Controlled trial of effect of documented cardiovascular risk scores on prescribingBMJ 2003; 326 doi: http://dx.doi.org/10.1136/bmj.326.7383.251 (Published 01 February 2003) Cite this as: BMJ 2003;326:251
- L M L Hall, medical student,
- R T Jung, consultant,
- G P Leese (), consultant
- Correspondence to: G P Leese
- Accepted 18 April 2002
Cardiovascular disease causes the death of around 80% of patients with type 2 diabetes.1 However, risk factors for cardiovascular disease in such patients are often untreated2 despite the proved benefits of intervention. 3 4 One way to help clinicians identify patients at high risk of cardiovascular disease is to use cardiovascular primary prevention risk tables. These tables integrate the multiple risk factors into a single score. We did a pilot study to test the hypothesis that documentation of a cardiovascular risk score in the case notes would improve the management of cardiovascular risk factors.
Participants, methods, and results
We recruited patients with type 2 diabetes who had no history of cardiovascular disease or renal disease. All patients were aged 35–75 years and attending a hospital outpatient clinic. We recruited 323 patients (167 men and 156 women) attending the clinic consecutively. Patients were seen by one of six doctors who were unaware of the ongoing project. We allocated patients alternately to the experimental and control groups (162 experimental, 161 control), and all doctors saw an equal number of experimental and control patients. The University of Dundee special study module subcommittee approved this project on behalf of Tayside Committee on Medical Research Ethics.
We calculated New Zealand cardiovascular risk scores for all patients.5 These were clearly documented at the front of the notes of patients in the experimental group only. Standard information, such as weight; glycated haemoglobin, urinary microalbumin, and total and high density lipoprotein cholesterol concentrations; and blood pressure was available for all patients in both groups.
Only 42 patients (13%) had a low risk for a cardiovascular event (<10% five year risk), with 113 (35%) having moderate risk (10-20% risk) and 168 (52%) a high risk (>20% risk). Overall, there were no significant differences between control and experimental groups in the primary outcome measures (table): change of diabetes treatment (36% v 42%), lipid lowering drugs (9% v 12%), or blood pressure drugs (10% v 16%) and referral to dietician (13% v 10%). There were no differences in other interventions between the control and experimental groups. Among high risk patients, however, those in the experimental group were more likely to be prescribed blood pressure and lipid lowering drugs than those in the control group (P<0.02, Mantel-Haenszel test). Despite this difference, the time until the next hospital outpatient appointment was the same in the two groups, with 24% in each group (39 in the experimental group and 38 in the control group) receiving an appointment in less than six months.
We found that clear documentation of a cardiovascular risk score in the notes increased prescribing of risk modifying drugs for patients with diabetes who are at high risk of cardiovascular disease. More high risk patients in the experimental group were prescribed both blood pressure lowering and lipid lowering drugs. However, there was no increase in prescribing for patients at relatively low risk.
Although individual risk factors such as blood pressure, smoking status, and lipid concentrations are generally available in clinics, integrated cardiovascular risk scores are often not calculated because of lack of time. This leaves the clinician with complex clinical data that can be difficult to interpret and are thus often not acted on. Our results indicate that it is worth developing clinical support systems that will calculate cardiovascular risk before the consultation.
We thank Dr S Ogston for statistical advice.
Contributors: GPL designed the study, wrote the protocol, and helped to write the paper. LMLH conducted the study and wrote the paper. RTJ helped analyse the data and write the paper.
Competing interests None declared.