This has been extensively debunked. Read the original methods of the BMJ article that you linked. They took every single minor error, like prescribing medicine 15 minutes late, and if the patient died, even of an aggressive cancer that they had already, it would be counted in the 'medical error that caused the death' statistic.
https://www.nytimes.com/2016/08/16/upshot/death-by-medical-e...
https://www.medscape.com/viewarticle/863788?scode=msp&st=fpf...