Association between infectious diseases consultation and mortality in hospitalized patients with Gram-negative bloodstream infection: a retrospective population-wide cohort study.

OBJECTIVES: Data supporting routine infectious diseases (ID) consultation in Gram-negative bloodstream infection (GN-BSI) are limited. We evaluated the association between ID consultation and mortality in patients with GN-BSI in a retrospective population-wide cohort study in Ontario using linked health administrative databases. METHODS: Hospitalized adult patients with GN-BSI between April 2017 and December 2021 were included. The primary outcome was time to all-cause mortality censored at 30 days, analyzed using a mixed effects Cox proportional hazards model with hospital as a random effect. ID consultation 1-10 days after the first positive blood culture was treated as a time-varying exposure. RESULTS: Of 30,159 patients with GN-BSI across 53 hospitals, 11,013 (36.5%) received ID consultation. Median prevalence of ID consultation for patients with GN-BSI across hospitals was 35.0% with wide variability (range 2.7-76.1%, interquartile range 19.6-41.1%). 1041 (9.5%) patients who received ID consultation died within 30 days, compared to 1797 (9.4%) patients without ID consultation. In the fully-adjusted multivariable model, ID consultation was associated with mortality benefit (adjusted HR 0.82, 95% CI 0.77-0.88, p < 0.0001; translating to absolute risk reduction of -3.8% or NNT of 27). Exploratory subgroup analyses of the primary outcome showed that ID consultation could have greater benefit in patients with high-risk features (nosocomial infection, polymicrobial or non-Enterobacterales infection, antimicrobial resistance, or non-urinary tract source). CONCLUSIONS: Early ID consultation was associated with reduced mortality in patients with GN-BSI. If resources permit, routine ID consultation for this patient population should be considered to improve patient outcomes.

Combination of Antistaphylococcal ß-Lactam With Standard Therapy Compared to Standard Therapy Alone for the Treatment of Methicillin-Resistant Staphylococcus aureus Bacteremia: A Post Hoc Analysis of the CAMERA2 Trial Using a Desirability of Outcome Ranking Approach.

Background: Desirability of outcome ranking (DOOR) is an emerging approach to clinical trial outcome measurement using an ordinal scale to incorporate efficacy and safety endpoints. Methods: We applied a previously validated DOOR endpoint to a cohort of CAMERA2 trial participants with methicillin-resistant Staphylococcus aureus bacteremia (MRSAB). Participants were randomly assigned to standard therapy, or to standard therapy plus an antistaphylococcal ß-lactam (combination therapy). Each participant was assigned a DOOR category, within which they were further ranked according to their hospital length of stay (LOS) and duration of intravenous antibiotic treatment. We calculated the probability and the generalized odds ratio of participants receiving combination therapy having worse outcomes than those receiving standard therapy. Results: Participants assigned combination therapy had a 54.5% (95% confidence interval [CI], 48.9%-60.1%; P = .11) probability and a 1.2-fold odds (95% CI, .95-1.50; P = .12) of having a worse outcome than participants on standard therapy. When further ranked according to LOS and duration of antibiotic treatment, participants in the combination group had a 55.6% (95% CI, 49.5%-61.7%) and 55.3% (95% CI, 49.2%-61.4%) probability of having a worse outcome than participants in the standard treatment group, respectively. Conclusions: When considering both efficacy and safety, treatment of MRSAB with a combination of standard therapy and a ß-lactam likely results in a worse clinical outcome than standard therapy. However, a small benefit of combination therapy cannot be excluded. Most likely the toxicity of combination therapy outweighed any benefit from faster clearance of bacteremia.

Overlapping Streptococcus pyogenes and Streptococcus dysgalactiae subspecies equisimilis household transmission and mobile genetic element exchange.

Streptococcus dysgalactiae subspecies equisimilis (SDSE) and Streptococcus pyogenes share skin and throat niches with extensive genomic homology and horizontal gene transfer (HGT) possibly underlying shared disease phenotypes. It is unknown if cross-species transmission interaction occurs. Here, we conduct a genomic analysis of a longitudinal household survey in remote Australian First Nations communities for patterns of cross-species transmission interaction and HGT. Collected from 4547 person-consultations, we analyse 294 SDSE and 315 S. pyogenes genomes. We find SDSE and S. pyogenes transmission intersects extensively among households and show that patterns of co-occurrence and transmission links are consistent with independent transmission without inter-species interference. We identify at least one of three near-identical cross-species mobile genetic elements (MGEs) carrying antimicrobial resistance or streptodornase virulence genes in 55 (19%) SDSE and 23 (7%) S. pyogenes isolates. These findings demonstrate co-circulation of both pathogens and HGT in communities with a high burden of streptococcal disease, supporting a need to integrate SDSE and S. pyogenes surveillance and control efforts.

