‘Go remote’, he said! The early promise of remote consultations
During the COVID-19 pandemic many of us reaped the seeming benefits of an increase in remote primary care consultations. My immunocompromised brother had less chance of catching COVID-19 and people who were less mobile could talk to their GP from home.
The benefits seemed so great that Matt Hancock became the remote consultation champion, stating unequivocally in July 2020 that ‘better tech means better health care’ and calling for all consultations to be remote unless there was ‘a compelling clinical reason not to’. In the early days of the pandemic, public opinion seemed to agree and early adopters in the system benefitted, with practices using remote approaches experiencing fewer decreases in consultations.
However, as the waves of the pandemic washed over us, individual reports of unintended consequences began to emerge. Some of these consequences were unanticipated and tragic, such as missed cancer diagnoses. Other consequences, including potential increases in antibiotic prescribing, had been anticipated prior to the drive to go remote.
A year before Hancock’s declaration, an opinion piece urged for further implementation of remote consulting to be ‘avoided until evidence of possible unintended consequences for antimicrobial resistance is clearer’. But it is only now thanks to novel research methods that the ramifications of anticipated consequences are being confirmed.
‘One of the biggest threats to global health.’ Antibiotic prescribing
A new cohort study by the Improvement Analytics Unit (IAU; a partnership between the Health Foundation and NHS England) challenges Hancock’s truism that better tech always means better health care.
The study used an approach called targeted maximum likelihood estimation (TMLE), which combines machine learning and causal inference (working out possible causes from data).
Analysis of a random sample of 600,000 patients in primary care using the Clinical Practice Research Datalink found that a remote consultation increased the likelihood of an adult being prescribed antibiotics by 23% compared to face to face consultation. This echoes views among GPs.
The study used a representative sample of almost 46,000 consultations. It controlled for patient, clinician and practice-level factors and used machine learning to explore relationships between those factors. This enabled the research to tackle questions of generalisability that previous studies left unanswered or inconclusive.
As a result, it provides new robust evidence that increases in remote consultations are associated with an increase in antibiotic prescriptions. WHO describes antibiotic resistance as ‘one of the biggest threats to global health’ and identifies overuse of antibiotics as a chief cause. The UK has a national target to reduce use in humans by 15% by 2024. Overall, primary care is currently achieving that target. However, this Health Foundation analysis suggests that despite national targets, when prescribing is examined by type of consultation, the decision to prescribe could be affected by whether the consultation is ‘in person’ or not.
‘Not routinely.’ When is it ok to prescribe antibiotics remotely?
The analysis compares prescribing rates of common antibiotic prescriptions in consultations where a person can be examined ‘in person’ against remote approaches (including via phone and video call) where that’s not possible. This comparison enables us to ask whether or not the type of consultation matters.
Despite Hancock’s assurances, determining this is tricky. The latest NICE guideline on suspected acute respiratory infection recommends that antibiotics should not be ‘routinely’ prescribed based on ‘remote assessment alone’, but the term ‘routinely’ leaves us uncertain.
The NICE guideline, along with other national guidance on remote consultations and prescribing, asks prescribers to determine how ‘confident’ they are in the need for antibiotics.
Ultimately, we are all left unclear as to what a ‘compelling reason’ for not conducting remote consultation might be.
‘Compelling’ reasons. Antimicrobial resistance
The Health Foundation analysis is thankfully less equivocal and indicates that the anticipated consequences of antimicrobial resistance should be considered a compelling reason.
Analysis also points to avenues for further exploration. Results showed no association between prescribing and type of consultation for children. However, children had more in-person consultations, possibly pointing to underlying reasons around risk, as other studies suggest. The more risk in a situation, as perceived to be with children, the more ‘compelling reasons’ to see someone in person. The more remote a consultation, the less certainty, and so clinicians may over prescribe to manage risk.
While the study generates stronger evidence of cause and effect, it cannot fully answer the question of why remote consultation is associated with higher antibiotic prescribing. To understand that, we need to dig into this data more deeply. Identifying encounters where prescribing may have been unnecessary and supplementing this quantitative data with qualitative research will help us to understand what it was about those remote encounters that led the clinician to decide there was a compelling reason to prescribe.
For antibiotic resistance at least, this retrospective analysis of data confirms sooner than the 15 years hindsight Perri says we need what the consequences of policy have been.
It also helps us prepare. Antimicrobial resistance has been described as an ‘endemic problem’ potentially compounding future pandemics. This research gives us a compelling reason to ensure we better understand the behaviour tech can generate and what changes might need to be put in place, in order to realise the benefits of better tech rather than potential unintended harms.
This content originally featured in our email newsletter, which explores perspectives and expert opinion on a different health or health care topic each month.