A tale of two testing strategies in Italy for covid-19

The first patient believed to be infected with covid-19 through community transmission in Italy was diagnosed at a hospital in Codogno, a village near Milan in the Lombardy region on 18 February 2020. Three days later 14 people in the area of Codogno and two in Vò Euganeo, a small village of the neighbouring region of Veneto, were confirmed to be infected.1 The epidemic spread quickly across North Italy, prompting the Italian government to initiate a progressive lockdown, and by 9 March it covered to the entire country.

Lombardy remained the epicenter of the epidemic, and by 24 April the number of infections had reached 71,256 and the apparent lethality rate was 18.4%.2 By contrast in Veneto, the epidemic progressed more slowly and had an apparently lower lethality: by 24 April there had been 17,229 cases and the apparent lethality was 7.2%.2

So, what could have been responsible for this difference?

The lockdown measures had been the same in both regions, but what had been different was the testing strategies employed. While most regions strictly followed the WHO’s and central health authorities’ indications and restricted viral RNA testing to symptomatic people, Veneto implemented much more extensive population testing.

After the first infection in the village of Vò Euganeo was identified, all 3,500 people living in there were tested for covid-19 as part of a pilot study, on the suggestion of Andrea Crisanti, director of microbiology and virology at the University of Padua.3 The results showed that 2.6% of the population of Vò Euganeo were infected, and that 43.2% of these were asymptomatic.3 

Those people infected were asked to self-isolate at home with their families. Follow-up testing 7-10 days later showed that asymptomatic patients had transmitted the infection to other family members.3 Quarantine of all positive patients, both symptomatic and asymptomatic, was effective in stopping viral transmission and no new cases were reported in Vò Euganeo from 13 March.3

The Vò Euganeo strategy – to test broadly and isolate rigorously – prompted debate among experts across Italy and around the world about the best testing strategy.3,4 The finding that asymptomatic and otherwise undetected infections could be drivers of the epidemic conflicted with messages coming from Chinese authorities and the World Health Organization.4,5 Other Italian regions, like Tuscany, followed a strategy similar to Veneto ramping up quickly the testing capacity; by 24 April total cases were 8,877, and new cases soon became stable below a hundred per day. By contrast, regions like Piemonte, that had few cases in the initial phase but focused testing on symptomatic patients, like Lombardy, had 23,822 cases and 682 new cases by 24 April.2

Now it appears clear that focusing attention only on symptomatic patients was a crucial mistake in the initial handling of the epidemic in Italy. The assumption by the central health authorities that symptomatic patients were primarily responsible for transmitting the disease meant that the containment measures were insufficient to control the spread.5

Use of masks by the population to reduce spread of the virus, including by people who were asymptomatic, was not recommended. The response was focused on hospitals, which filled quickly with symptomatic patients and these infected doctors and nurses werenot sufficiently prepared to deal with them. Hospitals, as well as retirement homes, became disease transmission hubs, with the infection transmitting from patients to staff and from staff to patients. The high prevalence of the infection among old and fragile people, where lethality for covid-19 is very high,6 as well as the underestimation of the prevalence of the infection likely explain the high lethality observed particularly in some regions.

The Italian experience clearly suggests that, although essential to reduce the exponential growth of the epidemics, lockdown measures alone are insufficient to control the pandemic, particularly long term. Extending testing to asymptomatic people, as performed not only in Veneto, but also in South Korea, Hong Kong, and Taiwan, is essential for controlling regional outbreaks and to set up an exit strategy now that a reduction of cases is observed. This does not mean that all the population needs to be screened, rather that the right categories are screened. These include staff in contact with patients, such as doctors and nurses, hospital support staff, pharmacists and workers in the retirement homes, where infection displayed the highest lethality.

The regular and extended testing of healthcare workers in hospitals and care workers in retirement homes is essential to avoid spread of infection and to reduce death rates not only from covid-19, but also from other conditions. This can be achieved by establishing hospitals dedicated to the treatment of patients staffed by personnel who is adequately protected and periodically checked or eventually, immune to the infection.

Capacity for rapid diagnostic testing for viral RNA, follow-up of all the contacts of every infected patient, and searching for asymptomatic carriers is crucial to control the epidemic, even beyond the lockdown measures.

Paola Romagnani is Professor of Nephrology at the University of Florence, Italy

Sergio Romagnani is Professor Emeritus at the University of Florence, Italy

Competing interests: none declared


  1. Coronavirus: 16 new cases in Italy – ANSA
  2. Covid-19 Italy situation report – intelworks
  3. Lavezzo et al. Suppression of COVID-19 outbreak in the municipality of Vo, Italy. medRxiv 2020.04.17.20053157
  4. World Health Organization: Novel Coronavirus Situation Report – 12 February 2020
  5. Coranavirus – Trasmissione, prevenzione e trattamento – L’epidemiologia per la sanità pubblica
  6. Ruan S. Likelihood of survival of coronavirus disease 2019. Lancet Infect Dis. 2020 Mar 30. pii: S1473-3099(20)30257-7