Inter-species gene flow drives ongoing evolution of Streptococcus pyogenes and Streptococcus dysgalactiae subsp. equisimilis.

Streptococcus dysgalactiae subsp. equisimilis (SDSE) is an emerging cause of human infection with invasive disease incidence and clinical manifestations comparable to the closely related species, Streptococcus pyogenes. Through systematic genomic analyses of 501 disseminated SDSE strains, we demonstrate extensive overlap between the genomes of SDSE and S. pyogenes. More than 75% of core genes are shared between the two species with one third demonstrating evidence of cross-species recombination. Twenty-five percent of mobile genetic element (MGE) clusters and 16 of 55 SDSE MGE insertion regions were shared across species. Assessing potential cross-protection from leading S. pyogenes vaccine candidates on SDSE, 12/34 preclinical vaccine antigen genes were shown to be present in >99% of isolates of both species. Relevant to possible vaccine evasion, six vaccine candidate genes demonstrated evidence of inter-species recombination. These findings demonstrate previously unappreciated levels of genomic overlap between these closely related pathogens with implications for streptococcal pathobiology, disease surveillance and prevention.

Follow-up blood cultures do not reduce mortality in hospitalized patients with Gram-negative bloodstream infection: a retrospective population-wide cohort study.

OBJECTIVES: The utility of follow-up blood cultures (FUBCs) in patients with Gram-negative bloodstream infection (GN-BSI) is controversial. Observational studies have suggested significant mortality benefit but may be limited by single-centre designs, immortal time bias, and residual confounding. We examined the impact of FUBCs on mortality in patients with GN-BSI in a retrospective population-wide cohort study in Ontario, Canada. METHODS: Adult patients with GN-BSI hospitalized between April 2017 and December 2021 were included. Primary outcome was all-cause mortality within 30 days. FUBC was treated as a time-varying exposure. Secondary outcomes were 90-day mortality, length of stay, and number of days alive and out of hospital at 30 and 90 days. RESULTS: Thirty-four thousand one hundred patients were included; 8807 (25.8%) patients received FUBC, of which 966 (11.0%) were positive. Median proportion of patients receiving FUBC was 18.8% (interquartile range, 10.0-29.7%; range, 0-66.1%) across 101 hospitals; this correlated with positivity and contamination rate. Eight hundred ninety (10.1%) patients in the FUBC group and 2263 (8.9%) patients in the no FUBC group died within 30 days. In the fully adjusted model, there was no association between FUBC and mortality (hazard ratio, 0.97; 95% CI, 0.90-1.04). Patients with FUBC had significantly longer length of stay (median, 11 vs. 7 days; adjusted risk ratio, 1.18; 95% CI, 1.16-1.21) and fewer number of days alive and out of hospital at 30 and 90 days. DISCUSSION: FUBC collection in patients with GN-BSI varies widely across hospitals and may be associated with prolonged hospitalization without clear survival benefit. Residual confounding may be present given the observational design. Clear benefit should be demonstrated in a randomized trial before widespread adoption of routine FUBC.

TMPRSS2 inhibitors for the treatment of COVID-19 in adults: a systematic review and meta-analysis of randomized clinical trials of nafamostat and camostat mesylate.

