Identification of and solution for false D‐dimer results

Abstract Background Clinically, D‐dimer (DD) levels are mainly used to exclude diseases such as deep venous thrombosis (DVT). In clinical testing, DD assays can be subjected to interference that may cause false results, which directly affect the clinical diagnosis. Our hypothesis was that the 95% confidence intervals (CIs) of the fibrin degradation product (FDP)/DD and fibrinogen (Fib)/DD ratios were used to identify these false results and corrected via multiple dilutions. Methods In total, 16 776 samples were divided into three groups according to the DD levels detected by Sysmex CS5100 and CA7000: Group A, DD ≥ 2.0 μg/mL fibrinogen equivalent unit (FEU); group B, 0.5 < DD < 2.0 μg/mL FEU; and group C, DD ≤ 0.5 μg/mL FEU. The 95% CIs of the FDP/DD and Fib/DD ratios were calculated. Six abnormal DD results were found according to the 95% CIs. For verification, we performed multiple dilutions, compared the results with those of other instruments, and tested the addition of heterophilic blocking reagent (HBR). Results The median and 95% CI of the FDP/DD ratio were 3.76 and 2.25‐8.15 in group A, 5.63 and 2.86‐10.58 in group B, 10.23 and 0.91‐47.71 in groups C, respectively. For the Fib/DD ratio, the 95% CIs was 0.02‐2.21 in group A, 0.68‐8.15 in group B, and 3.82‐55.27 in groups C. Six abnormal results were identified after multiple dilutions, by comparison with other detection systems, and after HBR addition. Conclusions The FDP/DD ratio is more reliable for identifying false results. If the FDP/DD ratio falls outside the 95% CI, it should be verified by different methods.

time, the fibrinolytic system is activated, and plasmin cleaves the substrate fibrin at a specific site, forming D-dimer. 2,3 Fibrin degradation product (FDP) is the general term for the degradation products produced after the decomposition of fibrin or fibrinogen under the action of plasmin during hyperfibrinolysis. 4 FDPs include fibrinogen degradation products (FgDPs) and cross-linked fibrin degradation products (FbDPs). 5,6 The former are the products of fibrinogen (Fib) and fibrin monomer (FM), while the latter include the products of D-dimer and other fragments. 6 An elevated level of FDP indicates hyperfibrinolysis activity, including primary hyperfibrinolysis and secondary hyperfibrinolysis.
Increased D-dimer formation indicates the presence of thrombosis and secondary hyperfibrinolysis in the body, such as disseminated intravascular coagulation (DIC). 7 Because D-dimer is highly sensitive and has a high negative predictive value, its measurement is used to exclude pulmonary embolism (PE), venous thromboembolism (VTE), and deep venous thrombosis (DVT). 8,9 With the development of modern medicine, many different D-dimer analysis methods have been developed. Generally, these assays use monoclonal antibodies to detect epitopes that are present in the factor XIIIa-cross-linked fragment D domain of fibrin, including methods based on fluorescence, hemagglutination, chemiluminescence, or other techniques. 2,10,11 The detection methods for D-dimer in the clinical laboratory mainly include immunoturbidimetry, enzyme immunoassays, and immunochromatography, with the most widely used being immunoturbidimetry. Although immunoassays are commonly used in clinical laboratories, laboratory workers often obtain immunoassay results that are inconsistent with the clinical symptoms. The reason may interference in the immunoassay, including high-dose hook effects and the presence of heterophilic antibodies, autoantibodies, and cross-reactive substances. 12  to identify these false results, which will be beneficial to correct these by multiple dilutions.  Institute (CLSI). 13 According to the 95% CIs, some abnormal results were identified, and the samples were then diluted multiple times.

| Abnormal plasma sample results
Patient I was a woman diagnosed with placental abruption, and patients II and III were both male and were diagnosed with thrombocytopenia and leukocytosis. After coagulation testing, D-dimer levels were found to be slightly increased in the plasma samples of these three patients, while FDP was also significantly increased, and the degree of increase in both parameters were abnormal ( Table 2).

TA B L E 1
The D-dimer level and the 95% CI of the FDP/DD ratio and the Fib/DD ratio in each group Note: Compared with group A, a P < .05; compared with group B, b P < .05; compared with group C, c P < .05.
Patients IV, V, and VI were all elderly women diagnosed with renal insufficiency, anemic dizziness, and asthma, respectively.
After the detection of coagulation, D-dimer was abnormally increased in the plasma samples from these three patients, while the FDP was normal or only slightly increased (Table 2).

| Dilution test
The abnormal plasma samples were subjected to serial dilutions (

| Comparative testing
The plasma samples of patients IV, V, and VI were simultaneously tested with a different instrument using a latex-enhanced immunoturbidimetric immunoassay (INNOVANCE D-dimer assay using a SYSMEX CS-5100 and HemosIL D-dimer HS assay using an ACL TOP700) to determine the D-dimer value.

