Sage Journals: Discover world-class research

Abstract

A rapid and accurate diagnosis of venous thromboembolism (VTE) in the elderly individuals represents a dilemma due to nonspecific clinical presentation, confusing laboratory results, and the hazards of radiological examination in this age-group. d-Dimer test is used mainly in combination with non-high clinical pretest probability (PTP) to exclude VTE. d-Dimer testing retains its sensitivity, however, its specificity decreases in the elderly individuals. Raising the cutoff level improves the specificity of the d-dimer test without compromising its sensitivity. The current study aimed to explore the reliability of higher d-dimer cutoff values for the diagnosis of asymptomatic VTE in a population of bedridden hospitalized elderly patients with non-high clinical PTP. This retrospective study included 252 bedridden hospitalized elderly patients (>65 years) who were admitted to the Ain shams University Specialized Hospital with non-high clinical probability and developed later reduced mobility; all underwent quantitation of d-dimer and Doppler examination. Considering the whole population (>65 years), the age-adjusted cutoff achieved the best performance in comparison with the conventional and receiver operating characteristic (ROC)–derived cutoffs. When stratified according to age, the age-adjusted cutoff showed the best performance in the age-group 65-70 and comparable performance with the ROC-derived cutoff in the age-group 71-80, however, its sensitivity compromised in those older than 80 years. In conclusion, it is recommended to use age-adjusted cutoff value of d-dimer together with the clinical probability score in elderly individuals (65-80 years).

Keywords

age-adjusted cutoff DVT elderly venous thromboembolism

Introduction

Deep vein thrombosis (DVT) and its possible life-threatening consequence, pulmonary embolism (PE), are leading causes of preventable morbidity and mortality following the hospitalization of elderly people.¹ Accurate and timely diagnosis of venous thromboembolism (VTE) can be lifesaving especially in this age-group due to short-term mortality.² However, in bedridden elderly patients, the diagnosis of DVT may be challenging because of the high prevalence of asymptomatic cases.^3,4

Nowadays, the diagnostic workup for suspected VTE has evolved into an integrated approach that includes pretest probability (PTP) assessment and d-dimer testing in combination with imaging.^5,6

d-Dimer assay is the best recognized biomarker for the initial assessment of suspected VTE due to high sensitivity in comparison with poor specificity; false-positive d-dimer results have been noted in inflammation, pregnancy, malignancy, and the elderly individuals.⁷

A high proportion of elderly patients with a non-high clinical probability undergo unnecessary burdensome diagnostic investigations, in whom the yield of this imaging is relatively low. Hence, d-dimer tests are often used to rule out VTE (PE or DVT) in these patients in contrast to those with a high clinical probability in whom imaging examination is warranted to confirm or refute the diagnosis irrespective of the d-dimer results.⁸

d-Dimer levels significantly increase with age possibly due to a higher incidence of comorbidity. Although the incidence of VTE increases with age, the proportion of the patients with a d-dimer level lower than the conventional cutoff value (500 μg/mL for the most available commercial assays) and in whom the diagnosis of VTE may be ruled out by the test is reduced.⁹

Recently, many studies evaluated different approaches to improve the specificity of d-dimer test in the elderly individuals, using either different cutoff values in the elderly population, in combination with the PTP score, or adopting different age-dependent cutoff values for patients older than 50 years (patient’s age × 10 µg/L fibrinogen equivalent units [FEU]). Many studies were proposed on the basis of retrospective studies regarding patients with suspected DVT or PE,^{8,10

–22}

The aim of our study was to assess different d-dimer cutoffs combined with clinical PTP for better specificity of the d-dimer without compromising sensitivity, that is, safer exclusion of VTE.

Participants and Methods

This study was an observational retrospective study that was conducted from June 2013 till January 2014 in the Ain Shams University Specialized Hospital.

Inclusion Criteria

Age ≥65 years.

Hospital admission for acute medical illness other than VTE.

Life expectancy of at least 30 days.

Reduced mobility persisting for at least 7 days that was reported by either the attending physician or an informant.

Exclusion Criteria

Patient presenting with classical signs and symptoms of DVT or PE.

History of recent thromboembolism (up to 1 month).

Patients receiving oral anticoagulant therapy or low-molecular-weight heparin for any reason.

Any patient with the history of recurrent DVT.

All patients in our study were subjected to the following:

Clinical assessment: for exclusion of any VTE signs and symptoms using the Wells scoring system, selecting those with non-high clinical probability.

