Clinical Scorecard: Diverse Approaches for Defining Antibiotic Exposure in Clinical Studies: A Practical Examination of Health Care–Associated Clostridioides difficile Infection

At a Glance

Key Highlights

• Antibiotic exposure is complex and can be operationalized by multiple characteristics including type, dose, duration, and timing.
• Three analytic methods demonstrated for modeling antibiotic exposure: factor analysis, logistic regression, and LASSO regression.
• Best predictive models included variables such as any antibiotic exposure, number of antibiotic classes, and proportion of hospitalization on antibiotics.

Guideline-Based Recommendations

Diagnosis

• Confirm CDI diagnosis with rapid enzyme immunoassay test after 72 hours of hospital admission.

Management

• Consider multiple antibiotic exposure characteristics when assessing risk for CDI in hospitalized patients.

Monitoring & Follow-up

• Use electronic health record data to track antibiotic courses, doses, and timing during hospitalization.

Risks

• Exposure to specific antibiotic classes (e.g., β-lactam–β-lactamase inhibitors, carbapenems, cephalosporins, fluoroquinolones) increases CDI risk.
• Longer proportion of hospitalization on antibiotics correlates with higher CDI risk.

Patient & Prescribing Data

Hospitalized adults with recorded antibiotic courses and CDI status

Models incorporating detailed antibiotic exposure variables improve prediction of CDI risk compared to simple any-exposure definitions.

Clinical Best Practices

• Operationalize antibiotic exposure using multiple dimensions (type, dose, duration, timing) tailored to research or clinical questions.
• Utilize advanced analytic methods (factor analysis, LASSO regression) to select relevant antibiotic exposure variables.
• Incorporate antibiotic exposure characteristics into antimicrobial stewardship and infection prevention strategies.

References

• Diverse Approaches for Defining Antibiotic Exposure in Clinical Studies