Wednesday, March 17, 2021

(still unpublished) interview 3.3.2021 - On COVID-19 data


1.  An overview: How has the government fared in terms of publishing Covid-19 data and figures? Is it transparent enough?

To be honest, if the public would like to understand the general COVID-19 trends and development in the country, state or even district level, the current data provided by the government might be sufficient for general public consumption. Just in one go, let me try to illustrate what kind of daily data provided by our government: new cases and active cases, from national down to mukim level, new and existing cluster updates, number of people tested, number of death, number of patients admitted to ICU and those who need ventilator, imported cases and country origin, cumulative cases and risk level by colour code for all districts based on past 14 days local infection cases, non-citizens statistics updates, daily surveillance cases, as well as R values.

As a health data researcher myself who manually key in some of the covid-19 data everyday, one big inconvenient fact is that, many of the data provided is in image format, it is hardly convenient for people to do any sort of data analysis. Most of the data also are not provided in timeline format, so unless one does data input like me everyday or get the data from such person, it is impossible to see the long term trends yourself.

Of course, government could also provide further breakdown of case data according to characteristic: I refer to Australia’s Victoria State and Hong Kong’s health department , they also provide breakdown to cases by age and gender, mode of transmission or mode of case detection; I refer to Singapore’s Ministry of Health, they provide breakdown to cases by symptomatic and asymptomatic (which our MOH used to provide in certain surveillance groups); I refer to New Zealand’s Ministry of Health, they provide breakdown to active cases by current location (how many are in managed facilities or hospitals, at home or self-isolation, or in ICU). Not to say that Malaysia is doing badly because we do have some data categories some other countries do not provide in daily updates.

There are also instances it is debatable to publish certain granular and specific data such as Hong Kong they publish current location of active cases down to building level, Singapore they name the company or location of the infected person is working or living or having activities.

So, if you ask me about whether the government is transparent enough, that sounds like if the question is about whether the government has some data to hide and can justify so. Given the public stigmatization on covid-19 is so great, the identity exposure of the group of people, workplace or public space probably would not help containing the local infection but hurting the community for a considerable period.  So, first the authority should ask if the data release would be helpful in their disease containment strategies and better health protection; what could be the downside of such action, for example, creating negative social impacts, unnecessary public panicky and witchhunting.  Some data are difficult to understand or worse if one does not understand properly the implication or interpretation of data, this would risk certain people twisting and interpreting the fact wrongly, this could create public fear and anxiety. Thus, for certain data, it is actually wise to share with only relevant authorities and competent data researchers via different channels, not in the open.

 

2.  The channels they have opted to use are Telegram, Instagram, KKM website, MySejahtera. And also Live PCs by the Health DG. Could there be better, and more coordinated methods that the government could use, in light of how our culture/society uses technology?

Currently almost all COVID-19 info is aggregated and coordinated in the COVID-19 dedicated website covid-19.moh.gov.my , including state-district-mukim levels daily data. This should serve as the main source of information. Telegram channel has a little more data on the case distribution map and new cluster illustration, but it is difficult to navigate using that app, especially if you want to track historical data.

For live PCs by the Health DG, these days he reduced the frequency of such PCs, it would be helpful for the MOH to communicate issues concerning them directly to the public, media can get to enquire and get further clarification and response on some matters. Even the PC videos are also uploaded on the dedicated website I mentioned.

For most urgent and important info, most of the population who has a mobile number will get updates from the National Security Council (MKN) via SMS. And some health information will get broadcast on TV and radio. Therefore I think the government has used plenty of right channels at different technology levels to disseminate the info catering to many layers of society, so the bigger issue is more about the messaging clarity and effectiveness.

 

3. There have been arguments about how the data published is not granular enough. Is there a need for granular Covid-19 data and what can it be used for?

I have somehow mentioned on this issue when I responded to your earlier question. Some form of most granular Covid-19 data should be restricted to the usage of relevant authorities and competent researchers, for example mySejahtera contact tracing data. The authority should have responsibility to protect the identity of patient, workplace, or public places but leave the decision to the premise or the individual to make public announcement themselves. Relevant authorities should have such data to act upon, whether to do contact tracing, closure or sanitization of premise.

 

4. Dr Jemilah mentioned that data scientists have been hired to go through the government's Covid data to see how it can be useful towards the fight against Covid. In your opinion, how can the data on hand be useful in the fight against Covid?

We need to understand the limitation of data technology too. For example, the MySejahtera app data can tell the authority who were at the same site during certain period, it couldn’t tell the authority how the people interact. Even if the app is working as best as the South Korea’s app measuring the period and distance of contact between one another, it still couldn’t tell you whether the two persons if they wear mask, if they hug or shake hands, or if the measurement is just between the mobile phones lying on the dining table next to each other.

For contact tracing, while the granular data could be still useful as a reference even though it may be very noisy data, it still has to go back to the health personal who is trained to do contact tracing via personal interview to find out who could carry the bigger risk.

To me, it would be useful to look at the local transmission data and analyse the pattern significance, then it would be very informative for the local health authority to take action to respond timely, whether to do targeted mass screening at certain locality or improve the SOP there. This is where the granular data could help.

 

5. JKJAV has begun announcing daily vaccination dose figures & registration numbers. Is this necessary, and why?


In my opinion, the daily vaccination dose figures and numbers would help and is necessary, so that public would know the progress of our National COVID-19 Immunisation Programme. That is the bare minimum for transparency to keep the public informed, give them the confidence about the programme and perhaps rally or excite them to take part. But whether the government should publish the data down to district level or Vaccination Administration Centres at this moment, I guess this could be debatable, I think maybe not necessary now.

 

6. In an ideal world, what would be the best way to disseminate news about the vaccination programme to the whole nation? How could it be pulled off?

 

The programme publicity cannot be all top-down, one cannot underestimate the power of peer-influence. Besides the current variety of channels and technology platforms the government is using, the government should rally the support from the grassroot local community organisations and groups to help promoting the programme and get the people registered. It is always most persuasive if the people in your family and neighbourhood start talking about the programme and vaccine, show willingness to participate the vaccination. 

 

 

No comments: