This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Pregnancy is a hypercoagulable state due to the normal maternal physiologic changes during pregnancy. Virchow's triad of thrombosis formation consist of changes in blood constituents, blood flow and the vessel walls happened in pregnancy due to these physiological changes. Venous stasis is caused by the increase in the capacitance level of the vein via the increasing action of circulating progesterone and the reduction in venous flow in the pelvic vein and inferior vena cava by the pressure effect of enlarging uterus during pregnancy. These changes might also injure the vessel walls. Changes in the blood constituents include increases in the production of clotting factors and decreases in fibrinolytic activity following the decrease in protein S, tissue plasminogen activator and activated protein C1, 2, 3.
Even though these changes have a protective effect by decreasing the risk of post partum haemorrhage, they can also lead to the formation of deep vein thrombosis and ultimately pulmonary embolism. In pregnancy, the incidence of venous thromboembolism (VTE) is increased four-fold compared to the incidence in the non pregnant population4. In developed countries, deaths from thromboembolism remain as the main causes of maternal death and the rate has been unchanged since 1997 up till 20055. Two studies show, the incidence of VTE increased across gestational age with two thirds of cases were in the antepartum period especially in the later gestation and one third of cases were in the post-partum period 6, 7. However, it was reported in the confidential enquiry into maternal and child health (CEMACH) that 10 of the 22 women who died from pulmonary embolism was during the first trimester of their pregnancy. 6 of them have suddenly died without prior warning while another 4 who were symptomatic complained to their respective doctors but were never investigated accordingly 4. In another study, 47 out of 95 cases of DVT were detected before 15 weeks compared to only 28 cases after 20 weeks. Due to the long duration which was covered by this study, the true incidence of deep vein thrombosis might have been underestimated as the clinical features are only apparent in one third of the patients7. These results show that VTE in the first trimester were often missed and can be fatal. Substandard care of patients due to failure to recognize and implementing early treatment of VTE were one of the main causes of mortality due to VTE4.
Quantitative methods of diagnosis are more reliable compared to clinical signs, as the physiological changes of pregnancy produce symptoms that mimic PE and DVT, such as dyspnoea, limb swelling, and chest pain. So reliance only on clinical signs causes overestimation of the cases8, 9. Overestimation of these cases becomes a burden on economic capacity of the HSE due to unnecessary expensive investigations to exclude PE, as well as admission and prolonged length of stay in hospital.
There are a range of diagnostic procedures to exclude Deep Vein Thrombosis and Pulmonary Embolism; from the non-invasive test like Doppler ultrasound, D-dimer Assay, Compression Ultrasonography, and Plain chest radiography to the invasive 'gold standard' diagnostic procedures such as CT pulmonary Angiography, Radionuclide lung scan and Venography9. However, all of these tests have only been tested extensively in non-pregnant women, and are yet to be validated for use on pregnant women. Even though the risk of inappropriate use of anticoagulation and the risk of missing the diagnosis of PE outweighs any risk of radiation imposed on mother and fetus, most clinicians and mothers-to-be are unwilling to expose the fetus to the unknown risks of ionizing radiation and contrasts of diagnostic imaging test10. The vulnerability of the fetus to radiation increases especially in the first trimester of pregnancy where all the important organs such as the heart and central nervous system are developing11. All the imaging procedures are also expensive, as shown in one study of cost-effectiveness comparing between CT scan, VQ scanning and compression ultrasound scanning. The most cost-effective procedure was CT scan, with $17,208 per life saved12. Hence, these tests are no longer used as first-line investigation. Therefore more research and studies are needed to assess and also validate non invasive tests such as Doppler Ultrasound, compression Ultrasonography and D-dimer assay. Compression Ultrasonography has been used as the primary diagnostic test in acute Deep Vein Thrombosis13 which usually produces symptoms as this test has sensitivity of 97% for detecting asymptomatic DVT. The disadvantage of this test is that it cannot detect thrombi in the calf, iliac and femoral vein and the vena cava, but in pregnancy there is a higher incidence of isolated deep vein thrombosis in the iliac vein10,11,14. Hence, lower limb ultrasound cannot be used to exclude PE in pregnant patients with highly suspicious presentation of PE but without symptoms suggestive of DVT.
The D-dimer assay has been used together with pre-test probability of clinical model to refute the diagnosis of DVT with its high negative predictive value of 99.5% (CI 99.1% to 100%) in non pregnant patients15,16. However, due to the elevated D-dimer levels even in normal pregnancy, its false-positive rate is increased, and its specificity is decreased. In Chan et al, the specificity of the D-dimer (SimpliRED) assay was 60% (CI, 52% to 68%), but the sensitivity was 100% (CI, 77% to 100%), and the negative predictive value was 100% (CI, 95% to 100%)17. Another study also substantiates this result where sensitivity of 100% and specificity of 27.3% for detection of VTE was found for diagnosing DVT18. These results mean that D-dimer has a very big potential to be used to exclude the diagnosis of DVT in pregnant patients. One limitation of these two studies is the number of patients diagnosed with DVT is very small thus decreasing the validity of the results. Both of the studies suggest that a study to validate a reference range for each trimester must be conducted to further increase the credibility of D-dimer assay as an exclusion test for DVT in pregnancy.
Aim and objectives
The aim of this study is to define the reference values for D-dimer concentration in the first trimester in normal pregnancy as a baseline value. The new cut-off value derived from this study should be treated as a hypothetical measurement and can only be used in clinical setting after being further tested on the specificity and sensitivity.
To observe the trends of the distribution of D-dimer level across gestational age in the first trimester
To investigate the association between D-dimer level and miscarriage
To investigate the effects of smoking, high BMI, parity and combined oral contraception use on D-dimer level
Methods and participants
Before recruiting and performing venesection to obtain blood sample from participants, application for project ethical approval was sent to the Cork Research Ethics Committee. Expedited ethical approval was granted on 21st April 2009 and data collection began in June 2009 and was finished in November 2009.
This is a cross-sectional study observing the level of D-dimer level in the population of pregnant women in the first trimester. 164 healthy women with singleton pregnancy of less than 12 weeks were recruited using random sampling from the early pregnancy clinic, pre-booking clinic and emergency department of Cork University maternity Hospital. Participations were on voluntary basis where women who were approached have the right not to participate in the study. Our exclusion criteria are women carrying twin or higher order pregnancies or on any anti-coagulation medication such as aspirin and low molecular weight heparin, history of thrombophilia or having a positive previous history of venous thromboembolism in the past. Women with chronic disease such as diabetes mellitus, chronic renal failure and chronic heart disease were excluded as well.
This study was funded by the internal research funding in the haematology lab, Cork University hospital.
Study sample size
In order to maintain the same specificity observed for the reference value out of pregnancy, Arkin and Wachtel's inferential approach was used to derive the number of samples needed for this study19. It has been shown that 85% of healthy, non pregnant women will have D-dimer level of less than 0.50 mg/L (MDA turbodimetric assay) in a few small previous studies. Thus by using the inferential approach to set the sample size to narrow the 95% confidence limits around this 85% proportion to less than 5%, 160 patients will be required in this study. As my study is using a different D-dimer assay (Auto-Dimer assay) with different cut-off value of 224 ng/ml, the result obtained from this study must not be compared directly with other studies except those that use the same assay and cut-off value.
Written consents were obtained first before taking their demographic details and venesection were performed afterwards to obtain the blood sample. Demographic details collected from patients include: maternal age, gestational age, parity, method of delivery for previous pregnancy, current BMI, smoking history, history of COCP (combined oral contraceptive) use, alcohol intake, current medication, past medical history, and family history of VTE. After taking the history, blood was drawn for D-dimer level testing in the Cork University Hospital Laboratory using rapid automated quantitative immunoturbidimetric D-dimer assay (Auto-Dimer) with the normal reference range of 0-224ng/mL. Venous blood was collected in 5 ml vacuum tubes (Becton-Dickinson, Franklin Lakes, USA) containing sodium-citrate (3.8%), and centrifuged at 3600 -g for 10 minutes at 4 C within 24 hours of collection. At the end of the study, all participants were followed up for status of viability of pregnancy before 24 weeks of gestation.
Abstract and presentation submissionStudy Timeline
Project Approval received
Start collecting data
Submit written research project
Finished Data Collection
Project Ethical Approval from CREC
Project Approval Form submitted to UCC Med School
Collected data was safely stored in a password protected database and was analysed using Statistical Package for the Social Sciences application 17 (SPSS). Descriptive statistics were used to calculate the mean, standard deviation (SD) and range of D-dimer. The range of D-dimer was calculated using the 95th percentile value of the D-dimer level in the first trimester of pregnancy. This calculation was based on the evaluation report on eight D-dimer assays including auto-dimer assay by Medicine Health and Regulatory Agency in 200320. The cut-off value for Auto-dimer test in the general population was calculated using the 95th percentiles value and the specificity and sensitivity of the cut-off value was determined as well. To test for statistical difference in the D-dimer level between viable pregnancy group and the miscarriage group, non-parametric test, Mann-Whitney U test for 2 group variables and Kruskall-Wallis test was used to see for any difference in D-dimer level in relation to smoking, combined oral contraception used, maternal age, body mass index, alcohol consumption, family history of thromboembolism and parity. P values of less than 0.05 were considered statistically significant. The variables with significant p value will be further tested using Spearman correlation coefficient to analyse the strength of the relationship. This test was also used to investigate the relationship between gestational age and D-dimer level in the first trimester of pregnancy.
164 women between 4 weeks to 12 weeks gestation participated in this study in the course of 2 months. The mean of gestational age is 8.9 weeks. The mean age of participants is 31.5 years old with the range of 18 years old to 44 years old. This shows that the distribution of maternal age from this study represents the pregnant women population attending CUMH.
The results from the D-dimer test show a significant increase of D-dimer concentration during the first trimester of normal pregnancy of mother with no apparent history of coagulopathy. The mean of D-dimer value is 165.9 ng/mL. The values of D-dimer at the 5th, 50th (Median) and the 95th percentiles are 37.75 ng/mL, 101 ng/mL and 511.75 ng/mL respectively. Using the normal D-dimer cut-off value of 224 ng/mL , 133 of 164 (81%) of the participant had a normal D-dimer level. From the number of participants with D-dimer level above the normal reference value, 8 patients had a D-dimer value above 511.75 ng/mL. This show the significant shift of the normal ranges of D-dimer value in pregnancy compared to the normal non-pregnant population (0-224 ng/mL). This also shows that there will be an increase in the number of false positive test in the first trimester if the conventional cut-off was used.
Based on the calculation made to derive the normal range in the normal population, the 95th percentile was used as the reference point in first trimester. There are 4 outliers who have more than 800 ng/ml level of D-dimer in this study. After filtering these results, the new cutoff derived from the 95th percentile is 429.5 ng/ml. Thus, the proposed range of D-dimer level is 0-429.5 ng/mL.
1st trimester (n:164)
1st Trimester (n:160) (w/o outliers)
429.5*Taken from MHRA report 2003
# Manufacturer cut-off value
By looking at the D-dimer distribution from 4 weeks to 12 weeks gestation, there was no significant trend of D-dimer level observed. Spearman Rank correlation coefficient for non parametric variables was used to investigate the correlation between gestational age and D-dimer level. The correlation value is only 0.019 with p value of 0.807 which show there is no significant relationship between the values of D-dimer in relation to gestational age in the first trimester.
Miscarriage and D-dimer level.
31 out of 164 participants (19%) had a miscarriage during this study. The level of D-dimer between the two group of patient with ongoing pregnancy and miscarriage were compared using Mann Whitney U test and no significant statistical difference was found (p=0.073).
Potential confounding factor; Body mass index, smoking and parity
The association between potential confounding factors such as body mass index, smoking history and parity of participants were investigated to further validate the range of D-dimer proposed as these factors are known to increase risk for venous thromboembolism. There were no significant statistical difference of D-dimer level were found in relation to BMI, parity and smoking status of participants using Kruskall Wallis test. The p values for each variable were 0.649, 0.138, and 0.056 respectively. Family history of VTE also has no significant association with D-dimer level with p value of 0.269. Significant p value was found only between D-dimer and oral contraceptives usage before pregnancy, but a negative Spearman correlation 0f -0.20 was found.
This study is not the first study that has been done in the recent years to produce a new reference range for D-dimer value in pregnancy. But it is never been adequately done.
The aim of my study is then to investigate the extent of increase of D-dimer level in the first trimester of normal pregnancy as the first step towards further research in the future. This is a relatively small study compared to the previous study in the past as my study focused only in the first trimester, up till 12 weeks to obtain a large sample of 165 participants in a short period of time. The fresh blood sample were sent to be analysed using automated D-dimer test, Auto-dimer with the reference value of <224 ng/ml in non-pregnant population.
The result of my study agreed with previous studies confirming that there is a shift towards hypercoagulable state in the first trimester when compared with the non-pregnant population D-dimer level. Only 81% of the participants had a normal D-dimer value if based on the conventional reference value of 224ng/ml. But due to the various assays that were used, the numerical values of each study cannot be directly compared with my result.
In a study by Kline, the mean D-dimer concentration was found to be 0.43mg/L at preconception and 0.58, 083, and 1.16 mg/L in the first, second and third trimester 21. This study used a different D-dimer assay, MDA immunoturbodimetric Assay, Organon Teknika with reference range of 0.50 mg/L in non-pregnant population. The number of samples in first trimester was 32 which were enough to achieve the 95% confidence limits. 50% of the samples submitted at 12 weeks had a D-dimer level above the cut-off point of 0.50 mg/L which is quite similar to our result.
Choi et al has conducted a study with a big sample size of 436 healthy women with exclusion criteria of complications during pregnancy, evidence of recent infection, and any history of blood transfusion. Ultrasonography and the date of the last menstrual period were used to determine the gestation. There were 107 patients in the first trimester group (1.0-12.0 weeks), second trimester (12.1-24.0 weeks, n=65), and third trimester (24.1-40.0 weeks, n=123). A control group consists of age-matched healthy women who had no history of pregnancy or recent diseases were also recruited. The D-dimer was analysed using Quantitative automated latex D-dimer assays. The result showed that the mean value of D-dimer in first pregnancy was 0.34±0.16 µg/ml compared to the control group with mean value of 0.27±0.13 µg/ml. This study was conducted in Korea therefore the results might not represent the Irish population here22.
Again, a study by Morse has also sought to search for reference range of D-dimer in each trimester where the results yielded the upper value of normal range at 16 weeks to be 216 ng/ml. Nevertheless, he also showed that there were no significant difference statistically between the result of the control group and the first trimester group at 16 weeks using two samples t-test23. Bear in mind, he also used a different D-dimer assay, IL D-dimer assay for testing all his participants. This study chooses to look for progressive changes in the level of D-dimer throughout the pregnancy by using the same sample for each trimester. The sample was taken at 16, 26, and 34 weeks. Only 48 pregnant women were taken aboard for this study but 34 healthy non-pregnant women were also recruited as the control group. The inclusion criteria were not mentioned except for the parity and the history of abnormal coagulopathy.
Another study that used IL D-dimer assay, Kovac M, et al also agrees with the reference range proposed by Morse for the second and the third trimester24. 89 patients were followed up from the first trimester up till 6 weeks postpartum with the same exclusion criteria like us with an addition a control group. The participants were tested with coagulation blood test before they were taken on board the study to ensure the validity of the study as they can be asymptomatic. The strength of this study is that it further compared the result of its study in pregnant patient with clinical suspicion of VTE. The reference range proposed by her is 286 ng/ml for the first trimester, 457 ng/ml in the second trimester and 644 ng/ml in the third trimester. There is also an increase up to 6.7-7.6 times of D-dimer level found in women who develop venous thromboembolism in the first trimester compared to the D-dimer level of the women in the reference group24.
In a Swiss study the 5th and 95th percentiles of D-dimer level was 139-602 ng/ml for the first trimester of pregnancy25. D-dimer is analysed using VIDAS D-dimer NEW assay. The blood samples were collected from 144 pregnant women with uncomplicated pregnancies but the inclusion criteria was not very stringent; patients with past Thromboembolic history were included. The number of sample for first trimester group (10-14 weeks) is only 27 which might not represent the true value of D-dimer in first trimester.
Another study conducted in 1994 in Italy revealed that by the end of the third trimester the level of D-dimer can reach up to 685 ng/L but the range of D-dimer values using ELISA assay in first trimester was only 43-211 ng/L26, hence there is no significant increase compared to the control group. The number of samples are much lesser compared to those stated above, where in total only 108 women and 26 patients were from the gestation 1-20 weeks. No demographical patterns of the patients were mentioned except that they were recruited by having a normal pregnancy. Therefore, the outcome of this study can be questioned further.
