Getting it right: automated surveillance of healthcare-associated infections

Abstract

Healthcare-associated infections (HAI) such as surgical site infections and central line–associated bloodstream infections are among the commonest adverse events of medical care and are estimated to affect 6.5% of patients in acute care hospitals on any given day [1Hibbert P.D. Molloy C.J. Hooper T.D. Wiles L.K. Runciman W.B. Lachman P. et al.The application of the Global Trigger Tool: a systematic review.Int J Qual Health Care. 2016; 28: 640-649PubMed Google Scholar, 2Suetens C. Latour K. Kärki T. Ricchizzi E. Kinross P. Moro M.L. et al.Prevalence of healthcare-associated infections, estimated incidence and composite antimicrobial resistance index in acute care hospitals and long-term care facilities: results from two European point prevalence surveys, 2016 to 2017.Euro Surveill. 2018; 23 (Erratum in: Euro Surveill 2018;23): 1800516Crossref Scopus (204) Google Scholar, 3Cassini A. Plachouras D. Eckmanns T. Abu Sin M. Blank H.P. Ducomble T. et al.Burden of six healthcare-associated infections on European population health: estimating incidence-based disability-adjusted life years through a population prevalence-based modelling study.PLoS Med. 2016; 13e1002150Crossref PubMed Scopus (271) Google Scholar]. Surveillance of HAI and feedback of their rates, in particular within surveillance networks, is a key component of successful infection prevention programmes that provides caregivers and policy makers with the necessary information to identify areas of improvement and guide interventions [4Storr J. Twyman A. Zingg W. Damani N. Kilpatrick C. Reilly J. et al.Core components for effective infection prevention and control programmes: new WHO evidence-based recommendations.Antimicrob Resist Infect Contr. 2017; 6: 6Crossref PubMed Scopus (172) Google Scholar, 5Abbas M. de Kraker M.E. Aghayev E. Astagneau P. Aupee M. Behnke M. et al.Impact of participation in a surgical site infection surveillance network: results from a large international cohort study.J Hosp Infect. 2019; 102: 267-276Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar, 6Zingg W. Holmes A. Dettenkofer M. Goetting T. Secci F. Clack L. et al.Hospital organisation, management, and structure for prevention of health-care-associated infection: a systematic review and expert consensus.Lancet Infect Dis. 2015; 15: 212-224Abstract Full Text Full Text PDF PubMed Scopus (247) Google Scholar]. Conventional surveillance of HAI is done by manually reviewing patients' medical records and ascertaining the presence of HAI according to standardized surveillance definitions. This is a time-consuming and resource-intensive effort, and there are concerns regarding the uniformity of the surveillance results as a result of interrater reliability, differences in profiles of professionals conducting surveillance and the effort dependency of the surveillance process [7Mitchell B.G. Hall L. Halton K. MacBeth D. Gardner A. Time spent by infection control professionals undertaking healthcare associated infection surveillance: a multi-centred cross sectional study.Infect Dis Heal. 2016; 21: 36-40Abstract Full Text Full Text PDF Scopus (20) Google Scholar, 8Stricof R.L. Schabses K.A. Tserenpuntsag B. Infection control resources in New York State hospitals, 2007.Am J Infect Contr. 2008; 36: 702-705Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar, 9Birgand G. Lepelletier D. Baron G. Barrett S. Breier A.C. Buke C. et al.Agreement among healthcare professionals in ten European countries in diagnosing case-vignettes of surgical-site infections.PLoS One. 2013; 8e68618Crossref PubMed Scopus (24) Google Scholar, 10Dixon-Woods M. Leslie M. Bion J. Tarrant C. What counts? An ethnographic study of infection data reported to a patient safety program.Milbank Q. 2012; 90: 548-591Crossref PubMed Scopus (95) Google Scholar, 11Ehrenkranz N.J. Shultz J.M. Richter E.L. Recorded criteria as a ‘gold standard’ for sensitivity and specificity estimates of surveillance of nosocomial infection: a novel method to measure job performance.Infect Contr Hosp Epidemiol. 1995; 16: 697-702Crossref PubMed Google Scholar, 12Emori T.G. Edwards J.R. Culver D.H. Sartor C. Stroud L.A. Gaunt E.E. et al.Accuracy of reporting nosocomial infections in intensive-care-unit patients to the National Nosocomial Infections Surveillance System: a pilot study.Infect Contr Hosp Epidemiol. 