COVID-19, Data Protection and The Role of Emerging Technologies – An Interview with Dr. Behrooz Behbod

8 Sep 2020

Big data has been a buzz term for several years now and rightly so as we’ve seen it used to generate unbelievable new products, opportunities and wealth for company’s able to harness it. But equally we have seen the access to this data bring risks to consumers and take advantage of them. This is why data protection laws have needed to ramp up with the introduction of stringent legislation such as Europe’s GDPR introduced in 2018.

But within 18 months of its introduction, the world entered a crisis, the COVID-19 pandemic, where access to data has become of paramount importance to stem the spread of the virus. As the “invisible enemy” it is only health data (such as test results) which provides us the ability to “see” the virus – albeit just a shadow.

We asked public health physician, Dr Behrooz Behbod, his views on the COVID-19 pandemic and how innovative technologies like Artificial Intelligence and Blockchain could have assisted and how patient’s data rights have been affected.

What are the main issues you have seen with gathering health data in the case of COVID-19?

As with any outbreak, some of the common issues with gathering data include:

1) Access to timely data to enable public health and clinical actions;

2) Ease and reliability of reporting the data to the public health agency;

3) Maintain confidentiality and abide by GDPR;

4) Validity so we can trust the data represents the truth;

5) Completeness so that all cases, contacts, hospitalisations, and deaths are identified; and

6) Linking data to capture a person’s journey through their illness and healthcare pathways.

How do you think a nation’s ability to respond effectively to the pandemic would have changed if we had more efficient and comprehensive data access?

Taking a step back, we need to consider why we are collecting data. This varies depending on the user. We must be clear how that information will be processed, analysed, interpreted, communicated and acted on. From a public health perspective, the efficient and comprehensive access to data facilitates: 1) surveillance to monitor the epidemiology of the wider pandemic; 2) identification and control of local clusters and outbreaks; 3) monitoring the impact of interventions; and 4) management of cases and their contacts (a.k.a., contact tracing).

We need to be able to collect information on people who have developed symptoms, any associated test results (including antigen, antibody, and whole genome sequencing), their demographic characteristics (e.g., age, sex, ethnicity, occupation), immunisation history, risk factors and comorbidities, travel and exposure history, where they have been prior to and following onset of illness, and whom they may have been in contact with. We will also need information on anyone whose illness requires admission to hospital or intensive care, or if they unfortunately died because of COVID-19. Then for all identified contacts (people exposed to anyone with infection), we need their phone number and address to advise them to quarantine, and to follow them up during their incubation period, arranging testing if they may have developed COVID19.

There are also additional sources of data that may be helpful beyond ‘health’. For example, if we identify a cluster or outbreak, it is useful to know whether they are all in certain areas at greater risk of transmission or poor prognosis, such as care homes, high-density housing, prisons, hospitals, etc…

Most importantly, there are several local, regional, national and international stakeholders involved in preparedness and response. It is vital that there is a single version of the truth, available to all that need the information in real-time, whilst maintaining confidentiality.

How could technologies such as AI and machine learning play a role?

Excellent question! There is no sector or setting where AI and machine learning cannot play a role. A few examples in public health or the clinical management of COVID-19 include:

1. At the individual level:

In the community, people could use wearables to monitor their signs and symptoms (e.g., temperature, heart rate, respiratory rate). Predictive analytics could identify anyone at risk and trigger alerts to the person to isolate and seek testing. AI could be used to automate the test request process.
Similarly, in the hospital, patients can be monitored for early signs of COVID-19, with trigger alerts sent to their clinicians. Machine learning could help risk stratify and provide more frequent or detailed monitoring of particular patients at greater risk of acquisition or poor prognosis.
Remote patient monitoring could support the earlier discharge of inpatients from hospital, reducing the risk of healthcare-associated COVID-19 infections.
Digital technologies could help identify potential exposures to people with known COVID-19, triggering messaging to contacts to quarantine, monitor their symptoms, and seek medical care / testing when needed. As mentioned above, such contacts could use wearables to automatically trigger test requests upon development of symptoms or signs of COVID-19. Their test results could then be immediately shared with clinicians and public health agencies in real-time.
AI has an emerging role in the discovery of pharmaceutical treatments.

2. At the population level

AI facilitates surveillance and the early detection and communication of clusters or outbreaks. Indeed, effective communication is perhaps one the most effective interventions, where AI can support targeted and tailored messaging.
Natural language processing and advanced analytics could translate the vast amounts of available qualitative and quantitative information into actionable intelligence.
The combination of AI with blockchain may enable verification of the authenticity of test results and immunisation certificates, as well as linkage to other sources of data to capture all the information required, as described earlier. This could include identification of people who have recently travelled back from countries with increased incidence.
Population health management systems can use AI and machine learning to help risk stratify and provide more targeted interventions. A combination of predictive and prescriptive analytics enable the practice of ‘precision public health’.

If you wish to learn more about how AI can help, take a look at Fountech: https://www.fountech.ai

When it comes to GDPR, a legal exception was made to allow public health teams to be notified about COVID19 results without the patient needing to consent. Do you feel there is a conflict between the rights and needs of patients, doctors and public health teams? Could technologies such as blockchain provide a solution to patient privacy without compromising public health?

COVID-19 is a notifiable disease, such that the public health agency must be aware to take any necessary precautions to minimise the risk of further spread of the infection. This is not always possible with a single clinical team, given the potential for wider spread. Public health teams include medical doctors and are part of the health system, ensuring patient privacy.

Finally, this pandemic has shone a light on our need to invest more in healthcare. How do you see the future of eHealth solutions from here? Which areas do you think should be prioritised for improvement?

COVID-19 is not simply an infectious disease issue that can be solved by health protection alone, but has wider implications on health improvement, indoor and outdoor air quality and environmental health, the economy and indeed all sectors beyond healthcare. COVID-19 has highlighted and exacerbated existing issues, but has also shown the value of public health. For example, public health interventions work with all sectors to target risk factors that are associated with poor prognosis and death following COVID-19, including obesity, diabetes, cancer, and cardiovascular disease. These same risk factors are also driving increased morbidity and mortality, reduced quality of life, and increased demand on healthcare systems. Digital or eHealth has shown its value in managing lifestyle risk factors (e.g., physical activity, diet), monitoring disease (e.g., blood glucose, blood pressure), and communicating alerts to both the individual person as well as their healthcare provider(s). Together with my colleagues, we wrote a chapter on opportunities in global health, available here: https://hashedhealth.com/blockchain-book/#

Dr Behbod helps entrepreneurs and startups whose vision is to improve the health and wellbeing of the population, and indeed the planet at large. If anyone has any further questions, you can learn more at https://www.entrepreneurs.doctor