BACKGROUND: Synthetic serine protease inhibitors block the cellular enzyme transmembrane protease serine 2, thus preventing SARS-CoV-2 cell entry. There are two relevant drugs in this class, namely, nafamostat (intravenous formulation) and camostat (oral formulation). OBJECTIVE: To determine whether transmembrane protease serine 2 inhibition with nafamostat or camostat is associated with a reduced risk of 30-day all-cause mortality in adults with COVID-19. DATA SOURCES: Scientific databases and clinical trial registry platforms. STUDY ELIGIBILITY CRITERIA, INTERVENTIONS, AND PARTICIPANTS: Preprints or published randomized clinical trials (RCTs) of nafamostat or camostat vs. usual care or placebo in adults requiring treatment for COVID-19. METHODS OF DATA SYNTHESIS AND RISK-OF-BIAS ASSESSMENT: The primary outcome of the meta-analysis was 30-day all-cause mortality. Secondary outcomes included time to recovery, adverse events, and serious adverse events. Risk of bias (RoB) was assessed using the revised Cochrane RoB 2 tool for individually randomized trials. Meta-analysis was conducted in the R package meta (v7.0-0) using inverse variance and random effects. Protocol registration number was INPLASY202320120. RESULTS: Twelve RCTs were included. Overall, the number of available patients was small (nafamostat = 387; camostat = 1061), the number of enrolled patients meeting the primary outcome was low (nafamostat = 12; camostat = 13), and heterogeneity was high. In hospitalized adults, we did not identify differences in 30-day all-cause mortality (risk ratio [95% CI]: 0.58 [0.19, 1.80], p 0.34; I2 = 0%; n = 6) and time to recovery (mean difference [95% CI]: 0.08 days [-0.74, 0.89], p 0.86; n = 2) between nafamostat vs. usual care; and for 30-day all-cause mortality (risk ratio [95% CI]: 0.99 [0.31, 3.18], p 0.99; n = 2) between camostat vs. placebo. CONCLUSION: The RCT evidence is inconclusive to determine whether there is a mortality reduction and safety with either nafamostat or camostat for the treatment of adults with COVID-19. There were high RoB, small sample size, and high heterogeneity between RCTs.

Antibiotics to eradicate Streptococcus pyogenes pharyngeal carriage in asymptomatic children and adults: A systematic review.

Streptococcus pyogenes (S. pyogenes) is a Gram-positive bacteria which causes a spectrum of diseases ranging from asymptomatic infection to life-threatening sepsis. Studies report up to 2000 times greater risk of invasive S. pyogenes disease in close contacts of index cases within 30-days of symptom onset. Despite this, there is variability in the management of asymptomatic carriage of S. pyogenes and those at risk of secondary cases of invasive S. pyogenes infection. OBJECTIVE: Our systematic review assessed the efficacy of different antibiotic regimens used for eradication of S. pyogenes from the pharynx in asymptomatic individuals. METHODS: We searched Pubmed, EMBASE (1974-), OVID Medline (1948-) and the Cochrane CENTRAL registry. We included randomised controlled trials (RCTs) with asymptomatic participants with >50% with pharyngeal cultures positive with S. pyogenes at baseline. Only studies with microbiological methods including culture (+/- polymerase chain reaction, PCR) were included. We included studies published in English. Each included study was assessed by two independent reviewers for data extraction and risk of bias. RESULTS: Of 1166 unique records identified, three RCTs were included in the review. Two of the three included RCTs found oral clindamycin for 10-days was the most efficacious regimen, compared to intramuscular benzathine penicillin G followed by 4 days of oral rifampicin, or monotherapy using benzathine penicillin, phenoxymethylpenicillin or erythromycin. Two RCTs were assessed as being at high risk of bias, with the third study demonstrating low/some risk of bias. CONCLUSIONS: Current available evidence for the optimal antibiotic in eradicating pharyngeal S. pyogenes carriage is limited. Future RCTs should include penicillin, first-generation cephalosporins, rifampicin, macrolides (such as azithromycin) and clindamycin.

Clinician perceptions of research priorities for the management of noncritically ill patients admitted to hospital with SARS-CoV-2 infection.

The changing phenotype of coronarvirus disease 2019 (COVID-19) may quickly render guideline-recommended interventions obsolete. We developed a 40-question clinician survey in consultation with the Australasian COVID-19 Trial site investigators. The survey was designed to assess clinician perceptions of the current treatment strategies and future research priorities in the management of non-critically ill patients admitted to hospital with SARS-CoV-2 infection. There were 84 complete responses from predominantly Australian and New Zealand clinicians. The perceived prevalence of patients with incidental COVID-19, nosocomial infection, underlying illness exacerbated by COVID-19, and/or immunocompromised status suggests new populations to target. The results highlighted clinician interest in antiviral therapies for future research in both immunocompetent and immunocompromised cohorts. These survey results underscore the need for ongoing surveillance of COVID-19 disease phenotypes and clinician and patient priorities for future research.

How I manage severe bacterial infections in people who inject drugs.

