Detecting, managing and monitoring haemostasis: viscoelastometric point‑of‑care testing (ROTEM, TEG and Sonoclot systems)

The diagnostics advisory committee reviewed the evidence available on the cost effectiveness of viscoelastometric testing to help detect, manage and monitor haemostasis in cardiac surgery and in the emergency control of bleeding after trauma and during postpartum haemorrhage.

The committee considered whether the ROTEM, TEG and Sonoclot systems included in this assessment could be considered equivalent for the purpose of this assessment. It noted that each device has different measures and includes different assays. The committee also noted that although none of the 11 randomised controlled trials (RCTs) included in the systematic review provided a direct comparison between TEG and ROTEM, the summary estimates for all outcomes were similar when stratified by viscoelastometric device. There was also no evidence to indicate a difference in effectiveness between the 2 devices. The committee also considered the results of the 3 prediction studies, which did not suggest a significant difference in the ability of Sonoclot and TEG to predict bleeding. The committee noted that equivalent clinical effectiveness had been assumed in the modelling but concluded that the level of evidence available for the Sonoclot system was not sufficient for the device to be considered equivalent to the ROTEM and TEG systems. The committee noted that the ROTEM and TEG systems comprise different tests and although both systems can be used to detect, manage and monitor haemostasis, the data from each system are not interchangeable. The committee therefore concluded that, based on the available evidence, TEG and ROTEM could be considered equivalent to each other but more evidence was needed on the clinical effectiveness of the Sonoclot system.

The committee considered the clinical evidence on the use of viscoelastometric testing in managing postpartum haemorrhage. It noted that the review did not identify studies that compared clinical outcomes among women with postpartum haemorrhage who were tested with viscoelastometric devices and those who were not. It also noted that lower levels of evidence in the form of 2 prediction studies, available only as abstracts, were included in the clinical effectiveness review. The committee agreed with the external assessment group's judgement that the available data are insufficient for constructing an economic model to assess the cost effectiveness of viscoelastometric devices in this population. The committee considered whether the nature of bleeding in postpartum haemorrhage is identical to that in trauma. Clinical experts informed the committee that such an assumption should be interpreted with caution because of the difference in clotting mechanisms. The committee concluded that more evidence is needed on the use of viscoelastometric devices in the management of postpartum haemorrhage.

The committee discussed the clinical evidence on the use of viscoelastometric testing in trauma patients. It noted that the review identified 1 ongoing RCT and a clinical trial that did not report numerical or statistical outcome data. It also noted that because of insufficient data from studies that evaluated differences between viscoelastometric‑tested and untested populations, 15 prediction studies were included in the review. The committee noted that the external assessment group had to assume similar probabilities of red blood cell transfusion (relative risk: 0.88), transfusion‑related complications and transfusion‑transmitted infections in trauma surgery as in cardiac surgery because data on how clinical outcomes vary in patients tested with viscoelastometric compared with conventional testing were not available to inform the economic model for trauma patients. The committee deliberated whether these assumptions are reasonable. Clinical experts informed the committee that the nature of coagulopathy is different in the 2 populations and that patients having trauma surgery are likely to have higher blood loss; they therefore have greater blood transfusion needs than those having cardiac surgery. The committee concluded that it is not appropriate to assume similar probabilities of red blood cell transfusion, transfusion‑related complications and transfusion‑transmitted infections in trauma surgery as in cardiac surgery.

The committee discussed the clinical evidence on the use of viscoelastometric testing in cardiac surgery. It noted that pooled estimates derived from the meta‑analyses of 6 RCTs that evaluated the ROTEM and TEG systems in cardiac surgery indicated that viscoelastometric testing is associated with a significant reduction in the numbers of patients receiving red blood cell transfusion (relative risk: 0.88, 95% confidence interval [CI] 0.80 to 0.96). The committee concluded that there is sufficient evidence to demonstrate the clinical effectiveness of using the ROTEM and TEG systems in cardiac surgery.

