Patient self‐testing and self‐management of oral anticoagulation with vitamin K antagonists: guidance from the British Committee for Standards in Haematology

The guideline Writing group was selected to include UK-based medical, scientific and laboratory representatives. Publications known to the writing group were supplemented with additional papers published between January 2005 and January 2014, i. e. , since the previous version of these guidelines (Fitzmaurice et al, 2005a), identified by searching MEDLINE/Pubmed using the keywords warfarin, Coumadin, self-testing and self-management. Publications known to the group included systematic searches undertaken by authors of this guideline, including a systematic review and meta-analysis of individual patient data analyses (Heneghan et al, 2012) which searched Ovid versions of Embase, Medline (1966–2009), Cochrane Central Register of Controlled Trials, the Cochrane Library and Cinahl. The Writing group produced the draught guideline, which was subsequently revised by consensus by members of the Haemostasis and Thrombosis Task Force of the British Committee for Standards in Haematology (BCSH). The guideline was then reviewed by a sounding board of approximately 50 UK haematologists, the BCSH and British Society for Haematology Committee and Anticoagulation in Practice and comments incorporated where appropriate. Criteria used to quote levels and grades of evidence are as outlined in: http: //www. bcshguidelines. com/BCSHPROCESS/EVIDENCELEVELSANDGRADESOFRECOMMENDATION/43GRADE. html. The objective of this guideline is to update previously published guidance (Fitzmaurice et al, 2005a) to provide healthcare professionals with information on the clinical effectiveness and health economics of patient self-testing and self-management of oral anticoagulant therapy. Patient self-testing (PST) refers to patients using a point-of-care testing (POCT) device to measure their own international normalized ratio (INR), but communicating the result to a healthcare professional for interpretation; patient self-management (PSM) refers to patients measuring their own INR, interpreting the result and making dosing decisions themselves. A Haematology audit template is available for this guideline (Appendix S1). Several devices are available for POCT by healthcare professionals (for a review see National Health Service NHS Purchasing and Supply Agency 2008). However, not all POCT devices are suitable for PST/PSM. Consideration for selection of a device for patient use should include: size; portability; ease of use; single analyte device; volume of capillary (fingerstick) blood required; test strips/cuvette storage; display and cost. Instruments evaluated for PST or PSM in clinical studies include the CoaguChek series of devices (Roche) (Fitzmaurice et al, 2005b; Braun et al, 2007), Protime (ITC) (Völler et al, 2007) and INRatio (Alere) (Völler et al, 2004). Guidance on the regulation of in vitro diagnostic devices for self-testing has been reported by the Medicines and Health Regulatory Agency (MHRA, 2012). Discussion should take place between patients and healthcare professionals overseeing their care regarding choice of device. Several point-of-care devices have been shown to give reliable and accurate INR values when compared to traditional laboratory methods (NHS) Purchasing and Supply Agency, 2008; Christensen McCahon et al, 2007). Evaluation of a POCT device will usually include an assessment of reproducibility (i. e. precision). Accuracy should also be determined; this presents a challenge given that the accepted model is comparison on a ‘split’ sample with a laboratory INR measurement. POCT and laboratory INRs may not agree for several reasons, and correlation is only expected for stable patients with INR results in the therapeutic range. For POCT devices to be considered acceptable in terms of precision and accuracy, performance should be comparable to that achievable with INR measurement systems in secondary care. Healthcare professionals accept quality assurance as an essential part of the laboratory process to ensure accuracy and reliability of a test. However, in the POCT setting, this issue is less widely understood, despite recommendations in national and international guidelines and reviews (Tripodi et al, 2001; Briggs et al, 2008; Perry et al, 2010). The introduction of the concept and importance of quality assurance should be part of the training process for PST/PSM. Quality assurance is composed of internal quality control (IQC) and external quality assessment (EQA), and both should be performed where possible. Internal quality control is a locally controlled process, whereby the user performs tests with the foreknowledge of a target result and acceptable limits. For POCT devices, there are three types of IQC available: electronic QC, on-board QC and liquid QC, and devices differ as to which of these they provide. Electronic IQC is used in place of a test strip, and confirms the optical/mechanical systems are operating correctly. If available, electronic IQC should be performed