To face the Covid-19 pandemic, most European countries, and in particular those analyzed in this study (France, Germany, Italy, Spain and UK), have adopted diverse non-pharmaceutical and restrictive interventions ranging from case quarantine and isolation of contacts to the lockdown of entire populations. Assessing the impact of lockdowns and identifying the optimal strategies to manage the health crisis beyond lockdowns is of critical importance. To avoid the healthcare facilities overload, the severity of clinical cases and many deaths, it is has proved necessary to plan the response, adopting measures such as case finding, contact tracing, isolations and lockdowns of nations [29] and in parallel the large-scale vaccination campaign. Although these restrictive measures are crucial in minimizing the transmission of the SARS-CoV-2, the risk of pandemic resurgence when restrictions are relaxed and societies go back to a “business as usual” lifestyle has remained high. It is important to find a compromise between public health requirement and the economic and social needs of the population, analyzing the different restrictive measures in order to improve these measures aimed at reducing the spread of SARS-CoV-2 [10].
The Oxford Covid-19 Government Response Tracker (OxCGRT) constantly collects information on measures taken by different governments to contain the effects of the pandemic and now has data from more than 180 countries. Surely this is an important tool for worldwide comparisons but, in our study, we thought it would be helpful to understand these comparisons across a limited number of European countries. As a result, we went for a tailored definition of all parameters for Containment Index definition.
Other studies have addressed the topic of quantifying the effect of containment actions by looking at a rather high number of countries: Talic S et al. performed a meta-analysis to research the effect of handwashing, mask wearing and physical distancing measures on incidence of Covid-19. They concluded that these several personal and social protective measures are associated with decrease in virus circulation [30].
Iezadi S et al. in their meta-analysis showed that NPIs has determined a decrease in the COVID-19 reproduction number, daily case growth rates, daily death growth rates and COVID-19 daily ICU admission. It emerged that early enaction of lockdown, which predates the explosion in the number of positive cases, has been followed by a shorter duration of the lockdown itself and a smaller increase in the rate of case growth in the period after the application of this maximum containment measure [31].
Li et al. conducted a modelling study across 131 countries [28] and Flaxman et al. studied the effect of major interventions across 11 European countries [12]. Conversely, our analyses aimed to describe, in a comparative way, different containment approaches implemented by the five countries and to qualitatively correlate them with the evolution of the pandemic. Our analyses were aimed at assessing whether mild measures or delays in adopting the different types of restrictive measures could have influenced the number of Covid-19 cases and the evolution of the epidemic curve, in order to be able to guide the near-future choices of governments, policy makers and health authorities in the best possible way. Consequently, based on the evolution of case numbers and the adopted restrictive measures in Europe, it is possible to make the following considerations.
Following the diagnosis of the first patient affected by Covid-19 in Germany on January 27th, Germany implemented much milder and shorter restrictive measures in comparison to the other four countries. All restrictions were applied to every German state and were kept in place for a total of 6 weeks but there was never a total closure and, unlike the other European countries, Germany never issued a curfew. It should be noted that during the lockdown Germany never adopted the stay-at-home requirements in an absolute way (C8 = 0.5). Germany would have had a reduction of new cases from an extension of the lockdown and likely from the application of more restrictive measures as well, given the trend of the T2 + 30 curve in our predictive model (− 3.3%).
As early as January 30th, the Italian government ordered the blocking of air traffic from China, the first European government to do so. This was the beginning of the application of a long list of restrictive measures in Italy. After the first 2 weeks of targeted measures, Italy adopted one of the longest periods of lockdown (10 weeks, like in the UK) compared to the other selected countries. The adoption of such stringent measures has made a unique and indispensable contribution to the ongoing fight against the pandemic. In fact, as we observed in our predictive model, it seems that the relaxation of the restrictive measures was introduced at the right time because there would have been no net reduction of daily cases if the country would have entered Phase Two later (T2 + 30 = 0,5%). This shows a big difference between Italy and the other four countries analyzed in our study regarding the entry of the countries to Phase Two.
As France experienced the first wave of the Covid-19 pandemic, a nationwide lockdown was implemented to curb the dramatic increase in the number of patients in critical conditions. The country is the only one to have maintained a total closure for all parameters (C.I. = 9) during the entire lockdown. However, this seems not to have been enough. In fact, considering our predictive model, France could have benefited from an extension of the lockdown duration, being able to obtain a further reduction in the number of new cases (T2 + 30 = − 4,9%).
Following the story of Italy, Germany and France, Spain adopted a long lockdown (9 weeks) but maintained the maximum level of restriction for the shortest time (the maximum C.I. = 9 was true for only 3 weeks, compared to Germany where the maximum C.I. = 7.25 was held for 4 weeks). Spain is also the only country to have applied all restrictive measures at the targeted level before Phase Two (i.e., during the lockdown). Both these attitudes could justify the trend of the T2 + 30 curve of our preventive model (− 2,4%). The country probably should have applied more stringent restrictive measures and should have delayed the application of restrictive measures at the regional level, considering that Spain is the country that entered Phase Two with the highest number of cases per 100 K population (496 total cases per 100 K population) among the five countries analyzed.
