Old people, Covid-19 vaccines, and the NNTV controversy
Dear Editor
On December 1, 2020 I received an email from the distinguished author of an American textbook of pediatric infectious diseases. He was severely critical of my November 13, 2020 rapid response (https://www.bmj.com/content/371/bmj.m4374/rr-4), particularly my calculation that 256 individuals would have to receive Pfizer’s vaccine to prevent a single Covid-19 illness. In the language of the clinical epidemiologist, 256 is the Number Needed To Vaccinate (NNTV). It is the vaccine equivalent of the Number Needed To Treat (NNTT), which is used to measure the real-world benefit of drugs and procedures for individuals. (Sackett et al, Clinical Epidemiology, 1991) Application of the concept to vaccines is controversial. Most public health authorities emphasize the population benefit of vaccines and herd immunity over individual benefit, and it is no surprise that manufacturers support this emphasis because it fosters greater vaccine uptake. Others believe that NNTV is a useful measure of the impact and cost-effectiveness of immunization programs. (Hashim et al, Vaccine 2015;33:753)
Fresh questions about safety and effectiveness have been aroused by a report of 33 deaths in elderly Norwegians shortly after their first dose of Pfizer’s Covid-19 vaccine. (Torjesen, BMJ 2021;372:n167, Jan 19) Our knowledge is limited, but NNTV is a straightforward way to quantitate individual benefit. The latest published information indicates that the absolute risk reduction (ARR) by the vaccine for any Covid-19 illness is 0.0046 for individuals 65 and older. (Calculated from Table 3 in Polack et al, NEJM 2020;383:2603, Dec 31) The NNTV is 1/0.0046=217. This means that 216 old people will derive no benefit but will be subject to adverse effects….There are at least two qualifications for this calculation: first, the NNTV for severe illnesses and deaths is likely to be much higher; second, NNTV is likely to change with time as the calculated ARR and vaccine effectiveness change with time.
I hope the foregoing will not be misquoted or used on social media (which I rigidly avoid) in a misleading way. Meanwhile, we still have a lot to learn about Covid-19 vaccines, something we should know from our experience with influenza vaccines (https://www.bmj.com/content/371/bmj.m4037/rr-3).
Rapid Response:
Old people, Covid-19 vaccines, and the NNTV controversy
Dear Editor
On December 1, 2020 I received an email from the distinguished author of an American textbook of pediatric infectious diseases. He was severely critical of my November 13, 2020 rapid response (https://www.bmj.com/content/371/bmj.m4374/rr-4), particularly my calculation that 256 individuals would have to receive Pfizer’s vaccine to prevent a single Covid-19 illness. In the language of the clinical epidemiologist, 256 is the Number Needed To Vaccinate (NNTV). It is the vaccine equivalent of the Number Needed To Treat (NNTT), which is used to measure the real-world benefit of drugs and procedures for individuals. (Sackett et al, Clinical Epidemiology, 1991) Application of the concept to vaccines is controversial. Most public health authorities emphasize the population benefit of vaccines and herd immunity over individual benefit, and it is no surprise that manufacturers support this emphasis because it fosters greater vaccine uptake. Others believe that NNTV is a useful measure of the impact and cost-effectiveness of immunization programs. (Hashim et al, Vaccine 2015;33:753)
Fresh questions about safety and effectiveness have been aroused by a report of 33 deaths in elderly Norwegians shortly after their first dose of Pfizer’s Covid-19 vaccine. (Torjesen, BMJ 2021;372:n167, Jan 19) Our knowledge is limited, but NNTV is a straightforward way to quantitate individual benefit. The latest published information indicates that the absolute risk reduction (ARR) by the vaccine for any Covid-19 illness is 0.0046 for individuals 65 and older. (Calculated from Table 3 in Polack et al, NEJM 2020;383:2603, Dec 31) The NNTV is 1/0.0046=217. This means that 216 old people will derive no benefit but will be subject to adverse effects….There are at least two qualifications for this calculation: first, the NNTV for severe illnesses and deaths is likely to be much higher; second, NNTV is likely to change with time as the calculated ARR and vaccine effectiveness change with time.
I hope the foregoing will not be misquoted or used on social media (which I rigidly avoid) in a misleading way. Meanwhile, we still have a lot to learn about Covid-19 vaccines, something we should know from our experience with influenza vaccines (https://www.bmj.com/content/371/bmj.m4037/rr-3).
ALLAN S. CUNNINGHAM 28 January 2021
Competing interests: No competing interests