Summary. Over a 1-year period, 34 998 prothrombin time tests were performed on 2379 patients aged 40-89 years with a recommended target International Normalized Ratio (INR) value of 2AE5. At least one INR value of ‡ 5 was found in 507/2379 patients (21AE3%). The elderly demonstrated higher maximum and lower minimum INR values, lower warfarin doses, and an increased number of tests performed.After adjustment for these factors, the risk of INR values ‡ 5 increased with age by 15% every 10 years (95% CI 4-28%). We conclude that age is a risk factor for more unstable prothrombin time results.Keywords: INR, elderly, anticoagulation, control, instability.Older age is associated with an increased risk of bleeding in patients treated with oral anticoagulants (Hutten et al, 1999). The risk of bleeding increases in parallel to the International Normalized Ratio (INR) (Van Der Meer et al, 1996) and cessation of treatment is required when the INR is ‡ 5 (Hirsh et al, 1998). However, such values have not been consistently associated with older age, and the increased risk of bleeding in the elderly might not be due to an increased risk of over-anticoagulation, but rather to an increased bleeding tendency (Panneerselvam et al, 1998). On the other hand, the small number of patients in previous studies has limited their power and might explain inconsistent findings.We studied 2379 patients on chronic oral anticoagulation therapy aged 40-89 years in order to determine the effect of age on anticoagulation control.
PATIENTS AND METHODSWe identified 2379 patients (1237 men/1142 women) aged 40-89 years, who had received warfarin treatment between 1 January and 31 December 2000, had received treatment for at least 2 months before the study's inception and had conditions requiring a target INR of 2AE5. Patients were treated by general practitioners without the aid of a computer algorithm. In this region, the health maintenance organization serves approximately 750 000 people.Blood was obtained from patients being treated with warfarin using commercially prepared vacutainer sodium citrate tubes of 3AE17 ml draw and 13 · 75 mm size (Becton Dickinson, Franklin Lakes, NJ, USA). The blood samples arrived within 3 h from the various outpatient clinics. Prothrombin times (PTs) were performed on automatic microprocessor-controlled coagulation analysers (SYSMEX CA-6000; TOA Medical Electronics Company, LTD, Kobe, Japan) using dried rabbit brain thromboplastin with calcium (Dade Thromboplastin IS; Dade Behring, Marburg, Germany). The coefficient of variation between tests was 1AE5%, between analysers was 1AE4% and between days was 1AE9% (Froom et al, 2001). The laboratory is a member of an external quality control programme [UK National External Quality Assessment Scheme (NEQAS)].Patients were divided up into age groups of 10 years increments and into two subgroups: those with and without at least one PT result of ‡ 5 INR units, which is the recommended value for temporary cessation of treatment (Hirsh et al, 1998). For continuous variables, the...