Tools and challenges in the use of routine clinical data for antimicrobial resistance surveillance

There is a large ecosystem of tools relevant to the collection, analysis, and sharing of AMR data in the context of routine clinical services in hospital settings. Rather than providing an exhaustive list, here we outline the key elements required and highlight examples of free (and mostly open-source) software tools that have been developed to address particular steps in the data journey from hospital laboratories to national and international surveillance systems.

A Laboratory Information Management System (LIMS) is software designed to support data collection and storage within a laboratory, tracking specimens as they move through workflows, recording and linking results from different assays through integration with laboratory instruments, and interfacing with other databases and information systems such as patient records. Clinical microbiology laboratories have specific requirements¹⁶, including the need to track multiple specimens from the same patient (e.g. a blood sample and a urine sample); multiple culture media and multiple microbial isolates derived from those specimens; and multiple assay results for each isolated pathogen (species identification, susceptibility to a panel of antimicrobials).

Whilst many commercial LIMS are available, the Surveillance and Epidemiology of Drug-resistant Infections Consortium (SEDRIC) in 2019 identified the need for a free and open-source, microbiology-focused LIMS suitable for deployment in LMICs. SEDRI-LIMS was subsequently developed and is now available at a range of scales from single workstations to local servers and cloud-based setups, making it suitable for diverse laboratory environments; source code is due to be released in April 2025 ( In addition to the requisite features outlined above, SEDRI-LIMS supports sample barcoding, interpretation of susceptibility test results into S/I/R categories, and data export (to WHONET format described below, with other formats planned). This complements the SILAB for Africa LIMS developed to support veterinary laboratories in Africa¹⁷.

WHONET is a free Windows-based application designed to support clinical microbiology laboratories with the management, analysis, and reporting of AMR data ( It can be populated by data extracted from a LIMS, or by direct entry of sample-level data on species identification and susceptibility test results (including data exported from automated susceptibility testing platforms, via the integrated BacLink tool). The software also supports interpretation into S/I/R categories, facility-level summaries and cluster alerts (e.g. increased AMR in a particular ward, supported by the integrated SaTScan software), and exporting of data in a variety of formats used by public health surveillance programs (including the ‘WHONET’ format which is required for submission of data to WHO GLASS). Training materials, online courses, and webinars are available to help ensure the software is sustainable and empower local and regional stakeholders to support tasks ranging from annual surveillance to real-time decision-making and policy advocacy.

The AutoMated tool for Antimicrobial resistance Surveillance System (AMASS, is a free and open-source software package designed to support the use of AMR data at facility level, as well as national and regional surveillance, through linkage of laboratory and clinical data¹⁸. It is an offline tool that takes as input both AMR data (extracted from a LIMS or WHONET) and electronic health records (in CSV or Excel format), links these records at patient level, and exports de-identified data suitable for transmission to national surveillance networks (CSV format). AMASS also generates facility-level summaries (e.g. bug-drug AMR prevalence, stratified by community vs hospital-acquired), which can be exported as PDF reports or CSV data summary files, and can generate automated reports on AMR and notifiable bacterial diseases for transmission to national authorities¹⁹. In addition to providing AMR statistics including AMR proportion, AMR frequency and case fatality rate and total number of deaths following AMR bloodstream infection, the software also reports quality indicators like contamination rates and infrequent antibiotic resistant profiles. Future goals include integrating microbiology, hospital admission, and pharmacy data. The utility of AMASS for country-wide AMR surveillance was assessed by the Ministry of Health in Thailand in 127 public hospitals nationwide^20,21; it is also being deployed in other LMICs^9,22. Future goals include integrating microbiology, hospital admission, and pharmacy data.

Whilst the above tools have some flexibility to deal with the often messy error-prone and non-standardised data from across different systems, supplementary tools may be needed for more specialised data tasks. One such tool is the AMR R package, which provides a range of statistical tools to standardise and facilitate analysis of complex routinely-collected AMR data²³. Key features include functions for applying EUCAST or CLSI guidelines to interpret assay results, and tools to correct errors and standardise microorganism and antibiotic nomenclature.

There are several configurations in which the tools outlined above can work together or be used interchangeably, and connect with other tools such as DHIS2 (a general open-source platform for data integration and visualisation, to support AMR surveillance. For laboratories without a LIMS, WHONET may be used as a primary tool for data entry, analysis, and onward sharing. Those with a LIMS might choose to export data to WHONET or AMASS for analysis and reporting, and this decision may be influenced by the need to interact with national reference centres and surveillance programs. Notably, each of these tools use different formats to store AMR data (with WHONET emphasising interoperability of formats via their BacLink tool), and implement their own code to interpret susceptibility assay data into S/I/R categories using EUCAST or CLSI guidelines (which need to be digitised, and frequently updated to keep pace with updated guidelines). These areas would benefit from the development of agreed standards to improve interoperability and robustness, reduce the burden of software development and maintenance, and facilitate cross-sector integration of AMR data across the One Health continuum. For example, the One Health AMR Surveillance (OHAMRS) system in Kenya reports collecting data from hospitals via a mix of LIMS (43%), WHONET (19%), and a custom MS-Excel template (38%); and from veterinary laboratories via the SILAB LIMS; which then had to be integrated via DHIS2 to create surveillance dashboards²⁴.