The UK, despite not having gone through a targeted period, has applied very stringent restrictive measures during the lockdown and the country has very gradually entered Phase Two. The UK lockdown lasted for the same duration as the Italian lockdown, but the country started to adopt the first restrictive measures 4 weeks later than Italy did (March 23rd in the UK compared to February 24th in Italy). In our preventive model, the country would have benefited greatly (− 5.9%) from a delayed entry into Phase Two. This can be justified, as in France, by a consistent circulation of the virus that was difficult to control, already existing at the beginning of the adoption of the restrictive measures in the country. The UK has decided not to take some measures on a regional basis, as has been done by France, Germany, Italy and Spain.
The easing of the measures was carried out in a more or less gradual and cautious way by all five countries: for Phase Two, Germany had reported the lowest number of total cases per 100 K population among the five countries of the study, followed by France, Italy, the UK, and finally Spain (which loosened with the majority of cases). We can say that, compared to the other four countries analyzed, Germany was the first country to adopt a relaxation of restrictive measures: Germany began its first steps to ease restrictions on April 20th (during lockdown). Schools were reopened in Germany (C1 = 0.75), and approximately 25% of the students were allowed to go to school. Schools were soon opened in France (on May 11th) but there was a maximum number of 15 students for each classroom (C1 = 0.5). During Phase Two in Spain and Italy, schools continued to be closed. Particular attention should be paid to the fact that when Germany entered Phase Two, it had not imposed any restrictions on internal movement of citizens. Further, Italy was the only country that renounced the stay-at-home requirements immediately when it entered the Phase Two.
Consistent evidence can be observed in our study and in others previously published on the impact of restrictions’ ease. In particular, our findings on restrictive measures’ effects are in line with the findings from Flaxman and colleagues [12] who assessed the impact of different NPIs among 11 European countries, but not their effects. Flaxman and colleagues reported clearly that several NPIs (e.g., school closure and public events ban) combined with lockdown had a large effect (81%) on reducing transmission. Our findings are also consistent with results from a 131 cross-country study by Li and colleagues [28]. They observed that individual NPIs are associated with a reduced SARS-CoV-2 transmission and that the effect of introducing and lifting NPIs is delayed by 1–3 weeks, with a longer delay occurring when NPIs are lifted: a resurgence in Covid-19 cases has been reported in some countries that lifted some of these NPIs.
Based on our observations, France and the UK had a small number of positive cases at the beginning of the pandemic and during lockdown, while Spain experienced high caseloads during the first wave of the pandemic and a constant increase of positive cases for SARS-CoV-2 during the second phase.
Further, looking at the evolution of the epidemic curves in the different countries, if we assume that lockdown containment measures were not eased (i.e., predictions until T2 + 30, the dotted line in the Fig. 3), we obtain interesting information when comparing countries. If the total blocking measures were prolonged, they would have led to a greater reduction of cases in the UK (− 5.9%), France (− 4.9%), Germany (− 3.3%) and Spain (− 2.4%), while Italy (+ 0.5%) would not have achieved substantial improvements (the zeroing target had already been achieved).
When speculating about possible reasons for the different effects in easing lockdown measures, we mainly observed different behaviors related to school measures. Schools remained closed in Phase Two in Italy and in Spain while the other three countries gradually reopened (C1 was 0.75 at the beginning of Phase Two in France, in Germany and in the UK). In Germany, only the last year of each school level restarted (approximately 25% of students). In France, kindergarten and primary school children returned to school on a voluntary basis, middle schools were gradually reopened in the green zones while they remained closed in red zones. Like France, the UK began easing restrictions on schools starting with the opening of kindergarten and primary school.
As for other parameters such as businesses and workplaces, shops and retail, hospitality, personal care activities, assembly and leisure, internal movement, stay at home and gatherings, only Italy adopted targeted restrictions before the lockdown and Germany only had partially adopted restrictions for shops and retail, hospitality, assembly and leisure, and gatherings. In Italy, the early partial - and then total - restrictions concerning all the mentioned parameters can justify the good result reached by the predictive model T2 + 30. Furthermore, based on the number of Covid-19 cases, Germany was the first to introduce targeted measures (at 0.02 total cases per 100 K population) and Spain the last one (at 4,94 total cases per 100 K population). The UK entered lockdown with the same number of cases as France (10 total cases per 100 K population), but the UK started applying the restrictive measures a week later than France and without previous targeted measures reporting a number of Covid-19 cases almost double than France at the beginning of Phase Two (407 cases per 100 K in the UK vs. 214 cases per 100 K in France).
The results of our model speculate the evolution of the pandemic curve if the containment measures were prolonged and suggest that all the countries except Italy would have benefited from longer restrictions at the time of Phase One, although it was not possible to clearly quantify the contribution different behaviors had. Overall, we can say that there were three main factors that affected the SARS-CoV-2 pandemic in the above mentioned European countries: 1) the time when containment measures were adopted according to the epidemic curve (the earlier were the better); 2) the duration of containment measure adoption (2 weeks of early restrictive measures before the lockdown gave better results than no restrictive measures before the lockdown); 3) the number of Covid-19 cases before easing containment measures (the fewer the better before concluding the lockdown). We agree with Iezadi S et al. that the NPIs has had successful impacts on containing the spread of SARS-CoV-2, despite the substantial impacts on economies and mental health. In addition to addressing issues regarding universal access to vaccines, considering the severe consequences of national lockdown and other restrictions, these interventions should be accompanied or mitigated by the adoption of other NPIs such as contact tracing, the use of face masks and suspected/patient case isolation strategies [31].