Governments detail gaps in their scientific readiness for a pandemic

“Our ability to recognize troubles looming on the horizon was, at best, inadequate.”

9 June 2020

zubada/Getty

Inadequate data sharing and a failure to exploit available technologies for rapid diagnosis and drug development were missed opportunities for suppressing the coronavirus pandemic, a survey of government scientists suggests.

The Organisation for Economic Cooperation and Development (OECD), an international policy forum, asked its 37 member countries and several non-members about the strengths and weaknesses of the response of their scientific systems to the pandemic.

Thus far, 38 have participated in the survey, which is closing this month. Major countries that have not yet replied include the United States, China, India, and Indonesia.

Around half of the respondents said governments and health systems could have used smart technologies, artificial intelligence (AI), or big data to inform their response to the pandemic, or should use them now and in future.

Half also said they intended to share or were already sharing data, and that this would have been or is critical to combating the virus internationally.

About one-quarter urged improved data sharing across national borders, within health systems and across policy units.

Italy said OECD countries “need to reflect on how to improve our foresight activities”, stating that, “we should recognize that our ability to anticipate troubles looming on the horizon was, at best, inadequate.”

Harnessing the potential of AI

Scientists from several countries, such as Canada, Italy, Germany, Portugal, Russia, and South Korea emphasized the potential of AI to provide new solutions to address the pandemic.

Had AI, big data, and super-computer processing power been used, these countries indicated that drugs could have been identified more quickly to treat patients affected by the SARS-CoV-2 virus, which causes COVID-19.

“Supercomputers and big data can effectively improve the simulation of disease spreading, or the automatization of the analysis of targets for new drugs and treatments,” Germany commented.

South Korea drew attention to healthcare companies that are already using AI to screen for new drug candidates and drug repositioning opportunities. This activity, it said, “would dramatically reduce the time for developing therapeutics”.

However, several respondents, such as Italy, Japan, the United Kingdom, and Switzerland, said better practice in data sharing, cooperation, and communication was needed.

“It is necessary to establish a framework for sharing knowledge and data on infectious diseases across countries,” stated Japan.

Germany emphasized the need for collaboration: “The immediate sharing of results and more urgently available data is crucial.”

Embracing ‘lab-on-a-chip’ technology

Fulvio Esposito, Emeritus Professor of Parasitology at Camerino University in Italy, who drafted the Italian reply, says ‘lab-on-chip’ devices are an example of a ‘smart technology’ that could have played a greater role.

Such devices use wearable technology to relay information about symptoms from patient to specialist and enable quick diagnosis in the critical early stages of an emerging new disease. Had this technology been in widespread use, it could have halted the pandemic, he says.

“Unfortunately, we didn’t take full advantage of the potential of technologies such as lab-on-chip before the virus,” says Esposito.

AI, which relies on machine learning to interpret subtly shifting patterns in large datasets to inform new responses, has already been used for illnesses like cancer and to combat the HIV virus.

It should now be applied to COVID-19, says Alan Bernstein, president and chief executive officer of CIFAR, a Canadian-based global charitable science research organization that supplied the Canadian response. CIFAR is using AI to identify existing drugs that may be effective in treating COVID-19.

A separate project uses both AI and medical imaging to forecast how sick someone will become following COVID-19 infection, while a third identifies at-risk populations and predicts the course of the disease at individual and population levels.

A super-computer partnership

A European Union project known as E4C is also employing AI to screen thousands of approved drugs to find candidates for treating COVID-19. It’s one of many new initiatives to combat the disease described in the survey.

E4C is engaging Europe’s most powerful computers in a public-private partnership of 18 institutions coordinated by Italian pharmaceutical company, Dompé. The smart platform at its heart, Exscalate, is set to boost daily molecule-scanning capacity to 10,000, compared to 100 in existing tools employed by research institutes.

Data abundance will be critical to making optimal use of such technologies, Bernstein points out, emphasizing the Canadian response.

“Sharing data between cities and countries is key. The more data you have the more you can achieve,” he says.

Meanwhile, lab-on-chip-style technologies could mean reconfiguring established routines in healthcare research and treatment, Esposito says. In the current healthcare environment, he observes a significant communication and coordination gap between primary care (general practice) and secondary care (hospital/care home).

“This can, and indeed has in the current epidemic, caused a harmful delay in the early detection of cases,” says Esposito, indicating that such errors could be repeated, if these gaps are not bridged.

For instance, a consultant expert on rare diseases might not hear about a patient’s symptoms arising from a new illness until it is too late. Esposito suggests lab-on-chip devices could communicate them, although the role of the general practitioner would first need to be reviewed.

Looking ahead, he says better ICT infrastructure is needed to deploy such appliances.

“Remote reading of bio-images by a super-expert, able to recognize an interstitial pneumonia within a minute, is definitely possible today,” says Esposito, “provided that the remote site (such as peripheral hospital or dispensary) where the images were taken and the 'reference centre' are broad bandwidth interconnected – which is not the case.”