Table 8
Author, date and countryPatient groupStudy type (level of evidence)OutcomesKey resultsStudy weaknesses
Legani C et al, Italy 1999 105 consecutive outpatients ?DVTProspective cohortROC curve to define max sensitivity of IL test230 ng/ml cut offPatients ?DVT not PE
sensitivity 100%High prevalence DVT. Would not reflect an average emergency department population
specificity 77.2%
van der Graaf F, 200099 ?DVT outpatients. Comparison of 13 different D-dimer testsProspective cohortIL testPatients ?DVT not PE
Sensitivity90%
specificity78%High prevalence of DVT
Villa P et al, Spain 2000 86 patients with a moderate or high clinical suspicion DVTProspective cohortSensitivity98.4%Patients suspected of having DVT not PE
Specificity33.3%
Iltest using 255 ng/ml cut offCohort had high prevalence DVT
Venography not used
Sensitivity95.2%
Specificity41.7%
Iltest using 292 ng/ml cut off
Sensitivity90%
Specificity57.1%
Gold EIA ELISA test
Harper P et al, New Zealand 2001 235 patients presenting to emergency department with ?DVTProspective cohortSensitivityAll patients presented as ?DVT not PE
IL test (250 ng/ml cut off) and94.1%
SimpliRED66%The gold standard venogram not used in diagnosis DVT
SpecificityAll patients underwent ultrasound (USS), but not all underwent more than one. Probable under-estimation of DVT prevalence
IL test and51.5%
SimpliRED75.6%
Patients were simply advised to return to the department if symptoms did not settle
Kovacs MJ et al, Canada 2001 All patients with suspected DVT (468 patients) or PE (525 patients), presenting to four hospitalsProspective cohortSensitivity ofResults combined for ?DVT and ?PE patients
SimpliRED,80%
IL test and91%Cut off level of 200 ng/ml was used for IL test (much lower than most labs)
Accuclot91%
Specificity of SimpliRED, IL test and Accuclot
79%
74%
76%