Free Solocat Essay Example

SUMMER PROGRAM (2014)

CAT Domain: Diagnosis
I hereby declare that this soloCAT exclusively contains original work
(Signature on hardcopy only – use ballpoint).

PDF of the article – import here

Clinical scenario:
A 62 year old Scandinavian man arrives in the emergency department of AZMC in Maastricht, complaining of pain in his right leg and palpitations. He tells me that he has just arrived on an especially cramped flight from Riyadh, Saudi Arabia where he could not move his legs. During the flight, he felt a pain in his right leg which he attributed to a muscle cramp. After landing, he noticed that the pain would not subside, and he had the sensation of his heart racing in his chest. His wife insisted on bringing him to the hospital.
He states that the pain is a deep throbbing that he has never had before, and denies any significant past medical history, including coronary artery disease, cancer, diabetes, hypertension, or stroke. He is a smoker, five cigarettes /day for five years, but has been trying to cut back, and only drinks socially.
On exam he is an overweight man of Swedish descent. On presentation BMI is 31 kg/m², blood pressure is 130/90 mmHg, heart rate 96 x’, and oxygen saturation is 100%. His right thigh appears slightly swollen compared to the left, and is tender on palpation. There is no redness or induration. Popliteal and femoral pulses are 2+ bilaterally. He has full strength and range of motion in all extremities. His lungs are clear to auscultation. Aside from tachycardia his heart sounds are normal. An abdominal and HEENT exam are also normal.
According to many studies in PubMed (CITATION NEEDED), the standard test to rule in/out DVT would be venography. This test is invasive, expensive and time-consuming tests. However, D-Dimer assay is noninvasive, relatively faster and cheaper.
While it is generally accepted that a low probability assessment and a negative D-dimer will exclude up to 50% of suspected DVT’s in the outpatient setting. I am wondering whether I can rely on clinical assessment and D-Dimer assay to rule out DVT, for population with higher disease prevalence like this patient, with enough certainty or not.
The potential risks of treatment and misdiagnosis of a DVT are such that false positives or false negatives may be fatal. An undiagnosed DVT may lead to a pulmonary embolism that could leads to death, while treatment regimens that generally involve anti-coagulant therapy may lead to bleeding complications (CITATION NEEDED). Furthermore, Inaccuracies in screening carry a high cost (CITATION NEEDED).

Clinical question:

Is the negative predictive value of a negative D-Dimer test high enough to exclude a DVT with enough certainty in a 62 year old Scandinavian male smoker, who has an intermediate probability for DVT due to tenderness and swelling, as it would be with a diagnosis using compression ultrasonography (CUS) or contrast venography ?

Literature search:

Disorder: Deep vein thrombosis (DVT)
Test: D-dimer
Methodological terms: pre-test probability, sensitivity, specificity
Patient Characteristics: male, smoker, Scandinavian descent, intermediate probability (tenderness and swelling).
“Reference test”: compression ultrasonography (CUS) or contrast venography
Search (((((((((suspected DVT) OR DVT) OR Deep vein thrombosis) OR ((Diagnosis/Broad[filter]) AND ("Venous Thrombosis"[Mesh]))) OR "Venous Thrombosis"[Mesh])) AND ((((Clinical probability assessment) OR D-dimer determination) OR D-dimer) OR "fibrin fragment D"[Supplementary Concept])) AND ((((pre-test probability) OR ("Sensitivity and Specificity"[Mesh])) OR ((Sensitivity and Specificity))) OR "Prospective Studies"[Mesh])) AND ((((((((((("Male"[Mesh]) OR male) OR smoking) OR smoker) OR "Smoking"[Mesh]) OR cigarette smoking) OR Scandinavian descent) OR intermediate probability) OR leg pain) OR tenderness) OR swelling leg)) AND multicenter management study Filters: Full text; published in the last 10 years; English; Male; Aged: 65+ years

Article selected:

