Correction: Shortening antibiotic therapy duration for hospital-acquired bloodstream infections in critically ill patients: a causal inference model from the international EUROBACT-2 database

When this article was first published it contained the following errors: 1. instead of "Kostoula Arvanti" the name of this co-author should be correctly "Kostoula Arvaniti" 2. The following sentence in the Statistical analysis section “ Clinically relevant covariates included catheter as source, source control status …, best available therapy (BAT), Staphylococcus aureus BSI (SAB), immunocompromised status, SOFA score at baseline, Delta-SOFA …, and combination therapy …” should correctly read: “Clinically relevant covariates included catheter as source, source control status …, delay until adequate treatment, best available therapy (BAT), Staphylococcus aureus BSI (SAB), immunocompromised status, SOFA score at baseline, Delta-SOFA …, and combination therapy …” 3. In the Results section, the most frequent infection source was incorrectly reported as catheter-related. It should have been respiratory. The corrected sentence is: “The source of the HA-BSI was most often a respiratory infection (n = 180, 33%).” 4. Accordingly, in Table 1 and the abstract, “catheter”, “respiratory” and “soft tissue” were mistakenly inverted. The corrected Table 1 is provided here. (Table presented.) Patients characteristics Overall, N = 5501 Long treatment 14–21 days, N = 3371 Short treatment 7–10 days, N = 2131 p-value2 64 (50–73) 64 (49–74) 64 (52–72) 0.7 351 (64%) 205 (61%) 146 (69%) 0.067 Respiratory 84 (15%) 48 (14%) 36 (17%) 0.4 Cardiovascular 117 (21%) 78 (23%) 39 (18%) 0.2 Neurology 88 (16%) 58 (17%) 30 (14%) 0.3 Metabolic* 196 (36%) 120 (36%) 76 (36%) > 0.9 Immunosuppression** 132 (24%) 80 (24%) 52 (24%) 0.9 1.00 (0.00, -4.00) 1.00 (0.00- 4.00) 1.00 (0.00–3.00) 0.2 8.0 (5.0, 10.0) 8.0 (5.0, 11.0) 7.0 (4.0, 10.0) 0.052 262 (47.6%) 157 (40.1%) 105 (49.3%) 0.59 0.7 Primary 91 (17%) 50 (15%) 41 (19%) Abdominal 75 (14%) 43 (13%) 32 (15%) Catheter 138 (25%) 87 (26%) 51 (24%) Respiratory 180 (33%) 116 (34%) 64 (30%) Soft tissue 21 (3.8%) 14 (4.2%) 7 (3.3%) Urinary 45 (8.2%) 27 (8.0%) 18 (8.5%) 212 (39%) 120 (36%) 92 (43%) 0.075 Non fermenting GN bacilli 112 (20%) 76 (23%) 36 (17%) 0.11 43 (7.8%) 19 (5.6%) 24 (11.3%) 0.026 Anaerobic and other GN 17 (3.1%) 6 (1.8%) 11 (5.2%) 0.026 50 (9.1%) 38 (11%) 12 (5.6%) 0.025 CoNS 49 (8.9%) 25 (7.4%) 24 (11%) 0.12 Streptococci, Enterococci*** 69 (13%) 42 (12%) 27 (13%) > 0.9 Polymicrobial 41 (7.5%) 30 (8.9%) 11 (5.2%) 0.10 94 (17%) 79 (23%) 15 (7.0%) < 0.001 201 (36.6%) 135 (40.1%) 66 (31.0%) 0.002 0.4 Not required 283 (51%) 167 (50%) 116 (54%) Required completed 249 (45%) 157 (47%) 92 (43%) Required not completed 18 (3.3%) 13 (3.9%) 5 (2.3%) Aminoglycoside 59 (11%) 29 (8.6%) 30 (14%) 0.043 Carbapenem 169 (31%) 112 (33%) 57 (27%) 0.11 Ureidopenicillin****** 115 (21%) 65 (19%) 50 (23%) 0.2 Glycopeptide 98 (18%) 68 (20%) 30 (14%) 0.069 Lipopeptide 7 (1.3%) 3 (0.9%) 4 (1.9%) 0.4 14.0 (9.0, 16.0) 16.0 (15.0, 18.0) 8.0 (8.0, 9.0) < 0.001 218 (40%) 155 (46%) 63 (30%) < 0.001 371 (67%) 237 (70%) 134 (63%) 0.071 < 0.001 Adequate therapy 9 (10%) 7 (9%) 2 (13%) Best available therapy******** 85 (90%) 72 (91%) 13 (87%) 6.0 (3.0, 8.0) 6.0 (3.0, 9.0) 5.0 (2.0, 7.0) 0.002 – 2.0 (– 4.0, 0.0) – 2.0 (– 4.0, 0.0) – 1.0 (– 4.0, 0.0) 0.2 In contrast, for BSIs caused by susceptible microorganisms, we included only patients who received at least one adequate antibiotic therapy. We excluded patients who did not receive appropriate treatment, as the susceptibility of the microorganism should have allowed for the administration of an effective antibiotic regimen SOFA sepsis-related organ failure assessment, ICU intensive care unit, GN Gram-negative, GP Gram-positive, CNS Coagulase-negative Staphylococci, HA-BSI Hospital-acquired Bloodstream infection, DTR difficult-to-treat, BMI body mass index, BC blood culture 1Median (IQR); n (%)2 Wilcoxon rank sum test; Pearson’s Chi-squared test; Fisher’s exact test *Include: diabetes, renal disease and connective tissue disease **Patients were considered immunocompromised if they presented any of the following conditions (organ transplant, IDS (not only HIV positivity, solid tumors, hematological malignancy (leukemia [acute of chronic] or lymphoma [any type]) or received any of the following treatments (steroids > 20 mg/day for at least 4 weeks or recent high dose steroids, chemotherapy / radiotherapy within 6 months, ongoing targeted cancer therapy), see more details in e-Table 2 in the ESM ***Streptococcus, Enterococcus and other GP ****DTR: a difficult to treat Gram negative is defined as due to a micro-organism Intermediate, or resistant to all reported agents in carbapenem, β-lactam, and fluoroquinolone categories (including additional agents when results available). For DTR A. baumanii, the strains should also be Intermediate or resistant to piperacillin-tazobactam and ampicillin-sulbactam. For other DTR than A. baumanii, the strain should be intermediate or resistant to aztreonam *****MDR was defined as acquired non-susceptibility to at least one agent in three or more antimicrobial categories ******Piperacillin – Tazobactam *******Durations were censored at day-28 ********In cases of BSIs caused by DTR microorganisms, we included patients who received adequate treatment. Additionally, we included patients who received the best available therapy (BAT), defined as any antibiotic regimen administered, even if it was considered inadequate or of uncertain efficacy. This decision was justified by the following: (i) Some centers might lack access to the most recent broad-spectrum antibiotics or advanced susceptibility testing techniques. (ii) Certain therapies might still exhibit partial activity against DTR microorganisms as they are usually active, but MIC was not available (i.e. tigecyclin, Colistin) and/or the breakpoint not defined (i.e.tigecyclin) 5. The abstract should have stated: “The most common infection source was respiratory (33%), most common microorganisms were Enterobacterales (39%).” The authors apologize for these errors and any inconvenience they may have caused. The Original Article has been corrected.