BACKGROUND: Injecting drug use is a risk factor for severe bacterial infection, but there is limited high-quality evidence to guide clinicians providing care to people who inject drugs. Management can be complicated by mistrust, stigma, and competing patient priorities. OBJECTIVES: To review the management of severe infections in people who inject drugs, using an illustrative clinical scenario of complicated Staphylococcus aureus bloodstream infection. SOURCES: The discussion is based on recent literature searches of relevant topics. Very few randomized clinical trials have focussed specifically on the management of severe bacterial infections among people who inject drugs. Most recommendations are, therefore, based on observational studies, extrapolation from other patient groups, and the experience and opinions of the authors. CONTENT: We discuss evidence and options regarding the following management issues for severe bacterial infections among people who inject drugs: initial management of sepsis; indications for surgical management; assessment and management of substance dependence; approaches to antibiotic administration following clinical stability; opportunistic health promotion; and secondary prevention of bacterial infections. Throughout, we highlight the importance of harm reduction and strategies to optimize patient engagement in care through a patient-centred approach. IMPLICATIONS: We advocate for a multi-disciplinary trauma-informed approach to the management of severe bacterial infection among people who inject drugs. We emphasize the need for pragmatic trials to inform management guidelines, including those that are co-designed with the community. In particular, research is needed to establish the comparative effectiveness, safety, and cost-effectiveness of inpatient intravenous antibiotics vs. early oral antibiotic switch, outpatient parenteral therapy, and long-acting lipoglycopeptide antibiotics in this scenario.

Evaluating the impact of a SIMPlified LaYered consent process on recruitment of potential participants to the Staphylococcus aureus Network Adaptive Platform trial: study protocol for a multicentre pragmatic nested randomised clinical trial (SIMPLY-SNAP trial).

INTRODUCTION: Informed consent forms (ICFs) for randomised clinical trials (RCTs) can be onerous and lengthy. The process has the potential to overwhelm patients with information, leading them to miss elements of the study that are critical for an informed decision. Specifically, overly long and complicated ICFs have the potential to increase barriers to trial participation for patients with mild cognitive impairment, those who do not speak English as a first language or among those with lower medical literacy. In turn, this can influence trial recruitment, completion and external validity. METHODS AND ANALYSIS: SIMPLY-SNAP is a pragmatic, multicentre, open-label, two-arm parallel-group superiority RCT, nested within a larger trial, the Staphylococcus aureus Network Adaptive Platform (SNAP) trial. We will randomise potentially eligible participants of the SNAP trial 1:1 to a full-length ICF or a SIMPlified LaYered (SIMPLY) consent process where basic information is summarised with embedded hyperlinks to supplemental information and videos. The primary outcome is recruitment into the SNAP trial. Secondary outcomes include patient understanding of the clinical trial, patient and research staff satisfaction with the consent process, and time taken for consent. As an exploratory outcome, we will also compare measures of diversity (eg, gender, ethnicity), according to the consent process randomised to. The planned sample size will be 346 participants. ETHICS AND DISSEMINATION: The study has been approved by the ethics review board (Sunnybrook Health Sciences Research Ethics Board) at sites in Ontario. We will disseminate study results via the SNAP trial group and other collaborating clinical trial networks. TRIAL REGISTRATION NUMBER: ClinicalTrials.gov Registry (NCT06168474; www. CLINICALTRIALS: gov).

Repurposing antibiotic resistance surveillance data to support treatment of recurrent infections in a remote setting.

In northern Australia, a region with limited access to healthcare and a substantial population living remotely, antibiotic resistance adds to the complexity of treating infections. Focussing on Escherichia coli urinary tract infections (UTIs) and Staphylococcus aureus skin & soft tissue infections (SSTIs) captured by a northern Australian antibiotic resistance surveillance system, we used logistic regression to investigate predictors of a subsequent resistant isolate during the same infection episode. We also investigated predictors of recurrent infection. Our analysis included 98,651 E. coli isolates and 121,755 S. aureus isolates from 70,851 patients between January 2007 and June 2020. Following an initially susceptible E. coli UTI, subsequent recovery of a cefazolin (8%) or ampicillin (13%) -resistant isolate during the same infection episode was more common than a ceftriaxone-resistant isolate (2%). For an initially susceptible S. aureus SSTI, subsequent recovery of a methicillin-resistant isolate (8%) was more common than a trimethoprim-sulfamethoxazole-resistant isolate (2%). For UTIs and SSTIs, prior infection with a resistant pathogen was a strong predictor of both recurrent infection and resistance in future infection episodes. This multi-centre study demonstrates an association between antibiotic resistance and an increased likelihood of recurrent infection. Particularly in remote areas, a patient's past antibiograms should guide current treatment choices since recurrent infection will most likely be at least as resistant as previous infection episodes. Using population-level surveillance data in this way can also help clinicians decide if they should switch antibiotics for patients with ongoing symptoms, while waiting for diagnostic results.

