The mythbuster: The government and BMA are both wrong about online GP access

The British Medical Association has threatened to work to rule if the government insists that GPs keep online access open all day as they are now required to do.

The BMA argue that unrestricted online access could unleash a huge wave of demand that GPs do not have the capacity to handle, and that this would increase the risk of urgent or important cases being missed. The government counters that it would be better and more convenient for patients, therefore increasing ease of access.

However, neither side has fully engaged with the dirty details that matter for patients or been entirely honest about the consequences of their positions. To properly grasp the problems, we need to understand what effective online access looks like.

For good and obvious reasons, the primary care team at NHS England began promoting online triage during the early days of the pandemic. Post covid, they continued to promote what had become known as “total triage”.

They pointed to evidence which showed that, done well, online total triage offered many advantages. Patients could make online requests asynchronously, therefore avoiding the 8am phone lottery. They could also often get useful responses without having to physically attend a GP. Urgent or important requests, where a face-to-face appointment was clearly needed, could be quickly and easily identified. All this, in turn, helped maximise the use of scarce GP time by better matching the primary care clinical response to specific patient needs. 

However, this stance is based on the false assumption that all online processes are equally effective. This attitude has more red flags than a Chinese Army military parade.

Not all triage systems deliver

The reality is that not all the software tools are equally effective, and simply installing some software doesn’t change inefficient GP processes.

NHSE, to their credit, know what a good online total triage system looks like and have produced some solid best practice guidance. But, perhaps excusably given the urgent need to increase online access during covid, it ignored the fact that many software solutions did not, and do not, work well for patients.

Some common online tools were built on the idea that the problem of access was caused by triage, which took up too much GP time. The tools therefore tried to automate triage using complex algorithms (and later using artificial intelligence).

These types of tools need large volumes of specific information about the patient to do a good job. To gain that information, patients have to fill out long, annoying questionnaires. Many, of course, refuse to do so unless forced.

Other tools instead asked a small number of questions about the patient’s problem. GPs (or skilled staff) would then manually triage the patients based on their written answers or request more details. These tools see communication, not triage, as the problem to solve. Patients were far happier with such systems, as they were easier and faster to use.

In the best cases, more than two-thirds of patients will volunteer to go online to make requests of their GP practice without any coercion. The majority will also prefer a faster and less costly remote response to those requests.

The risk GPs know about

The second problem, of failing to recognise that installing software is not itself a transformative act, is an example of the one-legged stool approach to policy-making that I wrote about in an earlier Mythbuster. 

Total triage only succeeds when GP practices fundamentally change how they work, which many of them do not. The situation is made worse by software vendors – even some of the better ones – deliberately choosing to ignore the need for effective implementation. They, of course, get paid when their tools are installed, and not when they have been proven to work.

So, even though most practices have installed online triage software, few have seen major benefits.

The BMA position on the issue of online access to GPs, as well as a great deal of social media discussion about it, has been based on a hostility to triage of any form.

There is a widely held view that triage is a waste of GP time – and that patients should be seen in the order they present themselves. The reality is that, done well, triage allows GPs to focus their efforts on the most important requests for the time, minimising clinical risk as a result.

The BMA’s concerns about being swamped by demand are not backed up by research into the effective implementation of online triage.

Also, consider the implications of the BMA position from the patient’s point of view. Limiting access does not meaningfully reduce risks for the patient.

Traditional phone systems – involving long waits to get through, during which some patients just give up – don’t limit clinical risk. They merely limit the risks the GP knows about. A “see no evil, hear no evil” approach is bad for patients.

The BMA needs to take a smarter approach that recognises the value of total triage for GPs and patients. It should switch its focus to better implementation, rather than restricting access.

The government should devote far more effort to effective implementation of total triage. Pushing for longer opening hours while still using badly implemented systems is futile. The Health Services Safety Investigations Body (HSSIB) report on online tools contains some valuable lessons about that.

Government should also pay far more attention to which tools offer the best value for money. GPs and integrated care boards have too often bought the cheapest, not the best, online tools.

Big improvements in GP access are possible. But both sides need to change their approaches if this is ever going to happen.