80% of total miscarriages happen in the first trimester. My study shows the rate of 19% which correlates to the normal rate of miscarriage in the population (15%-20%). Recurrent fetal loss is associated with anti-phospholipids syndrome, factor V Leiden and some other congenital or acquired thrombophilia. The management of recurrent fetal loss is via administration of aspirin or low molecular weight heparin even though the pathophysiology of recurrent fetal loss is still unclear. As pregnant women who have any history thrombophilia or clotting disorder and those who are on anti-coagulant and heparin were excluded, I tried to investigated on the difference of D-dimer level in relation to the status of pregnancy, either it is viable or not. The result shows that there is no significant association between D-dimer level and miscarriage. This correlate with a study done in Ukraine using ELISA to measure D-dimer where there is also no significant variation between the level of D-dimer in the normal pregnancy and the pregnancy with risk of fetal loss27. The Ukraine study has almost equal number of samples in the normal viable pregnancy group and the risk of fetal loss group. Bear in mind that the Ukraine study investigated the difference between normal pregnancy and pregnancy with risk of fetal loss not non-viable pregnancy. It was not mentioned in the study of the number of pregnancy that progress to miscarriage.
Based on the NICE guidelines, obesity (BMI >30 kg/m2), personal or family history of VTE, use of hormonal contraception and replacement therapy, and smoking will increase a person risk of developing VTE28. By looking for the association between D-dimer level and all these variables, I tried to eliminate possible confounding factors from this study. D-dimer is independent of all these variables except for use of oral contraception before pregnancy. The p value of Mann Whitney test between use of combined oral contraception and D-dimer level is 0.011. But the Spearman rank correlation coefficient value between combined oral contraceptives use and D-dimer level is only -0.20 means that the strength of the relationship is very little if any. The negative value of the correlation also shows reciprocal relationship between D-dimer value and combined oral contraception use. Thus, this will contradict the medical fact that usage of hormonal contraception will increase risk for venous thromboembolism.
The strength of my study would be that the size of my sample of 164 participants is larger compared to the previous studies on D-dimer level in the first trimester. My exclusion criteria are also quite stringent as we excluded women with high risk factor for VTE such as previous history of VTE and thrombophilia. Women who were on aspirin or low molecular weight heparin or treatment of recurrent miscarriages are excluded to ensure the validity of the result. Potential confounding factor was also eliminated by finding the variable which might has significant association and to eliminate them from the analysis. By using random sampling, my result can be said to represent the local population of pregnant women who are attending CUMH currently.
Among the limitations of my study would be that, due to time constraint I could not follow the patient up till delivery to further exclude patients with any complication in pregnancy such as pre-eclampsia or VTE itself. This study also focus on the first trimester of pregnancy, thus my result cannot be applied to the second and the third trimester of pregnancy. I also don't have a control group consist of non-pregnant women to compare my result with. Due to this limitation, I could not find the magnitude of increase of D-dimer level in the first trimester compared to the non-pregnant population. It would also be ideal to have the same sample size in the control group to further validate the range. I also did not perform a screening blood test for thrombophilia as they can be asymptomatic.
For future planning, a pilot study should be conducted on the new reference value on patients with clinical suspicion of VTE. Then, the sensitivity and specificity of D-dimer in the first trimester can be calculated by confirming the presence of VTE using diagnostic test such as Venous Doppler Ultrasound or Venography. It will best as well to conduct the improvised study on the second trimester, the third trimester and the post natal period as a continuation to my study. This new range can only be applied in the local hospital of the local population in Cork University Hospital. Comparison of numerical values in future studies can only be done if the same D-dimer assay and population were used.
In conclusion, it was shown that VTE occurs in first trimester of pregnancy, were often missed and can be fatal. The patients with DVT in first trimester are usually asymptomatic. It is very important to know if any non-invasive tests that carries no radiation risk to the fetus can be used as the screening or exclusion test for DVT to decrease the maternal mortality rate due to VTE. The new cut-off value of 429.5 ng/ml will improvise the recognition of venous thromboembolism related problem in the first trimester of pregnancy by reducing the false positive rate of D-dimer test. Patient with low clinical probability of VTE, D-dimer test with the new cut-off value will be useful in ruling out VTE but patient with high suspicion of VTE, radiological investigation must still be performed. Nevertheless, this will reduce the amount of invasive radiological investigation in the first trimester and also improve the diagnostic process of VTE in the first trimester of pregnancy.