Identification of risk factors for hospital-onset bacteraemia to inform a routine data based risk prediction: an umbrella review.

BACKGROUND

The newly proposed surveillance entity hospital-onset bacteraemia (HOB) is associated with significant morbidity, mortality and costs. Identifying and understanding risk factors is crucial for the development of reliable risk prediction models to guide targeted infection prevention and control strategies.

AIM: The aim of the present study was to identify and classify risk factors for the onset of HOB in the general patient population in Organisation for Economic Co-operation and Development countries.

METHODS

We conducted an umbrella review to synthesize the evidence from systematic reviews and meta-analyses. We searched CINAHL, PubMed, Cochrane Library, Web of Science and grey literature. Two researchers independently screened abstracts and full texts, extracted data and assessed quality with AMSTAR 2. Data were analysed descriptively. Risk factors were categorized according to whether they can be modified through infection prevention and control interventions.

FINDINGS: We included 19 systematic reviews and identified 43 risk and five protective factors, which we categorized into patient-related (e.g. prior bloodstream infection (odds ratio (OR) = 6.56, P = 0.004), male sex (OR = 2.18, confidence interval (CI) = 1.52-3.12), multiple comorbidities (OR = 1.66, CI = 1.35-2.49), smoking (OR = 1.26, CI = 1.01-1.57)), procedure-related (e.g. red blood cell transfusion (risk ratio (RR) = 4.82, CI = 2.9-8.08)) and setting-related (e.g. single room (RR = 0.64, CI = 0.53-0.76)).

CONCLUSION

Most of the risk factors identified are well recognized for Central Line-Associated Bloodstream Infection and plausible contributors to HOB. Although not all HOB incidents are preventable, an early identification of high-risk patients combined with risk-stratified targeted interventions can prevent cases and improve outcomes. To ensure real-world applicability with algorithmic support, relevant data must be automatically available, and changeable factors should be monitored.

PROTOCOL AND REGISTRATION: The review is registered in International Prospective Register of Systematic Review, and the protocol can be found under CRD42023480112.