Counting deaths 'with Covid' inflated mortality rate, B.C. data indicates

by George J. Dance

New data from British Columbia sheds light on one of the most contentious social media arguments that raged throughout the coronavirus pandemic, the dispute over 'deaths with Covid' versus 'deaths from Covid'. During the pandemic, many countries used the 'deaths with' standard: fatalities were reported as Covid deaths if they occurred within a reasonable time (usually a month) after a positive PCR test. The U.S., Sweden, and the U.K. all reported 'deaths with' Covid figures; not coincidentally, all have higher Covid mortality rates. Other countries, such as China, counted only 'deaths from': the only deaths included in the Covid count were those verified, by an investigation (a postmortem or even an autopsy), to be caused by the virus; again not coincidentally, China has one of the world's lowest reported Covid death rates. 

A big deal? Not to me. My main interest was in comparing countries, and so long as most most of them used 'deaths with', I could happily compare apples to apples. However, social media witnessed many bitter battles waged over it. Whichever standard one uses will affect the reported figures. Counting all 'deaths with' will give a higher figure, artificially high if it miscounts clearly non-Covid deaths (such as poisoning or accident wounds), and using it exaggerates or inflates the virus's impact. But how much higher?

Some Covid libertarians (opponents of Covid mandates) argued that public health authorities were using the 'deaths with' standard to deliberately exaggerate the death toll. In turn, the pro-mandate crowd argued that counting 'deaths from' would give an artificially low figure (by excluding uninvestigated deaths), and accused Covid libertarians of trying to trying to minimize the toll. The ongoing debate generated much heat, with one side accusing the other of alarmism, and the other accusing it in turn of denialism. However, for me it generated little light; whatever actual information emerged was merely drowned out in the tide of mutual recriminations. Many others also appeared to get little from the debate, as they argued in turn that it was a debate about nothing – that the distinction made no difference, as the differences between the two counts would end up being insignificant..   

Which is why I found recent news out of British Columbia worthy of comment. As per a May 26 story on CTN News, B.C.'s Center for Disease Control (BCCDC) "changed the way it counted COVID-19 deaths back in early April". The province had previously reported mortality using a 'death from' standard: "Before April 2, B.C.'s health authorities manually updated the pandemic death toll after investigating each death to determine if COVID-19 was a factor." After that date, the reporting standard became 'deaths with': "anyone who dies within 30 days of a COVID-19 diagnosis has been automatically flagged and reported as a death possibly caused by the disease."

"The province's Vital Statistics agency reviews each possible COVID-19 death and determines what the underlying cause was, a process that can take as long as eight weeks," CTV reporter Ian Holliday notes. Before April 2, then, it could take as long as eight weeks before a death was added to the provincial total. Switching to 'deaths with' makes for more immediate reporting, which is probably the reason most countries use it. The agency still investigates all deaths, though, and the data for those verified as being caused by Covid are released later. That gives a reader the chance to compare the two figures, and see how much higher (if any) the 'deaths with' figure is. 

Holliday reports: 

According to the BCCDC's weekly "situation report" released Thursday, there were 424 deaths between April 2 and May 14 that were flagged as potentially caused by COVID-19 because the person who died had tested positive within 30 days of their death. 
Of those, more than half (218) were still pending review by Vital Statistics to determine the underlying cause of death.
Of the 206 deaths for which an underlying cause had been determined, 94 were considered to be caused by COVID-19, while the remaining 112 had some other underlying cause.

As Holliday notes, that 94/206 ratio "works out to slightly less than 46 per cent of reported deaths between April 2 and May 14 for which an underlying cause has been determined." Or, as summed up in the headline: "Fewer than half of COVID-19 deaths reported since B.C. changed counting methods were caused by the disease."

These preliminary statistics (which may be the only ones that do get reported) allow us to compare the two counting methods: counting 'deaths with Covid' will give a figure at least twice as high as counting 'deaths from Covid'. That in turn allows one to compare reported deaths between jurisdictions which use different methods. 

Of course, these findings are preliminary. One should ideally look at all the data, from as many jurisdictions as possible. B.C.'s figures are only one data set, still incomplete, and possibly an anomaly. So this report's findings are hardly the last word on the subject. They are, though, an important first word, one which we have sorely lacked throughout the pandemic.  

Source:

Ian Holliday, "Fewer than half of COVID-19 deaths reported since B.C. changed counting methods were caused by the disease," CTV News, May 26, 2022. https://bc.ctvnews.ca/fewer-than-half-of-covid-19-deaths-reported-since-b-c-changed-counting-methods-were-caused-by-the-disease-1.5920782