• treatment
• disease activity
• rheumatoid arthritis

We read with interest the article by Dr Burmester on the digital technologies which have the potential to deeply impact healthcare delivery in rheumatology too.1 Technology has always been integral to modern medicine and without doubt it has played a big role in advanced diagnostics, drug development, treatment and rehabilitation. However, the personalised data-led approach is a recent phenomenon which is poised to change the way medicine has been practised traditionally since its evolution, as highlighted by Dr Burmester. In this context, some other aspects and concerns as we have outlined below may also be considered.

A burst of big data and the artificial neural network-powered artificial intelligence is already empowering the clinicians, patients and the care providers. We reckon chronic diseases especially rheumatology are ripe for disruption when it comes to patient care.2 We also envisage that by the time actual Rheumatology 4.0 fructifies in reality there may be few approved digital therapeutic options approved by the Food and Drug Administration in Rheumatology.2

A combination of data from multiple resources could also unearth various digital biomarkers, which are patient-generated physiological and behavioural measures used to explain, influence and/or predict the clinical outcomes.2 Also, the predictive analytics are expected to predict the individual outcome by pooling information from connected devices, genomic data, clinical course and various other sources.2

Personal data, particularly patient data, are extremely valuable and this fact is realised by all the stakeholders in current times. Data are an important economic resource which has multiple usage as well as applications in myriad forms. As more and more electronic medical records are deployed and data stored on cloud it is becoming critical to secure the data and safeguard the patient privacy from any form of breach and potential misuse by data hackers.

We would also like to add that genome-wide association studies (GWAS) in rheumatology would throw up interesting patterns now that the data are plentiful and genomic testing is more accessible. For example, in a recent paper, The Brainstorm Consortium examined multiple GWAS drawn from more than 200 000 patients for 25 brain-associated disorders and 17 phenotypes.3 Similar replication in rheumatology has tremendous potential and we believe that we are at the threshold of obtaining more in-depth and meaningful insights with ongoing and more studies in the future which will potentially help in better patient outcomes.

Machine learning, computational prowess and the artificial intelligence advances have enabled some industry leaders from the technology sector to predict that doctors would be becoming redundant very soon. In reality, the technology will always aid the clinicians with plethora of options but a machine can never replace the doctor in flesh and blood. Artificial intelligence algorithms require training from the human-generated models or examples as a mechanism of machine learning. Unless the correct answers are provided by the human beings the algorithms have various limitations. For instance, inability to understand speech or recognise objects in an image, and so on, without human guidance and training. We therefore completely agree with Dr Burmester’s conclusion that one may require new approaches to education and professional competence of clinicians. To paraphrase Dr Eric Topol (American cardiologist and digital medicine researcher)—The digital world has been in a separate orbit from our medical cocoon, and it’s time the boundaries be taken down.4