Introduction

Coronavirus disease 2019 (Covid-19) has been responsible to date for over 6.6 million deaths worldwide [1]. While vaccination is central to reducing the risk of hospitalisation and death due to Covid-19, there remains a significant proportion who are thought to be unable to mount a sufficient immune response to vaccination due to underlying health conditions [2]. A range of therapies have therefore been developed or repurposed to reduce the chance of progression of Covid-19 in high-risk individuals.

The two primary approaches used to reduce the risk of progression of early Covid-19 infection are treatment with neutralising monoclonal antibodies (nMAbs) and/or antiviral drugs. NMAbs and antivirals inhibit viral replication, and nMAbs also prevent SARS-Cov-2 entry into host cells, making these approaches attractive for use in early stages of Covid-19 infection when viral replication is the key driver of disease progression [2]. In the UK, non-hospitalised patients are eligible for nMAb or antiviral treatment if they have a confirmed SARS-CoV-2 infection, they are symptomatic and show no signs of clinical recovery, and are a member of the ‘highest’ risk group (Table 1) [3]. As of the UK National Health Service (NHS) interim clinical commissioning policy February 2022, first line treatment is nirmatrelvir/ritonavir (Paxlovid, antiviral) or sotrovimab (nMAb), however significant drug interactions and contraindications in liver and renal disease limit the use of Paxlovid [3]. Second and third line treatment are remdesivir and molnupiravir (antivirals), respectively [3].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Criteria for being “high risk” of adverse outcomes from Covid-19 [3].

https://doi.org/10.1371/journal.pone.0281915.t001

A single dose of 500mg sotrovimab IV has been reported to give a relative risk reduction of 79% of hospitalisation or death in high-risk patients with mild-to-moderate Covid-19 [4, 5]. Molnupiravir is an oral small molecule antiviral prodrug that is active against SARS-CoV-2 [6] and has been shown to give a 30% relative risk reduction of hospitalisation or death in non-hospitalised patients with mild-moderate Covid-19 and a risk factor for severe disease when commenced within 5 days of symptom onset [6]. As these trials were performed at a time when different variants of Covid-19 were circulating to those seen now, and in very specific patient groups, real-world evaluation of treatment for early Covid-19 is urgently needed.

Antiviral and nMAb therapy must be started within 5 days of symptom onset (except for remdesivir which can be commenced within 7 days) [3], therefore efficient identification of treatment candidates and delivery of indicated medication is required. This has led to the development of NHS Covid Medicine Delivery Units (CMDUs) for patient screening and the coordination of treatment. With the frequent emergence of new variants, and the inevitable variations between real-world patients and study populations, ongoing evaluation of CMDU activities and outcomes is essential to inform policy and practice. Here we present the experience of a major CMDU serving a large population in the East Midlands of the UK.

Methods

Results

Treatment decisions

During the study period, 1820 patients were screened by the CMDU, of whom 604 progressed to more detailed treatment discussions (33.2%) (Fig 1). Of these potentially eligible patients, 254 received treatment (42.1%), 70 declined treatment despite being eligible (11.5%), and 266 (44.0%) reported resolving symptoms and therefore were ineligible for treatment. It was unclear from the records whether 14 (2.3%) patients received any treatment, therefore these patients were excluded from further analyses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The Covid Medicines Delivery Unit referral pathway.

Flowchart showing the screening results of the 1820 patients referred to the Covid Medicines Delivery Unit (CMDU).

https://doi.org/10.1371/journal.pone.0281915.g001

In total, 249 patients received either sotrovimab or molnupiravir (76.9% of those eligible for treatment after full screening). Sotrovimab (169, 66.5% of those receiving treatment, and molnupiravir (80, 31.5%) were the main treatments prescribed. Another five patients received Paxlovid (2% of those receiving treatment), but were not included in further analyses because Paxlovid was not available for the entire study period. Details of the patients that received Paxlovid can be found in the supplementary information (S1 Table). During early 2022, evidence emerged of superior efficacy of sotrovimab compared with molnupiravir, and sotrovimab became the preferred treatment approach unless IV infusion or attendance at CMDU was not possible [3].