Laboratory assessment: quantitative determination of d-dimer using the INNOVANCE d-dimer assay, an automated, particle-enhanced, immunoturbidimetric assay for the cross-linked fibrin degradation products (d-dimers) in human plasma on Sysmex CA-1500 (Siemens Diagnostics, Germany). The results were expressed in mg/L FEU. Two milliliters of venous blood were collected from the patients on a trisodium citrate vacutainer (1 part of citrated sodium [0.109 mmol/L] with 9 parts of venous blood). Platelet-poor plasma was obtained by centrifugation at 3000g in room temperature, aliquoted, snap frozen, and stored at −70°C for further analysis; hemolysed samples were discarded.

Radiological assessment: Doppler examination of both lower limbs by a general electric logiq 9 device performed with a real-time, B mode, linear-array sonographic scanner at 5.0 to 7.5 MHz. The entire proximal deep vein system was explored for compressibility.

Study Design

Patients were analyzed by 2 approaches: whole population (≥65 years) and age-stratified groups—first group 65 to 70 years, second group 71 to 80 years, and third group above 80 years.

The following cutoff values were applied:

The “conventional” cutoff value for INNOVANCE d-dimer provided by the manufacturer was 0.5 mg/L FEU irrespective of the patient’s age.

The receiver operating characteristic (ROC)–derived cutoff value showing the best specificity, while preserving the 100% sensitivity.

An “age-adjusted” cutoff (age in years × 10 µg/L) proposed by Douma et al.^21,22

On correlating the results of both Doppler ultrasound examination and d-dimer regarding the 3 cutoff values, patients were stratified into true-positive, true-negative, false-positive, and false-negative cases, and the following were calculated for each cutoff level:

Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy.

The proportion in which DVT could be excluded (based on an unlikely clinical probability and a negative d-dimer test result using the different cutoff values), that is, true negatives.

The number of patients needed to undergo a d-dimer assay to exclude DVT in 1 patient (ie, the number of patients needed to test by dividing 1 by the proportion of true negatives).

Statistical Analysis

Data were analyzed using Statistical Package for the Social Science (SPSS; version 15), a statistical software package for the Windows XP operating system of IBM-compatible personal computer. The χ² test was used for comparison of qualitative data. Student t test was used for comparison of quantitative data. Receiver operating characteristic curves were prepared, and for each cutoff value, the area under the curve and the 95% confidence interval (CI) were calculated. We used the Wilson score method to compute the 95% CI for estimated proportions.

Results

Of a total 1123 patients older than 65 years who were admitted to the emergency department of Ain Shams University Specialized Hospital, 252 were serially included according to our inclusion and exclusion criteria (non-high PTP of VTE). Four hundred eight patients were not admitted to the hospital and left after the emergency department visit, 117 patients either refused our study or dropped out before completion of the desired tests, 100 patients were already on anticoagulation, 23 patients died after admission within 48 hours, 192 didn’t report any reduced mobility, 4 patients had previous recurrent DVT, 25 patients were presenting with signs and symptoms suggestive of DVT, and 2 patients’ radiological assessment wasn’t successful.

There were 127 (50.3%) males and 125 (49.6%) females with a male to female ratio of 1.01:1. Their ages ranged from 65 to 98 years with a mean of 73.3 years and standard deviation (SD) = 6.74. Frequency of different age-groups among our sample is shown in Table 1.

Table 1.

Frequency of Different Age-Groups Among Our Sample.

Age-Group	Frequency	Percentage
65-70	168	66.7%
71-80	58	23.0%
81 and above	26	10.3%
Total	252	100.0

They were divided according to the cause of admission with infection as the main cause of admission, representing 23.8% of the total sample.

Pearson χ² test was performed to determine the association between the cause of admission, presence and number of comorbidities, and the appearance of positive radiological findings. There was no statistically significant correlation between the medical grounds for admission and the prevalence of DVT (Table 2).

Table 2.

Different Causes of Admission in Relation With the Diagnosis of DVT.

