No abstract

Background There are multiple data sources relating to colorectal cancer (CRC) nationwide. Prospective clinical cancer databases, population‐based registries and linked administrative data are powerful tools in clinical outcomes research and provide real‐world perspective on cancer treatments. This study aims to review the different Australian data sources for CRC from the perspective of conducting comparative research studies using a PICO (patient, intervention, comparison, outcome) framework. Methods Data dictionaries from the different data sources were evaluated for the types of exposure and outcome variables contained to highlight their differing research utility. Results State or territory‐based cancer registries contain limited histology, cancer staging and treatment detail. They enable investigation of population‐level patterns in overall survival (OS) of cancer patients with different demographics. Prospective clinical cancer databases contain more detail, especially surgical. Their strength is in auditing short‐term surgical outcomes. They vary in the amount of data collected for other cancer treatments and completion of follow up data. Linked administrative databases have broad population coverage but less surgical detail. They provide population‐level data on treatment patterns, short‐term outcome measures and OS, as well as long‐term surgical outcomes such as identifying patients who did not undergo stoma reversal. These databases cannot assess disease‐free survival. Conclusion Of the various CRC data sources within Australia, linked administrative databases have the potential to provide the widest population coverage combined with the broadest range of exposures and outcomes, and arguably the most research utility.

Background Hospital acquired infections are common, costly, and potentially preventable adverse events. This study aimed to determine the effect of the COVID-19 pandemic-related escalation in infection prevention and control measures on the incidence of hospital acquired infection in surgical patients in a low COVID-19 environment in Australia. Method This was a retrospective cohort study in a tertiary institution. All patients undergoing a surgical procedure from 1 April 2020 to 30 June 2020 (COVID-19 pandemic period) were compared to patients pre-pandemic (1 April 2019–30 June 2019). The primary outcome investigated was odds of overall hospital acquired infection. The secondary outcome was patterns of involved microorganisms. Univariable and multivariable logistic regression analysis was performed to assess odds of hospital acquired infection. Results There were 5945 admission episodes included in this study, 224 (6.6%) episodes had hospital acquired infections in 2019 and 179 (7.1%) in 2020. Univariable logistic regression analysis demonstrated no evidence of change in odds of having a hospital acquired infection between cohorts (OR 1.08, 95% CI 0.88–1.33, P = 0.434). The multivariable regression analysis adjusting for potentially confounding co-variables also demonstrated no evidence of change in odds of hospital acquired infection (OR 0.93, 95% CI 0.74–1.16, P = 0.530). Conclusion Increased infection prevention and control measures did not affect the incidence of hospital acquired infection in surgical patients in our institution, suggesting that there may be a plateau effect with these measures in a system with a pre-existing high baseline of practice.