Patient demographics and indications for treatment

The demographic data and indications for treatment are shown in Table 2. The median age was comparable between the treated and untreated groups (51–52 years), with a trend towards a higher proportion of female patients in the treated group (62.1% of sotrovimab-treated and 69.4% of molnupiravir-treated) compared with the untreated group (57.1% of those that declined treatment, and 58.6% of those with resolving symptoms). Ethnicity data were not available for all patients, in particular the group with resolving symptoms, therefore a full analysis of the differences between ethnic groups could not be performed. However, there was an apparent trend towards more patients from white backgrounds receiving treatment than those from non-white ethnic groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographic details, indications for treatment, and vaccination status of patients screened within 5 days of symptoms onset.

https://doi.org/10.1371/journal.pone.0281915.t002

The distribution of criteria making our cohort of patients “high risk” is shown in Table 2. Immune-mediated inflammatory disorders (IMIDs) were the predominant indication for referral for treatment. There was a non-significant trend towards more patients having malignancy- (either haematological or solid organ) or transplant-related indications for treatment in the treated group.

Vaccination status was recorded for 180 patients across the “treated” and “declined treatment” groups. Of these, 172 (95.6%) had received at least two Covid-19 vaccination doses. It was not possible to further analyse the relationship between vaccination status, treatment decisions, and patient characteristics and outcomes because of a large amount of missing data and the small numbers in our study.

Patient outcomes

Hospital admissions and mortality data are shown in Table 3. 31 (5.2%) of the 599 patients screened within five days of symptom onset had an unplanned hospital admission within 28 days, of which 13 (2.2%) were Covid-19 related. Two patients from the sotrovimab group and one from the molnupiravir group (1.2% and 1.3% of each group, respectively) had a Covid-19-related hospital admission, compared with three of those that declined treatment (4.3%) and seven that reported resolving symptoms at screening (2.6%). This gives an overall 3.0% rate of admission within 28 days due to Covid 19 in the untreated group, compared with 1.2% in the treated group. These data suggest a trend towards a reduction in hospitalisation with treatment for early Covid-19 in high-risk patients but because rates of Covid-related hospital admission were low in all groups no significant effect of treatment was seen (odds ratio of Covid-19-related hospital admission of 0.40, 95% CI 0.12–1.36).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. 28 day hospital admission and mortality.

https://doi.org/10.1371/journal.pone.0281915.t003

Three patients who were screened within 5 days of Covid-19 symptom onset died within 28 days of screening (0.5%). All three deaths were related to Covid-19. One death occurred in the molnupiravir group, and two deaths occurred in the “resolving symptoms” group. There were no deaths within 28 days of screening in either the sotrovimab or “declined treatment” groups.

The presence of endogenous anti-SARS-CoV antibodies was analysed in patients receiving sotrovimab prior to their infusions. Results were available for 152 patients that received sotrovimab (90%). Of these, 21 (13.8%) had undetectable antibody levels, and 131 (86.2%) had detectable antibodies. In a subgroup of 68 samples where antibodies were quantified 32 (47.1%) only had low antibody levels. Of the 169 patients treated with sotrovimab, 123 had their vaccination status recorded and 121 (98.4%) of these had received at least two Covid-19 vaccine doses (Table 2), and these low antibody levels could represent a suboptimal antibody response to vaccination. However, given the small numbers of hospital admissions and mortality in this group, it was not possible to assess the effect of antibody status or vaccination on outcomes. However, these data indicate that many patients deemed at high risk of poor outcomes from Covid-19 can generate antibodies to either vaccination or natural infection, although in at least some these are at a lower level than in the general population.

