Attributable mortality from extensively drug-resistant gram-negative infections using propensity-matched tracer...

Area of Science:

• Infectious Diseases
• Health Services Research
• Pharmacology

Background:

• Extensively drug-resistant gram-negative infections (GNIs) pose a significant public health threat.
• Estimating mortality directly attributable to these infections is challenging using administrative data.
• Tracer antibiotic algorithms offer a potential method for such estimations.

Purpose of the Study:

• To investigate the utility of tracer antibiotic algorithms for estimating mortality attributed to extensively drug-resistant gram-negative infections (GNIs).
• To assess the accuracy of these algorithms through chart review.
• To examine factors influencing attributable mortality, including infection characteristics and sepsis severity.

Main Methods:

• Propensity score matching was used to compare adult inpatients with GNIs treated with colistin versus comparator antibiotics (non-carbapenem β-lactams or carbapenems).
• Attributable mortality was calculated as the difference in in-hospital mortality between propensity-matched groups.
• Algorithm performance was evaluated using positive predictive value and sensitivity via chart review.

Main Results:

• The study analyzed 232,834 GNIs, revealing a substantial mortality difference between colistin and non-carbapenem β-lactam cohorts (29.2% vs. 16.6%).
• Attributable mortality was estimated at 12.6%, varying significantly by infection site (e.g., urinary vs. respiratory) and onset (early vs. late).
• Mortality increased ninefold in patients with severe sepsis or septic shock, and the colistin algorithm demonstrated a positive predictive value of 60.4%.

Conclusions:

• Mortality attributable to treatment-limiting resistance in GNIs varies considerably based on infection site, onset, and severity.
• Tracer antibiotic algorithms provide a feasible method for estimating attributable mortality from administrative data.
• These findings highlight the critical impact of antimicrobial resistance and infection severity on patient outcomes.