Most clinicians and medical researchers believe the NHS should adopt an evidence-based approach to patient care. This will lead to greater effectiveness, fewer inappropriate interventions, greater consistency and less overuse or underuse of healthcare resources.

However, there is a sense that GPs are drowning in a sea of evidence-based knowledge and guidelines. Coupled with the varying availability of some drugs, it is likely some patients are not getting evidence-based treatment. Initiatives such as the quality and outcomes framework help to address this, but there is scope to improve our assessment of gaps in evidence-based practice.

We recently worked with a primary care trust to examine the application of evidence-based care, using a combined model developed jointly with the King's Fund. This exercise allowed us to identify where evidence-based care should be, but is not, being used.

For example, unless there are contra-indications, virtually all patients with coronary heart disease and/or congestive heart failure should be treated with beta-blockers.

Our analysis of beta-blocker usage in CHD/CHF patients in 27 GP practices in the PCT shows that there is an alarming number of patients who are not on this evidence-based medication, as well as revealing a significant variation between practices (see first chart).

We were able to get an even better insight by identifying underuse of evidence-based treatment for different risk groups (based on the risk of hospitalisation). We did this for the PCT by combining risk scores with clinical insights into what may be driving these scores to stop patients from moving into a higher-risk category.

We know that all those at the tip of the risk pyramid (second chart) are candidates for case management - but what are the specific opportunities to manage the care of those in the moderate risk category where it is harder to find cases but where opportunities for sustainable change exist?

We identified the 9.8 per cent of moderate risk patients who have CHD and, of those, around 30 per cent were not on beta-blockers. Similarly, we also looked at the proportions of moderate risk hypertension patients who have blood pressure above 150/90, and of patients with diabetes and CHD not getting lipid medication (third chart). Again, these are potential areas for evidence-based intervention.

This analysis provided clinical and risk-adjusted insights into how to best design intervention and target the right cases. We used these insights to prioritise how to close gaps in treatment.

There may be good enough reasons why some patients are still not receiving optimal treatment, but it is clearly best to identify these people, and the variations in use of evidence-based treatment, so that decisions can be made about the benefits of moving towards optimal treatment.

*080327/data briefing images