This is the first population-based study that analysed and compared the nationwide use of mAbs and antivirals for early COVID-19 treatment in outpatient setting during the period December 2021 to October 2022 in England versus Italy using publicly available data. A progressively increasing prevalence of use of mAbs and antivirals against SARS-CoV-2 in both Italy and England was observed. This finding was mostly determined by the marketing of the nirmatrelvir/ritonavir and molnupiravir as well as the extension of remdesivir use also for early treatment of COVID-19 outpatients, almost entirely replacing mAbs. In particular, in the most recent observation period, nirmatrelvir/ritonavir has been the most widely used antiviral, in line with the WHO clinical management of COVID-19 guidelines [20] as well as the European Society of Clinical Microbiology and Infectious Diseases guidelines [21]. A similar trend was observed in the USA, as reported on the US Department of Health & Human Services website [22]: in the period December 2021 to October 2022, mAb administration was replaced by nirmatrelvir/ritonavir and molnupiravir, due to the spread of new SARS-CoV-2 variants which were much less sensitive to mAbs anti-spike neutralization activity [23], in line with the National Institute of Health guidelines for COVID-19 treatment [24]. However, the prevalence of use of mAbs and antivirals is rather low as the highest reported prevalence of use within two weeks in patients diagnosed with SARS-CoV-2 infection was 3.8% in England and 2.6% in Italy. These findings are, however, explained by the fact that both mAbs and antivirals are approved specifically only for the early treatment of symptomatic adults or paediatric outpatients with specific comorbidities at high risk of developing severe COVID-19, which is only a minor proportion of all patients diagnosed with SARS-CoV-2 infection that were included as denominator in our study. On the other hand, as highlighted by Dal-Ré et al. [25], it was extremely challenging for all the national healthcare systems to be adequately adapted for the timely and correct use of these therapies as eligible patients have to be identified by general practitioners (GPs), referred to specialists in most cases for antiviral prescriptions and being dispensed and administered the drug within 5 days from the onset of symptoms. The availability for people who are eligible for treatment in association with the prescription process reflects several challenges for authorities. Therefore, the likely underutilization of mAbs and antivirals in at-risk patients with COVID-19 might be mainly due to a non-homogeneity of well-established networks between GPs (who identified and signalled the potential candidates to early treatment) and specialists (who confirmed the eligibility and prescribed the drug) causing a delay in access to treatment (Supplementary Table 2). Accordingly, in Italy we observed an increase in the use of nirmatrelvir/ritonavir since general practitioners have been directly authorized to prescribe this drug starting from 21 April 2022 [26]. Moreover, it is worth considering the method of administration of these new drugs, which also affects their use [4]. Monoclonal antibodies, as well as remdesivir, require intravenous (IV) infusion in healthcare facilities in which patients can be monitored during and for at least 1 h after administration (Supplementary Table 2). For this reason, in most cases nirmatrelvir/ritonavir is the preferred drug because of its oral administration, as well as its high efficacy [27]. However, the main disadvantage of nirmatrelvir/ritonavir is due to the high number of drug–drug interactions it has and therefore often cannot be administered.

Overall, our results showed that the increased use of mAbs during Delta predominance decreased with the spread of Omicron, while a more widespread use of sotrovimab, molnupiravir and nirmatrelvir/ritonavir was observed. This finding is consistent with the documented reduction of mAb efficacy against Omicron sub-lineages [23, 28] as it is known that mAbs act by binding to the SARS-CoV-2 spike protein, which varies according to the virus variant. On the contrary, antiviral therapies have a more unspecific mechanism of action which is unaffected by virus variants and the spike protein mutations. Several in vitro studies showed in fact that sotrovimab retained most of the activity against Omicron/BA.1 but was inhibited by Omicron/BA.2, while molnupiravir and nirmatrelvir/ritonavir consistently maintained in vitro activity against both BA.1 and BA.2 sub-lineages [29, 30]. However, despite the wide use of molnupiravir for early treatment of COVID-19 outpatients, on 24 February 2023, the EMA recommended against marketing authorization for molnupiravir for the failure to demonstrate a clinical benefit in terms of reduction of mortality, hospitalizations, duration of illness or time to recovery [8].