Cause of Admission	Total Number (N = 252)	Radiological Finding		Pearson χ²	Asymp. Sig. (2 Sided)
Cause of Admission	Total Number (N = 252)	Positive	Negative	Pearson χ²	Asymp. Sig. (2 Sided)
Infection	60 (23.8%)	21	39	.017	.895 (NS)
CVS-related causes	51 (20.2%)	20	31	.341	.559 (NS)
Liver dysfunction	27 (10.7%)	8	19	.488	.485 (NS)
Respiratory tract disease	26 (10.3%)	10	16	.095	.758 (NS)
CNS-related causes	24 (9.5%)	6	18	1.32	.249 (NS)
Fracture following fall	11 (4.4%)	4	7	.002	.963 (NS)
Other causes	53 (21.0%)	21	32	.559	.674 (NS)

Abbreviations: Asymp. Sig., asymptotic significance; CNS, central nervous system; CVS, cardio vascular system; DVT, deep vein thrombosis; NS, not significant.

Laboratory and radiological results were collected from the patients’ charts in Ain Shams University Hospitals.

d-Dimer Results

d-Dimer test performed on the participants showed values with a mean of 1.23, SD of 1.065, range of 0.06 up to 6.9 mg/L, and mode of 0.9 mg/L, with a mean value of males (1.24, SD 0.96) being higher than the mean value of females (1.17, SD 1.07).

For patients ages 65 to 70 years, d-dimer had a mean of 1.24, SD of 1.08 mg/L, and mode of 1.5 mg/L. However, for patients aged 71 to 80 years, D-dimer had a mean of 1.32, SD of 1.04 mg/L, and mode of 0.8 mg/L. d-Dimer in patients with ages 80 and above had a mean of 0.97, SD of 1.00 mg/L, and mode of 0.9 mg/L.

Radiological Data

Ninety (35.7%) of our participants showed compression changes in their lower limb by duplex sonography, diagnosed as having DVT with prevalence of 35.7% in the non-high PTP, whereas 162 (64.3%) showed normal patency and compressibility of blood vessels and normal blood flow inside it (proved to be free from DVT).

Patients who were found to have DVT by radiological examination had d-dimer results with a mean 1.98 ± 1.19 mg/L. Patients without DVT had d-dimer results with a mean 0.81 ± 0.69 mg/L.

Study of the Whole Sample Population

Table 3 shows that the best performance was for the age-adjusted cutoff in comparison with the conventional and ROC-derived cutoffs in the whole population.

Table 3.

Performance Criteria for Different Cutoffs in the Whole Population.

Whole Population	Conventional Cutoff	ROC-Derived Cutoff Value (0.6 mg/L; AUC, 0.868; 95% CI, 0.825-0.911)	Age-Adjusted Cutoff
Sensitivity	100%	100%	97.8%
Specificity	38.27%	42%	52.4%
PPV	47.3%	49%	53.3%
NPV	100%	100%	97.7%
Accuracy	60.3%	62.7%	68.6%
% Safely excluded	24.6%	27%	33.7%
No. needed to test	4.1	3.7	2.96

Abbreviations: AUC, area under the curve; CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; ROC, receiver operating characteristics [curve].

Study of Age-Specific Groups Revealed the Following Results

Table 4 shows that the age-adjusted cutoff achieved the best performance in the 65 to 70 age-group and with comparable performance to the ROC-derived cutoff in the 71 to 80 age-group, however, its sensitivity compromised in those older than 80 years in whom the conventional and ROC-derived cutoffs show comparable performance.

Table 4.

Performance Criteria for Different Cutoffs in the 3 Age-Groups.

Age-Groups	65-70 Years			71-80 Years			>80 Years
Cut offs	Conventional cutoff	ROC-derived cutoff value (0.6 mg/L)^a	Age-adjusted cutoff	Conventional cutoff	ROC-derived cutoff value (0.8 mg/L)^b	Age-adjusted cutoff	Conventional cutoff	ROC-derived cutoff value (0.550 mg/L)^c	Age-adjusted cutoff
Sensitivity	100%	100%	100%	100%	100%	100%	100%	100%	71.4%
Specificity	37.3%	41.1%	50.5%	25%	55.5%	52.7%	52.6%	57.9%	63.1%
PPV	47.6%	49.2%	46.5%	44.9%	57.9%	56.4%	43.7%	46.6%	41.6%
NPV	100%	100%	100%	100%	100%	100%	100%	100%	85.7%
Accuracy	60.1%	62.5%	68.4%	53.4%	72.4%	70.7%	65.4%	69.2%	65.3%
% Safely excluded	23.8%	26.2%	32.1%	15.5%	34.5%	32.7%	38.5%	42.3%	46.2%
No. needed to test	4.2	3.8	3.1	6.45	2.9	3.1	2.6	2.36	2.16

Abbreviations: CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; ROC, receiver operating characteristics [curve].