J Elf , K Strandberg , C Nilsson , P Svensson . Clinical probability assessment and D-dimer determination in patients with suspected deep vein thrombosis, a prospective multicenter management study. Thrombosis Research. 2009 Feb; 123 (4):612-616
Brief motivation of article selected: Article number 31. This article deals specifically with the effect of a higher prevalence of DVT in the outpatient population on the effectiveness of a clinical assessment score and D-dimer test as (index test). Also, the characteristics of the study population, mean age of 62, 138 (39%) were men, 66 (18%) smokers, BMI ranging from 21 to 33, 141 (39%) have intermediate probability for DVT, which all match with our scenario patient. Moreover, this is a large multi-centered study involving over 350 patients in seven hospitals resulting in narrow CI’s narrow, so it could generate precise results. Furthermore, it is accepted at 2008, although it is not considered recent, it is acceptable. This article is also in English, which is my preferred language. However, 37 articles were excluded either because they do not discuss DVT as a target disease (such as pulmonary embolism, DVT in upper extremities and thrombophilia), they do not discuss clinical assessment and D-dimer test as an index test, they addressed diagnosis in a post-operative setting, the study population uses medications such as NSAIDs or suffering from acute stroke, or did not investigate the effect of a higher prevalence population (Japan). Chiefly, I believe that this article is the most suitable article for my clinical question.

Critical Appraisal of the Article: “look to the appendix for more details”

(-)Reference tests would include contrast venography or compression ultrasonography. There was no direct comparison of D-dimer assay to a reference test. Instead, d-dimer was used to exclude patients in the low probability group from receiving further workup. In a sense, d-dimer was its own reference test.
(+)The study inclusion/exclusion criteria are that the patients be outpatients suspected of having their first DVT. The simplicity of this criterion helps eliminate many confounding variables and make the study widely generalizable. The study was done using a large population of patients (n=357) and the width of the confidence intervals is fairly narrow.
(+) Work up bias was avoided by performing the study in a prospective fashion and utilizing the Wells score to set clear guidelines for management. With definitive criteria there is little room for individual bias. All patients enter the study at the same point, e.g. as outpatients with, potentially, their first DVT. All patients are then submitted to the same tests, a clinical assessment in the form of a Wells score, and a D-dimer assay. There is no preference for one test over another or a tendency to alter workup based on clinical conditions.
(+) There is little risk of expectation bias as the D-dimer assay is performed separately from the Wells score. The emergency room physician’s expectation could not influence the result. The study is also done a prospective basis making it difficult for expectation to influence future results.
(-)The reliability and reproducibility of the locally used D-dimer was compared to the standard assay and determined to be adequate at 89% concordance. Reproducibility and inter-rater reliability of the Wells score is mentioned as a potentially issue during the authors’ discussion given the limited experience their physicians had with the scoring system.
(+) The study population is large and the resulting CI’s narrow. Narrow CI’s indicate a low degree of variability and strong confidence in the high sensitivities and specificities found. In particular the NPV for the low probability assessment plus a negative D-dimer approaches 100% making it extremely useful as a screening strategy.
(+) The optimal cutoff point was found to be a low probability Wells score with a negative d-dimer. At this point patients could be excluded. No further testing or treatment needed. This led to an exclusion rate of 30% allowing physicians to focus their resources elsewhere and allowing patients to avoid unnecessary interventions.
(+)The test was carried out at the initial point in the diagnostic process. Patients were assessed upon presentation and D -dimer tested upon receiving a low probability score. This is comparable to the real-world use of this strategy and is the optimal time to employ a screening strategy.
(+) Since the results of this study were consistent with other known results the study is considered valid.

Evidence Table: Results for Wells score assessment and D-dimer assay, all CI’s 95%

The strength of this screening strategy is the high negative predictive value, 98.2% for low probability patients and 94.3% for high probability patients. This allows the test to exclude many patients from further work-up and focus resources on patients more likely to benefit. The strategy is not perfect. The authors noted that one patient with both a low probability score and a negative D-dimer did in fact have a DVT. This DVT, however, was found on imaging and did not require treatment.
The 2x2 tables detail the diagnostic values and distribution of results for the study, the theoretical case of 20% prior probability, and the theoretical case of 15% prior probability. The negative diagnostic value in all three cases is over 90%. This indicates that the screening regimen would be effective on different populations with different pre-test probabilities. This is expected given prior literature.

