Development of a tool to assess evidence for causality in studies implicating sink drains as a reservoir for hospital-acquired gammaproteobacterial infection

Abstract

Background: Decades of studies document an association between Gammaproteobacteria in sink drains and hospital-acquired infections, but the evidence for causality is unclear.

Aim: We aimed to develop a tool to assess the quality of evidence for causality in research studies that implicate sink drains as reservoirs for hospital-acquired Gammaproteobacterial infections.

Methods: We used a modified Delphi process with recruited experts in hospital epidemiology to develop this tool from a pre-existing causal assessment application.

Findings: Through four rounds of feedback and revision we developed the 'Modified CADDIS Tool for Causality Assessment of Sink Drains as a Reservoir for Hospital-Acquired Gammaproteobacterial Infection or Colonization'. In tests of tool application to published literature during development, mean percent agreement ranged from 46.7% to 87.5%, and the Gwet's AC1 statistic (adjusting for chance agreement) ranged from 0.13 to 1.0 (median 68.1). Areas of disagreement were felt to result from lack of a priori knowledge of causal pathways from sink drains to patients and uncertain influence of co-interventions to prevent organism acquisition. Modifications were made until consensus was achieved that further iterations would not improve the tool. When the tool was applied to 44 articles by two independent reviewers in an ongoing systematic review, percent agreement ranged from 93% to 98%, and the Gwet's AC1 statistic was 0.91-0.97.

Conclusion: The modified causality tool was useful for evaluating studies that implicate sink drains as reservoirs for hospital-acquired infections and may help guide the conduct and reporting of future research.

Keywords: Cross-infection; Gammaproteobacteria; Gram-negative bacteria; Infection control; Sinks; Waste water.