Volume 122, Issue 11 , Pages 1050-1053, November 2009
D-dimer Testing in Patients with Suspected Pulmonary Embolism and Impaired Renal Function
Article Outline
Abstract
Background
Determination of pretest probability and D-dimer tests are the first diagnostic steps in patients with suspected pulmonary embolism, which can be ruled out when clinical probability is unlikely and D-dimer level is normal. We evaluated the utility of D-dimer testing in patients with impaired renal function.
Methods
D-dimer tests were performed in consecutive patients with suspected pulmonary embolism and an unlikely clinical probability. Creatinine levels were assessed as clinical routine. Glomerular filtration rate was calculated using the Modification of Diet in Renal Disease formula. Correlation between D-dimer level and renal function and proportions of patients with normal D-dimer in different categories of estimated glomerular filtration rate (eGFR) were assessed. Different categories of decreasing eGFR were defined as: normal renal function (eGFR >89 mL/min), mild decrease in eGFR (eGFR 60-89 mL/min), and moderate decrease in eGFR (eGFR 30-59 mL/min).
Results
Creatinine levels were assessed in 351 of 385 patients (91%). D-dimer levels significantly increased in 3 categories of decreasing eGFR (P = .027 and P = .021 for moderate renal impairment compared with mild renal impairment and normal renal function, respectively). Normal D-dimer levels were found in 58% of patients with eGFR >89 mL/min, in 54% with eGFR 60-89 mL/min, and in 28% with eGFR 30-59 mL/min.
Conclusions
The specificity of D-dimer testing in patients with suspected pulmonary embolism and decreased GFR is significantly decreased. Nonetheless, performing D-dimer tests is still useful because computed tomography scanning can be withheld in a significant proportion of these patients.
Keywords: Creatinine, D-dimer, Glomerular filtration rate, Pulmonary embolism, Renal function
The diagnostic workup of patients with clinically suspected acute pulmonary embolism has been greatly simplified. Two simple bedside tests can safely exclude pulmonary embolism from the differential diagnosis in 20%-40% of the patients, without the need for diagnostic radiological imaging, reducing costs, time, and exposure to radiation and contrast dye.1, 2 These tests include the determination of the patients' pretest probability for pulmonary embolism by means of a clinical prediction rule and a blood test measuring D-dimer concentration.1, 2, 3, 4 Pulmonary embolism can be ruled out in patients with an unlikely clinical probability (pretest probability less than approximately 15%-20%) in combination with a normal D-dimer test result and radiological imaging, and anticoagulant treatment can safely be withheld.1, 2, 3, 4 In patients with either an abnormal D-dimer test result or a likely clinical probability, the diagnosis should be established or excluded by computed tomography (CT) or ventilation/perfusion lung scanning.1, 2
D-dimer levels can be elevated in all clinical conditions that are associated with enhanced fibrin formation, including thrombosis. As a consequence, specificity of D-dimer testing for venous thromboembolism is, at most, 50%-70%. Certain patient categories, including elderly patients or patients with inflammation or malignancy,5 are very likely to have elevated serum D-dimer concentrations and, in the setting of suspected pulmonary embolism, the chances on a normal D-dimer test result will be significantly reduced. The percentage of normal D-dimer results, that is, the percentage of patients in whom pulmonary embolism can be ruled out without imaging, is the clinical utility of this test. If this utility is suspected to be decreased, clinicians, while considering the valuable diagnostic time, might bypass the D-dimer test and directly move on to radiological imaging.
Renal impairment is an additional clinical condition associated with elevated D-dimer concentrations.6, 7, 8 Especially in patients with kidney failure, CT scanning is relatively contraindicated because of the renal complications, due to allergic reactions to contrast dye. Whether diagnostic algorithms for pulmonary embolism can be used with equal efficiency in patients with decreased renal function is not known. To assess this efficiency, we have evaluated the utility of D-dimer testing in patients with suspected pulmonary embolism and impaired renal function.
Materials and Methods
Study Design
A post hoc analysis was performed, based on a recent multicenter prospective study in which the clinical effectiveness of a simplified algorithm using a clinical decision rule, D-dimer testing, and CT scanning in patients with suspected pulmonary embolism was evaluated.1 In consecutive in- and outpatients presenting with clinically suspected pulmonary embolism, the clinical probability of pulmonary embolism was assessed by using the clinical decision rule according to Wells et al.9 If this sum of the rule was 4 points or less, D-dimer tests were performed. In combination with a normal quantitative D-dimer test result, pulmonary embolism was excluded in these patients. In all patients with either elevated D-dimer concentrations or likely clinical probability (>4 points), the diagnosis of pulmonary embolism was confirmed or excluded by spiral CT scanning. Patients were excluded if they had received treatment with therapeutic doses of unfractionated or low-molecular-weight-heparin for more than 24 hours; had a life expectancy <3 months; were pregnant; had geographic inaccessibility, an age below 18 years, an allergy to intravenous contrast agents, or a known severe renal insufficiency (creatinine clearance <30 mL/min); or presented with hemodynamic instability. All patients were followed-up for 3 months, to assess the failure rate of the used diagnostic algorithm.