Comments:

This is a high quality article because it has a large study population and was done on a prospective basis. This produced high fidelity results with narrow CI ranges. The overall power and methodology of this study were sufficient to find the combination of assessment score and D-dimer assay safe and effective with a high degree of confidence
Deep vein thrombosis can produce serious complications if left untreated. Conversely unnecessary treatment also puts the patient at mortal risk. It is therefore necessary to have accurate reliable tests to determine which patients require treatment and which do not. The combination of clinical assessment and D-dimer has proven able to exclude 30-50% of patients with first suspected DVT in some populations. However, the prevalence of DVT varies with different populations. It is expected that a higher prevalence might degrade the screening tests ability to safe separate true negatives from false negatives. This study is designed to address this issue.
Our scenario patient fits with the study population in this article as he is a male, in his 60’s, with no previous history of DVT, and is from a population with higher prevalence than populations previously studied. He does not have any significant medical history; however, he does have risk factors for deep vein thrombosis. These risk factors include obesity and a history of smoking.
The study successfully used the Wells score to determine the pretest probability of DVT. With minimal training, physicians were able to distinguish low risk from intermediate and high risk patients. Combined with a negative D-dimer, this produced a 30% exclusion rate. Exclusion keeps these patients from being exposed to unnecessary tests and potentially unneeded anticoagulant therapy.
As the D-dimer test is non-invasive, it is not a significant inconvenience to the patient. If the D-dimer is positive the patients are categorized as intermediate probability and undergo imaging. The D-dimer’s true value is in ruling out patients who do not have DVT. The test has a high negative diagnostic value. This means that the test is effective at eliminating patients without disease from requiring further workup on treatment, screening negatives from potential positives.
The study has clear inclusion and exclusion criteria which help reduce the variability in the results and make them more reliable. In addition, it makes the results generalizable to high prevalence populations. The study is consistent with other similar studies. This provides internal validation. A number of earlier studies have demonstrated the utility of the D-dimer test under various conditions. Concern regarding the screening test’s fidelity with increasing pre-test probability prompted the current study.
Because of the high negative predictive value of the combination of assessment and D -dimer, the generalizability of the study, and its prospective nature, this article is suitable evidence for the management of our patient in this study. The diagnostic value of the test at different theoretical pre-test probabilities, indicates that the test could be used in various population settings.

Bottom line:

The combination of low probability assessment and negative D-dimer excluded approximately 30% of patients who did not require treatment or imaging. This combination was demonstrated to be safe and effective in a high prevalence population. With a negative predictive value of 98.2% (CI, 94 – 100) this is a safe and effective screening strategy ready for widespread usage.

