Only hospital deaths are counted, without testing daily caseload total is meaningless
In its daily briefing on Saturday, the Ministry of Health announced that 4,311 people had tested positive for Covid-19, and 116 more people had died in 24 hours from the pandemic.
Experts were encouraged by these figures, and took it as a sign that the second wave had peaked. Indeed, the daily confirmed cases had dropped to half the level a week ago, and so had the positivity rate. Fatalities were also down from nearly 200 a day earlier this month.
However, a new study shows that pandemic statistics the world over are fraught with inaccuracies, and the Nepal figures are also a gross underestimation.
The calculations show that new daily infections were probably closer to 120,000 on Saturday, and there were more than 700 deaths from Covid-19 nationwide—more than 30 times higher than official figures.
The study by the Institute for Health Metrics and Evaluation (IHME) of the University of Washington School of Medicine in Seattle goes beyond official confirmed figures to calculate ‘excess mortality’ and undetected Covid-19 infections.
“As terrible as the Covid-19 pandemic appears, this analysis shows that the actual toll is significantly worse,” said IHME Director Chris Murray, revealing the results of the analysis this month.
“Understanding the true number of Covid-19 deaths not only helps us appreciate the magnitude of this global crisis, but also provides valuable information to policymakers developing response and recovery plans.”
The official figures for Nepal are frightening enough. The country has the world’s second-highest bi-weekly increase in deaths at 291%, and the highest national test positivity rate of 40%.
Yet, Nepal conducts only 713 tests per million people, contact tracing is virtually non-existent, less than 2% of the 30 million population is fully vaccinated, and mask-wearing is at less than 65%.
IHME’s graphs for Nepal show that during the early May peak when new confirmed cases were above 9,000 every day, the actual daily number was probably closer to 350,000.
Nepal’s Ministry of Health’s total tally of fatalities is now 7,163, but the modelling shows that the actually total is 28,256.
Epidemiologists warn that if the official figures are so wrong, then the planning to meet the requirements for everything from test kits, hospital beds, ICU, oxygen requirement, ventilators, and even funerary planning would need to be re-evaluated and up-scaled.
“We have known that under-reporting is widespread. The figures are underestimates, but the impact cannot be underestimated,” says virologist Sher Bahadur Pun at Teku Hospital.
“For example, although we are just beginning to see a plateauing of the second wave, the increase in new cases is constant and the strain on the health system remains the same.”
The IHME uses the “excess mortality” model to calculate a more accurate figure for the infection and death rate from Covid-19 by using data for previous years, and factoring in new variables to calculate how many more people are actually dying during the pandemic compared to pre-pandemic years.
There are many factors affecting mortality during the pandemic: deaths directly due to Covid-19, increase in mortality because patients with pre-existing conditions are not getting the hospital care they need, or deaths due to mental disorders.
However, there is also a reduction during the pandemic of deaths caused by high air pollution, fatalities from other air-borne infections like tuberculosis, measles or influenza, and even a drop in highway-traffic accidents.
So, IHME modellers collected available data on excess mortality from Covid-19, added an estimate the additional deaths that were not registered to came up with the total deaths from coronavirus.
The results are shocking. The actual global death toll from Covid-19 is more than double the 3.5 million official figure.
In India, the government’s figure for total fatalities till Saturday stood at 325,998, but the projections show that the actual deaths probably at more than 1 million.
Although the death rate is going down, IHME says India’s cumulative deaths from Covid-19 could be 1.2 million by 1 September.
In Nepal, even official data shows a steep spike during the whole of May as Nepali workers started arriving from India with the fast-spreading B.1.617 variant.
But on 1 May when the confirmed daily Covid-19 deaths was showing only 32, the model shows that 117 people were already dying every day nationwide from the disease.
Within two weeks, even when the official total had soared to 205, the graphs show that the excess daily mortality was already 742.
The reason for the discrepancy was that mostly hospital deaths were being counted, and the death registration system is inaccurate and late.
Epidemiologist Lhamo Yangchen Sherpa says the graphs just prove what everyone knew all along: testing is inadequate and even the distribution of testing facilitates is lop-sided.
For example, 29 of 42 labs with RT PCR testing capacity in Bagmati Province are clustered within Kathmandu Valley. The whole of Gandaki province is reliant on three labs in Pokhara.
She adds: “People in remote parts of Nepal like Solukhumbu have to walk for days to give their swab sample, which is then sent to a lab in Biratnagar. By the time result is back, a week has passed by and many of them are already sick.”
Most of the returnees from India were never tested, and went straight home without quarantine, becoming data blind spots and skewing transmission numbers beyond Nepal’s limited testing capacity.
Even though the national case numbers have gone down, experts say it is raging across rural Nepal—but there is no way to quantify it because of the lack of testing.
Epidemiologist Sherpa recommends rapid turnaround of antigen tests to reduce time lost waiting for PCR results in rural Nepal.
She says: “Patients show symptoms in five days but are contagious from the third day onwards. To avoid further spread in that crucial 48 hours, we must increase access to tests but also bring down the time it takes to get the result.”
But targeted testing of symptomatic patients in remote districts like Rolpa and Humla have shown a staggering 90% positivity rate.
Because of the lack of hospital care and the absence of medical oxygen, many die at home and the numbers never feed into the daily count announced in Kathmandu every evening.
Virologist Pun at the Teku Hospital says effective contact tracing would provide the most accurate situation report, but that is not happening.
He says: “The 70-80 daily new cases during the first wave were mostly asymptomatic travellers needing mandatory checks and their traced contacts. But in the more severe second wave, there is no contact tracing at all.”
Experts like Moshfiq Mobarak of Yale University say that the most effective strategy for countries with fragile health systems like Nepal to reduce fatalities is a three-pronged approach that aggressively promotes mask-wearing, responds to localised surges, as well as last-mile vaccine delivery.
Mask-wearing alone has been proven in a trial in Bangladesh to show that a vigorous campaign for no-cost mask distribution, aggressive awareness drive, and punishment for non-wearers tripled the number of people using masks properly.
“This is an extremely cost-effective way to reduce transmission risk and save lives,” Mobarak said in an op-ed he co-wrote for Nepali Times last week.
Indeed, the IHME model also proves that direct impact of mask wearing in Nepal. The modelling shows that the total coronavirus death toll in Nepal by 1 September will be at 44,711 if mobility went up to pre-Covid levels.
However, if 95% of Nepalis wore masks in public at all times, the number of deaths would be reduced to 42,500. This means 2,211 lives would be saved just by universal mask-wearing in the next three months.
Even though Nepal’s second wave may be subsiding, experts say the country will take longer to come out of it than neighbours India or Bangladesh despite the lockdown.
This is because of the lack of vaccines, mask-wearing at less than 65%, as well as lack of mass testing. Nepal’s seven-day decrease in new caseloads post-peak is estimated to be around 7.45% compared to India’s 12.8%.