Flucloxacillin worsens while imipenem-cilastatin protects against vancomycin-induced kidney injury in a translational rat model.

BACKGROUND AND PURPOSE: Vancomycin is one of the most common clinical antibiotics, yet acute kidney injury is a major limiting factor. Common combinations of antibiotics with vancomycin have been reported to worsen and improve vancomycin-induced kidney injury. We aimed to study the impact of flucloxacillin and imipenem-cilastatin on kidney injury when combined with vancomycin in our translational rat model. EXPERIMENTAL APPROACH: Male Sprague-Dawley rats received allometrically scaled (1) vancomycin, (2) flucloxacillin, (3) vancomycin + flucloxacillin, (4) vancomycin + imipenem-cilastatin or (5) saline for 4 days. Kidney injury was evaluated via drug accumulation and urinary biomarkers including urinary output, kidney injury molecule-1 (KIM-1), clusterin and osteopontin. Relationships between vancomycin accumulation in the kidney and urinary kidney injury biomarkers were explored. KEY RESULTS: Urinary output increased every study day for vancomycin + flucloxacillin, but after the first dose only in the vancomycin group. In the vancomycin + flucloxacillin group, urinary KIM-1 increased on all days compared with vancomycin. In the vancomycin + imipenem-cilastatin group, urinary KIM-1 was decreased on Days 1 and 2 compared with vancomycin. Similar trends were observed for clusterin. More vancomycin accumulated in the kidney with vancomycin + flucloxacillin compared with vancomycin and vancomycin + imipenem-cilastatin. The accumulation of vancomycin in the kidney tissue correlated with increasing urinary KIM-1. CONCLUSIONS AND IMPLICATIONS: Vancomycin + flucloxacillin caused more kidney injury compared with vancomycin alone and vancomycin + imipenem-cilastatin in a translational rat model. The combination of vancomycin + imipenem-cilastatin was nephroprotective.

Hepatitis B genotypes in Aboriginal and Torres Strait Islander Australians: correlation with clinical course and implications for management.

BACKGROUND: The prevalence of chronic hepatitis B (CHB) in Aboriginal and Torres Strait Islander Australians in Far North Queensland (FNQ) is greater than twice that of the general Australian population. CHB is common in Torres Strait Islanders diagnosed with hepatocellular carcinoma (HCC) - and in Aboriginals with HCC living in the Northern Territory - however, Aboriginals diagnosed with HCC in FNQ very rarely have CHB. The explanation for this apparent disparity is uncertain. AIMS: To determine the HBV genotypes in the FNQ Aboriginal and Torres Strait Islander population and their correlation with clinical phenotype. METHODS: We determined the HBV genotype of Aboriginal and Torres Strait Islander Australians living with CHB in FNQ and correlated this with demographic and clinical findings. RESULTS: 134/197 (68%) enrolled individuals had a sufficient viral load for genotyping. All 40 people with HBV/D genotype had Aboriginal heritage, whereas 85/93 (91%) with HBV/C had Torres Strait Islander heritage (P < 0.0001). Individuals with HBV/D were younger than those with HBV/C (median (interquartile range) age: 43 (39-48) vs 53 (42-66) years, P = 0.0002). However, they were less likely to be HBeAg positive (1/40 (3%) vs 23/93 (25%), P = 0.001). All three HCCs developed in Torres Strait Islanders; two-thirds were infected with HBV/C14; genotyping was not possible in the other individual. All 10 diagnoses of cirrhosis occurred in Torres Strait Islanders, 6/10 were infected with HBV/C14, genotyping was not possible in the other four individuals. CONCLUSIONS: HBV genotypes in Aboriginal and Torres Strait Islander Australians in FNQ differ markedly, which could explain the significant differences in the clinical phenotype in the two populations and might be used to inform cost-effective CHB care in the region.

Avoiding misclassification of acute kidney injury: Timing is everything.

