Meeting Abstracts

82 www.thelancet.com

Effi cient national surveillance for health-care-associated infectionsBram A D van Bunnik, Mariano Ciccolini, Cheryl L Gibbons, Giles Edwards, Ross Fitzgerald, Paul R McAdam, Melissa J Ward, Ian F Laurenson, Mark E J Woolhouse

AbstractBackground The detection of novel health-care-associated infections as early as possible is an important public health priority. However, no evidence base exists to guide the design of effi cient and reliable surveillance systems. Here we address this issue in the context of a novel pathogen spreading primarily between hospitals through the movement of patients.

Methods Using hospital admission data from the year 2007, we modelled the spread of a pathogen among a network of hospitals connected by patient movements using a hospital-based susceptible-infectious model. We compared the existing surveillance system in Scotland with a gold standard (a putative optimal selection algorithm) to determine its effi ciency and to see whether it is benefi cial to alter the number and choice of hospitals in which to concentrate surveillance eff ort.

Findings We validated our model by demonstrating that it accurately predicted the risk of meticillin-resistant cases of Staphylococcus aureus bacteraemia in hospitals in Scotland in 2007. Furthermore, the model predicted that relying solely on the 29 (out of 182) sentinel hospitals that currently contribute most of the national surveillance eff ort results in an average detection time (time until fi rst appearance of the pathogen in a hospital) of 117 days. This detection time could be reduced to 87 days by optimal selection of the same number of hospitals. Alternatively, the same detection time (117 days) can be achieved with just 22 optimally selected hospitals. Increasing the number of sentinel hospitals to 38 (teaching and general hospitals) reduced detection time by 43 days; a decrease to seven sentinel hospitals (all teaching hospitals) increased detection time substantially to 268 days.

Interpretation Our results show that the present surveillance system used in Scotland is not optimal in detecting novel pathogens compared with a gold standard. However, effi ciency gains are possible by better choice of sentinel hospitals, or by increasing the number of hospitals involved in surveillance. Similar studies could be used elsewhere to inform the design and implementation of effi cient national, hospital-based surveillance systems that achieve rapid detection of novel health-care-associated infections for minimum eff ort.

Funding This research received funding from the European Union Seventh Framework Programme (FP7-HEALTH-2011-single-stage): Evolution and Transfer of Antibiotic Resistance (EvoTAR).

ContributorsBADvB participated in the design of the study, carried out the analysis of the patient data, performed the simulations and statistical analysis, and wrote the abstract. MC participated in the design of the study. CLG carried out the analysis of the bacteraemia case dataset. GE provided bacteraemia case data. RF, PRM, MJW, and IFL contributed to the design of the study and writing the abstract. MEJW conceived the study and participated in its design.

Declaration of interestsWe declare no competing interests.

Published OnlineNovember 19, 2014

Centre for Immunity, Infection and Evolution, University of Edinburgh,

Edinburgh, UK (B A D van Bunnik PhD,

C L Gibbons MSc, M J Ward PhD, Prof M E J Woolhouse PhD);

Department of Medical Microbiology, University

Medical Center Groningen, University of Groningen, Groningen, Netherlands

(M Ciccolini PhD); Microbiology Department, Scottish MRSA

Reference Laboratory, Glasgow, UK (G Edwards MD); The Roslin

Institute and Edinburgh Infectious Diseases,

University of Edinburgh, Edinburgh, UK

(Prof R Fitzgerald PhD, P R McAdam PhD); and Scottish

Mycobacteria Reference Laboratory, Department of Laboratory Medicine, Royal

Infi rmary of Edinburgh, Edinburgh, UK

(I F Laurenson MD)

Correspondence to: Dr Bram A D van Bunnik,

Epidemiology Research Group, Centre for Immunity,

Infection and Evolution, Ashworth Laboratories, Kings

Buildings, West Mains Road, Edinburgh EH9 3JT, UK

bram.vanbunnik@ed.ac.uk