Cases, cases and more cases...why do public health officials refer to Covid-cases so much? There are better indicators out there...

AboutAnything  | Greg McComb

Image by cromaconceptovisual from Pixabay

   It's the lead on the morning news; the afternoon news; and the evening news.....it's everywhere all the time. Updates on the number of Covid cases by city, province or country. Lately, it's been the inevitable march towards a second wave of the pandemic, as cases spike. It's making everyone nervous; some hysterical.

  The thing is, if you only hear about Covid-19 case counts; things do look a lot worse. That seems to be all that public health officials talk about, and what media reports on. Certainly, cases are spiking in some regions of Canada, and there are increased risks of infection. People need to remain vigilant and mask in public. However, there is a more nuanced story to be told about the pandemic by a diverse basket of indicators. The story these indicators tell is the second wave is definitely not as bad as the first. I will spend the remainder of this blog article explaining why.

           Older people taking cover during second wave

  Let's start with the age of infected people. We heard early on that youth have much stronger immune systems able to fend off the coronavirus with a strong 'first punch,' (see link) while the weakened immune systems of the elderly are easy prey, (click link). 

  Statistically, it was a pretty sharp divide during the first wave: 95% of the deaths in Ontario and 97% in Quebec were people over 60, (link to April blog).

 Things have changed a lot over the past seven months. In short, seniors are taking cover (staying indoors and masking outdoors) and fewer are getting infected. Take a look at the graph below prepared by the Ontario government. There are two sharp drop offs, identified with arrows. The number of people over 80 (orange line) testing positive dropped off

Chart from "Public Health Ontario Daily Epidemiologic Summary;" arrows inserted by author

from just over 30% of the total to under 5% while the 60 to 79 age group, (arrow, light blue line) fell from 50% of the total to 10% in late September.

   So, the composition (cohort) of the second wave has changed a lot. Only about 12% of the people infected are in a high-risk elderly group. Therefore, we can expect fewer hospitalizations and fatalities in the second wave...and that's what we've been getting so far.

  So, a more comprehensive suite of indicators could include a number that summarizes this change, such as percent of cases over and under 60-years-old. This would capture the age-related risk of the infected population.

    Yes, cases have been spiking but fatalities have not

  So, let's take a look at Canada's case-based pandemic curve; the one that's reported all-the-time in the media.  Quite frankly, this curve does look scary, (see below). It's currently at 2,000 cases-a-day, (Oct. 8th) which is about 200 more than peak first wave in mid-May. If we use the case curve 

Data from Public Health Agency of Canada, author constructed graph

 

 

 

as the sole indicator, we would expect hospitals to be full and daily fatalities in the hundreds, similar to the first wave.

 However, this isn't happening. Although cases starting trending upwards in late July, there was no corresponding uptick in fatalities. The death-based pandemic curve flat-lined until late last week, (see below) when we saw an uptick in fatalities across Canada.

         Data from Public Health Agency of Canada, author constructed graph

  Hospitalizations are the same. Although we hear all the doom-and-gloom that hospitals will soon reach capacity and be overwhelmed, the reality is quite different. Daily Covid-19 patients hospitalized in Ontario in the last month have been a fraction of what they were during the first wave, (see below), although there could be localized issues. 

Data from Ontario government data catalogue, author constructed graph

                Better treatments during second wave...

So, why is this happening?: lower hospitalizations and fatalities in the second wave. I already discussed the lower age-risk of the infected population. Another likely cause is that front-line doctors better understand the pathology of the coronavirus, and improved treatments have been developed.  Over the past seven months, pathologists have performed  hundreds of autopsies worldwide and dozens of drugs have been tested in clinical settings.  As a result, when a patient presents at an emergency room with advanced Covid-symptoms, doctors know better how to manage patients and administer life-savings drugs. An insightful Newsweek article (click here) outlines some of those treatments, such as the anti-coagulation drug heparin used to treat tiny blood clots discovered during autopsies. Another is the use of certain steroids and immune modulating drugs to prevent the 'Cytokine Storm' that attacks patients' respiratory systems. The result: fewer patients on respirators who have a 50/50 chance of surviving.

     So along with Covid-19 cases and a measure of age-risk, a more comprehensive set of indicators could include trends in hospitalized patients and fatalities. These indicators provide clues to important-positive changes to the pandemic such as better treatments that result in lower fatality rates and lower age-risk due to a shifts in demography. Not so much doom-and-gloom.

    Not a fan of Covid case counts; they are not very reliable

    So, I've taken a look at other indicators that could provide a more comprehensive snapshot of the pandemic. How about Covid-case counts; the statistic governments rely on so much when making decisions about opening or shutting down an economy?  Is it a good indicator? My short answer is no; it's not. It's not very reliable and depends critically on the number of people that are tested: testing intensity. I raised this issue in an earlier blog when I compared pandemics across countries, (see link). I argued that countries that did more Covid testing would have more cases and appear worse off; so I used fatalities to build pandemic curves.