^aAUC, 0.862; 95% CI, 0.808 to 0.916.

^bAUC, 0.902; 95% CI, 0.822 to 0.981.

^cAUC, 0.835; 95% CI, 0.668 to 1.000.

The pooled specificity of d-dimer testing decreased substantially with increasing age from 37.3% in patients aged 65 to 70 years to 25% in patients aged 71 to 80 years and increased to 52.6% in those >80 years when the conventional cutoff value was applied, whereas the pooled sensitivity hardly differed between the age-groups (Table 4).

The prevalence of DVT was comparable between groups, 36.3% (65-70 years), 37.9% (71-80 years), and 31.8% (>80 years), with the overall prevalence in the whole population being 35.7%

Discussion

In elderly patients, specificity of d-dimer is compromised owing to the raised d-dimer levels without any underlying thrombotic disease, which increases the false-positive results and hence raises the number of patients undergoing unnecessary radiological examination. Developing more accurate interpretation of d-dimer results will reduce the financial burden of unnecessary tests and ensure not missing the diagnosis, which may be fatal.

Using the conventional cutoff for d-dimer (0.5 mg/L) that is suggested by most manufacturers and has been adopted by many institutions, laboratory, and clinicians worldwide will lead to a high number of false-positive results in elderly patients. An obvious solution to this problem was raising the cutoff to a level that improves the specificity of d-dimer test without compromising its sensitivity.

The aim of the current study was to explore the reliability of 3 different d-dimer cutoff values: the conventional cutoff (0.5 mg/L), the ROC-derived cutoff value, showing the best specificity while preserving the 100% sensitivity, and the age-adjusted cutoff (age in years × 10 µg/L) proposed by Douma et al^21,22 for the diagnosis of asymptomatic VTE in a population of bedridden hospitalized elderly patients in whom the clinical presentation per se wasn’t decisive.

The current study was carried out on 252 bedridden hospitalized elderly patients with non-high PTP for VTE based on clinical decision rules applying Wells’ scoring system (≤2), with exclusion criteria, inclusion criteria, and study design similar to the studies by Granziera et al,²³ Cini et al,²⁴ Douma et al,²² Harper et al,⁹ and Schouten et al⁸ as this is the indicated population for the application of d-dimer tests as opposed to the studies by Lippi et al,²⁵ Haas et al,¹⁰ Han et al,²⁶ and Schouten et al²⁷ who included patients with high PTP for VTE. All patients in our study were subjected to clinical assessment, d-dimer testing, and Doppler examination.

d-Dimer testing in the present study was performed by INNOVANCE d-dimer assay using a CA-1500 Sysmex Siemens analyzer. Other studies^{8
–10,22
–24,26} assessed different assays on different analyzers: Tinaquant Roche, rapid ELISA Vidas d-dimer bioMérieux, and STA-Liatest d-dimer test, with an assigned cutoff at 0.5mg/L (FEU) as reported by the manufacturer. Also, assessed the Auto-Dimer BIOPOOL DASIT method (Trinity Biotech, Bray, Ireland) and the HemosIL d-dimer HS (Instrumentation Laboratory, Bedford, Massachusetts) with an assigned cutoff at 0.23 mg/L (d-dimer unit [DDU]). d-Dimer unit can be converted to FEU by applying the formula: FEU = 2 × DDU.

In our study, 252 patients were included with mean ages of 73.3 ± 6.74 years, showing infection as the main cause of admission with a frequency of 23.8% and prevalence of VTE as 35.7%, without significant correlation between any of medical grounds for admission and the prevalence of DVT. Similar, the study reported by Lippi et al²⁵ included 1647 patients with mean ages of 77 ± 15 years, showing the highest frequency for admission due to infection (15.6%) and prevalence of DVT (about 21.5%), and the study reported by Granziera et al²³ had a study population of 199 patients with mean age of 86.3 ± 6.7 years, showing 29.1% of patients were admitted due to infection and 12.5% due to the prevalence of DVT, and neither studies showed correlation between any medical grounds for admission and the prevalence of DVT.

d-Dimer values in the present study had a mean of 1.23 ± 1.065 mg/L that is close to the mean of d-dimer values reported by Granziera et al,²³ which was 1.2 ± 1.5 mg/L. Other studies didn’t report mean values of d-dimer.