1998; 19: 308-316Crossref PubMed Scopus (151) Google Scholar]. These shortcomings of conventional HAI surveillance have led to the development and use of automated surveillance (AS) systems for the identification surgical site infections, central line–associated bloodstream infections and other HAI [13Streefkerk H.R.A. Verkooijen R.P. Bramer W.M. Verbrugh H.A. Electronically assisted surveillance systems of healthcare-associated infections: a systematic review.Euro Surveill. 2020; 25: 1900321Crossref Scopus (8) Google Scholar, 14Verberk J.D.M. van Rooden S.M. Koek M.B.G. Hetem D.J. Smilde A.E. Bril W.S. et al.Validation of an algorithm for semiautomated surveillance to detect deep surgical site infections after primary total hip or knee arthroplasty – a multicenter study.Infect Contr Hosp Epidemiol. 2021; 42: 69-74Crossref PubMed Scopus (4) Google Scholar, 15Lin M.Y. Hota B. Khan Y.M. Woeltje K.F. Borlawsky T.B. Doherty J.A. et al.Quality of traditional surveillance for public reporting of nosocomial bloodstream infection rates.JAMA. 2010; 304: 2035-2041Crossref PubMed Scopus (135) Google Scholar]. AS uses routine-care data stored in electronic health records (EHR) to identify patients who (may) have developed a HAI. These systems reduce the workload of manual surveillance, thereby freeing up infection control practitioners' time; in addition, it can provide better standardization of surveillance results by facilitating data collection and standardizing case ascertainment [16van Mourik M.S. Troelstra A. van Solinge W.W. Moons K.G. Bonten M.J. Automated surveillance for healthcare-associated infections: opportunities for improvement.Clin Infect Dis. 2013; 57: 85-93Crossref PubMed Scopus (48) Google Scholar, 17Russo P.L. Shaban R.Z. Macbeth D. Carter A. Mitchell B.G. Impact of electronic healthcare-associated infection surveillance software on infection prevention resources: a systematic review of the literature.J Hosp Infect. 2018; 99: 1-7Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar, 18Lin M.Y. Woeltje K.F. Khan Y.M. Hota B. Doherty J.A. Borlawsky T.B. et al.Multicenter evaluation of computer automated versus traditional surveillance of hospital-acquired bloodstream infections.Infect Contr Hosp Epidemiol. 2014; 35: 1483-1490Crossref PubMed Scopus (11) Google Scholar]. These benefits of AS, combined with the increasing availability of data stored in EHR and the need for large-scale surveillance data, could motivate many hospitals, public health institutes and surveillance networks to transition to AS methods. The need for surveillance methods that are less reliant on tedious manual chart review, and possibly more timely, is illustrated by the current coronavirus disease 2019 (COVID-19) pandemic. Surveillance has been interrupted in many places to reallocate human resources to the pandemic response, thereby reducing case identification and resulting in potential lapses in infection control [[19]Statens Serum InstitutHAIBA.https://miba.ssi.dk/haibaDate accessed: January 9, 2019Google Scholar]. At the same time, COVID-19 has spurred the digitalization of healthcare; it may turn out to be the catalyst needed to realize the transition to large-scale AS. However, this transition to AS is not without risk. Guidance is lacking on how to best automate the surveillance process and ensure the delivery of surveillance data that is uniform and useful for improving the quality of care. In addition, automation in itself is not a guarantee of high-quality surveillance output, as small differences in implementation, underlying clinical care practices or coding procedures can greatly affect results [[20]Klein Klouwenberg P.M. van Mourik M.S. Ong D.S. Horn J. Schultz M.J. Cremer O.L. et al.Electronic implementation of a novel surveillance paradigm for ventilator-associated events: feasibility and validation.Am J Respir Crit Care Med. 2014; 189: 947-955Crossref PubMed Scopus (131) Google Scholar,[21]van Rooden S.M. Tacconelli E. Pujol M. Gomila A. Kluytmans J.A.J.W. Romme J. et al.A framework to develop semiautomated surveillance of surgical site infections: an international multicenter study.Infect Contr Hosp Epidemiol. 