Using cloud technology in health care during the COVID-19 pandemic

How AI use cases are evolving in the time of COVID-19 | Healthcare IT News

There is now increasing recognition of the potential of cloud technologies, which provide data storage and computing resources managed by external service providers, to help improve safety, quality, and efficiency of health care.


However, adoption of cloud technology has been variable across health-care organisations, hampered by concerns that the technology might not align with existing methods of quality assurance and governance of privacy, data integrity, and service reliability.



The compelling shared purpose and informational needs in the context of the COVID-19 pandemic have provided a powerful incentive to adopt and benefit from fast scale-up of cloud-based solutions. Now that implementers have established this momentum, it is important to be mindful of the compromises and risks associated with these solutions and their implementation at speed.

Here, we provide an overview of how health-care settings have used cloud technologies to allow fast deployment of applications in individual organisations and integration of data analytics across organisations during the COVID-19 pandemic. We also discuss potential unintended consequences emerging from the scale and speed with which cloud technologies have been deployed. These include privacy and data governance considerations, lock-in of data structures, data silos, and unintended implications for work practices and organisational functioning. Although there are several examples of the implementation of cloud applications (appendix pp 1–3), it is important to note that there have not yet been formal evaluations of these approaches, which makes it difficult to assess their effect on clinical outcomes. We therefore highlight the need to avoid drawing simple causal inferences in relation to relatively short-term experiences of using complex technological infrastructures such as cloud technology.

Individual provider organisations have drawn on cloud technologies to implement discrete COVID-19-related functionality for organisational and clinical processes including monitoring, diagnostics, testing, triage, and consultations. Some applications facilitate real-time monitoring of patients in high-risk settings for COVID-19 through generating overviews of data from several sources, some enable interactions between health-care staff and patients at a distance, and others allow the development of operational management dashboards facilitating workforce, resource, and care planning.
A key benefit of cloud-based services to individual organisations and specialties is that they allow fast implementation and upscaling across a range of settings, because they do not require the organisation to purchase additional hardware (such as servers needed for on-premises solutions) and they can be implemented remotely (provided that appropriate infrastructure exists). For example, Huawei Technologies report that they deployed a pneumonia diagnostic solution to a hospital in Ecuador in merely 14·h,


and the Oklahoma State Department of Health deployed an application for medical staff, designed to follow-up people with reported symptoms of COVID-19, in 48·h.

However, such rapid implementations, although tackling immediate challenges, could have unintended consequences for existing health-care professional work practices and patient safety, particularly when new functionality is implemented across multiple contexts on a large scale.


By contrast, on-premises solutions allow piloting and tailoring to contextual requirements as they permit a greater degree of organisational control. This is important, as existing work with local electronic health records has identified the need to adjust to rapidly changing challenges associated with COVID-19.

Health-care organisations should therefore consider prioritising low-risk cloud solutions consisting of add-ons to existing functionality (eg, an application or a module on an existing cloud-based platform to allow rapid shared access), as these are more likely to enable better integration with existing practices than complex applications connecting departments and organisations.
Interorganisational data sharing in health care is difficult, particularly when data are held in local servers as these might become data silos.


COVID-19 has introduced common and pressing informational needs surrounding incidence, high-risk patients, and testing activity. Health-care settings now increasingly use cloud technologies to share COVID-19-related information and provide intelligence through real-time integrated data analytics from various sources across organisations (appendix pp 1–3). New applications range from dashboards connecting cloud-based electronic health records to identify trends in high-risk patients and testing activity, to the establishment of data hubs facilitating near-real-time data aggregation and analysis to inform decisions around resources and clinical care associated with COVID-19 across hospital groupings. There are also many examples of COVID-19 portals that give overviews of national and international trends, hosted on cloud services, which are currently in development.

This degree of information sharing on a large scale is simply not possible with on-premises systems, where additional integration engines would need to be installed but would only allow relatively little information sharing between organisations through standardised messages. Cloud technologies seem to offer a way forward here. However, there is now a need to align purposes of existing cloud technologies, as there is a risk of overlap between clouds from different service providers. There is also a risk of data silos in individual clouds, and associated issues surrounding data ownership and lock-in of data structures (eg, patient or location identifiers, disease classifications). Collaborative efforts aligning activities of cloud providers could reduce this risk, but this collaboration needs to be carefully balanced with information security considerations (which become exacerbated with larger scale as data transcends organisational boundaries). For instance, people have warned that establishing large clouds at speed might increase the risks of a so-called cyberpandemic, resulting in potential additional unanticipated risks and costs.


The often fast-paced implementation of cloud applications for COVID-19 might also have compromised adequate negotiation around harmonising data structures and governance at the outset, leading to potential issues surrounding data integration.