each time the device is used, but should not be a substitute for other forms of IQC. On-board IQC, in which a control process is integrated into individual test strips, can be useful to monitor test strip integrity but, where possible, liquid IQC material should also be used. Liquid IQC material should have an INR in the range 2·0–4·0, and repeat measurements should both be within the target range prescribed by the manufacturer and within 0·5 INR units of each other (especially if a wide target range is provided by the manufacturer). Tests should be performed: If the INR result of an IQC test falls outside the target, a second IQC sample should be tested. If this sample also falls outside the range, then PST should be suspended and the patient should contact the healthcare provider responsible for their anticoagulant care. For POC devices that do not have liquid QC available, the PST/PSM may require more frequent EQA checks to be in place. Some form of external assessment of INRs obtained with POCT devices is recommended (Fitzmaurice et al, 2005a; MHRA, 2013). For healthcare professionals, EQA is available and established for several POCT devices (Tripodi et al, 2004; Kitchen et al, 2005, 2012). The SMART (Self-management of anticoagulation: a randomized trial) study (McCahon et al, 2007) explored three different models of EQA for PST/PSM: (1) enrolment in a proficiency testing programme, with EQA material provided directly to the patient for testing; (2) comparison of fingerstick samples with a master POCT device, itself controlled through participation in a proficiency testing programme and (3) comparison of a fingerstick sample with a venous sample tested with a laboratory system. The authors concluded that patients were capable of performing EQA and the previous version of these guidelines promoted all three models as suitable (Fitzmaurice et al, 2005a). However, direct interaction between EQA provider and patient, without involvement of the patient's healthcare provider, is problematic and an EQA scheme for patients is not currently available. One of the following alternative models for EQA should be adopted when a patient is enrolled in a PST/PSM programme: There is no evidence available to support a defined frequency of EQA testing. We suggest testing every 6 months. If a patient INR of between 4·5 and 8·0 is obtained with a POCT device, this should be repeated to ensure that the prolonged result is not a consequence of poor sample quality. The second result should be within 0·5 INR of the first. If a result of >8 is confirmed with repeat testing patients should seek medical advice promptly, and a venous sample should be sent to the laboratory, as clinical decisions may be affected by the degree of INR prolongation (Keeling et al, 2011). Since publication of the previous guideline (Fitzmaurice et al, 2005a), several reviews have assessed the clinical and cost-effectiveness of PST and PSM. Bloomfield et al (2011) reported a meta-analysis of 22 trials, comprising 8413 patients; lower total mortality (odds ratio OR 0·74 0·63–0·87), lower risk for major thromboembolism (OR 0·58 0·45–0·75), and no increased risk of major bleeding (OR 0·89 0·75–1·05) was observed in the PST/PSM groups compared to usual care. In eight out of 11 trials, patient satisfaction, quality of life or both was reported to be better with PST/PSM than usual care. Relevant evidence of effectiveness was reported in a Cochrane systematic review of 18 randomized trials with data for 4723 participants (Garcia-Alamino et al, 2010) and an individual patient data meta-analysis including 11 trials with data for 6417 participants and 12 800 person-years of follow-up (Heneghan et al, 2012). Overall, this evidence indicates that, compared to standard monitoring, patients who self-test or self-manage show improved quality of their oral anticoagulation therapy. In the 2010 Cochrane review (Garcia-Alamino et al, 2010), pooled estimates of both PST and PSM strategies showed significant reductions in thromboembolic events, relative risk (RR) 0·50 (95% confidence interval CI 0·36–0·69) and all-cause mortality, RR 0·64 (0·46–0·89). In those patients who undertook PSM, significant reductions in thromboembolic events (RR 0·47, 0·31–0·70) and all-cause mortality, RR 0·55 (0·36–0·84) were reported, but not for those patients self-testing alone: thromboembolic events RR 0·57 (0·32–1·00), mortality RR 0·84 (0·50–1·41). PST significantly reduced major haemorrhage, RR 0·56 (0·35–0·91) whilst PSM did not, RR 1·12 (0·78–1·61). A subsequent, individual patient data analysis (Heneghan et al, 2012), reported that PST and PSM are safe options for suitable patients of all ages. In the combined groups (PST and PSM), a significant reduction in thromboembolic events was reported, hazard ratio HR 0·51 (0·31–0·85), but there was no effect on major haemorrhagic events (0·88, 0·74–1·06) or death (0·82, 0·62–1·09). Subgroup analysis in this study showed that there were significant reductions in thromboembolic events in younger patients (aged <55 years), HR 0·33 (0·17–0·66), which corresponded to a number needed to treat (NNT) of 21 (17–42) to prevent one thromboembolic event at 1 year. For participants with mechanical heart valves, significant reductions in thromboembolic events were also reported, HR 0·52 (0·35–0·77), which translates to a NNT to prevent one event at 1 year of 55 (41–116) and a NNT to prevent one event at 5 years of 24 (18–50). PSM participants had significantly fewer thromboembolic events whereas PST participants did not. Ninety-nine patients aged ≥85 years were included in the analysis, which showed no significant adverse effects of the intervention for all outcomes. The type of control care did not affect the overall effectiveness of self-monitoring, with little difference reported between anticoagulant clinic care and primary care for thromboembolic events, major haemorrhage and death. At 1 year, 2·7% and 5·1% improvements were reported for TTR for heart valve and atrial fibrillation (AF) patients, respectively, compared to standard care. At 1 year, participants with a mechanical valve performed an average of 24 more tests (18·4–30·0), and AF patients undertook an average of 21 more tests (13·1–30·4) than the usual care groups. The evidence for PST and PSM demonstrates improved quality of oral anticoagulation therapy compared to standard monitoring (Garcia-Alamino et al, 2010; Heneghan et al, 2012). The current costs of suboptimal oral anticoagulation should not be underestimated and are substantial. A US economic model reported that if 50% of those currently receiving warfarin had optimal anticoagulation, 9852 emboli would be prevented at a saving of US 1·3 billion (Caro, 2004). Yet, despite the growing evidence of effectiveness for PST/PSM, it is often not a funded option across different healthcare systems. This partly reflects concerns over the costs of metres and test strips, which together may prevent wide-scale uptake (Shah et al, 2013). Over the last decade a number of economic assessments have been undertaken across a wide variety of settings. Connock et al (2007) carried out a Health Technology Assessment (HTA) on the clinical effectiveness and cost-effectiveness of different models of managing long-term oral anticoagulant therapy. From seven cost-effectiveness studies identified, only one (Jowett et al, 2006) provided relevant UK data, reporting that PSM was more expensive (£417 pa) than current routine care (£122 pa). De novo modelling for this study reported that the incremental cost per quality-adjusted life year (QALY) gained by patient self-monitoring was £122 365 over 5 years and £63 655 over 10 years. It was also reported that wide adoption of patient self-monitoring of anticoagulation therapy would cost the NHS an estimated additional £8–14 million per year. The HTA report concluded that, ‘for selected and successfully trained patients, self-monitoring is effective and safe for long-term oral anticoagulation therapy. In general, patient self-management (PSM) is unlikely to be more cost-effective than the current specialized anticoagulation clinics in the UK. ' More recently, a review by Plüddemann et al (2012) concluded that there is good evidence for the clinical effectiveness of PSM, significantly reducing the rate of thromboembolic events, though not the risk of major haemorrhage or death. The authors suggest that PSM of INR may be cost-effective based on recent estimates. The overall picture therefore differs depending on which studies are looked at. There is no doubt that utilization of resources is relevant when considering whether to recommend widespread use of PST and PSM. Higher costs are driven mainly by the high cost of test strips and increased testing frequency (Holbrook et al, 2012). Effectiveness compared to usual care may also depend on the quality of usual care, which has been shown to vary geographically (Wallentin et al, 2010). Some models fail to assess PST and PSM separately despite evidence of increased clinical effectiveness in the latter group (Healthcare Improvement Scotland, 2013). Most models, however, do not take into account the increased convenience and the potential for lower personal costs for individual patients, particularly if the individuals live remotely. Manufacturers should therefore focus on increasing the affordability of INR testing devices and test strips for patients, which would substantially affect any subsequent evaluations. There is no consensus on which patients should be selected for PST or PSM. However, the success of the training programme in the SMART study (Fitzmaurice et al, 2002) may be considered as a basis on which patient training programmes should be developed. Patient selection should be based on an indication of more than 12 months anticoagulation (Fitzmaurice et al, 2002). Willingness to take part in a training programme by the patient (or for dependant patients, both the patient and their carer) is important. Failure of patients to successfully complete a training programme and demonstrate competency has been reported to be related to age, history of cerebrovascular accident (CVA), poor cognition, low