Nepal experts reject Covid-19 model

Seven noted public health experts have described an Oxford University study three months ago that projected 50,000 Covid-19 deaths in Nepal by October as dangerously alarmist, and which could have led to flawed policy decisions. 

The study was commissioned by the British aid agency DFID and predicted a peak of the pandemic in Nepal from June to September with an astounding 846,000 new cases a day in the worst-case scenario.

Titled Modeling of COVID-19 Strategies in Nepal, the report was recently made public by UKaid, Oxford Policy Management and University of Oxford, and spread alarm when it was carried by the media. 

‘Nepal is fighting against Covid-19 pandemic within its capacity, and so far it is doing well compared to global average,’ read a statement by the public health experts. ‘Projections such as this are the basis for further planning to mitigate the risk and hence should be based on available data and rational assumptions close to reality. A wrong prediction will lead to misguided policies and strategies, resulting in failure to address the actual problem.’

Issued by Mahesh Maskey, Sharad Onta, Badri Raj Pande, Aruna Uprety, Sameer Dixit, Abhinav Baidhya and Rajendra BC, the statement questioned the validity of the study, calling it misleading and one that could have created widespread panic among the general public.

The spread of SARS-CoV-2 has gained tempo in the past month in Nepal with a record 1,228 new confirmed cases on Thursday, with 481 of them in Kathmandu Valley alone. There were 6 additional deaths, bringing the total to 257. However, there were also 768 recoveries, highest in a single day and the total recoveries now stands at 24,207.

Despite these figures, the projections in the study even for the best-case scenario seem to be way off. The model predicted:

In case of low disruption strategy, the peak of the epidemic will extend from early June to early September, reaching a maximum of nearly 846,000 new COVID-19 cases per day with less than 10% of cases (75,000) reported to the health system and 49,200 cumulative deaths due to COVID19 by the end of the year.

In case of medium disruption strategy, the peak of the epidemic will extend from mid-July to early October, reaching a maximum of nearly 548,000 new COVID-19 cases per day with less than 10% (50,000) of cases being reported to the health system and around 5,000 less deaths compared to low disruption strategy.

In case of high disruption strategy, peak of the epidemic will extend from mid-June to late October, reaching a maximum of nearly 531,000 new cases per day with less than 10% of cases (46,800) reported to the health system and 11,300 less deaths compared to medium disruption strategy.

Completed in May the modelling does not have monthly figures for new cases and deaths, but includes complicated graphs that indicate only daily case peaks. For example, as per the medium disruption scenario, there should be a total new 548,000 cases a day in the last week of August. 

And if we are to assume that less than 10% of these were actually reported per day, as suggested by the study, there should have been some 50,000 new cases a day, except this is 50 times higher than the actual cases reported per day last week.

The Nepali experts say that instead of helping Nepal mitigate the Covid-19 crisis, the study added to the burgeoning challenge of testing, tracing and treating coronavirus patients. 

They added: ‘The scale of panic aside, what if government had accepted this report and spent its scarce resources on preparing hospital beds and ventilators at a massive level? Would it be a proper use of resources or misuse?’

The experts also questioned the basis for assuming such a high incidence and mortality in Nepal, and for assuming that less than 10% of the total cases would be reported to the health system. 

‘Given that the predictions are widely off the mark so far, why should Nepal trust the projections for rest of the year?’ they ask, adding that the authors of the study should clarify the assumption before they cause further damage.