Accurate detection of acute kidney injury (AKI) in clinical trials is important. Using a 'baseline' creatinine from trial enrolment may not be ideal for understanding a participant's true baseline kidney function. We aimed to determine if a 'pre-trial baseline creatinine' resulted in comparable creatinine concentrations to a 'trial baseline creatinine', and how the timing of baseline creatinine affected the incidence of AKI in the Combination Antibiotic therapy for MEthicillin Resistant Staphylococcus aureus (CAMERA2) randomised trial. Study sites retrospectively collected a pre-trial baseline creatinine from up to 1 year before CAMERA2 trial enrolment ideally when the patient was medically stable. Baseline creatinine from CAMERA2 (the 'trial baseline creatinine'), was the highest creatinine measurement in the 24 h preceding trial randomisation. We used Wilcoxon sign rank test to compare pre-trial and trial baseline creatinine concentrations. We included 217 patients. The median pre-trial baseline creatinine was significantly lower than the median trial baseline creatinine (82 µmol/L [IQR 65-104 µmol/L] versus 86 µmol/L [IQR 66-152 µmol/L] p = <0.001). Using pre-trial baseline creatinine, 48 of 217 patients (22%) met criteria for an AKI at CAMERA2 enrolment and only 5 of these patients met criteria for an AKI using the CAMERA2 study protocol (using baseline creatinine from trial entry). Using a baseline creatinine from the time of trial enrolment failed to detect many patients with AKI. Trial protocols should consider the optimal timing of baseline creatinine and the limitations of using a baseline creatinine during an acute illness.

A blueprint for a multi-disease, multi-domain Bayesian adaptive platform trial incorporating adult and paediatric subgroups: the Staphylococcus aureus Network Adaptive Platform trial.

The Staphylococcus aureus Network Adaptive Platform (SNAP) trial is a multifactorial Bayesian adaptive platform trial that aims to improve the way that S. aureus bloodstream infection, a globally common and severe infectious disease, is treated. In a world first, the SNAP trial will simultaneously investigate the effects of multiple intervention modalities within multiple groups of participants with different forms of S. aureus bloodstream infection. Here, we formalise the trial structure, modelling approach, and decision rules that will be used for the SNAP trial. By summarising the statistical principles governing the design, our hope is that the SNAP trial will serve as an adaptable template that can be used to improve comparative effectiveness research efficiency in other disease areas.Trial registration NCT05137119 . Registered on 30 November 2021.

Transparent reporting of adaptive clinical trials using concurrently randomised cohorts.

Adaptive clinical trials have designs that evolve over time because of changes to treatments or changes to the chance that participants will receive these treatments. These changes might introduce confounding that biases crude comparisons of the treatment arms and makes the results from standard reporting methods difficult to interpret for adaptive trials. To deal with this shortcoming, a reporting framework for adaptive trials was developed based on concurrently randomised cohort reporting. A concurrently randomised cohort is a subgroup of participants who all had the same treatments available and the same chance of receiving these treatments. The reporting of pre-randomisation characteristics and post-randomisation outcomes for each concurrently randomised cohort in the study is recommended. This approach provides a transparent and unbiased display of the degree of baseline balance and the randomised treatment comparisons for adaptive trials. The key concepts, terminology, and recommendations underlying concurrently randomised cohort reporting are presented, and its routine use in adaptive trial reporting is advocated.

A statistical genomics framework to trace bacterial genomic predictors of clinical outcomes in Staphylococcus aureus bacteremia.

Outcomes of severe bacterial infections are determined by the interplay between host, pathogen, and treatments. While human genomics has provided insights into host factors impacting Staphylococcus aureus infections, comparatively little is known about S. aureus genotypes and disease severity. Building on the hypothesis that bacterial pathoadaptation is a key outcome driver, we developed a genome-wide association study (GWAS) framework to identify adaptive mutations associated with treatment failure and mortality in S. aureus bacteremia (1,358 episodes). Our research highlights the potential of vancomycin-selected mutations and vancomycin minimum inhibitory concentration (MIC) as key explanatory variables to predict infection severity. The contribution of bacterial variation was much lower for clinical outcomes (heritability <5%); however, GWASs allowed us to identify additional, MIC-independent candidate pathogenesis loci. Using supervised machine learning, we were able to quantify the predictive potential of these adaptive signatures. Our statistical genomics framework provides a powerful means to capture adaptive mutations impacting severe bacterial infections.