Risk factors for mortality in HSCT recipients with bloodstream infection: points to be addressed by future guidelines.

Abstract Background : In the last years, important epidemiological changes have been described in hematopoietic stem-cell transplant (HSCT) patients with bloodstream infection (BSI), with increase in Gram-negative bacilli and multidrug resistant (MDR)-Gram-negative bacilli. These changes have been linked to a worrisome increase in mortality. Objectives : We aimed to define the risk factors for mortality of HSCT patients experiencing BSI. Study design : All episodes of BSI in patients with HSCT were prospectively collected (2008-2017). Multivariate analyses were performed. Results : 402 BSI episodes were documented in 293 patients who had undergone a HSCT (75.4% allogenic, 32.3% autologous, 19.3% second HSCT). Median time from HSCT to BSI was 62 days (IQR 9-182). Gram-positive cocci accounted for 56.7% episodes and Gram-negative bacilli for 42%. Most common microorganisms were coagulase-negative staphylococci (30.6%) and P. aeruginosa (15.9%). MDR-Gram-negative bacilli caused 11.9% of all episodes. Clinical characteristics, source of BSI, etiology and outcomes changed depending on time since HSCT. Globally, 26.6% episodes received inappropriate empirical antibiotic therapy, being more frequent in BSI episodes caused by P. aeruginosa, MDR-P. aeruginosa and MDR-Gram-negative bacilli. 30-day mortality was 19.2%. Independent risk factors for mortality were: BSI occurring ≥30 days after HSCT (Odds Ratio 11.21, 95% Confidence Interval 4.63-27.19), shock (OR 7.10, 95% CI 2.98-16.94), BSI caused by MDR-P. aeruginosa (OR 4.45, 95% CI 1.12-17.72) and inappropriate empirical antibiotic therapy for Gram-negative bacilli or Candida spp. (OR 3.73, 95% CI 1.27-10.89). Conclusions : HSCT recipients experiencing BSI have high mortality related to host and procedure factors, causative microorganism, and empirical antibiotic therapy. Strategies to identify HSCT-recipients at risk of MDR-P. aeruginosa and reducing inappropriate empirical antibiotic therapy are paramount to reduce mortality.