The inpatient diabetes audit: A tool for change
Among the data presented at this year’s Diabetes UK Professional Conference in London were the 2017 results of the National Diabetes Inpatient Audit – NaDIA. medwireNews spoke to the national clinical lead for inpatient diabetes Gerry Rayman (Ipswich Hospital NHS Trust, UK) about the challenges of improving care, and why the audit itself may be the most powerful tool for change.
Auditing for change
The 2017 NaDIA data were broadly favorable, with positive trends seen for a number of measures. Among these was a reduction in glucose management errors, which fell to 18.4% having remained stubbornly at around 23% since 2011. Rayman also highlights the fall in patients requiring injectable medication to rescue them from hypoglycemic coma, from 2.1% in 2011 to 1.3%. In addition to these improvements, there were reductions in hospital-acquired foot ulcers and a rise in the proportion of patients being assessed by the diabetes team.
But the specific factors driving these trends are not obvious. Although the adoption of care improvement initiatives is on the rise, with the proportion of sites at least partially using electronic patient records, electronic prescribing, and remote glucose monitoring up to 67%, 39%, and 65%, respectively, only 12% of sites fully use all three, and the small increases from 2016 seem unlikely to account for the size of the fall in glucose management errors.
… everyone wants to deliver better care for their patients, so that’s what’s driven the improvements…
Rayman, however, takes a broader view, suggesting that the mere existence of the audit is itself a large driver for change, by highlighting, perhaps for the first time, problems people were aware of without realizing their extent. And he adds that the willingness of hospitals to participate, which involves collecting comprehensive data for 1 week in September, “indicated the desire of people to want to know and want to improve.”
In 2011, the first year of the full audit covering England and Wales, 206 of 213 eligible sites took part, and 208 participated in the 2017 audit.
“The fact that that hasn’t dropped off at all is so important, again demonstrating the benefits that people see in the audit – everyone wants to deliver better care for their patients, so that’s what’s driven the improvements, I think,” says Rayman.
The results of the audit have, he feels, led to “more attention to insulin safety and prescribing in general in trusts, so I think it’s all just raising awareness of the issues about having diabetes and prescribing in hospitals.”
In any case, picking apart the specific elements that improve diabetes care is a challenge in itself. One problem Rayman highlights is that the evidence underlying the effectiveness of, for example, specialist inpatient diabetes nurses, comes from studies that are 10 years old. At this time, the average length of hospital stay was around 10 to 11 days, whereas the pre-intervention length of stay at Rayman’s Ipswich Hospital was 7.5 days, “so to reduce that is quite a tough ask.”
Despite this issue, efforts at research seem to have fizzled out in recent years, bar a couple of recent high-profile trials of novel technologies, with people content to base their improvement drives on old data. “It's only purists like me who want to really know the answers,” Rayman laughs.
But purists who wish to conduct up-to-date research run into a different problem. Rayman originally conceived the Diabetes Inpatient Care and Education (DICE) project – with the associated comprehensive inpatient diabetes care pathway, implemented at Ipswich Hospital – as a cluster randomized trial, but struggled to find funding because potential sponsors felt it was “too generic.”
They complained that “there’s too many things that you’re doing, so how do you know it’s a blood glucose meter or how do you know it’s the use of a care plan?” he says. But “to me that doesn’t matter – it’s a whole-system change.”
The DICE project
This whole-system change grew partly from the efforts of a key group of hospital staff: the diabetes inpatient specialist nurses (DISNs). Rayman sees them as educators, but “their role is also to innovate and come up with new ideas, new ways of doing things and implementing change.”
In his hospital, he encourages DISNs to identify adverse events, to explore why they occur, and to consider how they can be prevented in the future. “And so as a result of that we’ve come up with a lot of good, novel ideas, and that’s what formed the DICE program.”
The DICE care pathway has a dual function, providing information on, for example, hypoglycemia symptoms and novel insulins, as well as helping to ensure safe bedside care for diabetes patients. For the latter purpose, it includes admission checklists, a scoring system to identify patients for referral to the diabetes team, insulin prescription and blood glucose monitoring charts, and a management pathway to identify/prevent foot ulcers.
Preliminary results from the project showed impressive results, based on the hospital’s 2012 and 2013 NaDIA data; for example, medication errors fell from 57% to 21%, with insulin errors falling from 31% to 7%, and the incidence of severe hypoglycemia declined from 15% to 10%. These reductions are all the more striking when you consider that the DICE project was initiated in August 2013 – just 1 month before the NaDIA data were gathered, suggesting that it had a rapid positive effect.
If people adapted what we did, I think we could reduce length of stay across the whole country.
More recently, but as yet unpublished, the DICE team analyzed background trends in key indices finding that although length of stay was trending downwards in the overall hospital population, the reduction was significantly larger among those with diabetes (from 7.5 to 6.7 days). And this did not come at the expense of more readmissions; although this was on a general upwards trend, the increase was less among diabetes patients.
“And then the other thing – and no one’s done this before – is that for every single patient over the study period, we looked to see if complications arose while in hospital, such as pneumonia, falls, urinary tract infections, myocardial infarction, and so on,” explains Rayman. “And the number of patients who developed a complication in hospital was significantly reduced in those who were in the DICE project.”
Set up to fail?
In the published DICE findings, the team also reported very positive effects on the knowledge and confidence of trainee doctors of all levels, the very people who, along with the nonspecialist nurses, are responsible for the majority of the errors identified in the audit.
