Is antibiotic prescribing done according to guidelines? An important question in the fight against resistant bacteria. The A-team (antimicrobial stewardship team) of Amsterdam UMC developed, together with RIVM, a method to find out via the electronic patient file. So that adjustments can be made, if necessary.

In order to prevent certain bacteria from becoming uncontrollable, it is important to prescribe antibiotics correctly. That means: the patient receives an antibiotic that suits the bacteria that make him sick, and he takes it for as long as necessary according to the guideline. If all of that goes well, there is the least chance that bacteria will become insensitive (resistant).

But how do you keep track of whether antibiotics are prescribed in accordance with the guidelines throughout the hospital? Currently, one way of doing this is by means of a so-called point prevalence study. Once a year, a few weeks are spent tracking which antibiotics are prescribed for a bacterial infection, and for what reason. Very laborious and very time-consuming.

Electronic patient file

The A-team of Amsterdam UMC has developed a more efficient method in collaboration with RIVM and SWAB (Stichting Werkgroep Antibiotica Beleid). The key to this is a link with the electronic patient file (EPIC). With the new method, as soon as the doctor prescribes an antibiotic course, he or she must indicate the reason for it in a drop-down menu.

"This system was introduced in 2019," explains Annemieke van den Broek, who as a PhD student with the A-team is investigating, among other things, how to promote the correct use of antibiotics. "First by way of a trial, where we looked at whether the type of antibiotic chosen corresponds to what the guideline recommends." When that proved successful, Van den Broek and her colleague - physician researcher Jara de la Court - went a step further. "We now also look at the length of the cure. Not only in patients in hospital, but also in patients who are allowed to return home after an admission." This is of great added value because for a number of infections, such as strep throat or ear infections, the antibiotic course is largely followed at home.

Filling in a menu of choices

"The purpose of this registration is to see where the prescription of antibiotics is out of step with what the guidelines recommend. In this way, we can promote the proper use of antibiotics," clarifies De la Court, who is part of the medical-microbiology data team within the A-team. Her PhD research focuses on evaluating and guiding antibiotic use using data from the electronic patient record.

First question both PhD students wanted answered: do doctors fill out the menu of choices correctly? "We need to know whether we can trust the data entered," says De la Court. They concluded that the data entry went quite well. Although there is always room for improvement: "For 17 percent of the courses, the indication entered in the menu did not correspond with the doctor's note in the patient file."


What have the entered data yielded so far? Are the antibotics guidelines being properly adhered to at Amsterdam UMC? Van den Broek: "There is room for improvement in patients who have contracted pneumonia outside the hospital. They are regularly prescribed a drug that works against many different types of bacteria. Whereas the guideline says that for such pneumonia you should use a more specific antibiotic, targeting fewer types of bacteria." Another area of concern: in 40 percent of cases, antibiotics are prescribed for too long. Which is especially worrisome in view of possible side effects. De la Court: "We always say: it doesn't help, it always hurts."

For the A-team, it is reason enough to investigate the causes of this high percentage. Why is this or that guideline not followed and how can it be improved? "It's quite possible that a doctor deviates from the guideline with good reason," De la Court hastens to nuance. "But if that happens in 40 percent of the cases, something is wrong. Sometimes a guideline has only recently changed, and the old rules are still in the back of everyone's mind."

Specific questions

The new EPIC data make even more possible. Van den Broek: "We can also use them to answer specific questions, and indicate to the A-team exactly which things need to change." For example, the team asked whether the rarely prescribed antibiotic gentamicin was still being used anywhere. No problem: at the push of a button the system could provide the required data. De la Court: "It turned out that the advice about the dosage had changed, and now we knew which departments we had to notify. Before, we couldn't get this sort of thing done - not on this scale."

What is the A-team?

Since 2014, every hospital must have an Antimicrobial Stewardship or A-team. This includes, by default, a medical microbiologist, an infectiologist and a pharmacist. These are the senior members. There are also junior members; at Amsterdam UMC they are AIOSs and two PhD students. Our A-team also has a data analyst.

The task of the A-team is to ensure compliance with the guidelines for antibiotic use. We work side by side with the Antibiotics Committee (the director of the antibiotics guidelines) and Infection Prevention. De la Court: "Together we try to find the balance between being careful with the medication and ensuring that the patient receives good treatment."

Text: Irene van Elzakker