Patients
Patients referred to our hospital, the Leiden Medical University Hospital (Netherlands), with an unlikely clinical probability were eligible for the present study. D-dimer measurements were performed in all patients using the VIDAS D-Dimer assay (BioMerieux, Marcy L'Etoile, France). A D-dimer concentration <500 mg/mL was considered normal. Creatinine levels, which were assessed as clinical routine, were measured by a colorimetric method (Roche Hitachi Modular P800 analyzer; Roche Diagnostics, Indianapolis, IN). D-dimer and creatinine test results were derived from the electronic database of our hospital. Only creatinine levels that were assessed simultaneously with, or with a maximum of 4 days before, the assessment of the D-dimer concentration, were used for this analysis. Glomerular filtration rate (eGFR) was estimated using the formula of the Modification of Diet in Renal Disease,10 which takes into account age, sex, and race, and has been validated in large populations.11, 12 The institutional review board of our hospital approved the study protocol, and written informed consent was obtained from all participants.
Data Analysis
Patients were categorized in different categories of decreasing eGFR according to the K-DOQI (Kidney Disease Outcomes Quality Initiative) guidelines:13 normal renal function (eGFR >89 mL/min), kidney damage with mild decrease in eGFR (eGFR 60-89 mL/min), and kidney damage with moderate decrease in eGFR (eGFR 30-59 mL/min). Patients with eGFR <30 mL/min were excluded from the Christopher study.1 Consequently, we cannot present data on patients with severe renal impairment. Correlation between D-dimer levels and renal function was evaluated. The proportions of patients with normal D-dimer concentration in different categories of eGFR were compared. We calculated the specificity of the D-dimer test in all categories of renal function with the data from the 3-month follow-up performed in all patients.1 Differences in study outcomes were assessed by one-way analysis of variance. Post hoc least significant difference testing was performed for parameters that proved statistically significant on analysis of variance. Kruskall-Wallis tests with Mann-Whitney test for pairwise comparisons (P values were multiplied by 3 to correct for multiple testing) were performed for parameters that were not normally distributed. P values <.05 were considered significant.
Results
In total, 385 patients with a Wells score of 4 points or less were included in the Christopher study in our hospital. Because the assessment of renal function was clinical routine but not standard in all patients, creatinine levels were not measured in 33 patients (8.6%). The characteristics of all patients are presented in the Table. Patients in whom creatinine levels were measured were significantly older and had a tendency to a more extended and serious medical history.
Table. General Characteristics of the Study Patients
| Renal Function Assessed (n = 351) | Renal Function Not Assessed (n = 34) | |
|---|---|---|
| Sex, male – n (%) | 150 (43.1) | 9(26.5) |
| Age – years (mean ± SD) | 48±16⁎ | 38±15⁎ |
| History of VTE – n (%) | 36(10.3) | 3(8.8) |
| Active malignancy – n (%) | 26(7.4) | 0(0) |
| History of cardiac failure – n (%) | 13(2.7) | 0(0) |
| Outpatient – n (%) | 329(93.7) | 33(97.1) |
⁎P = .015. |
Patients in whom creatinine levels were measured were categorized by eGFR (according to the Modification of Diet in Renal Disease formula). Of the 352 patients, 107 had an eGFR >89 mL/min, 206 had an eGFR between 60 and 89 mL/min, and 39 patients had an eGFR of 30-59 mL/min. The correlation between D-dimer levels and renal function was evaluated. D-dimer concentration increased significantly between the 3 categories of decreasing eGFR (P = .027 and P = .021 for moderate renal impairment compared with mild renal impairment and normal renal function, respectively, Figure 1).
Figure 1.
Correlation between D-dimer levels and renal function. Figure represents median, interquartile range, and 5-95 percentiles. GFR = glomerular filtration rate; n = number.
The clinical utility of D-dimer testing for excluding pulmonary embolism in the 3 categories of renal function was determined. Overall, normal D-dimer test results were established in 56% (216/385) of the patients. Normal D-dimer levels were found in 107 patients (58%, 95% confidence interval [CI], 48%-67%) with eGFR >89 mL/min, in 206 patients (54%, 95% CI, 47%-61%) with eGFR 60-89 mL/min, and in 39 patients (28%, 95% CI, 15%-45%) with eGFR 30-59 mL/min (Figure 2). In the patients with the most severe renal impairment, CT scans could be withheld in significantly less patients (28%) compared with the cohort with normal renal function (58%, P = .0025).
Figure 2.
Utility of D-dimer testing in excluding pulmonary embolism in different categories of renal impairment. Error bars represent 95% confidence interval. GFR = glomerular infiltration rate.
None of the patients with a Wells score of 4 points or less and a normal D-dimer test result was diagnosed with venous thromboembolism in the 3-month follow-up period. Specificity of the D-dimer test was 60.8% in patients with eGFR >89 mL/min, 60.7% in patients with eGFR 60-89 mL/min, and 33% in patients with eGFR 30-59 mL/min. Specificity in both the patients with normal renal function and with mild decrease in GFR was significantly superior compared with the specificity in patients with moderate decrease in eGFR (P = .0086 and P = .0042, respectively).
