Assessing the accuracy and cost-effectiveness of diagnostic tests
This study aims to assess the methods used to combine information from studies that investigate the accuracy of tests used in the diagnosis of disease. It also investigates how this information is used to determine whether the test is cost-effective. For tests used in primary care, the study will assess whether the evidence is based on studies carried out in primary care.
Why is this important:
Diagnosing conditions in primary care is often difficult, as many serious diseases are rare and patients may visit their GP before obvious disease symptoms have occurred. Diagnostic tests often do not work well for patients at early stages of disease. Sometimes the information about a particular test comes from studies carried out on patients in hospitals, where disease is more common and will usually have progressed further.
It is important to assess the accuracy and cost of different diagnostic tests so that doctors can use the one that is most appropriate for their patients.
There is often information about the accuracy of diagnostic tests in several different studies, carried out in different settings. The results of these studies need to be combined to provide information about the accuracy of the test overall. The accuracy of the test – how many people receive a correct diagnosis compared to how many receive an incorrect diagnosis – will also affect whether the test is cost-effective. It is important to assess the accuracy and cost of different diagnostic tests so that doctors can use the one that is most appropriate for their patients.
In this study, we use reports from the National Institute for Health Research Health Technology Assessment (HTA) library. We will assess reports that describe the accuracy and cost-effectiveness of tests used to diagnose disease.
Although statistical methods to combine information from different studies (meta-analysis) are widely used in studies that compare treatments, alternative methods are required in studies of diagnostic accuracy. For example, the predictive value of a test depends on the prevalence of the condition in question, and measures such as sensitivity and specificity depend on the threshold chosen to define a test result as positive.
We will summarise which statistical meta-analysis methods were used in the HTA reports, and why they were chosen. We will also identify which economic decision models were used in the reports, whether they were appropriate, and whether they used the diagnostic accuracy results appropriately. Tests used in primary care will be examined separately to investigate whether they are based on evidence obtained from the primary care setting.
How this could benefit patients:
The results will help to guide future studies of diagnostic tests and the methodologies required to support them. It will assess whether tests carried out in primary care are supported by a strong evidence base, and identify areas where further research is required to ensure that diagnostic tests are used appropriately.
We are currently writing up the results of this study for publication.