Studying the Sample Population as a Whole

Performance of the conventional d-dimer cutoff value of 0.5 mg/L in the present study was close to the results obtained by similar studies^{8
–10,23–24,26} characterized by compromised specificity, while keeping the sensitivity above 97%.

The best performance was for the age-adjusted cutoff, showing the highest specificity of 52.4%, accuracy of 68.6%, percentage safely excluded of 33.7% with an absolute increase of 9.1% and 6.7%, and the least number of patients needed to be tested to exclude DVT in 1 patient (2.96) in comparison with the conventional cutoff and the ROC-derived cutoff (0.6 mg/L), respectively, while keeping the sensitivity at 97.8% due to the loss of 2 cases (2 false negatives) in those older than 80 years.

Similar to Han et al²⁶ who reported age-adjusted cutoff formula to give sensitivity of 95% and specificity of 84.1% for his non-high (low and moderate) PTP population, Cini et al²⁴ reported the age-adjusted cutoff formula to give sensitivity of 100% and specificity of 67.3% with an absolute increase in percentage safely excluded (24.1%) and reduced number needed to test from 2.8 to 1.7 for his whole population and Schouten et al⁸ documented increased specificity from 42% to 47.8% and a decrease in the number of cases needed to test from 2.4 to 2.1 with an absolute increase in percentage of excluded cases (efficiency; 5.7%) when applied the age-adjusted cutoff formula.

In this study, the ROC-derived cutoff value showing the best specificity while preserving the 100% sensitivity was at 0.6 mg/L and gave specificity of 42%, which is lower than that reported by Haas et al¹⁰ who suggested a cutoff value of 0.75 mg/L, which gives a sensitivity of 100% and a specificity of 49.2%. Harper et al⁹ chose a point of 1.00 mg/L, improving the specificity to 55% and reported sensitivity of >98%, Granziera et al²³ suggested raising the cutoff value to 0.49 DDU mg/L (equivalent to 0.98 mg/L FEU), which increased the specificity from 20.1% to 39.1% while keeping the sensitivity as 100%, and Cini et al²⁴ suggested raising the cutoff value to 376 DDU ng/mL (equivalent to 752 ng/mL FEU), which increased specificity to 63.7% while keeping sensitivity as 100%, and the fixed cutoff (>750) adopted by Schouten et al⁸ with specificity being improved from 42% to 47.4%.

Stratifying the Sample Population According to Age

Studying those aged 65 to 70 years

In the present study, performance of the conventional cutoff (0.5 mg/L) showed sensitivity above 99% and compromised specificity, which was close to other studies done on the similar age-group.^8,24,27

The best performance in this age-group was for the age-adjusted cutoff showing the highest specificity of 50.46%, accuracy of 68.4%, and proportion of safely excluded patients of 32.1% with an absolute increase of 8.3% and 5.9% in comparison with the conventional cutoff and the ROC-derived cutoff (0.6 mg/L), respectively, while preserving 100% sensitivity with the least number of patients needed to be tested to exclude DVT in 1 patient (3.1). This is very close to the results obtained by Schouten et al²⁷ who showed 97.3% sensitivity and 49.5% specificity, Schouten et al⁸ who reported 98.5% sensitivity, 39.3% specificity, and number of cases needed to test 2.6, efficiency 6.5% for the similar age-group of 61 to 70 years, and Cini et al²⁴ who documented 100% sensitivity, 71.9% specificity, and number of cases needed to test 1.4. The number of cases excluded increased to 24.5% when applied the age-adjusted cutoff formula.

Analysis of the ROC curve revealed that the cutoff value showed the best specificity of 41.1%, while preserving 100% sensitivity at 0.6 mg/L with accuracy of 62.5%, percentage safely excluded of 26.2%, number needed to test as 3.8 in comparison with the study reported by Schouten et al⁸ who adopted a fixed cutoff (>750 ng/mL) with a sensitivity of 98.5%, a specificity of 42.1%, number of cases needed to test as 2.4, and efficiency as 9.3% and the study by Cini et al²⁴ using the modified cutoff (376 ng/mL DDU) with 100% sensitivity for this age-group, 73.7% specificity number of cases needed to test to be 1.4, and the number of cases excluded increased to 26.3%.

