Metrics for AMS Programme Evaluation

One of the most commonly used metrics to evaluate antimicrobial stewardship programs (ASP) is antimicrobial consumption metrics which is discussed in the previous step. In this step, we will discuss the qualitative measures (or metrics) that can be used to evaluate ASP. As briefly mentioned in step 2.14 these can be divided into the structure, process, and outcome metrics.

Structural metrics evaluate whether governance structures are in place for the antibiotic stewardship program. Things to consider include:

• What is the availability of a formal antibiotic stewardship team (AST)?

• Does the AST meet regularly?

• Are local protocols available and updated annually?

• Does the AST include certain staff (microbiology lab, nursing, ICU head, IT)?

• Is an antibiogram available and updated annually?

• Do you distribute updated antibiograms to prescribers every year?

• Is there a regular AST ward round?

• Is there a list of restricted antimicrobials?

Process metrics evaluate systems in place for the program. Examples of ways to measure this include:

• Percentage of adherence to documenting the indication
• Number of ASP interventions

• Acceptance of ASP recommendations

• Awareness of local antibiotic policy

• Percent of compliance with the guidelines

• Percent of patients receiving appropriate antibiotics

• Indication documentation

• De-escalation frequency

• Percent of IV to PO switch

• Appropriate discontinuation time of surgical antibiotic prophylaxis

• Time to appropriate antibiotic levels

• Percentage of appropriate cultures obtained before starting antibiotics

Outcome metrics can measure clinical outcomes such as:

• Infection cure

• Mortality (all-cause? 30-day? in-hospital mortality?)

• Re-admissions (30-day?)

• Length of stay

• Surgical site infections

• C. difficile infections

• Antibiotic resistance (which microbe and which antimicrobial?)

• Adverse drug reactions

• What are the costs (direct and indirect)?

The clinical outcomes are generally not easy to collect. The outcome metrics could also be microbiological such as the percentage of Pseudomonas resistance to meropenem.

Some concerns with resistance as a metric are that there are many confounding variables, including infection control activities, they are not immediately sensitive to change, and we have to consider breakpoint changes over time.

In addition, when resistance is reduced with a reserved agent, resistance could increase with recommended agents.

Do you currently carry out any qualitative measures of antibiotic consumption in your healthcare setting? If yes, what measures do you use?