The committee considered the lack of a formal assessment of publication bias in the systematic review. The external assessment group advised the committee that for RCTs, the number of studies was too small for such an assessment to be meaningful and that for prediction studies, there is no reliable method of assessing publication bias. The committee noted that the external assessment group's search strategy included a variety of routes to identify unpublished studies and resulted in the inclusion of a number of conference abstracts and the identification of 1 ongoing RCT.

The committee discussed the cost‑effectiveness modelling on the use of viscoelastometric testing in trauma patients. The committee considered the areas of uncertainty in the economic model produced by the external assessment group. It heard that the structure of the model may not be appropriate because the model is driven by a decrease in transfusion rather than outcomes and mortality. The committee therefore interpreted the results of the cost‑effectiveness analysis in trauma patients with caution, and concluded that more evidence is needed on the clinical effectiveness of viscoelastometric testing in the management of trauma.

The committee discussed the results of the base‑case analysis on the use of viscoelastometric testing in cardiac surgery. It noted that viscoelastometric testing dominated standard laboratory tests (that is, was more effective and less costly) producing more quality‑adjusted life years (QALYs); 0.0047 and costing less (‑£43 for ROTEM and ‑£79 for TEG). The committee also noted that results of other outputs from the model show that the use of viscoelastometric devices is associated with lower mortality, a reduced probability of experiencing complications, and less transfusion and hospitalisation. Based on the level of clinical evidence, the committee did not consider that the Sonoclot system was of equivalent clinical effectiveness to the ROTEM and TEG systems. The committee did not therefore consider the results of the cost‑effectiveness analysis for the Sonoclot system to be robust. The committee concluded that the use of the ROTEM and TEG systems to help detect, manage and monitor haemostasis in cardiac surgery is cost effective when compared with standard laboratory tests alone.

The committee discussed the implications of using risk of red blood cell transfusion as the main outcome in the economic model and considered whether other types of transfusion such as fresh frozen plasma and platelet transfusion should have been included in the model. The external assessment group informed the committee that it used these outcomes in the model because most patients receiving any transfusion receive red blood cell, and no data were available to inform the probabilities of complications from red blood cell, platelet or fresh frozen plasma transfusions individually, or any combination of these. In addition, the studies included in the review do not indicate what combination of blood products patients received, or in what percentage. The external assessment group stated that this approach is consistent with the only cost‑effectiveness study in the field, the Scottish Health Technology Assessment, and is also consistent with the study by Davies et al. (2006), on which the Scottish Health Technology Assessment was based. The committee noted that relative risk for mortality in patients receiving red blood cell transfusions compared with non‑transfused patients, obtained from data from a large cohort study in a UK setting (Murphy, 2007), was almost identical to the pooled estimate obtained from studies that reported short‑term mortality included in the clinical‑effectiveness review. The committee therefore agreed with the external assessment group's approach of using red blood cell transfusion as the main outcome in the economic model and concluded that including other types of transfusion would not significantly affect the results of the cost‑effectiveness analyses.

The committee was informed of a recently completed UK‑based multicentre RCT, named TITRE2, which assessed the impact of changing the haemoglobin level threshold at which red blood cell transfusion is performed. The external assessment group informed the committee of the results of the unpublished study which are considered academic in confidence by the authors and therefore cannot be reported here. The committee discussed the relevance of the results to this assessment and concluded that the impact of changing the haemoglobin level threshold represents a different clinical scenario to using viscoelastometric testing to guide transfusion decisions. The committee therefore concluded that the cost‑effectiveness estimates are unlikely to be affected by the results of the TITRE2 trial.

