Support Careflight save lives
view counter

eHealth records key to thrombosis risk control

Using eHealth records, hospitals can address a major in-patient killer and reduce preventable deaths, a US-based Professor told NSW Health specialists last week.

In hospitals across Australia, venous thromboembolism (VTE) affects around 30,000 patients each year, and with 5,000 deaths a year, it is among the leading causes of preventable hospital deaths.

Risk factors include major trauma, major surgery and prolonged immobility, so it’s not surprising that hospital patients are the main victims, but despite having tested best-practice guidelines in place, these are underutilised world-wide – with Australia no exception.

This month, NSW Health’s Clinical Excellence Commission invited Professor Greg Maynard to visit to share his effective methods in managing VTE for in-hospital patients.

Maynard heads the Center for Innovation and Improvement Science at the University of California, San Diego where he has been involved in a long-term project to optimise prevention of venous thromboembolism (VTE) (which includes deep vein thrombosis – DVT – and pulmonary embolism - PE).

He’s using similar methods to also improve glycemic control in hospital patients – and since the project started in 2005, with their first report published in 2008, the UCSD team estimate an ongoing reduction in DVT incidence of around 40 per cent.

Patients are assessed for DVT risks on admission, on transferring to a different level of care, and in the medication ordering process.

“We use different phases of clinical decision support, and a key aspect is building guidance into the ordering process,” he says.

That’s a step away from most NSW hospitals, where computerised physician order entry isn’t routine, he acknowledges.

“Our work then involves integrating a simple DVT risk assessment model into the flow of work, so as a patient is admitted or transferred, they always go through this risk assessment model, which seamlessly guides people towards selecting the correct prophylaxis for that patient’s profile.”

The process doesn’t stop there, he adds – and getting it right meant that built-in to the system is the awareness that clinicians won’t always get it right the first time.

“When patients get admitted or transferred, we try to get them on the right prophylaxis at that critical point when they are just entering the system or going through a different level of care,” he says.

“And since patients change in their bleeding and clot risk – and as we know, it is not 100 percent that people will get it right the first time, another part of the clinical decision support is to try and set up monitoring systems to look for patients who may not be on the right prophylaxis or whose situation may have changed.”

Maynard calls the technique measure-vention - a cross between a measurement and an intervention – and this is where eHealth records play a critical role.

“We perform an act of surveillance with the help of the electronic health record to identify those who might not be on the correct prophylaxis and then we triage them to see if that is indeed the case or not,” he says.

The team developed a ‘toolkit’ in 2008 which led to a series of multi-site collaborative improvement efforts across over 300 hospitals and all University of California hospital programs are involved in similar interventions, he says.

“We accomplished about a 40 percent reduction in hospital associated DVT and we had good results in maintaining that high prophylaxis rate which we can monitor electronically now.”

On a practical level, the measures are in-built into the medication order set.

“We try to make sure that any elements that we need to monitor are captured as discrete data elements that we can pull out later, as opposed to text which is more difficult to pull out,” he says.

“As doctors fill out the risk assessment form, they are capturing what level of risk they are declaring the patient to be and whether or not the patient has a contra-indication.”

Lab results feed into the record too, he adds – so contra-indications, like a high INR, low haemoglobin or a low platelet count are highlighted.

“We also capture their ambulation and track to see if they are being classified correctly,” he says, adding that nurses already track movement to help prevent pressur sores in patients.

“We then colour-code this report to show if a patient is on an anticoagulant, a mechanical prophylaxis or no prophylaxis.”

Patients who have an anti-coagulant medication are coded green, those with ‘mechanical prophylaxis’ (such as elastic bandages, pneumatic devices and / or exercising lower extremities) coded yellow and patients with nothing, coded red.

“If you got a report with three different patient types on a given ward, you would focus on those mostly in the yellow or the red zones. This can help focus efforts where they are most needed, to detect under-prophylaxis.”

The report captures not only whether a patient is on an anti-coagulant, but also why the anticoagulant is prescribed, he says.

“If they are on a therapeutic anti-coagulant, then we assume they are protected. If they are on a prophylactic anti-coagulant, we look to see if they are ambulating – because if they are fully ambulatory, on prophylactic anticoagulant, it might be a cue that the patient is over-prophylactic.”

Other patients might be coded orange to indicate that they are on mechanical prophylaxis, but if a contra-indication was captured, their treatment plan is appropriate, so more monitoring might be appropriate.

“Our job is to make that frontline monitoring very easy so by doing a risk assessment, and re-doing it every day on 30 patients, it’s not a burden because they are actually focusing most care on the four who might be questionable.”

Maynard says that his team have moved on to devising other real-time monitoring strategies to complement the traditional month-to-month reports.

“You can actually bundle a lot of these. If you can imagine, a dynamic dashboard which is basically a live checklist of quality measures that gets done in real time, we are actually building something like that now, so we might have 10 crucial measures, all on the same report for patients on award.”

Some hospitals are consolidating this type of screening and putting it in the hands of one person, he adds.

“Some systems I work with like the idea so much that they went from intervention of one thing that they have created what they are calling a measure-ventionist – that’s the person who looks at these ten things at once.”

Maynard says that the whole system revolves around taking care of quality deficits on the same day that they are detected and doing something about it immediately rather than getting a report three months later showing care was sub-optimal when it’s too late to do anything for the patient.