Patient deprivation may influence treatment decisions

Index of multiple deprivation (IMD) data were generated from the home postcodes of patients who progressed past the initial screening stage (Table 4). Postcodes were available for all patients except one in the “resolving symptoms group”. The median IMD score was lower in the treated groups, at 16.1 (range (1.1–73.3) and 16.2 (1.6–70.4) in the sotrovimab and molnupiravir groups, respectively, compared with 20.9 (range 1.3–70.4) and 21.9 (range 1.1–72.2) in those that declined treatment or had resolving symptoms. When patients that received any treatment (sotrovimab or molnupiravir) were compared to those that did not receive treatment (declined treatment or reported resolving symptoms), the deprivation score was significantly higher in those that were not treated (IMID 21.8 in the untreated group compared with 16.1 in the treated group, p = 0.02, Mann-Whitney test). These data imply that patients experiencing higher degrees of deprivation are less likely to access treatment for early Covid-19.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Index of multiple deprivation data of patients referred to the Covid medicines delivery unit.

https://doi.org/10.1371/journal.pone.0281915.t004

Discussion

The data presented here provide a valuable early insight into patient outcomes seen in a UK CMDU over a 10 week period in early 2022. Importantly, only 13.7% of patients initially screened received either molnupiravir or sotrovimab treatment. This implies that the current system for identifying potentially eligible patients is inefficient, particularly as 400 patients (22%) were found not to have a medical indication for treatment. This may be because the automated referral systems flagged patients to the CMDU that were deemed high risk of severe Covid-19, but may not have met the commissioning criteria for treatment at the time of this study [3], for example patients with diabetes. Furthermore, 390 patients (21.4%) were already outside of the treatment window by the time contact was made with CMDU. We are aware that, particularly in the early stages of the CMDU service, many patients were screened by CMDU beyond the first five days of symptoms, and were thus ineligible for treatment. This issue was multifactorial, and likely related to the introduction of new clinical and computer systems, lower awareness of treatments among the patients and healthcare professionals, and the timeliness of PCR results. Further work is required to publicise the eligibility criteria so that clinical teams and patients are aware of the indications for treatment, and to facilitate timely referral.

In our patient cohort, we had low rates of Covid-19-related hospitalisation (2.2% of all patients within 28 days of CMDU triage). This is lower than the 7% 28 day hospitalisation or mortality rate reported in the placebo arm of the sotrovimab trials [4, 5], and 10% reported in the molnupiravir trial [6]. Similarly, only three deaths occurred in the 604 eligible patients presenting within the treatment window (0.5% 28 day mortality rate across all groups). This may reflect the impact of vaccination even in these at risk groups and less severe disease seen with variants circulating in early 2022 in the UK (predominantly Omicron and BA1) [10]. These real-world data however raise the question of the cost-effectiveness of the treatments given in CMDUs. As of the week ending 1^st May 2022, 41649 patients have received either antiviral or nMAb treatment for Covid-19 [7], therefore is it imperative to understand how these treatments affect patient outcomes.

We observed a trend towards reduced hospitalisation rates in the treated patients compared to patients that did not receive treatment. This did not reach statistical significance, which is likely due to our study being underpowered for detecting such differences between treatment groups. Assuming equal matching between the treatment groups, power to detect a difference between groups of 80%, and an α value (probably of type I error) of 0.05, the required number of subjects to detect a significant difference in outcome based on the event rates we observed would be 1990. Therefore, it is not possible to conclude whether treatment for early Covid-19 reduced the rates of hospitalisation in our study population, which was significantly smaller than this value.

In our study, 266 patients were excluded from treatment because they reported resolving symptoms. Whether symptoms are “resolving”, “persistent”, or “worsening” is subjective from both the patient and healthcare professional perspectives. Differences in symptom reporting and healthcare seeking habits of different patient groups could result in significant differences in access to treatments for Covid-19. While there were no statistically significant differences in outcomes between treated and untreated groups in our study, the two deaths and seven Covid-related hospital admissions in those with “resolving symptoms” suggests that this may not be an appropriate indication for exclusion from treatment for Covid-19. Alternative non-subjective scoring systems for the persistence or resolution of Covid-19 symptoms are required to address this issue.