| Heterophilic antibody blocking reagent (HBR)
The samples of patients IV, V, and VI were treated with a heterophilic blocking reagent (HBR, Scantibodies Laboratory Inc) following the manufacturer's instructions. Briefly, the HBR was immediately thawed in a water bath, gently shaken in an upright position and then added directly to the experimental sample. After incubation at the temperature indicated in the instructions, a D-dimer assay was performed. The control samples were identical to the test samples except that no HBR was added.

| RE SULTS
Since the FDP/DD ratio and the Fib/DD ratio did not conform to  Figure 1A, Table 1). Figure S1 shows the distribution as a scatter diagram (described in the Supplementary Files). By calculating the Fib/DD ratio, we could also calculate the 95% CI based on the 2.5 and 97.5 quantiles ( Figure 1B, Table 1). As shown in  (Figure 2A, Table 2).
Moreover, the FDP values of the three patients also increased after appropriate dilution ( Figure 2B, Table 2). In contrast, the initial  Figure 2C, Table 2). The Fib/DD ratios are similar, as shown in the Table 2. For specimens with increased pseudomorphic D-dimer levels, we used different instruments to perform comparative experiments. The plasma samples of patients IV to VI measured by latex-enhanced immunoturbidimetric immunoassay were within the normal reference range (Table 3). We also detected D-dimer levels after adding HBR to eliminate interference. The results showed that the D-dimer levels of the three samples were significantly decreased after HBR addition ( Figure 2D).

| D ISCUSS I ON
As a specific degradation product of cross-linked fibrin, D-dimer is a specific indicator of thrombosis and secondary fibrinolysis. If the determined D-dimer level exceeds the recommended cutoff value of the kit, it cannot be used as the only criterion for the diagnosis of acute PE, DVT, and DIC; rather, it must be comprehensively analyzed in combination with the clinical conditions. 14 The reason is that as long as there is activated thrombosis, namely, active fibrinolysis, in the blood vessels of the body, the D-dimer levels will be increased. results, we could not preliminarily judge whether the results were reliable according to the FDP/DD ratio. Therefore, we also calculated the Fib/DD ratio to determine whether this ratio could be used instead of the FDP/DD ratio as a reference. However, we found that for some abnormal results, the Fib/DD ratio did not exceed the 95% CI range, indicating that the Fib/DD ratio cannot be used. Therefore, to identify false D-dimer results as early as possible, one needs to detect D-dimer and FDP together. During the statistical analysis, we noticed that most of the abnormal results came from the ICU and the clinical departments of hematology, rheumatism, neurology, and infectious diseases.
Therefore, we selected several samples for specific analysis to determine whether the results were reliable and how to identify false-positive or false-negative results. In patients I to III, the initial D-dimer levels were significantly lower than the FDP results, and the results after dilution were increased by ten-fold or more. In these patients, the D-dimer test produced false-negative results. Similar cases have been reported in the relevant literature. 20 Because the principle of D-dimer detection relies on antibody recognition, false negatives easily arise because of inappropriate proportions of antigens and antibodies, termed the hook effect, which is mainly due to antigen overabundance. A solution to this problem is to dilute the plasma sample and retest at multiple dilutions. As shown in Table 2 (Table 3).
However, the cause of this discrepancy remains unclear. 21  judge the reliability of the results and whether the elimination of additional sources of interference is needed.
Taken together, when significant abnormalities in the FDP/DD ratio are encountered, the results should be verified to eliminate sources of unnecessary interference with clinical symptoms. If necessary, some corresponding problems can be solved by eliminating interference by heterophilic antibodies. We also suggest that each laboratory establishes its own reference ranges for the FDP/DD ratio so that it could be used to monitor whether the D-dimer level is a false result. However, when laboratory conditions are limited regarding the determination of false-positive or false-negative results, the simplest and most straightforward solution is to serially dilute the samples.

ACK N OWLED G M ENTS
This study was supported by the National Natural Science Foundation of China [grant numbers 81302540].

CO N FLI C T O F I NTE R E S T
No conflict of interest exists in the submission of this manuscript, and the manuscript has been approved by all authors for publication.

AUTH O R S ' CO NTR I B UTI O N S
Xianyan Zhang and Xuexuan Zhang participated in writing the article and performing the statistical analysis of the experimental data.
Jialong Xu and Tengyi Huang are mainly responsible for data collection. Ying Wu was responsible for the heterophilic antibody experiment. Yeru Yang and Huanbin Zhou were responsible for the dilution test and the comparative test. Yinge Wu was responsible for the design of the experiment and the final assembly of the article.