Exploiting large-scale linked electronic health records to investigate causes of current changes in Clostridioides difficile infection trends, contemporaneous risk factors for infection and severe disease, and treatment outcomes

Research question To exploit large-scale linked electronic health records to comprehensively investigate potential causes of current changes in Clostridioides difficile infection (CDI) trends, contemporaneous risk factors for infection and severe disease, and treatment outcomes. Background CDI is caused by a highly transmissible bacteria and affects over 15,000 people in England annually. CDI typically occurs after antibiotic treatment, causing diarrhoea. The disease can be severe, with a 30-day all-cause mortality of 13.8% observed in England in 2022/2023. NHS initiatives focused on infection prevention and antimicrobial stewardship successfully reduced CDI incidence and mortality in the early 2000s. However, after a period of stability, CDI rates began increasing in 2020/2021, with no clear cause identified. Beyond pandemic-related disruptions to infection prevention and management, changed population characteristics, increased case ascertainment, and shifts in antibiotic prescribing have all been hypothesised as potential drivers. Objectives My objectives are, in a regional data warehouse including positive and negative C. difficile tests, hospital prescribing and multiple potential biomarkers, to: Establish trends in CDI since 2000 (classified by onset/origin) and investigate whether current increases are driven by changes in case ascertainment/testing behaviour Estimate contemporaneous risk factors for CDI, CDI recurrence, C. difficile carriage and CDI mortality post-pandemic, and compare with pre-pandemic Compare hospitalised patients' biomarker response to different antibiotic treatments for CDI, to identify whether these could be more sensitive effectiveness measures and, in national linked prescribing and microbiology data, Quantify and compare the effects of different community-prescribed antibiotics on individual-level risk of developing community-onset/associated CDI. Methods The first phase of the study will address long-term trends in CDI using Oxfordshire hospital data, estimating testing rates, incidence and severity over time. The next phase will aim to identify contemporary risk factors for different CDI outcomes, considering a large list of variables including demographics, patient characteristics, hospital exposures, biomarkers, vital signs, and prior hospital prescriptions. This analysis will first be conducted for the post-pandemic period (2020-2023), and then compared to the period immediately pre-pandemic (2016-2019), to assess the potential impact of COVID-19. This analysis will identify baseline risk factors associated with different CDI severity outcomes, which will be used to adjust models investigating treatment responses to different antibiotics. The final phase of the study will use national-level linked data from the UKHSA to assess effects of community antibiotic prescriptions on risk of subsequently developing CDI. Anticipated impact Understanding the drivers behind recent changes in CDI trends is necessary to prevent the re-emergence of the major problems caused by CDI in the early 2000s. Contributing to the knowledge gap in this area may help improve patient outcomes through informing clinical treatment decisions and antimicrobial prescribing guidelines. Findings from this study can also shape public health policies around prevention of hospital-acquired infections and antimicrobial stewardship in the NHS. Dissemination Research outputs will be presented at conferences, and published in scientific journals, as well as communicated to clinical and academic contacts in the NHS and UKHSA. Findings will be communicated to patients and the public through blogs and newsletters.