The 4 things your organisation can do to improve evidence-based practice
Patients expect to receive the most effective care based on the best available evidence. But keeping up with evidence in today’s world is near impossible — to do so, consultant physicians might be expected to read 19 articles per day, 365 days per year. Clinical guidelines take this body of evidence for synthesis and evaluation with the result being, an up to date and evidence-based recommendation that assist clinicians on treatment decisions in various clinical scenarios.
“We’re finding that our increased compliance scores with evidence-based practice are correlating with improved patient outcomes an financial outcomes” Kelly Philiba, physician informaticist at Health Quest
Despite it’s potential to improve quality of care, standardise medical treatment and reduce costs, physician adherence to guidelines is low.
How do we use behavioural science to improve clinician adherence to evidence-based guidelines?
Whilst patients are under general anaesthesia, mechanical ventilation is required to keep patients breathing. There is increasing awareness that higher tidal volumes in mechanical ventilation can lead to lung injury.
Responding to new evidence, Emory Healthcare reduced the ventilator’s default tidal volume settings to 400ml and sent regular email feedback on the departments level of compliance. Doing so resulted in a 376% increase in the odds of compliance with lung protective ventilation strategies.
Defaults play on people’s bias towards the status quo where the current state of affairs is preferred and exploits our disposition to inertia — a tendency to do nothing.
2. Active Choice
Cardiovascular disease is the leading cause of death in the United States. Statins, a cholesterol lowering drug, has significantly reduced the risk of cardiovascular events; yet, clinicians don’t prescribe them to 50% of the patients that could benefit.
The University of Pennsylvania Health System implemented a system whereby physicians were sent a list of patients who met guideline criteria for statin therapy but had not been prescribed. The system asked clinicians to select, to prescribe a statin or not to prescribe a statin; if the decision was not to prescribe a statin, the system asked for a justification. This led to an increase in patients being prescribed a statin from 2.6% to 6.7%.
By removing default options but still forcing clinicians to choose, it increases the salience of a particular decision and achieves a higher level of perceived responsibility. Similar strategies to promote active choice which prompt clinicians to accept or cancel an order for mammography have increased breast cancer screening by 22.2%.
3. Social Norms
In the same study to improve rates of statin prescribing, the University of Pennsylvania also sent performance feedback comparing prescribing behaviours to peers. This resulted in an increase in the number of physicians submitting prescriptions (25% vs 13%).
We have seen the power of social benchmarking before. It leverages on our desire to fit into a group. Clinicians in particular want to feel that their practice is in line with that of their peers.
4. Make it Easy
Humans prefer performing simple behaviours rather than hard ones. Shakespeare et al changed the inpatient medication chart for post caesarean section patients to include highlighted, pre-printed medication orders for regular and breakthrough analgesia as well as postoperative nausea and vomitting treatment. This resulted in an increase in the prescribing of guideline concordant analgesia from 40% to 90%.
When interventions reduce friction to performing a behaviour, compliance with said behaviour increases.
Improving compliance with evidence-based medicine is possible. Small changes in the choice architecture can respect clinician autonomy but also nudge physicians towards evidence-based choices. Our patients deserve it.
Defaults — Which clinician behaviour responds best to defaults? How can I use defaults without compromising clinician autonomy?
Active Choice — How do I frame choices in a way that gets clinicians to think about loss aversion?
Social Norms — What clinician behaviours would change if I showed clinicians how they performed in comparison to others?
Make it easy — What are the barriers that make it harder for my employees to perform a certain behaviour? Can I redesign their workflow to make it easier?