A second wave of covid-19 could have started in several European countries [1]. Is that however a wave of the same disease? While it is apparently - besides some genetic changes - the same virus, the number of hospitalizations and deaths is surprisingly small in regard to the number of people found to be infected by the virus. How is that possible? Surveillance bias and overdiagnosis help understand the current trend.
First, a surveillance bias explains the changing epidemiology of covid-19. Such a bias occurs when a condition is searched with differential intensity according to the setting of care or the type of patients [2, 3]. Covid-19 infection is a condition sensitive to the density of inquiry and screening activities. Currently, the tests to detect the SARS-CoV-2 are much more frequently done in many countries compared to the period of the first wave. Further, the patients tested are younger than previously. As a result, the typical covid-19 cases diagnosed have changed, moving from relatively old people with symptoms, tested positive, and at high risk of complications to younger low risk people with a polymerase chain reaction (PCR) positive result alone, regardless of symptoms [1].
Second, and this is related to the first point, one could argue that Covid-19 is overdiagnosed. Overdiagnosis occurs when people are diagnosed with a condition not associated with a substantial risk of adverse outcomes or for which no intervention reduces substantially this risk [3, 4]. Counting asymptomatic or paucisymptomatic cases of SARS-CoV-2 infection as Covid-19 cases is a form of overdiagnosis.
Overdiagnosis of Covid-19 does however not imply overdetection of SARS-CoV-2. On the contrary. Detection of all infections and tracing of close contacts is key to stop spread and contain outbreaks [5]. In that sense, there is no overdetection because it is useful, at a population level, to detect all persons with the virus.
References
1. Mahase E. Covid-19: the problems with case counting. BMJ 2020;370:m3374
2. Haut ER, Pronovost PJ. Surveillance bias in outcomes reporting, JAMA, 2011;305: 2462-3
3. Chiolero A, Santschi V, Paccaud F. Public health surveillance with electronic medical records: at risk of surveillance bias and overdiagnosis. Eur J Pub Health 2013; 23(3): 350-1.
4. Moynihan R, Doust J, Henry D. Preventing overdiagnosis: how to stop harming the healthy, BMJ 2012; 344: e3502
5. Baraniuk C. Covid-19 contact tracing: a briefing. BMJ 2020; 369: m1859
Competing interests:
No competing interests
06 September 2020
Arnaud Chiolero
Professor of Public Health and Epidemiologist
Population Health Laboratory, #PopHealthLab, University of Fribourg, Switzerland
Rapid Response:
Surveillance Bias and Overdiagnosis of Covid-19
A second wave of covid-19 could have started in several European countries [1]. Is that however a wave of the same disease? While it is apparently - besides some genetic changes - the same virus, the number of hospitalizations and deaths is surprisingly small in regard to the number of people found to be infected by the virus. How is that possible? Surveillance bias and overdiagnosis help understand the current trend.
First, a surveillance bias explains the changing epidemiology of covid-19. Such a bias occurs when a condition is searched with differential intensity according to the setting of care or the type of patients [2, 3]. Covid-19 infection is a condition sensitive to the density of inquiry and screening activities. Currently, the tests to detect the SARS-CoV-2 are much more frequently done in many countries compared to the period of the first wave. Further, the patients tested are younger than previously. As a result, the typical covid-19 cases diagnosed have changed, moving from relatively old people with symptoms, tested positive, and at high risk of complications to younger low risk people with a polymerase chain reaction (PCR) positive result alone, regardless of symptoms [1].
Second, and this is related to the first point, one could argue that Covid-19 is overdiagnosed. Overdiagnosis occurs when people are diagnosed with a condition not associated with a substantial risk of adverse outcomes or for which no intervention reduces substantially this risk [3, 4]. Counting asymptomatic or paucisymptomatic cases of SARS-CoV-2 infection as Covid-19 cases is a form of overdiagnosis.
Overdiagnosis of Covid-19 does however not imply overdetection of SARS-CoV-2. On the contrary. Detection of all infections and tracing of close contacts is key to stop spread and contain outbreaks [5]. In that sense, there is no overdetection because it is useful, at a population level, to detect all persons with the virus.
References
1. Mahase E. Covid-19: the problems with case counting. BMJ 2020;370:m3374
2. Haut ER, Pronovost PJ. Surveillance bias in outcomes reporting, JAMA, 2011;305: 2462-3
3. Chiolero A, Santschi V, Paccaud F. Public health surveillance with electronic medical records: at risk of surveillance bias and overdiagnosis. Eur J Pub Health 2013; 23(3): 350-1.
4. Moynihan R, Doust J, Henry D. Preventing overdiagnosis: how to stop harming the healthy, BMJ 2012; 344: e3502
5. Baraniuk C. Covid-19 contact tracing: a briefing. BMJ 2020; 369: m1859
Competing interests: No competing interests