As for remdesivir, during the study period the public reports from England documented a much wider use in hospitalized patients (18,503 doses) than outpatients (203 doses) [10], as probably the result of its IV infusion for three consecutive days, as well as national health policies, while tixagevimab/cilgavimab in England is authorized only for COVID-19 prevention. Similarly, in Italy during the same observation period, overall higher use of remdesivir in hospitalized patients as compared with outpatients was reported, albeit with no major differences in terms of doses administered between the two settings (23,893 versus 21,402 doses) [31].

Similar findings were observed in a drug utilization study conducted in Scotland in the period between December 2021 and September 2022. Of 11,465 COIVD-19 outpatients, 47.4% were treated with nirmatrelvir/ritonavir, 26.7% with sotrovimab and 24.4% with molnupiravir [32].

The use of mAbs/antivirals for early treatment of patients diagnosed with SARS-CoV-2 infection varies between countries on the basis of different dates of approval and market availability. For instance, at the loco-regional level in Italy at the beginning of the Omicron wave, just before nirmatrelvir/ritonavir approval, a shortage of sotrovimab was experienced, and for this reason molnupiravir was mainly prescribed.

This study has several limitations that should be acknowledged. Only aggregated data were available in the public reports, thus not allowing stratification by gender, age and comorbidities to explore inequalities in prescribing across different patient categories. For the same reason, we could not identify specifically the patients with COVID-19 who were eligible for receiving the antiviral treatment. Therefore, for calculating the prevalence of mAb/antiviral use we considered as denominator all patients diagnosed with SARS-CoV-2 infection, including the asymptomatic ones, identified during the study period, even if only a portion of them were likely to be eligible for mAb/antiviral treatment. In addition, the number of positive tests is strictly dependent on the country-specific public health indications, including screening campaign in high-risk population. Indeed, a higher number of positive diagnostic tests was reported in Italy as compared with England during the study period. This difference may be due to the fact that free SARS-CoV-2 infection testing has been stopped in England since 1 April 2022 [33] and this probably affected the denominator used for calculating the prevalence of mAb/antiviral use. Accordingly, in England a substantial reduction in the number of diagnostic tests performed and, consequently, in the patients diagnosed with SARS-CoV-2 infection was observed in the period 1 April to 31 October 2022, as compared with the period 1 December to 31 March 2022 (average monthly performed positive tests: 2,258,790 versus 8,895,396) [11]. Free SARS-CoV-2 infection testing continued to be mostly warranted for symptomatic subjects in high-risk settings, and as such, this may have contributed to the observed increased prevalence of use of mAbs/antivirals in England versus Italy starting from 1 April 2022. On the other hand, SARS-CoV-2 infection-diagnosed outpatients enrolled in the ongoing clinical trials of antivirals were not included in the public reports. As the number of users enrolled in clinical trials may differ across countries, this may have further affect the comparison in prevalence of use of antivirals by country. As an example, in the UK the Panoramic trial enrolled > 25,000 COVID-19 outpatients [34] and the study start date is overlapping with that of our study.

Another potential limitation is that as the numerator we considered the number of administered doses (i.e. treatment cycles) instead of number of users. As for all mAbs and antivirals for early COVID-19 treatment in outpatient setting, only one treatment cycle is administered to each patient with COVID-19, and as there is limited opportunity for switching to other antivirals due to lack of efficacy, we can assume that number of administered doses corresponds to the number of mAb/antiviral users. On the other hand, the number of administered doses is expected to be reasonably accurate as special intensive registry-based monitoring for each approved antiviral for early COVID-19 treatment in Italy and in England was implemented.