The committee discussed the external assessment group's decision to model the type of assays and number of tests for viscoelastometric testing based on the combination of assays and numbers of tests used in the trials so that the costs included in the model correspond to the source of the effectiveness data. The committee was informed that each device is available with different numbers of channels and runs different assays that are not directly comparable between devices. The committee considered whether the results found in the trials would also be applicable to different assay combinations and numbers of tests used in clinical practice. It noted that the results of the scenario analyses showed that varying the number of tests, which could also be a proxy for assay combinations, did not alter the conclusions in terms of cost effectiveness – that is, ROTEM and TEG continued to dominate standard laboratory tests. The committee concluded that when the combination of assays and numbers of tests is varied, viscoelastometric testing with the ROTEM or TEG system remains cost effective when compared with standard laboratory tests.

The committee considered the assumption in the base case that 500 tests would be run on each viscoelastometric device per year. It noted that, although the length of time for which a device is used and the average number of tests run per machine per year influences the material cost of a test, the number of tests had to be very low before viscoelastometric testing was no longer cost effective. The committee noted that in the scenario analyses the incremental cost‑effectiveness ratio (ICER) of viscoelastometric testing compared with standard laboratory testing is around £30,000 per QALY gained when the number of tests per year is reduced to 152 (from 500 tests per device per year in the base case). Clinical experts informed the committee that 600 to 800 tests could be run per device per year. The committee considered that it was likely that on average more than 152 tests would be run on each viscoelastometric device per year and therefore concluded that using viscoelastometric testing would be cost effective in routine practice when compared with standard laboratory tests.

The committee discussed the 1‑year time horizon used in the model. It heard from the external assessment group that because the ROTEM and TEG systems were shown to be both more effective and cheaper than standard laboratory tests at 1 year (with a probability of at least 0.68 of being cost effective), effectiveness would only increase and costs would be likely to decrease over a lifetime. The committee acknowledged that the expected increase in effectiveness is based on minimising long‑term complications such as stroke, which is likely to be avoided by fewer transfusions, and would also imply lower cost. The committee concluded that the 1‑year time horizon used in the model was appropriate.

The committee discussed the use of viscoelastometric testing as an add‑on, or as a replacement for, standard laboratory tests in cardiac surgery patients. It noted that 3 of the RCTs (2 of TEG and 1 of ROTEM) included in the systematic review reported the effectiveness of using viscoelastometric testing and standard laboratory tests together, compared with using standard laboratory tests alone. For all outcomes assessed, the results of these studies were similar regardless of whether viscoelastometric testing was used alone or in combination with standard laboratory tests. The committee concluded that performing standard laboratory tests in addition to viscoelastometric testing is unlikely to give further benefit over that provided by viscoelastometric testing alone. However, the committee heard from clinical experts that standard laboratory tests provide additional information such as prothrombin time and fibrinogen levels, which are useful when deciding on the dose of prothrombin complex concentrate or fibrinogen concentrate to be administered. The committee noted that the external assessment group had only used the viscoelastometric testing alone scenario in the economic model and discussed whether the additional cost of standard laboratory tests would affect the cost‑effectiveness results. It noted that additional scenario analyses carried out by the external assessment group showed that using viscoelastometric testing in addition to standard laboratory tests is cost effective when compared with standard laboratory tests alone. Viscoelastometric testing with the ROTEM system and standard laboratory tests resulted in an ICER of £7,487 per QALY gained compared with standard laboratory tests alone. Viscoelastometric testing with the TEG system and standard laboratory tests dominated standard laboratory tests alone. The committee concluded that viscoelastometric testing should be used in conjunction with standard laboratory tests in cardiac surgery.

The committee discussed whether viscoelastometric testing should be carried out intraoperatively, postoperatively or both during cardiac surgery. It noted that, although the timing of the viscoelastometric test varied across the 11 RCTs included in the review, the clinical effectiveness of viscoelastometric testing did not vary according to the time the viscoelastometric test was administered. The committee concluded that viscoelastometric testing should be used both intraoperatively and postoperatively.

The committee noted that the ROTEM and TEG systems need maintenance and quality control procedures to be in place to ensure the clinical effectiveness of the systems in use.