Studying those aged 71 to 80 years

Both the ROC-derived cutoff (0.8 mg/L) and the age-adjusted cutoff values showed comparable specificities (55.5% and 52.7%, respectively), accuracies (72.4% and 70.7%), and proportions of safely excluded patients (34.5% and 32.7%), which are higher than those of the conventional cutoff specificity (25%), accuracy (53.4%), and proportion of safely excluded patients (15.5%) with an absolute increase of 17.5% using the age-adjusted cutoff and 19% using 0.8 mg/L cutoff than using conventional cutoff, while preserving 100% sensitivity with the least number of patients needed to be tested to exclude DVT in 1 patient, 2.9 and 3.1, respectively, versus 6.45 for the conventional cutoff.

In accordance with the study by Schouten et al,⁸ comparable performance of both age-adjusted cutoff and their fixed cutoff for this age-group showed 100% sensitivity, 45% specificity, number of cases needed to test to be 2.2, and 14.4% efficiency for the former and 100% sensitivity, 45.9% specificity, number of cases needed to test to be 2.2, and 15.3% efficiency for the latter. Similarly, Cini et al²⁴ reported comparable performance of both age-adjusted cutoff and their modified cutoff with 100% sensitivity, 50% specificity, number of cases needed to test to be 2.0, and proportion of cases excluded increased to 29.3% for the former and 100% sensitivity, 48.8% specificity, number of cases needed to test to be 2.0, and proportion of cases excluded increased to 28.1% for the latter.

Also in a close proximity to Schouten et al,²⁷ results showed 97.3% sensitivity and 44.2% specificity for the age-adjusted formula.

Studying those aged 81 years and above

Both the ROC-derived cutoff (0.550 mg/L) and the conventional cutoff values showed comparable specificities (57.9% and 52.6%, respectively), accuracy (69.2% and 65.4%%), proportions of safely excluded patients (42.3% and 38.5%), and number of cases needed to test (2.36 and 2.6), while preserving 100% sensitivity. The age-adjusted cutoff did show an increase in specificity (63.16%) however at the expense of sensitivity (71.4%), making the age-adjusted formula not suitable to replace conventional cutoff (0.5 mg/L).

Similar studies^8,24,27 showed the age-adjusted cutoff to have the best performance in this age-group. However, our findings should be cautiously interpreted due to the small number of patients >80 years (n = 26) in comparison with the large cohort of patients in the study by Schouten et al²⁷ (n = 1269), Schouten et al⁸ (n = 62), and Cini et al²⁴ (n = 57).

In conclusion, in bedridden hospitalized elderly patients (65-80 years), combining the clinical probability for VTE with an age-adjusted formula ²¹ for the interpretation of d-dimer results improves the clinical utility of d-dimer with safer exclusion of DVT. However, further prospective impact studies are needed before application in daily clinical practice.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Aguilar

del Villar

. Diagnostic performance of d-dimer is lower in elderly outpatients with suspected deep venous thrombosis. Br J Haematol. 2005;130(5):803–805.

Siccama

Janssen

Verheijden

. Systematic review: diagnostic accuracy of clinical decision rules for venous thromboembolism in elderly. Ageing Res Rev. 2011;10(2):304–313.

Piazza

Seddighzadeh

Goldhaber

. Deep vein thrombosis in the elderly. Clin Appl Thromb Hemost. 2008;14(4):393–398.

Oger

Bressollette

Nonent

. High prevalence of asymptomatic deep vein thrombosis on admission in a medical unit among elderly patients. Thromb Haemost. 2002;88(4):592–597.

Huisman

Klok

. Diagnostic management of acute deep vein thrombosis and pulmonary embolism. J Thromb Haemost. 2013;11(3):412–422.

Wells

. Integrated strategies for the diagnosis of venous thromboembolism. J Thromb Haemost. 2007;5 suppl 1:41–50.

Kesieme

Jebbin

. Deep vein thrombosis: a clinical review. J Blood Med. 2011;2:59–69.

Schouten

Koek

Oudega

. Validation of two age dependent d-dimer cutoff values for exclusion of deep vein thrombosis in suspected elderly patients in primary care: retrospective, cross-sectional, diagnostic analysis. BMJ. 2012;344:e2985.

Harper

Theakston

Ahmed

Ockelford

. d-Dimer concentration increases with age reducing the clinical value of the d-dimer assay in the elderly. Intern Med J. 2007;37(9):607–613.

10.

Haas

Schutgens

Biesma

. An age-adapted approach for the use of d-dimers in the exclusion of deep venous thrombosis. Am J Hematol. 2009;84(8):488–491.

