Coronamonomania Lasts Forever, Part 234

Excess death analysis is rife with pitfalls for serious error.  How many deaths would really have been expected in a period in the absence of a particular event, like the epidemic?  What other events might be occurring at the same time that affect the number of deaths or the trend in deaths?  To what can a particular death be attributed?  If there are excess deaths in a period, how does that affect deaths in subsequent periods?  This Lancet study discusses several of the issues, starting with just using a five-year average, instead of using a longer period, using average not trend and determining trend by age group and cause of death.  According to one model, the UK continues to experience excess deaths following the peak of the epidemic through June of this year about 8% more.  50 to 64 year olds experienced the greatest rate of excess deaths.  Heart failure, liver diseases, other cardiovascular disease and diabetes contributed to the excess deaths.  In that 50 to 64 age group, heart diseases were the biggest contributors with diabetes not far behind.  Those are diseases particularly reflective of missed regular care.  A big proportion of the excess occurred at home, with fewer deaths in nursing homes or hospice, also a reflection of potential declines in the speed of ambulance response.    (Lancet Article)

Many viruses find ways to persist in the body after an acute infection, so it isn’t surprising that CV-19 may also have this capability and a recent study in primates suggests that the virus may even use some types of immune system cells as a refuge.    (Nature Study)

It is so easy to make mistakes with data and statistics and lots of people intentionally take advantage of that to claim that research or analyses support a point of view they already hold.  The epidemic has provided ample opportunity for that practice.  Here is a short post on one common source of error or bias in statistical analyses on vaccine safety.  (WATN Post)