From evidence to practice: consensus in cardiovascular risk assessment and diabetes

Abstract

Rigorously developed evidence based practice guidelines are necessary to promote clinical interventions that are consistent and most effective. Recommendations for treatment should be linked clearly to proven reduction in absolute risk, expressed as number needed to treat, rather than being based on the more commonly quoted and potentially misleading reduction in relative risk. Data published since 2003 has been incorporated in a 2009 revision of the New Zealand guideline for assessment and management of cardiovascular risk. Important changes are the decrease in optimal levels of LDL-C from <or= 2.5 mmol/L to <or= 2.0 mmol/L, the provision of a risk trajectory tool for younger subjects, and the removal of the metabolic syndrome from risk assessment. A review of emerging risk factors concluded that none are suitable for use in routine risk assessment, confirming the statement in the 2003 guideline. The revision allows HbA1c as a screening test for diabetes and data to allow the use of this analyte in cardiovascular risk assessment is being gathered. The use of HbA1c as a diagnostic test for diabetes was seriously considered and this foreshadowed very recent international recommendations in this area. Such a use will be the subject of vigorous debate in the immediate future.

Figure 1

Hierarchy of evidence related to risk of bias.

Figure 2

Illustration of absolute risk, absolute risk reduction, number needed to treat and relative risk reduction. The top panel shows 50 subjects at risk with 2 (4%) having an event over 5 years. The bottom panel shows the effect of treatment for 5 years. One subject has an event (2%). The ARR is 2% and NNT is 50. The RRR is 50% (2/4×100).