Flattening the curve? Not yet.

At the time of penning this piece, Pakistan’s confirmed positive COVID-19 cases had just crossed the 10,000 mark. There is noticeable alarm, and rightly so, in reaching this grim milestone. Some context is needed at this stage, to interpret the numbers with care.

There is universal consensus that until mass infection or mass vaccination (66% of the population) takes place (the former of which is an unacceptable outcome and the latter seemingly a long-term possibility), the virus won’t die down on its own. Therefore, the only thing worth trying in these harrowing times is to slow the spread of the virus. That’s the whole point of social distancing and lockdown measures, albeit with grave consequences for living standards for many years to come.

When the epidemic is following an exponential function, looking at absolute numbers can be misleading. Pakistan is still in its early stage of outbreak, but the data collected thus far can still be analyzed from a few angles. For instance, how long is it taking for the confirmed positive cases to double? And is this rate of doubling increasing or decreasing?

Data from the government’s covid.gov.pk website show that Pakistan had crossed 2,000 cases on March 31. After that, it took 7 days to double the count to 4,000, on April 7. Those cases took 11 days to double to 8,000, on April 18. (It is unclear which factor is more at work here: the cross-country lockdown that lasted until April 14 or the limited level of testing). If the next doubling to 16,000 cases took more than 11 days (April 30), it will be significant. But as of yet, it is premature to say yet that the spread has stabilized.

Analyzing the data since April 1, a time after which case reporting had streamlined and local transmission became significant, the daily growth rate in new cases averaged 7.8 percent. More recent growth rate is just above 7 percent. If the daily cases continued to grow between 7 to 8 percent, Pakistan would cross the mark of 25,000 cases around May 4, it will clock 50,000 cases around May 14, and there will be 100,000 cases around May 24. Keep watching the daily growth rate in the month of Ramazan!

A different picture emerges when one analyses weekly data, which is more reliable because testing and reporting delays are somewhat mitigated in this frame. There were 2,035 new cases reported in the first week of April; some 1,911 cases came out in the second week of April; and much higher 3,764 cases cropped up in the third week of April. It is tempting to attribute the spike in third week of April to the easing of the first lockdown that ended April 14. However, there is no data to support that leaning yet.

In reality, the third week of April was also when testing picked up pace – the average daily tests more than doubled to almost 6,000 tests per day, compared to fewer than 3,000 daily tests averaging in the preceding two weeks. So the number of new cases was bound to get higher, too. On an encouraging note, in fact, the share of confirmed positive cases in total tests conducted had declined to 9 percent in the week ending April 21, compared to 9.5 percent for the week ending April 14 and 12 percent for week ending April 7. The current infection rate is within the WHO’s recommended limit of 10 percent.

Be that as it may, the real scale of infections will remain hidden until large-scale testing takes place. (For more on the need for active, random testing, read “Get real and test more,” published April 16, 2020). Otherwise, an uneven testing pattern will continue to throw up erratic numbers on daily additions to positive case, resulting in many small peaks before things really get ugly.

Early evidence suggests that countries that have managed to tame SARS-COV-2 all have in common the tactic of aggressive testing from the onset. Pakistan is now almost two months into the throes of COVID-19 since the first case was reported, but numbers-wise the country is still at the start of this outbreak. However, data from Oxford University’s Our-World-in-Data dashboard show that Pakistan’s average of 12 tests per confirmed case metric is very low compared to countries that had a testing head start.

As coronavirus has a basic reproduction number of 3 on the high side, perhaps the good folks at NCOC should have a random testing target that is 3 times the number of total projected infections on a going basis. (Besides ramping up random testing, the idea is to test at least 3 close contacts for all positive cases). That would require conducting at least 30,000 tests per day from tomorrow. It’s an ambitious target, but the federal and provincial governments need to act fast if Pakistan is to get ahead of the curve.