Quality of vitamin K antagonist treatment during the last year of life

Abstract

Limited data exist on the quality of anticoagulation in patients approaching the end of life. This study evaluated vitamin K antagonist (VKA) anticoagulation control during the last year of life, using nationwide data from Statistics Netherlands, linked to anticoagulation clinics' data and the Netherlands Cancer Registry. We included prevalent VKA users who were hospitalized with a severe medical condition and died between January 1, 2013, and December 31, 2019. Anticoagulation control was assessed using time in therapeutic range (TTR), time above therapeutic range (TAR), and time below therapeutic range (TBR) and the international normalized ratio (INR) variance growth rate (VGR), which reflects INR variability. Anticoagulation control was examined by two approaches: (1) over four intervals (0-12 months, 0-9 months, 0-6 months, and 0-3 months preceding death), and (2) in 3-month intervals (9-12, 6-9, 3-6, and 0-3 months before death) to describe temporal changes. Among 6874 VKA users in their last year of life (median age 82 [Interquartile range: 76-87] years, 46.9% female), the most prevalent severe medical conditions were heart disease (60.4%), cancer (16.2%), and hip fracture (15.2%). As death approached, TTR and TBR decreased, while TAR and mean VGR increased, particularly in the last 3 months of life. This decline was more pronounced in cancer patients and acenocoumarol users. In conclusion, the quality of VKA anticoagulation declined in the last year of life in severely ill patients, marked by reduced TTR and increased TAR and VGR, suggesting an increased bleeding risk. These findings highlight the importance of reassessing VKA use and considering discontinuation in patients approaching the end of life.

Figure 1

Flowchart of the selection of our study cohort. A severe medical condition is defined as “a diagnosis that carries an increased risk of mortality, hospitalization and emergency room visits.” DOAC, direct oral anticoagulant; INR, international normalized ratio; VKA, vitamin K antagonists. Created in BioRender. Kruip, M. (2025), https://BioRender.com/a20d177.

Figure 2

Anticoagulation control in the last year of life. Bar charts with 95% confidence intervals of the mean time below therapeutic range, time in therapeutic range, and time above therapeutic range, along with the mean variability of INR measurements expressed as the variance growth rate, calculated across four distinct intervals leading up to death: 0–12 months (A), 0–9 months (B), 0–6 months (C), and 0–3 months (D) before death.

Figure 3

Anticoagulation control in two subcohorts of VKA users stratified based on survival time since severe medical condition diagnosis. Anticoagulation control is expressed as the percentage of time spent within the therapeutic range (green line), above the therapeutic range (red line), and below the therapeutic range (blue line). The dotted lines represent the 95% confidence intervals for each metric. The variance growth rate is depicted with gray bars, accompanied by 95% confidence intervals.

Figure 4

Anticoagulation quality in two subcohorts of VKA users stratified based on survival time since severe medical condition diagnosis and type of severe medical condition. Anticoagulation quality is expressed as the percentage of time spent within the therapeutic range (green line), above the therapeutic range (red line), and below the therapeutic range (blue line). The dotted lines represent the 95% confidence intervals for each metric. The variance growth rate is depicted with gray bars, accompanied by 95% confidence intervals. Anticoagulation quality is displayed in patients with cancer surviving ≥365 days (A) and 180–364 days (B) following cancer diagnosis and in patients with non‐cancer severe medical conditions surviving ≥365 days (C) and 180–364 days (D) following severe medical condition diagnosis.