2020; 41: 194-201PubMed Google Scholar]. Importantly, the transition to AS entails more than converting a manual process to an automated process; it will affect surveillance targets, definitions, methods and interpretation of data, and it thereby runs the risk of losing clinical buy-in [[22]van Mourik M.S.M. Perencevich E.N. Gastmeier P. Bonten M.J.M. Designing surveillance of healthcare-associated infections in the era of automation and reporting mandates.Clin Infect Dis. 2018; 66: 970-976Crossref PubMed Scopus (32) Google Scholar]. As AS is implemented on a larger scale, it is paramount to explicitly define the purpose of surveillance – for example, quality improvement or, in some settings, pay-for-performance – and involve the correct stakeholders. Ideally, all implementation efforts should be coupled to an assessment of whether the surveillance method delivers results that contribute to reducing HAI rates, for example by observing the effect of HAI prevention programmes on measured infection rates or by assessing whether surveillance results are used to trigger interventions. AS has been applied in the research setting within hospitals [[13]Streefkerk H.R.A. Verkooijen R.P. Bramer W.M. Verbrugh H.A. Electronically assisted surveillance systems of healthcare-associated infections: a systematic review.Euro Surveill. 2020; 25: 1900321Crossref Scopus (8) Google Scholar] and sparsely in the setting of large-scale AS [[19]Statens Serum InstitutHAIBA.https://miba.ssi.dk/haibaDate accessed: January 9, 2019Google Scholar]. However, many of the currently available AS systems were developed in individual institutions with specific local preferences and conditions, and hence they are diverse in their aims, definitions, design and methods used. This solitary, stand-alone development of AS systems leads to heterogeneity and poor interoperability between systems. In addition, not all healthcare facilities can make a transition to AS on their own, leading to a segregation in healthcare facilities that collect surveillance data manually and automatically. These factors jeopardize the possibility of using surveillance data for comparison and subsequent quality improvement. The PRAISE network (Providing a Roadmap for Automated Infection Surveillance in Europe) was initiated in 2019 to develop guidance support the transition to large-scale AS; the products of this collaboration are presented in this supplement. The roadmap offers high-level conceptual guidance for the development of surveillance systems and details two possible approaches to their implementation, so-called locally implemented and centrally implemented AS. It pays extensive attention to the selection of surveillance targets and definition and the design of AS systems (including selection of data sources and algorithms), and it discusses implementation considerations, validation, maintenance and areas of future research. The accompanying article on the information technology (IT) aspects of large-scale AS focuses specifically on the IT requirements for implementation of AS, including the (re)use of healthcare data from EHR, standardization, interoperability, IT architecture and secure data transfer [[23]Behnke M. Valik J.K. Gubbels S. Teixeira D. Kristensen B. Abbas M. et al.Information technology aspects of large-scale implementation of automated surveillance of healthcare-associated infections.Clin Microbiol Infect. 2021; 27: S29-S39https://doi.org/10.1016/j.cmi.2021.02.027Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar]. The article on governance discusses governance aspects that are of particular importance in large-scale AS systems, including engagement of stakeholders, transparency of algorithms and accountability and the legal and ethical principles regarding the reuse of personal data for the purpose of surveillance [[24]van Rooden S.M. Aspevall O. Carrara E. Gubbels S. Johansson A. Lucet J.C. et al.Governance aspects of large-scale implementation of automated surveillance of healthcare-associated infections.Clin Microbiol Infect. 2021; 27: S20-S28https://doi.org/10.1016/j.cmi.2021.02.026Abstract Full Text Full Text PDF Scopus (3) Google Scholar]. Surveillance networks and hospitals can use this roadmap and the guidance documents to develop a harmonized automated approach to surveillance that suits their local situation and ideally that results in data supporting comparison and quality improvement. A first step in going forwards will be to choose an approach to implementation and achieve consensus on the targets for automated HAI surveillance and their definitions, as this will form the basis of all further development efforts. This network has been supported under the 7th transnational call within the Joint Programming Initiative on Antimicrobial Resistance (JPIAMR), Network Call on Surveillance (2018) and was thereby funded by ZonMw (grant 549007001 ). The author has no conflicts of interest to report. I would like to thank Anthony D. Harris for his critical review.