The Value of Hospital Data for covid-19 Pandemic Surveillance and Planning

Publicly available hospital data for covid-19 are challenging to find in Canada and elsewhere well into the pandemic. Of thirteen covid-19 data aggregate portals cited by the Research Data Alliance only one had hospital census counts and none had hospital admissions.\cite{RN3,RN1,RN5,RN7} China and Belgium report admission data, but data for other G7 or OECD countries are absent from public health websites.\cite{Belgium,RN2} The lack of new covid-19 hospital admission data as part of surveillance reporting is an indicator or long-standing poor data integration between hospitals and public health.
Covid-19 will continue to affect our society, health, and hospitals for at least the next year. Short-term forecasts of hospital use are vital to planning for covid-19 surges and waves. Long-term covid-19 projections are required to assess the effectiveness of different control strategies related to physical distancing, healthcare planning, contract trancing or economic policies. Hospital data are the best source of information for short-term forecasts of health care capacity, and they improve long-term covid-19 models that are typically based on cases.
We describe the role of hospital data, why admission data are hard to find, and how improved data infrastructure can meet surveillance and planning needs over the next year. There are promising signs that hospital-based data are being organized for research, and open data standards for covid-19 are maturings.\cite{RN3,RN9} These new covid-19 data initiatives support the use of individual hospital patient data, as opposed to the current use of aggregate hospital census counts. Routinely-generated lists of hospitalized covid-19 patients should be shared and combined with public health patient case data. The combined and integrated data provide robust support covid-19 surveillance, modelling and planning.