An investigation if the cost-effectiveness of medication monitoring in UK primary care: a case study of amiodarone

Talk Code: 

The problem

Amiodarone is an antiarrhythmic agent used to treat ventricular arrhythmias and atrial fibrillation in the community setting. Due to the high risk of thyroid dysfunction associated with the drug, treatment decisions involving amiodarone must weigh therapeutic benefit against the risk of ADEs in patients. Regular monitoring of thyroid function during the course of therapy is recommended in guidelines in order to reduce the risk amiodarone-induced hypothyroidism (AIH) or thyrotoxicity (AIT). Despite this, reported compliance with monitoring guidelines in primary care ranges between 20-70%. The evidence of the effect of amiodarone monitoring on the risk of ADE, cost and utility is sparse. This study used electronic records from the Clinical Practice Research Datalink to investigate the cost-effectiveness of alternative frequency of monitoring amiodarone therapy in UK general practice.

The approach

An observational cohort study using electronic patient records from CPRD was conducted to investigate the effect of monitoring frequency on the risk of ADEs linked to amiodarone therapy. The cohort was categorised according to frequency of monitoring (recommended, low frequency, high frequency) and time-to-event analysis with propensity score adjustment was used to estimate the treatment effect of monitoring. A Markov model was developed to simulate the cost of treatment and patient utility across the comparators over the course of follow-up.


Patient records between 2003 and 2012 were extracted for 8349 eligible patients with evidence of ongoing amiodarone therapy. 66.4% of patients had a record of monitoring during the observed period, with 37.3% monitored at recommended frequency or more. Compared to the baseline group with recommended frequency, the low frequency group had a significantly higher risk of AIH (HR 1.72 (95 C.I. 1.25-2.37), AIT (HR 1.35 (1.02-1.79). No statistically significant (5% sig. level) difference in the risk of ADE was found between patients with high frequency and recommended monitoring. The deterministic cost-effectiveness analysis comparing all 3 strategies found the low frequency monitoring strategy to be dominated (higher incr. cost and lower incr. utility relative to the other options); the incremental cost-effectiveness ratio (ICER) of the recommended strategy compared to the high frequency strategy was £1788.91/QALY.


This study has found that monitoring amiodarone therapy less frequently than recommended in guidelines is associated with a higher risk of both AIH and AIT. Despite this, 3 out of 5 patients in the cohort did not receive appropriate monitoring. Moreover, monitoring below recommended frequency led to higher cost of treatment and lower utility. When comparing recommended to high-frequency monitoring, the recommended frequency option yields both higher incremental cost and utility, with the ICER below the standard threshold for cost-effectiveness of £20,000/QALY. This preliminary result supports the current published guidelines which recommend 6-monthly TFT monitoring of amiodarone therapy.


  • Vladislav Berdunov
  • Anthony Avery
  • Rachel Elliott