Trusting Diagnostic Test Results
As healthcare providers, we routinely rely on monitors, lab results, imaging studies, and a myriad of other diagnostic tests to diagnose new medical problems and monitor responses to therapy. Before interpreting the conclusions or results of these tests, we must ask a simple question – can you trust the test?
This applies to anything! Are low voltages on an electrocardiogram (ECG) due to pericardial effusion, an obese patient? Are areas of atelectasis on a chest x-ray (CXR) due to poor inspiratory effort when the x-ray was captured? Is my pulse oximeter reading 85% because of methemoglobinemia or nail-polish? Does the complete blood count (CBC) show a low platelet count because of isolated thrombocytopenia or platelet clumping causing a lab artifact? Is the thyroid stimulating hormone (TSH) a little low because of subclinical hyperthyroidism or because the patient is pregnant?
Every diagnostic test has implicit error related to the modality. This error is compounded by other variables like suboptimal conditions (ie, morbidly obese patients having poor transthoracic echo windows) and procedural errors (ie, poor lead placement on an ECG).
Please consider the results of your tests in terms of likelihood and probabilities in the context of your patients’ comorbidities, and always account for confounders!