GS: <No, you are not correct. The results are not poor because it is early stage cancer. The following is in the same study for stage IV cancers, i.e, late stage.
Stage IV (D) sensitivity 68% specificity 97% cancers vs benign You <Stage IV (D) sensitivity 68% specificity 97% cancers vs benign >
Your mistake is to assume that the WORSE set of results is the right one. Common sense indicates that you cannot have better results than what is possible and therefore, if two studies differ in performance, the best one - not the worst one - has to be considered as the attainable performance of the test.
If you measure the max speed of a car and in one trial you get 200 mph and in the next you get 180 mph you cannot say "this car can only do 180 mph", you have to say that the car CAN go up to 200 mph. Same with this test.
If the study in 2006 showed 68% sensitivity for stage IV that means that the test performed below its attainable performance (as shown in 2004).Thus, at the attainable performance, the sensitivity at ALL stages will be improved which means that the sensitivity at EARLY stages can only be higher.
That is excellent news, thank you for pointing it out.