Personalising blood sugar targets

“What’s my HbA1c doc?” asks John right after he sits down. As usual he’s no-nonsense and straight to the point. We’ve been managing his type 2 diabetes for the last 2 years. It’s a condition that affects over 1.7 million Australians and is characterised by persistently elevated blood sugars.

He’s referring to a blood test that gives an estimate of his average blood sugar over the last three months. Essentially, it’s the amount of red blood cells with sugar stuck to them. The higher the average sugar, the higher the HbA1c and the higher his risk of long-term damage to his blood vessels. This can mean heart disease, strokes, blindness and kidney failure.

He’s right to be concerned.

The hope is that the more his HbA1c is lowered the more we can lower his chance of these serious complications. Unfortunately for the majority of Australians with type 2 diabetes their HbA1c is still too high.

“7.8 per cent,” I say. It’s lower than it was before, but still not at the generally accepted target of seven per cent. I can see my patient mulling this over further, not fully satisfied.

He pauses and then asks, “…. what should it be?”

And this is a tricky question.

How aggressive should we be in reducing John’s blood sugars?

There have been multiple large studies looking at low HbA1c targets. One earlier trial showed that people with type 2 diabetes who were treated with a lower blood sugar target did have a reduced rate of heart attacks, but this benefit was only seen after 10 years of follow up. Also, the patients in this trial were younger, in their 50s, and newly diagnosed.

Unfortunately, two large follow up trials in older patients either showed no benefit with regards to heart disease or an increase rate of death with aggressive treatment. That last trial was stopped early. Some reductions in eye disease and kidney disease were seen however.

And there are a whole host of other factors that determine if a patient might benefit from aggressive blood sugar lowering or not.

As I say, it’s tricky.

But there’s also some psychological and economic factors to consider. Trying to achieve a lower HbA1c can result in more doctor visits, medications and tests. It can also mean people become more distressed and anxious. It can also cause more falls by forcing the sugar level so low that the patient becomes dizzy.

Therefore, most guidelines around the world suggest that the Hba1c target be personalised to the individual. The clinician and the patient consider the benefits of a lower target, then consider the potential harms, to decide on a course of action.

My mind returns to the consult.

John is in his mid-50s. His only medical issue is type 2 diabetes.

He’d like to retire early and go sailing and he is the first to point out he can’t do these things if he develops multiple complications of diabetes later in life.

He decides to aim for a lower HbA1c target.

A new era for medicine

We are now in the age of personalised medicine. Previously John’s target would have been dictated to him by a one-size-fits-all guideline. Now the decision comes at the end of discussion between him and his GP.

But how can we, as researchers, better help patients and GPs to make these very personalised decisions?

At the University of Melbourne, we have developed an electronic tool that is embedded in a general practitioner’s practice software. The idea is that this aids the shared decision-making process, leading to better outcomes and patient engagement.

In fact, there is already good research that supports the use electronic clinical decision support tools. However, there is still a long way to go to make them practical and helpful to GPs and patients.

Our support tool, which we have called GlycASSIST, has been developed with input from diabetes nurses, general practitioners, endocrinologists and people with type 2 diabetes. Its development has been funded by the Royal Australian College of General Practitioners, Diabetes Australia and the Networked Society Institute.

It’s built on evidence-based algorithms and draws data automatically from electronic medical records to assist two main decision-making processes.

Firstly, it helps the GP and patient decide on a personalised target HbA1c result. This target represents the best way to lower the risk of diabetes complications whilst also minimising the burden and side effects of treatment.

And secondly it helps guide the discussion about how to add or change the medications the person takes to achieve this result. Recently there has been an explosion of new medications to treat type 2 diabetes. It can be difficult for GPs to keep up.

GlycASSIST gives a list of sensible medication options based on the patient’s allergies and kidney function. It then provides a list of important facts about the medication to help the patient choose between them based on their preferences for weight loss, the frequency of dosing, the side effects, their comparative HbA1c reduction potential, their chance of reducing risk of cardiovascular disease and whether they would prefer a tablet or injectable medication.

John chooses a once weekly injectable medication. He’s not so happy with the idea of needing to inject himself but he chooses it anyway as he likes the idea of losing some weight as well.

We know that decision aids at the point of care can help both the GP and their patient. We also know that, despite multiple government interventions, most people with type 2 diabetes are not being treated to target.

We hope GlycASSIST will start to narrow this treatment gap and provide more patient-centred care for people with type 2 diabetes.

Decision support tools are not limited to helping patients with type 2 diabetes.

We are building on these technologies to provide advanced support for the care of patients with other chronic diseases like chronic kidney disease and cardiovascular disease.

GlycASSIST is being developed in collaboration with Associate Professor John Furler, Dr Jo-Anne Manski-Nankervis, Dr Breanne Kunstler, Associate Professor Douglas Boyle and Sean Lo.

We are currently testing a second phase prototype with GPs in a computer simulation environment, and with focus groups comprised of people with type 2 diabetes.

Banner image: Getty Images