Treatment Pathways for Patients With Atrial Fibrillation

J. A. Hodgkinson; C. J. Taylor; F. D. R. Hobbs


Int J Clin Pract. 2012;66(1):44-52. 

In This Article


The study consisted of 67,857 AF patients. About half of patients were overweight or obese, and 74% had CHADS2 scores of 0–2 (low or moderate stroke risk). Fifty-one per cent were male patients and mean age was 69.5 years (54% were aged 60–79 and 35% were 80 or over).

The results compared the treatment that AF patients started on against the average time over a period of 5 years and a period of 1 year during which they were on the range of treatments analysed. Thus, we can estimate the transition probability from one group of drugs to another, projected across 5 years. The results do not show the pathway that patients took from one treatment to the next during the analysis periods, and so we cannot provide any indication of the sequence of drugs that patients were prescribed.

Probability of Treatment Change

The states of treatment that were included in the Markov model of transition probabilities are all indicated in Table 1a, which details the amount of time patients remained on their original treatment regime in the period under analysis.

CHADS2 score did not correlate with likelihood of treatment continuation, regardless of initial treatment category. For example, those with CHADS2 scores of 5–6 were less likely to remain on anticoagulants alone (24.0% of time remaining on original treatment regime over 5 years) than those at low risk of stroke with scores of 0 (31.4% of time).

The proportion of time patients remained on their original treatment regime during the 5-year analysis period (Table 1a) varied from a high of 44.1% (anticoagulants + rate + rhythm) to a low of 13.8% (antiplatelets + rhythm). 38.1% of patients newly diagnosed with AF and prescribed no treatment continued to receive no treatment over the subsequent 5-year period.

Results from the 1-year analysis showed that treatment for AF was less likely than not to change within the first year after diagnosis, except if the original treatment incorporated antiplatelets (Table 1b; the proportion remaining on original treatment ranged from antiplatelets + rhythm 45.8% to anticoagulants + rate + rhythm 76.2%). Approximately 60% of those beginning on anticoagulants, on rate control or on rhythm control, including combinations of these treatment strategies other than antiplatelets, remained on their original treatments within the first year, but less than 50% of those started on antiplatelets did so (e.g. anticoagulants only 60.6%; antiplatelets only 49.7%).

Table 2a–c compare the proportion of time remaining on or converting to anticoagulant or antiplatelet therapy (as monotherapy or in combination with other treatment types) in the immediate years after diagnosis of AF, using initial treatment regime category and CHADS2 score.

There was a clear drop-off in the amount of time on anticoagulants between the first year and the average over 5 years for those who began with anticoagulant treatment (Table 2a; e.g. time on anticoagulants only declines from 95.7% to 66.3%). As much as one-third of treatment time of all those starting on a therapeutic approach involving anticoagulants featured no use of anticoagulants (either as monotherapy or in combination) over the 5-year period. Contrary to advice around suitable therapy for high- when compared with low-risk patients, there was minimal variation in the trends by CHADS2 score (between 63.1% and 69.2%).

Similarly, there was a clear drop-off in the amount of time on antiplatelets between the first year and the average over 5 years, for those who began with antiplatelet treatment (Table 2b). Indeed, of those starting on antiplatelet therapy, on average, a higher proportion of treatment time had been spent on anticoagulant therapy than antiplatelet therapy after 5 years [29.6% of time remaining on antiplatelets (Table 2a) as against 34.8% of time converting to anticoagulants (Table 2b)]. However, again this did not vary by CHADS2 score.

Over 5 years, those starting on anticoagulants were on no antithrombotic treatment (Table 2c) almost a quarter of the time (e.g. anticoagulants only 25.6%), whereas those starting on antiplatelets were on no antithrombotic treatment for almost a third of the time (e.g. antiplatelets only 32.5%). Although a higher proportion of time was spent on no antithrombotic therapy by those with a CHADS2 score of 0 than in patients with higher CHADS2 scores, regardless of treatment combination started with, arguably this difference by CHADS2 group remained relatively marginal.

Sensitivity Analyses

The results were found to be similar when the cohort was restricted by excluding patients with an index date prior to 1990, 1996 and 2002.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: