Article type
Year
Abstract
Background: A recent meta-analysis found that patient self-monitoring (PSM) of oral anticoagulation improved quality of oral anticoagulation, and reduced the risk of thromboembolic events and deaths. However, our independent meta-analysis on this topic provided a different interpretation.
Methods: MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched and references of
retrieved articles examined. Meta-analyses were conducted by the random-effects model using risk difference as the outcome
statistic. Heterogeneity, potential bias and generalisability of findings of individual trials were investigated by narrative description and subgroup analyses.
Results: PSM was better than poor quality controls but was as effective as good quality specialized anticoagulation clinics in maintaining the quality of anticoagulation therapy. Meta-analyses found no significant difference in the risk of major bleeding events (-0.002, 95% CI -0.011 to 0.007) and deaths (-0.005, 95% CI -0.017 to 0.008) between PSM and usual care controls. Pooling of results from trials found that PSM was associated with significantly fewer thromboembolic events (-0.014, 95% CI -0.026 to -0.003). However, trials conducted in the UK failed to confirm findings of trials in other countries about improved anticoagulation control and the reduction in major complications and deaths by PSM. Systematic errors could not be excluded from many trials and findings from trials in other countries may not be generalisable to the UK setting.
Conclusions: Research evidence from randomized trials may be interpreted differently in meta-analysis partly because of additional data identified, different outcome statistics, assumptions about missing data, heterogeneity and generalisability investigations. Patient self-monitoring of oral anticoagulation is as effective as high quality specialized anticoagulation clinics.
Methods: MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched and references of
retrieved articles examined. Meta-analyses were conducted by the random-effects model using risk difference as the outcome
statistic. Heterogeneity, potential bias and generalisability of findings of individual trials were investigated by narrative description and subgroup analyses.
Results: PSM was better than poor quality controls but was as effective as good quality specialized anticoagulation clinics in maintaining the quality of anticoagulation therapy. Meta-analyses found no significant difference in the risk of major bleeding events (-0.002, 95% CI -0.011 to 0.007) and deaths (-0.005, 95% CI -0.017 to 0.008) between PSM and usual care controls. Pooling of results from trials found that PSM was associated with significantly fewer thromboembolic events (-0.014, 95% CI -0.026 to -0.003). However, trials conducted in the UK failed to confirm findings of trials in other countries about improved anticoagulation control and the reduction in major complications and deaths by PSM. Systematic errors could not be excluded from many trials and findings from trials in other countries may not be generalisable to the UK setting.
Conclusions: Research evidence from randomized trials may be interpreted differently in meta-analysis partly because of additional data identified, different outcome statistics, assumptions about missing data, heterogeneity and generalisability investigations. Patient self-monitoring of oral anticoagulation is as effective as high quality specialized anticoagulation clinics.