Effectiveness of COVID-19 Medications in High-Risk Groups

In December 2021, new units called COVID-19 medicine delivery units (CMDUs) started operating in England. Their goal was to provide antiviral medications and neutralising monoclonal antibodies (nMABs) to people with COVID-19 who were not in the hospital but were at high risk of serious health issues. Before February 2022, two main medications, Sotrovimab and Molnupiravir, were commonly used. These medicines were approved based on results from two major clinical trials that involved people who had not been vaccinated.

As new COVID-19 variants appeared and population immunity changed, it was important to keep checking how effective these medications remained. Comparing treated individuals to those who did not receive treatment can have many biases. It is vital to control for various factors that could affect who receives treatment and who does not. Another concern is "immortal time bias," which happens if the time between diagnosis and treatment is not accounted for properly. This can result in people who had severe outcomes being wrongly placed in the untreated group.

This study used data from real-world observations to mimic a randomised trial involving high-risk groups in England. We looked at the effectiveness of either sotrovimab or molnupiravir compared to no treatment at all.

#Data Source

Data came from electronic health records managed by TPP SystmOne, through the OpenSAFELY platform. This system allows for secure linking and analysis of health data without revealing personal information. We checked various data types, including patient vaccination records, hospital visits, COVID-19 test results, and treatment records. To protect privacy, detailed patient data were not shared openly.

#Study Design and Patient Selection

Two main analyses were performed, focusing on the BA.1 and BA.2 periods when different Omicron variants were prevalent. We identified all adults over 18 who tested positive for COVID-19 in the community during the study periods. Eligible patients had specific high-risk health conditions, like Down syndrome or various organ diseases.

We ensured that the groups used for treatment were distinct to allow accurate comparisons. Patients needed to have no previous treatment for COVID-19 and could not have been hospitalised on the day they tested positive.

On February 10, 2022, Paxlovid, another antiviral medication, was introduced, prompting us to exclude patients who received it during the BA.2 period to keep consistency with BA.1 analysis.

#Outcome and Follow-Up

The main outcome we looked for was COVID-19-related Hospitalisation or death within 28 days of a positive test. Certain hospital admissions related to treatment were not counted as negative outcomes. We followed patients from the day they tested positive until they were either hospitalised, passed away, deregistered from their GP, or until 28 days had passed.

#Treatment Strategies and Target Trial

We aimed to compare the following treatment options among high-risk patients:

1. Starting either sotrovimab or molnupiravir within 5 days of testing positive for COVID-19.
2. No treatment within the same 5-day period.

#Statistical Analysis

We used statistical methods to compare the likelihood of hospitalisation or death between treated and untreated patients. To avoid biases, we employed a technique that creates copies of data while censoring it based on treatment. This way, we ensured that the time before treatment didn’t skew the results.

In our untreated group, we used statistical models to estimate how many patients would remain uncensored up to 5 days. We adjusted for factors like age, sex, region, and vaccination status to minimize biases.

Following the adjustment, we used advanced statistical models to estimate the risk of hospitalisation or death for the patients. Additionally, we examined survival rates over 28 days across the groups.

#Patient Characteristics

Out of more than 33 million individuals registered in the OpenSAFELY system, nearly 19.5 million were aged 18 and older. Among this group, over 35,000 and 39,000 patients tested positive during the BA.1 and BA.2 periods respectively. We also noted the number of patients treated with each medication within 5 days of their positive test.

#Key Findings

During the BA.1 period, around 3.4% of patients experienced serious outcomes, and in the BA.2 period, it was around 3.0%. The results showed that sotrovimab was linked to a lower risk of severe outcomes after infection, although this effect appeared weaker during the BA.2 period. There was no significant evidence that molnupiravir provided any benefit in either period.

#Subgroup and Sensitivity Analyses

Since certain medications might help patients with specific conditions like blood disorders or organ transplants, we performed analyses focused on these groups. Results were similar to the main analysis. We also conducted sensitivity checks to ensure our findings were robust and explored the influence of various treatment timelines.

#Limitations of the Study

While this study used extensive healthcare data, there remain some limitations. Even though we controlled for many patient characteristics, there may still be hidden biases influencing the results. For example, some patients who appeared to be eligible for treatment may not have been at high risk or had mild symptoms.

Also, the study could not adjust for ongoing symptom severity within 5 days post-infection, which might have influenced the treatment received. The nature of treatments changed between the two periods, which could have impacted patient characteristics and outcomes.

#Context of Findings

Our findings support previous research. For instance, other studies found that molnupiravir was ineffective in preventing severe COVID-19 outcomes, and there were mixed results for sotrovimab against newer variants. This illustrates the challenges in evaluating COVID-19 therapies due to the rapid evolution of the virus.

#Implications for Policy and Research

Guidelines for using sotrovimab differ across countries. In the UK, it is recommended for patients who cannot take Paxlovid, while other countries may have withdrawn support for it due to concerns about its effectiveness.

Our study underlines the value of using linked health data to evaluate treatments efficiently. Future research could benefit from aligning data collection with the necessary information to conduct robust analyses quickly.

#Conclusion

In summary, the use of sotrovimab compared to no treatment showed a reduced risk of serious COVID-19 outcomes during the BA.1 period, but evidence was less strong during the BA.2 period. There was no benefit seen with molnupiravir. The overall rates of severe outcomes were low among high-risk patients, but careful consideration of potential biases is needed to fully interpret the findings.