Appendix

Criterion 1:
Reference tests would include contrast venography or compression ultrasonography. There was no direct comparison of D-dimer assay to a reference test. Instead, D-dimer was used to exclude patients in the low probability group from receiving further workup. Patients with a low probability assessment and a negative D-dimer were followed prospectively for any complications and recurrence of symptoms. The authors were comparing the use of D-dimer and clinical assessment in the diagnosis DVT for their population to those generated in other countries. In a sense, D-dimer was its own reference test.
The D-dimer test had been established as a screening test in other populations. However, there were concerns regarding the effectiveness of the test in populations thought to have higher prevalence of disease. Scandinavia is one of those populations. The authors sought to address this issue.
Criterion 2:
Patients in this study presented with their first suspected DVT in an outpatient setting. The simplicity of this criteria helps eliminate many confounding variables and make the study widely generalizable. Symptoms included tenderness, swelling in the extremity, and pitting edema. Given the relatively high prevalence of DVT in this particular population the diagnosis may be expected but is unknown given that the first suspected case. Consequently, the test is used in the correct domain. The mean age of 62, 138 (39%) were men, 66 (18%) smokers, BMI ranging from 21 to 33, and 141 (39%) have intermediate probability for DVT.
The study was done using a large population of patients (n=357). Spread over seven regional hospitals. The width of the confidence intervals is fairly narrow. Therefore, it could generate precise results. The overall power and methodology of this study were sufficient to find utility in combining D-dimer with clinical assessment.
Criterion 3:
Work up bias was avoided by performing the study in a prospective fashion and utilizing the Wells score to set clear guidelines for management. With definitive criteria there is little room for individual bias. All patients enter the study at the same point, e.g. as outpatients with, potentially, their first DVT. All patients are then submitted to the same tests, a clinical assessment in the form of a Wells score, and a D -dimer assay.
Criterion 4:
There is little risk of expectation bias as the D-dimer assay is performed separately from the Wells score. The emergency room physician’s expectation could not influence the result. The study is also done a prospective basis making it difficult for expectation to influence future results. There was concern for inclusion bias as the emergency room physician determined which patients were enrolled in the study. However, the physicians involved had limited familiarity with the Wells score which would limit this bias. Over time, as the probability of this bias increased with the success of the study, the study stopped enrolling patients. Patients were followed by either phone interview, return visit to a different physician, or tracked through medical records. This makes it unlikely for expectation to interfere.
Criterion 5:
The reliability and reproducibility of the locally used D-dimer was compared to the standard assay and determined to be adequate at 89% concordance. Reproducibility and inter-rater reliability of the Wells score is mentioned as potentially issue during the authors’ discussion given the limited experience their physicians had with the scoring system. There is the possibility for a great deal of variation between different scorers particularly in the low and intermediate probability groups. However, the consistency of the results, in particular that only 1 low probability patient had a DVT on follow up, indicates that inter-rate reliability was not an issue. This is likely due to the relative ease of use of the scoring system.
Criterion 6:
The study population is large and the resulting CI’s narrow. Narrow CI’s indicate a low degree of variability and strong confidence in the high sensitivities and specificities found. In particular the NPV for the low probability assessment plus a negative D-dimer approaches 100% making it extremely useful as a screening strategy. The narrow CI’s indicate a great deal of certainty and reliability in the result. This is a high-powered study from a statistical standpoint due to its large study population.
Criterion 7:
Wells score was used to estimate pre-test probabilities and combined with the d-dimer to determine management. Wells score divided patients into 3 groups, low, intermediate, and high probability. The optimal cutoff point was found to be a low probability Wells score with a negative d-dimer. At this point patients could be excluded. No further testing or treatment needed. This led to an exclusion rate of 30% allowing physicians to focus their resources elsewhere and allowing patients to avoid unnecessary interventions. The ability to exclude true negatives is of critical importance in any screening test and this strategy was very successful in this regard.
Criterion 8:
The test was carried out at the initial point in the diagnostic process. Patients were assessed upon presentation and d-dimer tested upon receiving a low probability score. This is comparable to the real-world use of this strategy and is the optimal time to employ a screening strategy.
The test is highly comparable to the clinical scenario. An outpatient with their first suspected DVT is the focus of the study and the description of the patient in the scenario.
Criterion 9:
Sensitivity and specificity of the low probability/negative d-dimer combination are relatively high making this a strong screening strategy. The combination of a sensitivity of 85.7% (CI, 57-98%) and specificity of 74.4% (CI, 67-81%) create a test with a high NPV which is most useful in a screening test. The test is internally validated by the results. Only one patient excluded by low probability assessment and negative d-dimer was found to have DVT on follow-up, and that DVT did not require treatment. The test is further validated by its consistency with similar studies done in Scandinavia. Since the results of this study were consistent with other known results the study is considered valid.
Se= 12/13 =92.3% PV- = 109/110 =99.1%
Sp = 109/146 = 74.65% PV+ = 12/49 = 24.5%
Applicability of D -dimer test:
Is D-dimer test identifying a treatable disease?
Deep vein thrombosis is a treatable condition. Anticoagulation can be used to resolve the blood clot. Failure to identify and a DVT may lead to a more serious conditions such as pulmonary embolism (PE). A PE is a potentially fatal complication.
What is the additional value of D-dimer test compared to the test already used?
D-dimer test can be used fast and inexpensive. The study’s negative predictive value rules out disease in low probability patients. This prevents these patients from receiving unnecessary tests or treatments.
Is D-dimer test expensive?
No it is not. It is cheaper than other tests such as compression ultrasonography (CUS) or contrast venography. One of the main values of the D-dimer as a screening test is that it is inexpensive.
Will D-dimer test be accepted by the patient?
Yes, the D-dimer test is a quick and painless blood test. It is administered in much the same way one tests for electrolyte levels. The patient is as likely to consent to this test as any other blood test. It is not invasive.
Can D-dimer test make a difference with regard to the medical procedure?
Yes, if the result of D-dimer test was negative, the patient would not be undergone further diagnostics since DVT is excluded with enough certainty. It can shorten work up and eliminate unnecessary tests in potential DVT patients who test negative.