Covid by Numbers review – how to make sense of the statistics
Along with successive waves of infection, the coronavirus pandemic has provided us with a tsunami of data and graphs. Thanks to the Public Health England dashboard and websites such as Our World in Data, every internet user can access accurate and timely information on Covid cases, deaths, hospitalisations and vaccines, broken down by age, gender and location.
However, while this wealth of information can be immensely valuable, it can also cause problems. Taken out of context and spun in a misleading way, raw coronavirus numbers can be a source of disinformation, which through social media can spread as efficiently as the virus itself. A simple fact, such as the median age of coronavirus victims (83) actually exceeding UK life expectancy at birth (81) can lead to governments and the public not taking Covid as seriously as they should. (Having lived to 83, one would ordinarily expect to live longer still – what matters is life expectancy conditional on having reached this age.)
Pieces of coronavirus data are like the ingredients for a meal: they can be combined in a variety of ways, with some recipes being more palatable than others and some being actively harmful. The right way to think involves not raw data but its analysis via the academic discipline of statistics. Here the inherent uncertainty and randomness behind the numbers is acknowledged and calibrated, along with issues of sampling, lag, under-reporting and sometimes differences in measurement methods.
There can be few better people to do this than David Spiegelhalter, a former president of the Royal Statistical Society, and Anthony Masters, the Society’s “statistical ambassador”. Readers of their Observer column will not be surprised that in this book they give a clear and extremely readable guided tour of the pandemic, mostly from a UK viewpoint. They present well-chosen numbers and graphs, backed up with exhaustive footnotes and references, with technical jargon kept to a minimum (though more details are available with the aid of an extensive glossary).
The book is divided into seven sections, with the first four representing a sequence starting with the virus itself, through diagnosis and cases, to the severity of disease, concluding with deaths. Two further sections consider the role of interventions such as lockdowns (studying their impact on the disease, the economy and on physical and mental health) and vaccines (for example, explaining the difference between efficacy in trials and in the real world, working out the frequency of side-effects and justifying the UK’s age-based strategy and spacing out of doses). The final part considers the key role played by mathematical modelling in our response to the coronavirus, and makes a series of valuable recommendations to government and journalistsfor more transparent data handling and numbers-based interrogation of policy.
The authors give a detailed and balanced account of the subtleties involved in measuring the impact of the virus. For example, in discussing deaths, Spiegelhalter and Masters describe the issue of reporting fatalities “from” and “with” Covid, compare risk by age and other underlying factors, consider the thorny issue of how to judge outcomes in different countries according to their data, and make comparisons between the impact of Covid and of other historical events.
It would have been easy for this kind of numbers-based account to become dry and miss the essential human context. ButSpiegelhalter and Masters never present Covid as a mathematical abstraction, always as a real disease with very personal consequences. So, rather than simply giving hospital admission data, they relate the numbers to the human effects on patients and their families, and to the exhaustion of healthcare workers, making it clear that “statistics alone cannot convey these sacrifices”.
Similarly, they delve into death data by age, occupation, ethnicity and medical circumstances, showing how headline numbers hide the very unequal individual impact of coronavirus. They do not shy away from the uncomfortable fact that through lockdowns and other restrictions young people “have sacrificed so much while being at low risk themselves”.
There is an issue, of course, about the right time to publish such a book in the midst of a global pandemic. Indeed, the text was finalised early on in the UK Delta outbreak, meaning some recent developments are not covered and some of the numbers themselves may already be slightly out of date. However, as arguments about vaccinations, past lockdowns and herd immunity rage on, this book represents an extremely timely contribution. It gives not just proper context for these discussions, but also suggests the right way to think about future events. If journalists, politicians and the public were all provided with a copy then the debate would be vastly better informed, with much more light than heat.
They do not shy away from the fact that young people 'have sacrificed so much while being at low risk themselves'