Research published in the Future of Healthcare Journal shows that models developed by UK healthtech start-up Gendius can determine which patients with type 2 diabetes are most likely to have renal impairment, allowing healthcare systems to prioritise the highest risk patients for earlier screening.
The results of the study show that it is possible to reduce Chronic Kidney Disease screening by 46.3%, while still detecting over 80% of those with abnormal kidney function using Gendius’ models.
People with type 2 diabetes are at a higher risk of developing Chronic Kidney Disease (CKD), and it is recommended that they undergo annual screening. Gendius’ predictive models could help to highlight those patients at the highest risk of having an abnormal CKD screening result, so that they can be prioritised for screening. This is particularly important in healthcare systems where screening resources are limited, especially with rapidly rising numbers of patients diagnosed with type 2 diabetes worldwide.
Dr. Camilla Sammut-Powell, Senior Data Scientist at Gendius, said, “We worked closely with clinical experts to determine the model inputs and considered what might be available to a clinician when applying the model. We developed two versions of the model for each marker of kidney function: a minimum-resource model containing only information available at a face-to-face encounter and a second version, encompassing additional information from laboratory test results available from the health record.
“For one of the markers, the minimum-resource model showed an 86.3% improvement, meaning that we could reduce the testing population by 46.3% while still detecting over 80% of those with abnormal kidney function. This is extremely promising for lower-to-middle income countries where laboratory test information may be lacking.”
Diabetes is the leading cause of kidney failure both in the UK and across the world and is responsible for roughly one third of people needing dialysis or a pre-emptive kidney transplant[1]. Early diagnosis is key; if CKD is recognised early, its progression can be effectively slowed down[2]. However, CKD is usually asymptomatic until it has progressed, making early intervention difficult without regular screening.
Rory Cameron, Gendius CEO, commented, “Our models can be used in two ways as part of a screening programme: where resources are limited, it can be used to reduce the screening population; or, if there is enough resource to screen everyone, it can be used to make sure those at a higher risk are screened first. Prioritising patients could result in earlier identification of CKD, enabling clinicians to intervene early to preserve kidney function.”
Gendius will be presenting further details on its chronic kidney disease pre-screening models at the European Association for the Study of Diabetes (EASD) conference in September 2022.
[1] https://www.kidneycareuk.org/about-kidney-health/conditions/diabetes/
[2] https://www.diabetes.org.uk/professionals/resources/shared-practice/kidney-care
