Do COVID-19 containment measure work? Learning from ASEAN Countries

Risk managers have been in the limelight as organisations contingency plans have been tested during this pandemic. Typically a risk manager's job is a very strange one, you would only know they are doing a bad job if things go South. Similarly there is no standard rule book on how to contain the COVID-19 pandemic except that hoping for herd immunity over a short time-frame does not work. The stages of decisions have mainly followed similar stages: nudging for improved hygiene, social distancing, closure of schools and public gatherings, and lock-down (or partial lock-down). Vietnam implemented these measures in almost a perfect fashion to date: from a catchy song to nudge hygiene (below) to aggressive lock-downs and disinfection of infected areas.

This article follows a series discussing the measures taken in different ASEAN countries (Brunei & Cambodia, Indonesia, Lao PDR & Myanmar, Malaysia, The Philippines, Singapore, Thailand and Vietnam.) and a summary. Here, I try to discuss the effect of stopping all public gatherings, closing schools, closing borders and a lock-down based using the experience of ASEAN countries. We have used ASEAN countries due to their proximity to China and therefore they can be considered as being a few weeks ahead.

Measuring the Impact of each decision

The impact of any decisions would be known subject to a delay of a few days. As the virus can take up to fourteen days to manifest after exposure, a decision taken for total lock-down in the middle of the month would only start reaping results at the end of the month as those infected before the lock-down would continue to manifest sickness later. However, the manifestation can take as little to a few hours and a study based on over 1000 cases reported that about three fourths would have shown symptoms in seven days.

Hence the simple analysis here focused on four stages of growth. The percentage increase is split into four weekly periods. One period is just before the decision was taken while the other three are after. Growth is defined in percentile terms. For example if there were 100 cases seven days before, 150 on the day and 250 seven days later; the rates of growth would be: Rate A = 50% (from 100 to 150) and Rate B = 66.67% (from 150 to 250).

Assuming a delay in manifestation by seven days, one would expect

• Rate B to be higher than Rate A. The reason is that the decision was possibly taken at the moment that the number of cases was growing significantly. On taking the decision to limit cases, there would need to be a further week of those already exposed to manifest.
• Once the effect of the decisions is being seen, fewer people would be infected over the seven days after the decision and thus less showing the symptoms a week later. We would expect rate C to be lower than B (and A).
• Continuous improvement should lead to rate D being lower.

Results

Banning public gatherings, closing schools and closing borders all seem to have had the same pattern as can be seen in the table on the right.

The closing of schools seems to have had the highest effect as the improvement in median improvement in rate is actually over 200% between rates B and C. Consider that not all countries may have had all rates - for example Lao PDR closed all schools before the first case.

Yet in every circumstance, a growth rate in which essentially the number of cumulative cases was at least tripling per week was lowered significantly. While this may sound drastic, a daily increase of 20%, which is in line with the experience worldwide, would lead to an increase of 250% (more than a triple).

There are many limitations to this summary. Firstly the rates may have naturally gone down once the decision is taken, especially if the growth has a natural peak (more about why this may also be wrong here). Secondly, decisions would have been taken in tandem. For example the Philippines took the three decisions on the same date. Finally this did not apply to all countries - while every rates C & D were higher than B, some rates D showed a deterioration from C (in simple terms the improvement was not sustained).

Despite all this, it is impossible for the natural peak to have occurred exactly when these decisions were taken for all ASEAN countries.

This puts into perspective the decisions taken in USA to allow for public gatherings as a form of liberation or giving exemptions to religious activities as .... comical, sad and criminal.