guideline; 2) switch from intravenous to oral therapy (IV-oral switch); 3) perform bedside consultation in patients with a Staphylococcus aureus bacteremia (SAB).

From September 2017 until December 2017, a pilot was performed in ten hospitals to evaluate the feasibility to extract data on the performance of these stewardship objectives from the hospital electronic medical records (EMR) and associated lab systems. In addition, participating hospitals were offered the opportunity to provide data manually in a web-based portal. The data were used to compute process indicator performance scores.

Results

Eight of the ten hospitals that participated in the pilot could provide data. One hospital had insufficient IT support and in one hospital the board of directors did not grant approval to participate. Three hospitals provided data only electronically, three only manually, and two hospitals both electronically and manually. The pilot ended 31st of December 2017.

Restricted antimicrobials

It proved to be technically feasible to extract antimicrobial prescriptions from the EMR. Also, it was feasible to collate the individual prescriptions to one antibiotic course in case of dose adjustments or changes in route of administration. An example from one hospital is shown in Figure 5.2.1. Assessment of the appropriateness of these prescriptions by the A-team was performed for a part of these courses.

Data on these assessments could not be extracted from the EMR, for technical reasons. Therefore, this information was added manually to the stewardship monitor, leading to the quality indicator

performance as shown in Figure 5.2.1.

IV-oral switch

As a proxy for timely IV-oral switch, two types of indicators could be calculated from data provided electronically: first, the mean duration of intravenous administration of an antibiotic course and second, the percentage of patients that were switched to oral administration within 72 hours after initiation or in whom the antibiotic treatment was stopped. An example from one hospital is shown in Figure 5.2.2.

Bedside consultation

Positive blood cultures with S. aureus could be automatically extracted from the lab systems. The challenge to count a patient with several positive blood cultures as one episode of SAB was overcome.

However, whether or not bedside consultation was performed was not documented as discrete variable in this hospital, and had to be manually completed in order to calculate quality indicator performance (80% [16/20]).

Figure 5.2.1 Number of antibiotic courses (total column), number of antibiotic courses assessed (green + violet column), of antibiotic courses assessed as appropriate (green column), of antibiotic courses assessed as inappropriate (violet column) for carbapenems, piperacillin-tazobactam, and

fluoroquinolones. The percentages show the performance of the quality indicator “prescribe restricted antimicrobials according to the local guideline”.

Figure 5.2.2 Number of antibiotic courses that were administered intravenously for >72 hours (blue column) or ≤72 hours (purple column; percentage). The mean duration of the intravenous administration for the different antibiotics is shown above the columns.

Carbapenems Piperacillin-tazobactam Fluoroquinolones

Penicillins 3rd generation cephalosporins Carbapenems

0

2.9 days 3.3 days 4.5 days

Discussion

This pilot has yielded valuable information on 1) the possibilities of data-extraction for the purpose of surveillance and 2) barriers (and solutions) for its implementation.

Dutch hospitals engage actively in antimicrobial stewardship. This inquiry shows that 60 of the 64 responding hospitals (94%) have established an A-team and that almost all monitor the quality of antimicrobial use. A-teams focus on the use of restricted antimicrobials (91%), followed by IV-oral switch, and bedside consultation (~60%). These activities come on top of baseline functions of the three core specialties, and complement those with proactive and persistent efforts to measure the quality of antimicrobial use, facilitating improvement strategies and feedback to prescribers. There is, however, room for improvement. Not all stewardship objectives are covered, and structural

documentation facilitating analysis of the quality of antimicrobial use needs to be improved. A likely explanation is the lack of human resources, since only 41% of the A-teams were financially supported by the hospital boards of directors. If budget was provided (median of 0.5 FTE), this was significantly less than the recommended national staffing standards.

The quality improvement cycle is the core of antimicrobial stewardship. For this purpose, data on the quality of antibiotic use and its determinants are crucial. To facilitate this process, the SWAB performed a pilot to test the feasibility of automated data-extraction from the EMR and lab systems. This pilot shows that relevant data indeed can be extracted for the calculation of quality indicators. Half of the hospitals could provide data electronically, although only for part of the data. Data requiring an assessment by the A-team, such as the appropriateness of antimicrobial use, could only be provided manually, since these are not available as discrete data in the EMR. Thus, standardized documentation in the EMR is crucial. Another prerequisite for data-extraction is IT support. This was insufficient for the hospitals that could not provide data electronically. This is in line with the results of the survey. Only nine percent of the A-teams had structural IT support.

This pilot illustrates that data extraction from EMR and lab systems is feasible, but faces several challenges. Efforts are made to include missing variables in EMR and A-teams should claim structural IT support in their hospitals.

We pursue a steady increase in hospitals participating in the antimicrobial stewardship monitor, with ultimately an interactive dashboard showing the quality of antimicrobial use. Collected data from multiple hospitals will enable benchmarking, but challenges remain, for example to correct for the selections made for the prescriptions monitored by the A-teams. For the performance of bedside consultations in SAB, currently only the performance of bedside consultation is reported. As of 2018, the stewardship monitor will be integrated with the national initiative of a SAB registry, and more aspects of care for patients with SAB will be included.

MARAN 2018