Warfarin can be safe and effective in the extreme elderly

SIR—We read with interest the article by Vidya Perera and colleagues that frail elderly patients were unlikely to receive warfarin as they generally have adverse clinical outcomes (some of which may be unrelated to anticoagulation) [1]. We do not consider age to be a barrier if clinical benefits outweigh the obvious risks and assess each patient on their own merit. The risk of emboli in those with atrial fibrillation without anticoagulation is around 4.5% per annum that rises with advanced age [2]. As a district general hospital catering to a population of 150,000, we have a significant proportion of non-agenarians (age >90 years) on warfarin and run three anticoagulation clinics per week to ensure target international normalized ratio (INR), and dosing records are up to date. We gathered data on these patients who had been on warfarin between the years 2001 and 2006 and identified warfarin-related fatalities and non-fatal hospital admissions. A total of 69 patients (15 males and 54 females; age range 90–99 years) had been on warfarin for a total of 261.25 years (median duration on warfarin, 3.78 years). A total of 51 had non-valvular atrial fibrillation, 3 recurrent deep venous thromboses, 12 recurrent pulmonary emboli, 1 had an intracardiac thrombus and 2 suffered from recurrent cerebrovascular accidents. Warfarin was stopped in two patients, one due to poor control of INR and one for recurrent falls. There were five deaths that could have been linked to warfarin and included three gastrointestinal and two intracranial bleeds (annual mortality 1.9%). Both patients who died from intracranial bleeds had INRs in the therapeutic range at admission. A total of 22 patients died from conditions unrelated to use of warfarin. There were 15 warfarin-related admissions—2 with epistaxis, 1 per vaginal bleeding, 4 gastrointestinal bleeding and 8 were admitted due to significantly raised INR. All seven patients admitted with bleedings had INRs in the therapeutic range. No thrombotic or embolic events were recorded in any patient. All our patients were provided with written patient educational information (in large print where required), and we ensure that treatment is adhered to the safety indicators recommended by the British Committee for Standards in Haematology (BCSH) [3]. This is possibly the first survey that shows warfarin to be effective in preventing embolic/thrombotic events in this elderly population with an acceptable risk of 1.9% mortality per annum. Strategies into VKORC1 (vitamin K epoxide reductase) or CYP2C9 genotyping help identify patients at high risk of bleeds, potentially more out-of-range INRs and high sensitivity [4], but these are not available routinely at present. Until then, successful and safe anticoagulation in the extreme elderly would require knowledge of patient home environment, patient education, review of medication in outpatient visits (drug interactions), assessment of compliance (including ensuring availability of multiple strengths of warfarin tablets 1 mg/3 mg/5 mg) but most importantly, effective communication. None declared.