11.

Leng

Sitaraaman

. Application of age-adjusted d-dimer threshold for exclusion thromboembolism (PTE) in older patients: a retrospective study. Acute Med. 2012;11(3):129–132.

12.

Penaloza

Roy

Kline

. Performance of age-adjusted d-dimer cut-off to rule out pulmonary embolism. J Thromb Haemost. 2012;10(7):1291–1296.

13.

Hamblin

Cairns

Keeling

. The use of age-dependent d-dimer cut-off values to exclude deep vein thrombosis. Reply to using an age-dependent d-dimer cut-off value increases the number of older patients in whom deep vein thrombosis can be safely excluded. Haematol. 2012;97(11):1507–1513.

14.

Raviv

Israelit

. Shifting up cutoff value of d-dimer in the evaluation of pulmonary embolism: a viable option? Possible risks and benefits. Emerg Med Int. 2012;2012:517375.

15.

Jaffrelot

Le Ven

Le Roux

. External validation of a d-dimer age-adjusted cut off for the exclusion of pulmonary embolism. Thromb Haemost. 2012;107(5):1005–1007.

16.

van Es

Beenen

Gerdes

Middeldorp

Douma

Bossuyt

. The accuracy of d-dimer testing in suspected pulmonary embolism varies with the wells score. J Thromb Haemost. 2012;10(12):2630–2632.

17.

van Es

Mos

Douma

. The combination of four different clinical decision rules and an age-adjusted d-dimer cut-off increases the number of patients in whom acute pulmonary embolism can safely be excluded. Thromb Haemost. 2012;107(1):167–171.

18.

Vossen

Albrektson

Sensarma

Williams

. Clinical usefulness of adjusted d-dimer cutoff values to exclude pulmonary embolism in a community hospital emergency department patient population. Acta Radiol. 2012;53(7):765–768.

19.

Linkins

Bates

Lang

. Selective d-dimer testing for diagnosis of a first suspected episode of deep venous thrombosis: a randomized trial. Ann Intern Med. 2013;158(2):93–100.

20.

Andro

Righini

Le Gal

. Adapting the d-dimer cut off for thrombosis detection in elderly outpatients. Expert Rev Cardiovasc Ther. 2013;11(6):751–759.

21.

Douma

le Gal

Söhne

. Potential of an age-adjusted d-dimer cutoff value to improve the exclusion of pulmonary embolism in older patients: a retrospective analysis of three large cohorts. BMJ. 2010;340:c1475.

22.

Douma

Tan

Schutgens

. Using an age-dependent d-dimer cutoff value increases the number of older patients in whom deep vein thrombosis can be safely excluded. Haematol. 2012;97(10):1507–1513.

23.

Granziera

Rechichi

De Rui

. A new d-dimer cut off in bedridden hospitalized elderly patients. Blood Coagul Fibrinolysis. 2013;24(2):109–112.

24.

Cini

Legnani

Frascaro

. d-dimer use for deep venous thrombosis exclusion in elderly patients: a comparative analysis of three different approaches to establish cutoff values for an assay with results expressed in d-dimer units. Int J Lab Hem. 2014;36(5):541–547.

25.

Lippi

Bonfanti

Saccenti

. Causes of elevated d-dimer in patients admitted to a large urban emergency department. Eur J Intern Med. 2014;25(1):45–48.

26.

Han

Zhao

Cheng

. The performance of age-adjusted d-dimer cut-off in Chinese outpatients with suspected VTE. Thromb Res. 2015;136(4):739–743.

27.

Schouten

Geersing

Koek

. Diagnostic accuracy of conventional or age-adjusted d-dimer cutoff values in older patients with suspected VTE: systematic review and meta-analysis. BMJ. 2013;346:f2492.

Performance Evaluation of Different d -Dimer Cutoffs in Bedridden Hospitalized Elderly Patients

Abstract

Keywords

Introduction

Participants and Methods

Inclusion Criteria

Exclusion Criteria

Study Design

Statistical Analysis

Results

d-Dimer Results

Radiological Data

Study of the Whole Sample Population

Study of Age-Specific Groups Revealed the Following Results

Discussion

Studying the Sample Population as a Whole

Stratifying the Sample Population According to Age

Studying those aged 65 to 70 years

Studying those aged 71 to 80 years

Studying those aged 81 years and above

Footnotes

Declaration of Conflicting Interests

Funding

References