Atrial fibrillation and heart valve disease: self-monitoring coagulation status using point-of-care coagulometers (the CoaguChek XS system)

The External Assessment Group did not carry out a formal evaluation of the performance of the CoaguChek system or the INRatio2 PT/INR monitor with regard to INR measurement because it was outside the scope of this assessment. However, an objective 'true' INR remains to be defined and INR determined in the laboratory is regarded as the gold standard to which all other measurement methods should be compared. Information on the precision and accuracy of these point‑of‑care coagulometers was therefore gathered from the available literature.

A systematic review by Christensen and Larsen published in 2012 assessed the precision and accuracy of currently available point‑of‑care coagulometers including CoaguChek XS, INRatio and ProTime. The authors found that the precision of CoaguChek XS varied from a coefficient of variation of 1.4% to 5.9% based on data from 14 studies. The precision of INRatio and ProTime varied from 5.4% to 8.4% based on data from 6 studies. The coefficient of correlation for CoaguChek XS varied from 0.81 to 0.98, and that for INRatio varied from 0.73 to 0.95. The review concluded that the precision and accuracy of point‑of‑care coagulometers were generally acceptable compared with laboratory‑based INR testing. The same conclusions were drawn by the Canadian Agency for Drugs and Technologies in Health report published in 2012 on point‑of‑care testing. Similarly, the international guidelines prepared in 2005 by the International Self‑Monitoring Association for Oral Anticoagulation stated that 'Point‑of‑care instruments have been tested in a number of different clinical settings and their accuracy and precision are considered to be more than adequate for the monitoring of oral anticoagulation therapy in both adults and children'.

Six studies compared the performance of CoaguChek S with that of CoaguChek XS in relation to conventional INR measurement. The studies showed a good agreement between the 2 CoaguChek models and conventional laboratory‑based testing results. However, the CoaguChek XS showed more accurate and precise results than CoaguChek S in both adults and children, especially for higher INR values (>3.5).

The systematic review identified 12 relevant economic evaluations. All of these evaluations compared INR self‑monitoring strategies with standard care and were assessed against the NICE reference case by the External Assessment Group. The results of the studies included in the systematic review varied widely and showed that the cost effectiveness of self‑monitoring was dependent on a number of key factors.

The adopted perspective and the initial costs associated with self‑monitoring appeared to substantially affect the cost effectiveness. Self‑monitoring strategies appeared more favourable than standard care when a wider societal perspective was adopted, as a result of lower time costs associated with fewer health service contacts. The size of the estimates of effect applied to self‑monitoring in reducing thromboembolic and bleeding events compared with those applied to standard care also appeared to affect cost effectiveness. The 2 UK‑based evaluations applied effect estimates consistent with small or negligible differences between self‑management and usual care with respect to time in therapeutic range and adverse thromboembolic and haemorrhagic events. This resulted in a low probability of self‑monitoring being cost effective. Several studies that applied large effect estimates resulted in a high probability of self‑monitoring being cost effective.

The 2 UK‑based economic evaluations were based on data from the same trial. One evaluation adopted an NHS and wider societal perspective, and the other adopted an NHS and personal social services perspective. Self‑monitoring strategies appeared to increase the costs of INR monitoring in the short term and because these 2 evaluations applied small effect estimates, consistent with those seen in the largest UK‑based trial of patient self‑management, self‑monitoring of INR appeared unlikely to be cost effective. However, no UK‑based trials have been sufficiently powered to detect a statistically significant difference between standard INR monitoring and patient self‑monitoring in terms of major thromboembolic or haemorrhagic events. Therefore, the External Assessment Group carried out a meta‑analysis of relevant trials including evidence from a number of European trials in which standard care is similar to that provided in the UK in terms of approach, frequency of testing and the level of INR control achieved.

The External Assessment Group developed a de novo economic model designed to assess the cost effectiveness of self‑monitoring (self‑managing and self‑testing) coagulation status using 2 different point‑of‑care coagulometers: the CoaguChek XS system and the INRatio2 PT/INR monitor.

The structure of the Markov model was based on the review of published models of INR self‑monitoring and previous models evaluating the cost effectiveness of new anticoagulant drugs compared with warfarin therapy in people with atrial fibrillation. A further unpublished economic model of INR self‑monitoring was provided by the manufacturer of CoaguChek XS, and this model was also used to inform the structure of the new economic model.

