184 Chapter 8 Discussion In hospitalised patients with predominantly MSSA bacteraemia, five sub-phenotypes can be identified using routinely-available clinical data. These sub-phenotypes differ in survival and microbiologic outcomes. In a hypothesis generating secondary analysis of the ARREST trial, differential treatment effects were observed. Adjunctive rifampicin was associated with increased 84-day mortality in one sub-phenotype (nosocomial intravenous catheter SAB) and an improved microbiologic outcome in another (community-acquired metastatic SAB). Our findings permit several observations about SAB from an unbiased standpoint. Sub-phenotype B (nosocomial intravenous catheter SAB) represents patients at low risk of adverse outcomes. One hundred and thirty two (28·8%) patients were predicted to belong to this sub-phenotype in the Edinburgh cohort, whereas 71 (15·5%) met the inclusion criteria for the SABATO trial28,29 and 83 (18·1%) met the Figure 3: Effect of adjunctive rifampicin in SAB sub-phenotypesComparison of outcomes of patients randomised to placebo or adjunctive rifampicin when SAB sub-phenotypes considered separately. Treatment outcomes within each sub-phenotype were compared using Fisher’s exact test. Two comparisons were made within each sub-phenotype so the significance level was set at 0·025 (alpha=0·05, n=2).
RkJQdWJsaXNoZXIy MTk4NDMw