Discussion
Our data show that the specificity of D-dimer testing in patients with clinically suspected pulmonary embolism and impaired renal function is significantly decreased compared with patients with a normal renal function. However, almost one third of the patients with an eGFR beneath 60 mL/min still have a normal D-dimer test result. A recent study reported an incidence of laboratory-defined contrast nephropathy of 4% in a heterogeneous outpatient population undergoing CT scanning for pulmonary embolism using nonionic, low osmolar iodinated contrast, although none of these patients developed irreversible end-stage renal failure.14 Because contrast nephropathy is recognized as one of the most important causes of renal failure in hospitalized patients and is associated with high in-hospital mortality and poor long-term survival,15, 16, 17 it seems essential to limit the number of CT scans to a minimum. Considering this, preventing CT scans in 28% of patients with modest renal function impairment is, in our opinion, a considerable proportion.
Our study has 2 limitations. First, creatinine levels were not evaluated in all study patients. Nevertheless, patients in whom data on creatinine levels were missing were generally healthier, average D-dimer levels were lower, and the amount of normal D-dimer tests was significantly greater compared with the patients in whom we could calculate the GFR. Consequently, if our data are subject to confounding, we believe our results tend to be an underestimation. Second, we have no data on patients with an eGFR of 30 mL/min or less. Of note, administration of iodinated contrast in these patients is relatively contraindicated. Although specific prophylactic therapies designed to reduce contrast toxicity have been suggested,18, 19 it seems to be beneficial to perform an ultrasonography of the legs before proceeding to CT scanning in patients with renal failure. Deep vein thrombosis can be diagnosed in 7.0% (95% CI, 5.0-9.4) of the patients with a low/intermediate clinical probability, but an elevated D-dimer concentration,3 thereby creating an indication for anticoagulant therapy. In these cases, CT scans can be withheld. Finally, our study population consisted mainly of outpatients (94%). We could not determine whether our results can be applied to an inpatient population.
Data from this post hoc analysis show that the specificity of D-dimer testing in patients with suspected pulmonary embolism and decreased eGFR is significantly decreased. Nonetheless, performing D-dimer tests in patients with kidney damage with moderate decrease in eGFR is still useful because CT scanning can be withheld in one third of the patients. The clinical implication of our findings is that D-dimer tests to exclude the presence of pulmonary embolism should be performed in all patients with suspected pulmonary embolism and an unlikely clinical probability, also in the setting of renal function impairment. Large prospective studies are needed to substantiate our findings and to evaluate whether our results can be applied to different D-dimer tests and an inpatient population.
References
- Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA. 2006;295:172–179
- Computed tomographic pulmonary angiography vs ventilation-perfusion lung scanning in patients with suspected pulmonary embolism: a randomized controlled trial. JAMA. 2007;298:2743–2753
- Diagnosis of pulmonary embolism by multidetector CT alone or combined with venous ultrasonography of the leg: a randomised non-inferiority trial. Lancet. 2008;371:1343–1352
- Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med. 2001;135:98–107
- . Monoclonal antibodies to crosslinked fibrin degradation products (XL-FDP) (II Evolution in a variety of clinical conditions). Br J Haematol. 1988;68:91–96
- . Abnormalities of coagulation in hypertensive patients with reduced creatinine clearance. Am J Med. 2000;109:556–561
- Elevations of inflammatory and procoagulant biomarkers in elderly persons with renal insufficiency. Circulation. 2003;107:87–92
- Renal function and cardiovascular mortality in elderly men: the role of inflammatory, procoagulant, and endothelial biomarkers. Eur Heart J. 2006;27:2975–2981
- Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with simpliRED D-dimer. Thromb Haemost. 2000;83:416–420
- A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation (Modification of Diet in Renal Disease Study Group). Ann Intern Med. 1999;130:461–470
- Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247–254
- . Assessing kidney function–measured and estimated glomerular filtration rate. N Engl J Med. 2006;354:2473–2483
- . K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39(Suppl 1):S1–S266
- . Contrast nephropathy following computed tomography angiography of the chest for pulmonary embolism in the emergency department. J Thromb Haemost. 2007;5:50–54
- Hospital acquired renal insufficiency: a prospective study. Am J Med. 1983;74:243–248
- . The effect of acute renal failure on mortality (A cohort analysis). JAMA. 1996;19:1481–1494
- Acute renal failure after coronary intervention: incidence, risk factors, and relationship to mortality. Am J Med. 1997;103:368–375
- Prevention of radiographic contrast agent-induced reductions in renal function by acetylcysteine. N Engl J Med. 2000;343:180–184
- Ascorbic acid prevents contrast mediated nephropathy in patients with renal dysfunction undergoing coronary angiography or intervention. Circulation. 2004;110:2837–2842
Funding: This study was supported in part by unrestricted grants from our hospital.
Conflict of Interest: None.
Authorship: All authors had access to the data and a role in drafting and writing the manuscript.
PII: S0002-9343(09)00492-6
doi:10.1016/j.amjmed.2009.03.032
© 2009 Elsevier Inc. All rights reserved.
Volume 122, Issue 11 , Pages 1050-1053, November 2009
