Impact of COVID-19 lockdowns on hospital presentations and admissions in the context of low community transmission: evidence from time series analysis in Melbourne, Australia.

BACKGROUND: Melbourne, Australia, successfully halted exponential transmission of COVID-19 via two strict lockdowns during 2020. The impact of such restrictions on healthcare-seeking behaviour is not comprehensively understood, but is of global importance. We explore the impact of the COVID-19 pandemic on acute, subacute and emergency department (ED) presentations/admissions within a tertiary, metropolitan health service in Melbourne, Australia, over two waves of community transmission (1 March to 20 September 2020). METHODS: We used 4 years of historical data and novel forecasting methods to predict counterfactual hospital activity for 2020, assuming absence of COVID-19. Observed activity was compared with forecasts overall, by age, triage category and for myocardial infarction and stroke. Data were analysed for all patients residing in the health service catchment area presenting between 4 January 2016 and 20 September 2020. RESULTS: ED presentations (n=401 805), acute admissions (n=371 723) and subacute admissions (n=15 676) were analysed. Substantial departures from forecasted presentation levels were observed during both waves in the ED and acute settings, and during the second wave in subacute. Reductions were most marked among those aged >80 and <18 years. Presentations persisted at expected levels for urgent conditions, and ED triage categories 1 and 5, with clear reductions in categories 2-4. CONCLUSIONS: Our analyses suggest citizens were willing and able to present with life-threatening conditions during Melbourne's lockdowns, and that switching to telemedicine did not cause widespread spill-over from primary care into ED. During a pandemic, lockdowns may not inhibit appropriate hospital attendance where rates of infectious disease are low.

Readability of online health information: implications for health literacy.

Accessibility is one of six quality criteria articulated by the European Commission in its code of conduct for health websites. Readability plays an integral part in determining a website's accessibility. Health information that is hard to read may remain inaccessible to people with low health literacy. This study aimed to calculate the readability of websites on various causes of disease. The names of 22 health conditions were entered into five search engines, and the readability of the first 10 results for each search were evaluated using Gunning FOG, SMOG, Flesch-Kincaid and Flesch Reading Ease tests (n=352). Readability was stratified and assessed by search term, search term complexity, top-level domain and paragraph position. The mean reading grade was 12.30, and the mean FRE was 46.08, scores considered 'difficult'. Websites on certain topics were found to be even harder to read than average. Where conditions had multiple names, searching for the simplest one led to the most readable results. Websites with .gov and .nhs TLDs were the most readable while .edu sites were the least. Within texts, a trend of increasing difficulty was found with concluding paragraphs being the hardest to read. It was also found that some of the most frequent search results (such as Wikipedia pages) were amongst the hardest to read. Health professionals, with the help of public and specialised libraries, need to create and direct patients towards high-quality, plain language health information in multiple languages.