The Markov model compared the alternative monitoring strategies for a hypothetical cohort of people with atrial fibrillation or an artificial heart valve, and was used to simulate the occurrence of thromboembolic and bleeding events over a 10‑year period. People with atrial fibrillation or an artificial heart valve represent the majority of people on long‑term vitamin K antagonist therapy. The model simulated transitions between the discrete health states, and accumulated costs and quality‑adjusted life years (QALYs) on a quarterly (3 month) cycle. Within each cycle, the simulated cohort was exposed to a risk of the adverse events as well as death from other causes. The adverse events included in the model were ischaemic stroke (minor, non‑disabling, and major, disabling or fatal), systemic embolism, minor haemorrhage, and major haemorrhage (intra‑cranial haemorrhage, including haemorrhagic stroke, gastrointestinal bleed, and others). A constraint was applied whereby the simulated cohort in the model could only experience 1 event per cycle.

The model was populated using data derived from the systematic clinical effectiveness review, other additional focused reviews to inform key parameters (for instance baseline risks), routine sources of cost data, and where necessary some study‑specific cost estimates based on expert opinion.

Data on the resource use and costs associated with the alternative monitoring strategies were informed by published literature, existing guidance, expert opinion, manufacturers' and suppliers' prices, and other routine sources of unit cost data. Some costs were informed by expert opinion where suitable data from other sources were not available.

The baseline utility value for people with atrial fibrillation or mechanical heart valve who were stable was taken as the baseline EQ‑5D value from trial data, 0.738. This value was applied to 65–70 year old people and adjusted by the External Assessment Group to estimate age‑specific baseline utilities in the model.

Utilities associated with acute events were applied for the 3‑month period after the event. For post‑event states with associated ongoing morbidity, the appropriate health state utilities were applied for all subsequent cycles spent in these states. Half‑cycle corrections were applied, by assuming that people experienced events on average at the mid‑point of the cycle. Thus a patient starting off in the well state and experiencing a major stroke in a given cycle of the model would accrue 6 weeks at the utility value for well and 6 weeks at the utility value for major stroke.

For the purposes of decision‑making, the incremental cost‑effectiveness ratios (ICERs) per QALY gained were considered. The following assumptions were applied in the base‑case analysis:

• 66.45% of standard care monitoring was done in primary care by practice nurses.

• 60% of the cohort had atrial fibrillation and 40% had an artificial heart valve.

• The average age of the cohort was 65 years, and 55% were male.

• 50% of people who self‑monitored did self‑testing and 50% self‑managed.

• The increase in the number of tests done per year with self‑monitoring was 23 (that is, 35 tests compared with 12 tests in standard care).

• Relative treatment effects were estimated and applied separately for self‑testing and self‑management.

• 15% of participants did not start self‑monitoring after training (training failure).

• 10% of participants stopped self‑monitoring within a year of starting.

• Self‑monitoring device costs were annuitized over 5 years.

• 75% of devices were reused by another patient when a patient stopped self‑monitoring.

The results indicated that over a 10‑year period, introducing self‑monitoring would reduce the proportion of people experiencing a thromboembolic event by 2.5%, while slightly increasing the proportion having a major haemorrhagic event by 1.4%.

The predicted monitoring costs were higher with self‑monitoring compared with standard monitoring, but the total health and social care costs were similar and in some cases lower. The QALY gains were greater for self‑monitoring than standard monitoring. For all of the self‑monitoring coagulometers there was a QALY gain of 0.027 compared with standard monitoring. Self‑monitoring with the INRatio2 PT/INR monitor was £29 cheaper than standard monitoring. Self‑monitoring with the CoaguChek XS system was £37 more expensive than standard monitoring. Therefore, in the base‑case scenario, the self‑monitoring strategies compared favourably with standard care. The INRatio2 PT/INR monitor dominated standard monitoring in the analysis because it was less costly and more effective. The ICER for the CoaguChek XS system was £319 per QALY gained compared with standard monitoring. The lower cost of the INRatio2 PT/INR monitor and testing strips, coupled with the assumption of equivalent clinical effectiveness, meant that the INRatio2 PT/INR monitor also dominated the CoaguChek XS system. However, it should be noted that no direct evidence of clinical effectiveness was identified exclusively for the INRatio2 PT/INR monitor from the systematic review.

