Family Practice Advance Access originally published online on July 8, 2005
Family Practice 2005 22(5):579; doi:10.1093/fampra/cmi062
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Correspondence |
Limited value of patient history and physical examination in diagnosing deep vein thrombosis in primary care
Department of General Practice/Family Medicine, University of Marburg/Germany, D-35033 Marburg, Email: norbert{at}med.uni-marburg.de
Ruud Oudega and colleagues are to be lauded for their paper1 on the effectiveness of the history and physical findings in diagnosing deep vein thrombosis (DVT). Contrary to previous work they have recruited a sufficient number of consecutive patients at the primary care level to examine the diagnostic performance of 17 different predictors. They conclude, that these do not differentiate to an acceptable degree between patients who have DVT and those who have not.
However, I would like to take issue with this conclusion. I would even suggest that their data show how well the history and physical signs predict the occurrence of DVT.
The overall prevalence of DVT in their sample was 29% which is well in the range of secondary or tertiary care samples. Given that the study was conducted in general practice this is an very high proportion. I can imagine that the diagnosis was—however remotely—considered in a much larger number of patients. The participating GPs have succeeded in raising the prevalence of DVT from below 1% (my guess for the prevalence of DVT among all practice attenders) to 29% by selecting patients for inclusion in the study. Of course the only criteria available to them were the history and physical findings.
It should not come as a surprise, that in the sample thus generated the discrimination by historical and physical items was poor. One could say, that these had done their job already. In fact, participating GPs have produced a sample in which the traditional diagnostic means had been exhausted and further tests, such as D-dimer, were indicated.
Rather than specifying "suspected DVT" as the entry criterion, which seems to represent the end rather than the start of GPs' reasoning process, patients should had been selected on the basis of a specific symptom, e.g. pain in one leg or swelling (or both), irrespective of the probability of DVT assumed by the GP.2
I agree with the authors that the history and physical findings will not give a definite answer in a large proportion of patients. However, to define the population in need of further testing there is only one means—the history and physical findings.
References
1 Oudega R, Moons KGM, Hoes AW. Limited value of patient history and physical examination in diagnosing deep vein thrombosis in primary care. Fam Pract 2005; 22: 86–91.
2 Donner-Banzhoff N, Kunz R, Rosser WW. Studies of symptoms in primary care. Fam Pract 2001; 18: 33–38.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||