Rayman observes that the DISNs “audit the drug charts, and so they’ve got a very good eye for spotting errors, and are pretty au fait at not making the errors themselves, because they’ve learnt by spotting errors.”
General ward staff, by contrast, lack this experience and, moreover, are under extreme pressure in the current NHS. “I’ve got a real issue with what’s going on in the health service at the moment, and the winter this year has been just awful,” says Rayman. He is disparaging about the effectiveness of the winter planning, and while acknowledging the apology issued by NHS England to patients, feels that this should be extended to staff, who have been placed under such extreme pressure as to make increased errors almost inevitable. “It’s fine thanking them,” he says, “but actually an apology is necessary.”
I’ve got a real issue with what’s going on in the health service at the moment, and the winter this year has been just awful.
Yet specialist support for these staff is patchy. The 2017 audit showed that 28% of hospitals have no DISNs and 73% have no specialist dietetic support. And even where they do exist, specialists are available only 9 to 5, and, in the vast majority of hospitals, are unavailable at the weekend.
“And diabetes doesn’t go away at the weekend – nor does it go away in the middle of the night,” says Rayman.
And while it might seem ideal for specialist nurses to be on hand 24/7, Rayman argues otherwise, saying: “The last thing we want is our specialist nurses to do the work, otherwise the others will become downskilled. And this includes the junior doctors.”
His own experience suggests that empowering specialist nurses to manage patients’ insulin leads to the junior doctors becoming disempowered, leaving it in the hands of the nurses and thus being ill-equipped to deal with problems outside of the specialist nurses’ working hours. “So it’s a real danger.”
For this reason, he believes the most critical part of a DISN role is educational. “I think they need to identify patients who are likely to get into a major problem, and focus on those, but educate the rest of the staff to deliver good diabetes care.”
New tools for change
A lesson from the success of the DICE project is that all the tools needed to deliver effective and safe inpatient diabetes care already exist and have done for many years – it is just a matter of using them. That said, one of the very few recent major studies to look at inpatient care did demonstrate the benefits of a technological innovation: remote glucose monitoring.
“I think remote monitoring is fantastic,” says Rayman. He believes the data need not be burdensome, and require constant monitoring. Instead the DISNs look at the data on a daily basis, identify patients who have had a first hypoglycemic event, and try to uncover the underlying factors to prevent further episodes. Indeed, being able to examine recurring themes within the data for multiple patients has helped them to identify the circumstances that frequently lead to a first event and take steps to avoid it.
I think remote monitoring is fantastic.
The data confirmed that hypoglycemia was more frequent overnight. Rayman says: “I think it’s partly because patients come in, we give them their usual insulin, but at home they’re having a much bigger evening meal than they get in the hospital. Not only that, at home they’ll eat at 7, whereas in the hospital they’re fed at 5 for catering purposes to get staff off early.”
And on top of that, “they're not given a hot meal – they’re actually just given a little box of food. So if you do that and give them the same dose of insulin that they have at home, you’re almost inevitably going to produce hypoglycemia. If you give them the same dose of sulfonylurea – the same problem.”
As a result, when diabetes patients are admitted to Ipswich Hospital, basal insulin, if used, is reduced by 20%, sulfonylureas are stopped if the patient is on a low dose and high doses are halved. In addition, patients are now offered a bedtime snack, “whereas there were no snacks available prior to us doing this.”
Another recent study looked at the potential for the artificial pancreas to improve glycemic control for insulin-dependent patients with type 2 diabetes. Rayman describes this as “a very worthwhile study,” but notes that the patients were all cared for by specialists and very closely monitored; he believes the general wards “would get into so much difficulty” attempting to use a closed-loop system. It is, however, early days for closed-loop systems; future systems that function dependably with little specialist input could, if affordable, save nursing time and improve patient outcomes.
Auditing for change (reprise)
Technological innovations aside, a project such as DICE must be delivered by people – and not just the front-line medical staff.
“You need to have leadership, you need to have a project manager, you need to have the involvement of management and IT,” says Rayman. “So it needs to be a project which is going to be delivered by the trust and people are going to sign up to do it, and they need to be meeting every week and they need to be monitoring what they’re doing, so they get feedback.”
He believes that, broadly speaking, the DICE project is transferable to other hospitals, although some adaptations to suit local conditions may be needed. “But all the DICE says is that if you concentrate on trying to solve the problems you can improve the care.”
There are things which will work, and then you turn your back on it and it slips up again, so you need to keep that continual audit to keep driving it.
Rayman is working with Diabetes UK to produce a suite of tools for inpatient diabetes care, which other hospitals can adopt and adapt. “If people adapted what we did, I think we could reduce length of stay across the whole country.”
And here we return full circle. The audit kick-starts the process, by identifying and quantifying what needs to be addressed, but becomes equally essential after the tools for change have been applied.
Rayman says: “The most important thing [other hospitals] need to adopt, and this is what’s driven me all the way along, throughout my service development – you need to audit it, you need to know that [when] you’ve put something in place: is it working? And you will not know that unless you audit it.”
And auditing must be an ongoing process to be effective, he stresses. “There are things which will work, and then you turn your back on it and it slips up again, so you need to keep that continual audit to keep driving it. And then some things work really well and then people get bored with it, and you need to change it and come away with something different for them to do. It is constantly looking at the service.”
medwireNews is an independent medical news service provided by Springer Healthcare. © 2018 Springer Healthcare part of the Springer Nature group