Infection on Frail Patients in the Intensive Care Unit: Insights From the PalMuSIC Study.

ABSTRACT

BACKGROUND: Along with population aging, frailty is also increasingly common in the intensive care unit (ICU). However, the impact of frailty on the infection incidence, the risk of multidrug-resistant (MDR) microorganisms, and the potential benefits of broad-spectrum antibiotics are still poorly studied. METHODS: This is a multicentric, prospective, observational study collecting data for 15 consecutive days of all consecutive adult patients admitted in each participating ICU. Exclusion criteria included admission for less than 24 hours or failure to obtain informed consent. The Clinical Frailty Score (CFS) was calculated both by the doctor and by the nurse in charge, and the patient's next of kin. Patients were considered frail if the mean of the three measured scores was ≥5. This is a post hoc analysis of the PALliative MUlticenter Study in Intensive Care (PalMuSIC) study. The Hospital de Vila Franca de Xira Ethics Committee approved the study (approval number 63). RESULTS: A total of 335 patients from 23 Portuguese ICUs were included. Frailty was diagnosed in 20.9%. More than 60% of the patients had a diagnosis of infection during their ICU stay, either present on admission or hospital-acquired. This included 25 (35.7%) frail and 75 (28.3%) non-frail (p=0.23) patients diagnosed with infection. In 34 patients, MDR microorganisms were isolated, which were more common in frail patients (odds ratio (OR) 2.65, 95% confidence interval (CI) 1.3-5.6, p=0.018). Carbapenems were started in 37 (18.1%) patients, but after adjusting for frailty and severity, no clear mortality benefit of this strategy was noted (odds ratio for ICU mortality 1.61, 95% confidence interval 0.49-5.31, p=0.43; odds ratio for hospital mortality 1.61, 95% confidence interval 0.61-4.21, p=0.33). CONCLUSION: Frail patients had similar rates of infection to non-frail patients but were more prone to have MDR microorganisms as causative pathogens. The use of empirical therapy with large-spectrum antibiotics should be based on microbiological risk factors and not simply on the host characteristics.