FPIN Journal Club
SYSTEMATIC REVIEW WORKSHEET
SPEAKER NOTES
Title:It’s time to use an age-based approach to D-dimer
Journal Club Author: Corey Lyon, DO University of Colorado Department of Family Medicine
PURL Citation:Urban K, Kirley K, Stevermer JJ. PURLs: It's time to use an age-based approach
to D-dimer. J Fam Pract. 2014 Mar;63(3):155-8. PubMed PMID: 24701602.
Original Article:Schouten HJ, Geersing GJ, Koek HL, Zuithoff NP, Janssen KJ, Douma RA, vanDelden JJ, Moons KG, Reitsma JB. Diagnostic accuracy of conventional or age adjusted D-dimer cut-off values in older patients with suspected venous thromboembolism: systematic review and meta-analysis. BMJ. 2013 May 3;346:f2492. doi: 10.1136/bmj.f2492. PubMed PMID: 23645857
Definitions
Systematic review:
A review in which evidence on a topic or research question has beensystematically identified, appraised and summarized according topredetermined criteria.
Meta-analysis:
A statistical technique. Summarizes the results of several studies intoa single estimate, giving more weight to larger studies.
1. What question did the study attempt to answer?
Patients - Patients over 50 years with suspected VTE but with non-high clinical risk for VTE
Intervention - age adjusted (age X 10 ug/L) D-dimer
Comparison - conventional D-dimer (500 ug/L)
Outcome–sensitivity/specificity in diagnosis of DVT/VTE
Did the study address an appropriate and clearly focused question Yes No
2. Determining relevance:
a. Did the authors study a clinically meaningful
and/or a patient oriented outcome? Yes No
This is a diagnostic study, so sensitivity and specificity are the primary outcomes. However, the authors employ a hypothetical cohort to estimate the number of unnecessary imaging studies that are avoided and the number of cases of VTE that would be missed using age-specific cut-offs.
b. The patients covered by the review similar to your population Yes No
Patients >50 with non-high clinical risk for VTE are commonly seen in the ED or primary care setting.
3. Determining validity:
a. What type of studies are included in the review?
Diagnostic retrospective cohort trials; only trials that enrolled consecutive patients
b. The literature search is sufficiently rigorous to identify all the relevant studies?
Look for Yes No
· Which bibliographic databases were used – Embase and Medline
· Follow up from reference lists
· Personal contact with experts
· Search for unpublished as well as published studies
· Search for non-English language studies – yes; translated
c. Did the review’s authors do enough to assess the
quality of the included studies? Yes No
· The authors need to consider the rigor of the studies
they have identified.
Detailed description provided on quality assessment and risk of bias assessment
d. If the results of the review have been combined, was it
reasonable to do so? Yes No
Consider whether
· The results were similar from study to study
· The results of all the included studies are clearly
displayed
· The results of the different studies are similar
· The reasons for any variations are discussed
4. What are the results?
a. What is the overall result of the review?
5 trials with 13 cohorts and over 22,000 patients included in the review.
Using the conventional cut-off values; the pooled specificity of D-dimer testing decreased substantially with increasing age from 66.8% in patients aged less than 50 years to 14.7% in patients aged more than 80 years when the conventional cut-off value was applied .
The pooled sensitivity was not different between the age groups.
Review Table 2
Specificities for conventional cut-off versus age-specific cut-off:
-age 51-60: 57.6% v 62.3%, P= 0.005
-age 61-70: 39.4% v. 49.5%, P<0.001
-age 71-80: 24.5% v. 44.2%, P<0.001
-age >80: 14.7% v. 35.2%, P<0.001
Sensitivities decreased slightly with age-specific cut offs, but remained greater than 97% and was not significantly different.
Specificities still remain low, but notice how it doubles in patients > 80 years old.
Using a hypothetical model, the authors calculated that applying age adjusted cutoff values would exclude VTE in 303/1000 patients >80 years, compared with 124/1000 when using the conventional cutoff.
b. Consider;
Are we clear about the reviews ‘bottom line’ results Yes No
Are the results presented with confidence intervals, Yes No
NNT, odds ratio, etc
Presented in sensitivity and specificity.
5. Applying the evidence:
a. If the findings are valid and relevant, will this change
your current practice? Yes No
b. Is the change in practice something that can be done in
a medical care setting of a family physician? Yes No
c. Can the results be implemented? Yes No
d. Are there any barrier to immediate implementation? Yes No
You will have to do the calculations. Physicians can easily apply an age-adjusted D-dimer cutoff as they interpret lab results by multiplying the patient’s age in years × 10 mcg/L. This is not something that the labs will do for you.
Of note, this is only for patients at low risk of DVT.
6. Teaching Points
Understanding sens/spec/likelihood ratios;
Sensitivity - SnOut (Sensitivity rules Out dz) – so a high sensitivity tells us if the test is negative, we can believe that the patient doesn’t have the disease
Specificity - SpIn (Specificity rules In disease) – so a high specific test tells us that if the test is positive, we can believe the patient has the disease (or in this case, IPV)
With D-dimer, we have a high sensitivity, which means if it is negative, we can feel comfortable that the patient does not have a DVT and no additional testing needed.
By adjusting the cut off values for a “positive” result (by increasing the cut off value), we will potentially have more patients that have a “negative” test and will not require additional testing.
Although, not presented in this study, it would be good to review likelihood ratios, as these are becoming more and more common in the literature, and they can be easier to understand.
Likelihood ratios
These ratios tell you how much the odds of a disease increase when a test is positive (LR+) or how much the odds of a disease decrease when a test is negative (LR-)
-Likelihood ratios are independent of prevalence
-Can be calculated from the sensitivity and specificity (formulas below)
Positive likelihood ratio -– How likely is the disease present if a test is positive
- It should be > 1, the higher the better
- LR + > 10 is an ideal test
- LR+ < 2 is no change
- You can calculate it out from the sens/spec; LR+ = sens / (1-spec)
Negative likelihood ratio– How likely is the disease not present with a negative test
- It should be < 1, the lower the better
- LR- < 0.1 is an ideal test
- LR- > 0.5 is no change
- LR− = (1-sens) / spec
From our study;
In patients ages 51-60; the age adjusted specificity was 62.3% and sensitivity was 99.4
LR + = 0.994/(1-0.623) = 2.6
LR - = (1-0.994)/0.623 – 0.01
So, here, you can see if the test is negative (below the cutoff point), we can feel confident that the patient does not have a DVT.
If the value is above the cutoff point, the LR is low, so we will need to do additional testing to rule in, or rule out a DVT.