Survival of Patients With Staphylococcus aureus Bacteremia

TO THE EDITOR—We read the article by López-Cortés et al [1] with great interest because the authors found improved outcomes among patients with Staphylococcus aureus bacteremia (SAB) after a quality-of-care bundle for managing SAB was implemented. The authors found better outcomes in 14-day mortality, although they did not detect improvements in 30-day and 90-day mortality. The authors also found several risk factors for death, including older age and higher Pitt scores. Unfortunately, the authors did not specify whether improved quality-of-care management changed outcomes among patients with risk factors for SABattributable death. We would therefore like to add some data from our own experience. We implemented routine infectious diseases consultation at our department of internal medicine, using almost the same bundle of standards of care for patients with SAB, and analyzed a 3-year data set involving patients for whom SAB was diagnosed before (39 patients) or after (90 patients) implementation of routine consultation. We also found a significant survival benefit with routine infectious diseases consultation for patients with SAB. However, in contrast to LópezCortés et al, we did not find a survival benefit at day 14 but observed a survival benefit from day 40 until the end of the 12-month observation period, defined as 12 months after blood culture (Figure 1). In contrast to our cohort, the cohort examined by López-Cortés et al had more vascular catheter–related episodes in the intervention period. This might explain the 14-day survival benefit and higher rate of source control in the intervention group, because removing a vascular catheter is more easily achieved than source control of deep-seated SAB foci. Although López-Cortés et al observed improvements in individual quality-ofcare management parameters, information regarding the aggregate bundle of all management parameters during the preintervention and intervention periods is missing. Increases in the frequency of performing follow-up blood cultures or echocardiography without concomitant increases in the use of adequate treatment (cloxacillin in the Spanish study) or treatment duration will have no impact on long-term survival. In our cohort, use of all quality-of-care management parameters was achieved more often in the intervention group, compared with the preintervention group (73/90 vs 2/39; P < .0001). In contrast to López-Cortés et al, we also found a significant reduction of SAB recurrence among discharged patients in the intervention group, compared with those in the preintervention group (0/75 vs 6/30; P < .0001). In our cohort, intervention had no effect in patients with a high Pitt bacteremia score (>3), a high Charlson comorbidity index (>7), and an age of >80 years. Whereas our subgroup analysis might be hampered by sample size, the influence of improved quality-of-care management on the survival of patients with a high Pitt bacteremia score and older age (>60 years in the Spanish study) derived from the large study by López-Cortés et al would be of considerable interest. Although the Spanish study did not detect improvements in 30-day and 90-day mortality, we hope that these findings will not lead to the assumption that establishment or continuation of quality-of-care management is not justified in patients with SAB. Analysis of large SAB cohorts, such as that evaluated byLópez-Cortés et al, to determine the impact of management bundles on survival of specific patient populations (eg, elderly patients with underlying comorbidities)