Session K: Microbiological Disease Surveillance, present and future A discussion paper for the Path IT Summit 17/18 Nov 2008 Prepared by James Freed, Jon Green and Andrew Chronias Introduction Microbiological disease surveillance aims to provide information to inform public health action, ranging from the management of the contacts of a single case of infection (e.g. meningococcal meningitis), to the investigation and control of an outbreak, to the development and implementation of national policy e.g. a new immunisation programme or a change in national guidance on antimicrobial therapy. Resulting public health actions may also include decisions to stop particular interventions, on the basis that surveillance has demonstrated a lack of effect or has provided evidence of the elimination of the problem. Surveillance might also identify risk factors for disease that can be the focus of research studies into disease aetiology or might identify cases for inclusion in research studies. Microbiological disease surveillance is currently conducted by routine diagnostic laboratory reporting (CoSurv/ LabBase), enhanced surveillance through paper or electronic data capture (e.g. Vaccine Preventable Diseases, HIV, HCAI), sentinel surveillance through schemes such as European Antimicrobial Susceptibility Surveillance (EARSS) or the referral of specimens for reference confirmation or typing (e.g. Salmonella, VTEC, Invasive Pneumococcal Disease) plus a combination of all of the above. It is very difficult for routine surveillance through a generic system like CoSurv (mentioned below) to serve the detailed needs of the surveillance of each infection. For these reasons enhanced surveillance is in place for a number of infections to address the specific needs of that infection, to capture additional information such as clinical, risk factor and outcome information and to ensure completeness of reporting. Enhanced surveillance may be triggered by the routine report or by clinical reporting and results in additional data being requested by means of paper forms or web based data entry. Examples of this would be vaccination status and follow up of vaccine preventable diseases, additional information to determine the source and risk factors associated with Healthcare Associated Infections and HIV surveillance where clinical information such as CD4 counts are required. Reference laboratory services provide the ability for almost complete referral of positive isolates for confirmation and typing purposes e.g. Salmonella to determine the strains circulating and whether these are associated with clusters that might indicate an outbreak. Sentinel surveillance is also used to examine a particular issue using standard methods and techniques such as monitoring changes in antimicrobial susceptibility patterns. Reference services are predominantly provided by the HPA so these data are captured in HPA surveillance schemes and where appropriate linked with some of the other data sources mentioned. The Present (an example) CoSurv collects positive laboratory reports of infections of clinical and public health significance from each NHS and HPA laboratory. In over 80% of microbiology labs the CoSurv system receives a defined file output of positive reports from the Laboratory Information Management System which passes through an interface (Lablink+) which maps the fields used by each individual lab and translates local coding to a national surveillance coding scheme. Data is transmitted from CoSurv modules in each lab onto regional epidemiology and local Health Protection Units in the HPA, for onward transmission to the national database LabBase. The Lablink+ interface and the mechanisms for reporting to regional epidemiology is/ will also be used to feed routinely generated susceptibility data to AmSurv as labs sign up. Electronic provision of data to CoSurv has taken significant investment by laboratories and the HPA, including the maintenance of code translation tables for each laboratory. It is recognised that this distributed system would benefit from standardisation in terms and coding and a more centralised model. The greatest challenges of the current situation are: gathering all of the information that is required for health protection purposes whilst minimising the burden on data providers collating and linking the various available datasets that make the complete picture when reliable common identifiers are not available interpreting these data due to differences in screening, testing and laboratory techniques conducting all of this within a suitable information governance framework to ensure confidentiality whilst providing data back to those who need it Looking forward: The changing landscape In looking forward, we need to consider the changing landscape in which we operate. Factors may include: Likely changes in the commissioning and provision of diagnostic laboratory services. Is increasing involvement and dependence on the private sector likely to impact on our ability to undertake surveillance? Could the increasing use of molecular approaches drive the integration of pathology services such that the specific microbiological expertise is diminished? Molecular characterisation and typing of microbial strains provides an important adjunct to more traditional phenotypic characterisation methods. However, these data (for example, DNA sequence data) have particular handling and analysis needs which current systems may not accommodate. The increasing expectation of ‘end-users’ of surveillance data i.e. those with a role in infectious disease control is that data will be aggregated, analysed and interpreted in ‘real-time’ . Specific infectious disease informatics and bioinformatics systems are required therefore not just to collate data but to provide timely analysis and context information in order to support infectious disease control and management. Where are we going? The key challenge now and in the future is that of data integration using operational data sources. In order to pull information from laboratory and other health systems and stitch it together to produce a nationwide view of the incidence and trends of infectious disease, we need our computer systems to speak the same language. We can either do this by mapping all existing coding systems to a single national standard or we can all use that single standard in the first place. If all pathology systems used the same National Catalogue and shared pathology data using national messaging standards, the process of collating data from hundreds of labs across the country would prove to be easy and would result in fewer errors. The current overhead of maintaining a mapping table for each individual laboratory system is simply incorrect and untenable going forward. The use of standards will support more effective re-use of data. The Health Protection Agency collects laboratory data for population-level (often anonymised) surveillance and patient- centric studies. Similar data often needs to be collected in different ways because of the inflexibility of the way that information is coded. Linkage of that data to a rich repository of coded clinical data will dramatically reduce the effort spent in collecting that information as well as improving timeliness, quality, data protection, coverage, completeness, richness and ultimately public health. Establishing these mechanisms is the platform on which all subsequent developments and refinements will be built.