Several scenario analyses were done by the External Assessment Group:

• exclusive self‑testing or self‑management compared with standard monitoring in primary and secondary care

• exclusive primary or secondary care clinic testing compared with self‑monitoring in primary and secondary care

• different pooled risk estimates applied.

For the exclusive self‑management strategy, the INRatio2 PT/INR monitor and the CoaguChek XS system dominated standard monitoring under the base‑case assumptions, whereas for the exclusive self‑testing strategy, the ICERs were above £2 million per QALY gained compared with standard monitoring. The results also showed that for a mixed self‑monitoring strategy (50% self‑testing, 50% self‑management), the CoaguChek XS system and the INRatio2 PT/INR monitor dominated standard monitoring when exclusively carried out in secondary care. When applying the pooled relative risk estimates for adverse events (derived from all self‑monitoring studies) to both self‑testing and self‑managing participants, the cost savings and QALY gains associated with self‑monitoring increased.

The External Assessment Group carried out alternative non‑base‑case scenarios, to assess the impact of using self‑monitoring to replace standard monitoring tests (that is, no increase in the number of tests done annually). It was assumed that there was no difference in clinical effectiveness between self‑management, self‑testing and standard care. Under most of these scenarios, standard monitoring was found to be less costly than self‑monitoring. However, self‑testing and self‑management with the INRatio2 PT/INR monitor and the CoaguChek XS system dominated standard monitoring when carried out exclusively in secondary care.

Subgroup analyses showed the cost effectiveness of self‑monitoring compared with standard care, stratified by indication (atrial fibrillation and artificial heart valves) and cohort age. Self‑monitoring in a '65 years old with atrial fibrillation' cohort was estimated to cost £2574 per QALY gained when using the INRatio2 PT/INR monitor and £4160 per QALY gained when using the CoaguChek XS system, compared with standard monitoring. For a '65 years old with artificial heart valve' cohort, self‑monitoring with the INRatio2 PT/INR monitor and the CoaguChek XS system was found to be more effective and less costly (dominant) compared with standard monitoring.

A further analysis was carried out for the atrial fibrillation cohort using the baseline risks seen for participants with better INR control in standard care, assuming a constant relative risk reduction for thromboembolic events associated with self‑monitoring. As the INR time in therapeutic range increased in the control group, and the baseline risk of thromboembolic events consequently dropped, the cost effectiveness of self‑monitoring also decreased. However, the ICERs for the CoaguChek XS system and the INRatio2 PT/INR monitor only rose above £20,000 per QALY gained when the baseline time in therapeutic range was set at greater than 72.6%.

Deterministic sensitivity analysis showed that the model‑based findings were most sensitive to the baseline risk of thromboembolic events and the effectiveness of self‑monitoring for preventing these events. The ICERs for the self‑monitoring strategies rose above £30,000 per QALY gained when the baseline risk was set to 1.15% and the upper confidence limit for the relative risk of thromboembolic events associated with self‑management (RR 0.69) was applied. The same was found when the lower baseline risk of thromboembolic events was coupled with the upper confidence limit of the pooled relative risk for self‑monitoring (RR 0.89). It should be noted however that self‑management on its own remained cost saving under the former combined scenario.

A sensitivity analysis was also conducted to approximate the cost effectiveness of self‑monitoring for a cohort of children with an artificial heart valve on long‑term vitamin K antagonist therapy. For this analysis, the cohort age was set to 10, the baseline risk of thromboembolic events was reduced to 1.4%, and the standardised mortality ratio for all‑cause mortality after a stroke was set at 14.5. Under this scenario, self‑monitoring with the CoaguChek XS system and the INRatio2 PT/INR monitor dominated standard monitoring. However, it should be noted that the standardised mortality ratio estimated for an 18–55 year old cohort of people with artificial heart valves was applied because no robust data were identified to appropriately adjust the risk of death from all causes in children with an artificial heart valve.

Probabilistic sensitivity analyses of the base case were done to examine the uncertainty in the cost effectiveness of self‑monitoring. Self‑monitoring with the CoaguChek XS system and the INRatio2 PT/INR monitor were estimated to have an 80% and 81% probability of being cost effective if the maximum acceptable ICER was £20,000 per QALY gained, respectively. However, it should be noted that there is no direct randomised controlled trial evidence to show the clinical effectiveness of the INRatio2 PT/INR monitor.