Recently the Health Services Safety Investigation Body (HSSIB) published a report exploring potential patient safety risks associated with the use of online consultation systems in general practice. The report contained a few statements that we’d like to take a moment to address.
Before we get started, it’s important to note that, as the creators of an online consultation system ourselves, we take patient safety extremely seriously.
Our product, Patchs, has been created with input from clinicians, and continues to be monitored by clinicians – in fact, our Chief Medical Officer, Dr Ben Brown, is himself a practising GP and research scientist who brings a real-world understanding of the complex intersection between patient and practice staff needs to Patchs. Added to this, our Artificial Intelligence (AI) is an award-winning, UKCA marked Class I Medical device, registered with the MHRA.
Keenly aware of the advantages and disadvantages involved in the use of online consultation, Dr Brown has co-authored a number of research papers which explore their use and safety. Two of these are made explicit reference to in the report in question:
- Understanding How the Design and Implementation of Online Consultations Affect Primary Care Quality at the Journal of Medical Internet Research, and
- Patient experiences of an online consultation system: a qualitative study in English primary care post-COVID-19 in the British Journal of General Practice.
Patchs has used the findings of these reports to inform Patchs’ design, and we continue to work closely with researchers from The University of Manchester: for example, we have recently begun an NIHR-funded research project into the use of AI in primary care.
Examining the Report’s Findings
The report contains a number of findings, some of which we’d like to examine more closely and with reference to Patchs’ specific capabilities and design.
-
Harm can result to patients where they are unable to use an online consultation tool, which may result in inequitable access to care
This is a central concern: the whole point of online consultation is to broaden access for patients, making it more equal and efficient. That means for all patients, not just for those who possess a high degree of technological literacy – or indeed, those for whom English is a native language. That’s why we have:
- incorporated Patchs Translate, which enables patients to use Patchs in over thirty of the most commonly spoken non-English languages in the UK.
- developed Patchs Telephone Assistant, which provides patients who cannot use a computer or smartphone to contact the same opportunity to contact their GP on the telephone, and have their request dealt with in the same way as those made online.
-
Questions of design: ‘free text’ vs ‘structured’ interface types each have their pros and cons
This is a point upon which Patchs agrees: free text forms, which allow patients to describe their condition in their own words, are easier to use and more flexible for the patient – but they run the risk of not gathering enough information for staff to make decisions. The alternative approach – a structured question-and-answer format – provides staff with controlled information, but it can be inflexible for patients to use.
In Dr Brown’s paper in the Journal of Medical Internet Research (a summary of which can be found in this blog post), the research team found that any pitfalls associated with online consultation can be best avoided by:
- using the free text format
- incorporating Artificial Intelligence (AI)
- integrating seamlessly with an organisation’s existing software.
This is because – in spite of the acknowledged drawbacks of the free text format – a structured question-and-answer format provides patients with limited response options. It leads many to feel shoehorned towards a particular outcome, as if the system wants to ‘slot them in’ to a particular algorithm or pathway.
The incorporation of AI is therefore crucial: it mitigates any ‘cons’ associated with offering patients’ free text answers. With AI, Patchs offers the best of both worlds: patient users can describe their symptoms in their own words, while from the staff perspective the AI works to identify both the topic and urgency level of incoming requests. Patchs AI can also be configured to automatically gather relevant clinical information from the patient, making initial triage decisions easier and safer – and response times faster.
For example, Patchs AI Care Navigator is capable of automatically triggering a UTI questionnaire if a patient mentions relevant symptoms, or a PHQ-9 questionnaire if a patient mentions symptoms of depression. This means that, contrary to the idea that the free text approach provides staff with less information, with the addition of AI they in fact receive as full a picture as possible of a patient’s condition, at the moment the request hits their inbox.
-
Regarding the safety of Artificial Intelligence
The report states that there is a ‘lack of clarity around whether software (including AI) is a medical device.’ It also states that ‘if something is not registered as a medical device, scrutiny of its safety may be less.’
While the second point is not in question, we would like to make clear that Patchs AI is a Class I Medical Device. This is because it offers ‘triage and signposting of next steps based on filters by severity and probability of a match’, but without ‘direct diagnosis’. Patchs AI is registered as a medical device with the Medicines and Healthcare products Regulatory Agency (MHRA) and is UKCA marked. This means that we undertake rigorous clinical evaluations and safety assessments before releasing Patchs AI modules to customers, and conduct ongoing post-market surveillance after they are released. This process includes multiple stages of testing including on historic patient requests, newly collected patient requests, interviews with staff and patients, and direct comparisons with GPs.
Most online consultation systems that claim to use AI will use it for triage, and to help staff and patients make decisions related to care and diagnoses. In our opinion they should therefore also be registered as medical device. If they aren’t then questions should be asked as to why not.
-
In summary
We understand that newer technologies can be foreboding for those who have only ever handled patient request through traditional means, and we applaud a rigorous approach to the adoption of anything new. However, we also believe that with ever-growing patient demand and limited GP practice capacity, digital transformation is necessary if primary care is to realistically address its problems.
In the second paper cited above (and summarised here), Dr Brown and his team of researchers found that the majority of patients do in fact enjoy using online consultation systems: they appreciate the ease and efficiency of using them. The paper also acknowledges that these positive feelings are, however, dependent upon several factors, including the flexibility of the design. In particular, patients do not enjoy being forced to select a problem from a limited set of options, as some other online consultation systems out there require them to do.
So long as an online consultation system is properly regulated and complies with clinical safety standards, we believe that the advantages of online consultation systems outweigh the disadvantages. Not only do they offer an important communication and demand management tool, but with the addition of AI – which is always improving – they can be a valuable safety and efficiency tool.
If you have any queries or concerns regarding the findings of the HSSIB report, please reach out to a member of the Patchs team: features@patchs.ai