A further explanation for the lower hospitalisation and mortality rates we observed may be because of differences between the study populations. In the original sotrovimab and molnupiravir trials [4–6], the patient population was deemed “high-risk” because of age (over 55 years), obesity, diabetes, heart failure, or chronic lung disease [4, 5]. Our study population was deemed “high-risk” according to the interim clinical commissioning policy (Table 1) [3], which consist of different criteria to those used in the clinical trials. These differences in baseline mortality may therefore reflect a difference in patients’ underlying risks. Alternatively, these differences could be driven by different uptake of vaccination, as being unvaccinated was a requirement for inclusion in the molnupiravir trial [6]. We could not fully analyse vaccination status in our population, as there was a significant amount of missing data, however many patients had received at least one vaccination. Studies of other CMDUs in the NHS, and similar units across the world, will help to clarify the baseline risk of hospitalisation and death in these “high-risk” patients.

The commonest complications of sotrovimab were nausea and dizziness, in keeping with the product literature [3], and no severe reactions occurred. These data support the finding of clinical trials and real-world studies that sotrovimab is a safe treatment for early Covid-19 [4, 5, 11, 12].

We also assessed baseline demographic differences between those accepting treatment or not receiving treatment. Although our study was small there was a trend towards a higher proportion of patients receiving treatment being female and from white ethnic backgrounds. This may reflect differences in treatment-seeking behaviours, underlying disease, or symptom manifestation between the sexes and different ethnic groups. The deprivation data suggest that patients receiving treatment had lower levels of deprivation than those that were not treated. Unfortunately, the reasons for declining treatment were infrequently recorded. Possible explanations include patients from more deprived background being less likely to be able to attend for a sotrovimab infusion, being more reluctant to travel for treatment, or being more likely to minimise their symptoms and decline treatment. Furthermore, patients from more deprived backgrounds may have had different prior experiences of healthcare, and may be less trusting of and therefore less willing to accept the relatively new treatments offered for Covid-19. This is a concerning finding given the universal healthcare model in the UK’s NHS, but is consistent with other reports that patients from more deprived areas may receive poorer quality healthcare [13]. Studies in the NHS and other healthcare systems may clarify the reasons behind this. Further consideration is required to ensure that hard to reach groups are not excluded from treatment for early Covid-19.

Our study is limited by being a retrospective review of the medical records of a local population, and missing data has limited our ability to address issues such as antibody status. Expansion of this work to include other areas in the UK would strengthen our conclusions, and routine prospective gathering of the data analysed here will facilitate future service evaluations.

Omicron became the dominant variant in England mid December 2022 and is likely to have been the predominant Covid-19 variant during our study [10]. While sotrovimab retained efficacy against the original omicron strain, the BA.2 subvariant exhibited increased resistance to sotrovimab in vitro [14]. Although the now predominant BA.4 and BA.5 strains are not thought to have further reduced susceptibility to sotrovimab [15], this drug is no longer recommended for use by the US Food and Drug Administration (FDA) [16]. This highlights the need for ongoing monitoring of nMAbs and antiviral treatment for this rapidly evolving virus, and the continued development and assessment of new treatment strategies.

Conclusion

Current referral pathways for outpatient treatment of early Covid-19 in the UK are inefficient, and ongoing service evaluation is required to ensure that all eligible patients are offered treatment if the service continues. Hospitalisation with Covid-19 was rare in our population regardless of whether treatment was received. The population highlighted as high risk by NHS clinical commissioning has different underlying disease profiles and vaccination uptake rates to those assessed in the sotrovimab and molnupiravir clinical trials. More work is required to ensure hard to reach groups, which may include ethnic minority populations and deprived individuals, have equitable access to treatment of early Covid-19.

Supporting information

Acknowledgments

We would like to thank Ms Judith Palmer (Chief Pharmacist, Nottingham University Hospitals) for advice on obtaining the CMDU data, and Dr Jeremy Lewis (Consultant Physician, Nottingham University Hospitals) for assistance in identifying patients for this work.