literacy or poor manual dexterity (Fitzmaurice et al, 2002; Matchar et al, 2005). UK studies have demonstrated that training programmes can be successfully completed by 75% of patients (Fitzmaurice et al, 2002). Dolor et al (2010) evaluated ability to PST and demonstrated competency in 80% of subjects. However, in other studies over 50% of patients were excluded because of physical limitations, inability to demonstrate competence with POCT devices, apprehension about self-care, or patient refusal (Bloomfield et al, 2011). The American College of Chest Physicians (ACCP) guidelines (Holbrook et al, 2012) recommend PSM for patients who are motivated and can demonstrate competency in self-management strategies. Therefore the major criteria in the decision on PSM/PST are the motivation of the patient and the ability to complete and demonstrate competence in a valid self-testing/self-management training programme. The patient training programme in the SMART study (Fitzmaurice et al, 2002) was based on the programme established by the Association of Self-Management of Anticoagulation (Arbeitsgemeinschaft Selbstkontrolle der Antikoagulation e. V. ASA) in Germany (Morsdorf et al, 1999), where more than 400 000 patients manage their anticoagulation therapy and where a nationally approved, formalized, training programme is in place. In the SMART study, patients randomized to PSM attended at least two training sessions, which were practice-based. The aims of the training were to ensure theoretical understanding of oral anticoagulation and INR monitoring; that patients (or carers) were able to measure INR reliably using a POCT system; and were able to interpret the INR in terms of appropriate warfarin dose. Competency was assessed, and patients not meeting the assessment criteria were offered an additional training session. Those not considered capable of PST/PSM were then asked to return to their usual care. Patient training programmes should include the following: The practical should include: will be for patients where self-management is adopted in to Competency using a dosing be One was reported to TTR in a anticoagulation clinic et al, 2010), and to be as effective as dosing et al, 2012), but did not the TTR by primary care et al, There concerns and about the standard of training provided to the patient outside as a competence to care. It is that although there is patient in there are about PSM in some (Shah et al, 2013). review is on an individual of testing should be by the responsible It is recommended that the patient is at least every 6 months. INR results and quality control results and any should be used widely by anticoagulant clinics can be used by the patient to INR results and of In a the of the Society of Thrombosis and et al recommend that PST over PSM be considered for receiving therapy The authors reported a TTR of to in patients who were the number of receiving warfarin is INR monitoring is in This is mainly for and with heart who require long-term The of self-testing include a reduction in time to anticoagulant the ability to to or in and improved quality of It is when care is to that this anticoagulant The process should ensure that are in the use of warfarin, and the effects of and that they are with testing and have to an of support in terms of dosing advice and QC their anticoagulant that by directly a are available. There is good evidence that for patients with AF or venous thromboembolism these are to warfarin et al, et al, et al, 2010; et al, et al, et al, et al, and these have National for Health and for in AF and for the and secondary of are not suitable for patients with mechanical heart oral target or In the clinical trials these with warfarin, the were given in with no monitoring and no For patients for the issue with warfarin is to have blood to a is an alternative to PST on be better on warfarin using PST/PSM rather than on a on long-term warfarin for secondary of or AF may PST/PSM over to a will be a better choice for patients with significant or One of warfarin is the ability to the for a patient who has had a on therapeutic anticoagulation can have their anticoagulation by increasing their target INR to range for the are in they are not available at and some patients may warfarin they there will be those patients who a that has been for given that long-term adverse events for be the cost of the is than the cost of warfarin usual UK monitoring when and for in AF and for the and secondary of they all to be cost effective less than 000 per as compared to This of on the cost of warfarin At the time of the of the analyses for in estimated the cost of warfarin in the UK to be PST/PSM is more expensive than routine care the though more expensive than PST/PSM, are to be cost-effective as compared to PST/PSM in the the writing group on this and the on available and Quality Kitchen the the on available and Quality Heneghan the on and cost the on selection and the on and the authors to the and to of all of these The is not responsible for the or of any information by the than should be to the for the