   Something different is happening now. As countries open up around the world, they are testing a lot more, sometimes double or triple what they were a few months ago. Mostly, health authorities are doing this so they can track localized outbreaks, and take action such as closing down a school and contact tracing infected people.  In Canada, (see graph below) daily testing increased from about 20,000 a day in March and April to the current 70,000 tests per day. Other countries increased their testing even more.

Chart generated from website: https://ourworldindata.org/coronavirus

   So, how is this affecting case numbers; are we inadvertently inflating case counts as we trace local outbreaks? We probably are. The main reason is the coronavirus is a sneaky virus; it lays low in the bushes. Most people who have it actually don't know they do; they have few or no symptoms. This is why the respected Centers for Disease Control estimates that ten times as many cases are out there than reported by testing, (see link). If a country tests more, it will simply pick-up the other ninety percent of cases, even though community transmission is about the same.

  So, how do you test this? I developed a quick-and-dirty model that holds Covid-testing constant (the same) over the first and second waves, and multiplied this by the actual daily positive rate for testing.  Let's see if I can explain this more simply. Pretend that instead of staffing up dozens of new testing facilities in community centers, clinics and drug stores we had the same number of testing facilities that we had back in April and May.  What would our second wave look like?  The simple answer is: quite a bit lower. I first ran the model for the United Kingdom and it shrunk their second wave by less than half, (see orange wave in graph below).

Data obtained from website: https://ourworldindata.org/coronavirus; chart constructed by author

  I then ran the model using Canadian data and got similar results although not as pronounced because our testing push has not been as big in Canada, (see graph below, orange).

Data from Public Health Agency of Canada, author constructed graph

   The second wave looks a lot less scary, although there is a certainly reason for concern that cases and the risk of infection are on the rise.  

  Percent positive Covid tests...a reliable indicator of community transmission

  While setting up the adjusted-case model, I stumbled across data for 'percent positive Covid tests,' something we don't hear a lot about in the media or discussed by public health authorities. Simply put, if we test 1000 people-a-day and ten are positive, the positive rate would be 1%, and community transmission would be deemed low, (see link for explanation). If 50 or more test positive, the positive percent would be five or higher and the risk of community infection would be cause for concern.

  We can see that early on in Canada's pandemic, (see graph below) our positive rate was around 10%, so community transmission was very high, double the threshold. After several months of quarantines, the positive rate dropped to below one and it is now inching up to three, which is cause for concern but still below a risky threshold, (see red line). Certainly, not as scary as second-wave case data...

 I really like this indicator and I don't know why health authorities don't use it more: it's reliable because testing intensity doesn't affect it, and public health experts agree it provides a good indicator of community transmission. As a sampling technique, percent positive tests also provide insight into how big the elusive population of infected people is - something that will never be known.

Data obtained from website: https://ourworldindata.org/coronavirus; chart constructed by author

                                 What next.....

   As discussed in the opening, the daily march of ever increasing cases has become a 'policy tonic' for governments to act: to dial-back openings by closing bars, restaurants, theatres and gyms. Just as the Quebec government has recently done in three hot spots including Montreal and Quebec City.  My argument isn't that I oppose shutting down regional hot spots to contain the virus; it's that we should have a more diverse and comprehensive set of indicators from which to base decisions. Actual or raw cases should be the lead indicator, but governments and media should also view the age-risk in a community, local hospitalizations, an estimate of fatality rates and how well prepared local hospitals are with cutting-edge treatment. Add to that a reliable measure of community transmission (positive Covid tests) and cases adjusted for testing intensity, and you have a solid set of indicators. Simply shouting 'claims are spiking; hospitals will soon be overrun' isn't good enough. It simply adds to the hysteria. 

   The other side of the coin in this pandemic is the economy (see my June blog). Every time a government closes it down, we can expect more-and-more bankruptcies and a more protracted recession and greater public debt. As lock downs last longer and longer, mental health issues also become more prevalent. Dentists have even noticed a record number of patients with cracked teeth due to teeth grinding, a sign of restlessness and anxiety,(see link). Early in the pandemic, (click for April blog) I commented that it will be the job of governments and health authorities to take measures to eradicate this scourge, while balancing the needs of health against economic costs, essentially on the head of a pin. My view is that governments in Canada have so far done a pretty good job. 

   My concern is this final lap before a vaccine is found; that we gradually ease restrictions backed by a good set of scientifically-backed indicators. In recent months, most of the messaging has been around the 'scare of a second wave,' constructed with case data that's not very good. Based on this broad scan of indicators, my view is the second wave won't be nearly as bad as the first... 

  Please prove me wrong!

Read more »