The coronavirus disease 2019 (COVID-19) pandemic has sparked novel research and insights, but also concerns and anxiety regarding established practices. Early into the pandemic, public and scientific concern was raised regarding the role of renin–angiotensin–aldosterone system (RAAS) inhibitors on the susceptibility to COVID-19 given their effect on angiotensin-converting enzyme 2 (ACE-2), the host receptor for the virus. This gathered media attention globally, despite several health boards encouraging the ongoing use of these medications. We aimed to investigate whether, despite advice supporting continued use of these medications, there was a change in prescribing practices for RAAS inhibitors in general practice. Data were collated from the NHS digital platform, which provides monthly practice-level prescribing information for all primary care practices in England. We performed an interrupted time-series analysis on national-level prescribing data comparing time-series coefficients pre- and post-March 2020 with metformin used as a control. We find that from March to December 2020, prescribing rates of RAAS inhibitors were reduced relative to the previous time-series trend. This finding persisted after adjustment for rates of metformin prescription. This suggests that there was a change in prescribing behaviour during the COVID-19 pandemic, which may be linked to the public and scientific concerns during this time.
Introduction
From December 2019, SARS-CoV-2, a novel coronavirus, sparked a global pandemic and rapid scientific responses to the new coronavirus 2019 (COVID-19) disease. Rapid identification showed the angiotensin-converting enzyme 2 (ACE-2) as the host receptor for SARS-CoV-2. Given this, concerns were raised that renin–angiotensin–aldosterone system (RAAS) inhibitors, such as ACE inhibitors and angiotensin-receptor blockers (ARBs), which may increase the expression of ACE‑2, could negatively influence COVID-19 outcomes. This led to significant media attention and anxiety about ongoing use of these medications. However, there has been no scientific evidence to support modifying the use of RAAS inhibitors during COVID-19, with studies showing no evidence of a difference in hospital admission survival, COVID-19 progression or risk of intensive care admission with ongoing use. Therefore, the recommendation from multiple health associations was to continue the use of RAAS inhibitors during the pandemic. Given these concerns, we aimed to retrospectively identify whether there was any impact on the prescribing of RAAS inhibitors across GPs in England during the pandemic.
Materials and method
We used routinely available data from the NHS Digital platform, which provides monthly practice-level prescribing information for all primary care practices in England, obtained from pharmacy claims. Monthly practice list sizes were obtained from the same source and used to calculate estimates for daily prescriptions per 1,000 individuals for each practice. Data were downloaded from January 2019 to December 2020 and national monthly prescribing rates calculated. Ethical approval was not required.
Using the ‘ITSA’ module in Stata V 16.1 we performed interrupted time-series analysis, on national-level RAAS inhibitor prescribing data. Using Google’s ‘trends’ we identified that the frequency of internet searches for ‘ACE inhibitor’, ‘angiotensin converting enzyme inhibitor’ and ‘angiotensin blocker’ spiked in mid-March 2020. We, therefore, selected March 2020 as the ‘interruption’ or ‘intervention’ time point, and we estimated and compared time-series coefficients pre- and post-March 2020.
Given that multiple factors could account for changes in prescribing patterns during the COVID-19 pandemic, we ran a secondary ITSA model controlling for metformin prescriptions in multiple-group comparisons analysis. Metformin was selected as it is a long-term medication and both hypertension (most common indication for RAAS inhibitor prescription) and type 2 diabetes (most common indication for metformin prescription) are established as risk factors for mortality in COVID-19, with both medications typically obtained on regular monthly prescriptions. Therefore, changes in patient or clinician behaviour related to the COVID-19 pandemic (such as longer prescription intervals or travel restrictions) should impact RAAS inhibitors and metformin prescribing rates similarly. However, given metformin was not frequently implicated in COVID-19 susceptibility and there was no change in online search rates during the analysis period, we hypothesise that any differences in prescribing rates between metformin and RAAS inhibitors could reflect behaviour change specifically related to the concerns about RAAS inhibitors.
Results
Interrupted time-series analysis shows a step-change drop in rates of RAAS inhibitor prescribing after March 2020 relative to before (p<0.001), but no change in the slope of the time-series plot (p=0.99). Similar results were obtained for metformin with an immediate step-change reduction in prescribing rates (p=0.006) but no change in slope of the time-series plot (p=0.63).
In a multiple-group comparison model (figure 1) accounting for changes in metformin prescription, interrupted time-series analysis showed an independent immediate reduction in RAAS inhibitor prescription rates (p=0.004), and reduction in time-series slope gradient (p=0.02).
Discussion
We find that from March to December 2020, prescribing rates of RAAS inhibitors were reduced relative to the previous time-series trend. This finding persisted after adjustment for rates of metformin prescription. These findings could imply that the prescribing behaviour of clinicians, patients, or pharmacists was influenced by public and scientific concerns about the possible effects of RAAS inhibitors on COVID-19 infection, despite there being, at the time, no scientific evidence to support a change in behaviour.
Given that studies have shown that chronic heart failure and hypertension are associated with worse prognosis and increased mortality from COVID-19, as they are two of the most common causes of the prescription of RAAS inhibitors, the reduction in prescribing we have observed may have implications on the risk of mortality or serious disease in those patients.
Our findings highlight the evolving role of the internet and social media in sharing health information. There has been an increase in health seeking using the internet and social media potentiated by the COVID-19 pandemic, but information has been limited by the accuracy and quality of the resources available. We found an increase in Google searches for RAAS inhibitors in our study, which suggests a growing role for health providers in ensuring the spread of correct medical information.
A strength of our analysis is the use of national prescribing data covering all GP practices in England, incorporating a wide range of demographics. By using a control drug, we were able to assess whether our results were specific to RAAS inhibitors and not a consequence of a change in prescribing practices during the pandemic generally. However, there are several confounding variables that were not included in the analysis, which could impact the results observed. Further limitations of our study are that we cannot infer causality given the use of observational data and the data do not allow analysis on an individual level. Further, by using items dispensed as a measure of prescriptions, it does not give any indication on the length of course prescribed, which may be variable.
Our findings have clinical implications. ACE inhibitors are the first-line antihypertensive recommendation in under 55s by the National Institute for Health and Care Excellence (NICE). Given that hypertension has been shown to be a risk factor for increased COVID-19 severity and mortality, stopping the medication could have implications for those at-risk.
Conflicts of interest
None declared.
Funding
None.
References
Available on request from the authors.
