Destiny: D-dimer. Investigating D-dimer Levels Using Point-of-Care (POC) Testing at Primary Care. (Destiny)

The diagnosis of conditions that cause hypercoagulability result in economic burden for the healthcare industry. The non-specific symptoms incur clinical problems for diagnosing conditions such as Deep Vein Thrombosis (DVT). DVT is a common condition in the UK, with more than 140,000 patients presenting to primary care with suggestive symptoms each year (Thrombosis UK, 2020). The reduction of blood flow caused by a blood clot in a deep set vein can lead to a pulmonary embolism and infarcts throughout the circulation, which is potentially fatal if not treated quickly.

The purpose of this study would therefore be to safely exclude DVT in a primary care setting using Point of Care (POC) testing as a means of reducing the number of referrals to secondary and tertiary care. The use of a small protein fragment as a biomarker utilising POC testing will be investigated to determine the clinical effectiveness and accuracy of ruling out DVT in primary care. The current practise however relies on referral from a primary care setting of all patients with symptoms of hyper coagulation, for diagnostic testing with ultrasonography and laboratory generated D-Dimer analysis at secondary care. Although ultrasonography is safe and easily available, it requires an additional patient visit and increases the need for clinical and diagnostic resources. The use of a D-dimer POC test could therefore be used to exclude a DVT at the initial presentation in the primary setting and reduce the economic burden on the NHS and be more convenient for the patient.

Venous Thromboembolism (VTE) encompasses the conditions deep vein thrombosis (DVT) and pulmonary embolism (PE). VTE is the third most frequent cardiovascular disease globally and can be lethal in the acute phase and potentially lead to a chronic disease and disability. The potential to prevent a chronic disease from developing however is highly possible with the use of a rapid and reliable diagnosis. Only around 15% of patients with suspected DVT will have a confirmed ultrasonography scan, with an annual incidence of 1 per 1000 (Wells, Anderson and Rodger, 2004). Using ultrasonography to diagnose DVT is reliable, although currently the only recognised method post initial assessment. A more precise diagnosis at primary care can reduce a large percentage of ultrasound referrals to secondary care with an advantage to both patients, clinical staff and efficiencies within the health care service, keeping patients and their care in the primary care setting.

D-dimer is a small protein fragment that is the result of endogenous fibrinolysis and can be found in healthy individuals in negligible amounts of around 100-200ng/ml (Lippi, Bonfanti, Saccenti and Cervellin, 2014). D -dimer can be detected using monoclonal antibody immunoassay based lab methods as the current method in practice. Current problems faced using D-dimer as a diagnostic biomarker are that it is highly non-specific. D-dimer levels may rise in conjunction with many other conditions such as cellulitis, inflammation, liver conditions and in pregnancy (Wells, Anderson and Rodger, 2004). D-dimer therefore cannot solely be used to diagnose a venous thromboembolism (VTE), further imaging testing (ultrasonography) would likely be needed to confirm or exclude. However, a negative test is useful as it is allows clinicians to reliably exclude VTE on the basis that clinical probability is low (Lippi, Bonfanti, Saccenti and Cervellin, 2014). Clinical probability has been generated for DVT using a points scoring 'decision rule' to categorise risk variables in order to reduce referrals to secondary care for a D-dimer laboratory test.

The two main points scoring rule systems currently in place are the Wells Rule and Primary Care Rule. The Wells Rule combines patient history and a physical examination which is followed by a D-dimer assay to determine patients for ultrasonography. Clinicians are able to risk stratify patients of a low risk score (Wells Score <2) and a negative D-dimer to be excluded from ultrasonography (van der Velde et al., 2011). A 2006 study concluded that patients with a low clinical probability have prevalence of DVT >5%, suggesting the Wells Rule is capable of categorising patients with low, moderate and high probability (Plüddemann et al., 2012). However, the Wells Rule has been criticised for its inaccuracies with prevalence of thrombosis in as many as 2.9% of patients in primary care and with a normal D-dimer (Antovic et al., 2012).

Uncertainties surrounding the Wells Score Rule led to the development of the Primary Care Rule. The only difference from the Wells Scoring Rule, being that it does not include the estimated probability of an alternative diagnosis. Analysis of data from a study of 1086 patients using both scorings systems in conjunction with a D-dimer POC test revealed that a thromboembolic event occurred in follow up of a negative D-dimer and low risk score in both Wells Rule (1.6%) and Primary Care Rule (1.4%) (Plüddemann et al., 2012). The results of this study suggests a thromboembolic event cold be ruled out safely with either decision rule in combination with D-dimer POC test at rates comparable to ultrasonography. The Primary Care Rule has also been praised as a robust rule as adding external variables or adjusting scoring does not improve the efficiency or safety of the rule (van der Velde et al., 2011). Using decision rules in combination with D-dimer POC testing allows a reduction of referrals by approximately 50%, significantly reducing the cost of further appointments (Plüddemann et al., 2012).

