Positive D-dimer results in a private hospital emergency department : an audit of patient management and outcomes (posed study)
Rudolph, Adriana Josina
Background: The D-dimer test is commonly requested in an emergency department to confirm or exclude clinically suspected venous thrombosis. In this setting, the D-dimer is used in conjunction with the Wells clinical score and other investigation such as duplex doppler, V/Q scan or CT scan to make a definitive thrombosis diagnosis. In a typical South African private hospital setting, such as Flora Clinic, it has not yet been established as to how many of the requested D-dimers are positive, what the impact and contribution of this result is on patient management and outcome in an emergency department. In the era of diagnostic test cost cutting and evidence based practice, this knowledge is important in informing rational decision making by emergency practitioners and also as part of auditing and improving the implementation of diagnostic pathways. The aim of this study was to establish the prevalence of D-dimer positivity in an emergency department, to determine in how many patients venous thrombosis was part of the documented differential diagnosis and if patients with a positive D-dimer test were managed in accordance with published clinical management recommendations. Materials and methods: This study is a retrospective cross-sectional analysis. All D-dimer tests requested at Flora Clinic emergency department from 1 November 2008 to 31 May 2009 were reviewed. The percentage of positive results was calculated from the total D-dimer tests conducted by the laboratory, as requested from the emergency department during the defined study period. Clinical notes and records of those patients with positive results were de-anonymised and de-identified, using a data capture sheet. The clinical notes of the patients with a positive D-dimer were reviewed to see what impact the D-dimer result had in the management and outcome of these patients in the emergency department. Results: There were 365 D-dimer tests conducted in the Flora Clinic emergency department, within the study period. Out of these 145 patients had a positive D-dimer test, of which 144 met further inclusion criteria. The study population had a mean age of 62 years with 53% men in the study sample. Twenty two patients‟ patient records were missing. They still formed part of the intent-to-treat analysis, although the data were not available beyond demographic data collection. The prevalence of a positive D-dimer test, out of all D-dimer tests conducted, between 1 November 2008 and 30 May 2009 within the Flora Clinic emergency department was 39%. The diagnosis of venous thrombosis was considered in 25% of patient records. The higher the D-dimer level the more likely the emergency practitioner was to admit the patient to hospital. No association could be found between age or sex and D-dimer value. No patient record had a formal pre-test probability evaluation recorded, although some elements of the Wells Score were recorded in all the clinical notes. Among the special investigations requested in the emergency department, 11% were in accordance with published treatment protocol. Out of these 3% involved chest imaging. In 10% of the patient records the management initiated in the emergency department was according to recognised treatment protocol. Conclusion: The study results suggest that the assessment of pre-test probability is not done in a formalised manner and that positive D-dimer tests are either misinterpreted or that the emergency practitioners do not have sufficient knowledge regarding the correct management of a positive D-dimer test. The study identified the current trend of practice among emergency practitioners in the South African environment. Further studies evaluating the reasons for and methods to improve the under-utilisation of both clinical prediction rules and established treatment protocols will be useful.