Age, among other factors has a significant effect on the safety and efficiency of diagnosis. The incidence of a Venous Thromboembolism (VTE) increases with age, with 450-900 cases per 100,000 people in ages over 80 in comparison to less than 5 amongst children (Schouten et al., 2012). The use of age dependant cut-off values for a D-dimer test can improve the efficiency of excluding a venous thromboembolism. Using higher cut-off values for those in different age groups respectively, clinicians are able to accurately refer less patients to secondary care. However, by increasing the cut-off value for elderly patients develops a risk of a higher percentage of false negatives. A study conducted using age dependant cut off values reported an increase of 15.3% of patients in the age bracket of 70-80 that were excluded from referral, with no increase in false positives (Schouten et al., 2012). The reliability of such finding should be scrutinised as the accuracy of the diagnosis directly correlates with the accuracy of the reader.

The use of the Roche Cobas H 232 POC D-dimer test has been supported by a significant amount of clinical data in the evaluation of D-dimer whole blood levels. The Cobas H 232 test makes use of a venous blood sample to provide a quantitative test result in 12 minutes or less (Dempfle et al., 2006). The Cobas H 232 test uses a handheld device, which provides clinicians with immediate results and quick patient throughput. A comparative study concluded that using the Cobas H 232 test reduced the use of ultrasonography by up to 29% (Schutgens et al., 2003). Statistically the Cobas H 232 test is comparable to the laboratory immunoassay D-dimer test. The POC D-dimer assay displayed a high sensitivity of 96.9% for the diagnosis of DVT and specificity of 60.8% at a pre-specified cut-off of 500ng/ml. In comparison, the laboratory D-dimer is slightly lower, with a sensitivity of 94.9% and specificity of 64.8% (Dempfle et al., 2006).While using a handheld POC test prevents the transport of blood specimens to secondary care, a venous blood sample still needs to be obtained. This may be inconvenient for the patient and may incur unnecessary time commitments of both the patient and clinicians.

The Lumira Dx D-dimer test is a new POC test that introduces the opportunity to measure D-dimer levels in capillary blood, providing quantitative results. The lumira Dx kit boast the capability of producing a quantitative results in under 10 minutes (Gray, 2020). A fingerpick sample would reduce the time taken for a venous blood sample be more convenient for patients and clinicians. A statistical evaluation of both POC tests will be made in this study and will also investigate the clinical significance of capillary versus venous blood samples.

The emergence of the recent COVID -19 pandemic has led clinicians to investigate D-dimer as a biomarker of the disease. Initial data from Wuhan suggests that higher mortality can be associated with a high D-dimer test result (Reyes Gil, 2020). Association of high D-dimer and COVID-19 may however be non-specific. Reliability of the relationship is highly scrutinised due to the low specificity of D-dimer as a biomarker and relationship with other diseases. Confounding laboratory issues may also cause elevated results, such as hemolysis and lipid content which would present analytical interference. However, recent studies have suggested that biomarkers such as low lymphocyte count and high levels of CRP, LDH, and D-Dimer are associated with the more severe COVID-19 cases (Ali et al., 2020). Further clinical evaluation of the relationship between elevated D-dimer levels and COVID-19 is required to confidently make a diagnosis of COVID-19, although biomarkers such as D-dimer could currently be used as a subsidiary to COVID-19 diagnosis. The research does however highlight the importance of scrutinising the specificity of D-dimer as a biomarker and signifies the importance of developing clinically relevant cut off values as a result.

The non-specific nature of D-dimer level elevation highlights the need for thorough clinical examinations when investigating DVT's. D-dimer levels may rise in conjunction with other conditions that a patient may present upon initial assessment at a GP Practice. An inflammatory response is triggered in a range of diseases such as the digestive condition IBS or infections such as tuberculosis. In a recent study of 205 patients utilising D-dimer testing, 9 patients were diagnosed with pancreatitis and 14 patients with cirrhosis. The study concludes that D-dimer levels directly correlate with CRP levels, WBC count and C reactive protein levels, and are inversely proportional to lymphocyte count (Bao et al., 2017). Retrospective studies such as this highlight the need for full clinical examination for patients suspected of DVT. It is therefore important that the precision of the DVT pathway ensures no false positive or negatives and that ultrasonography is used to confirm DVT in cases of elevated D-dimer.