Regardless of whether you want to assess the effect of an intervention or rather would like to investigate the effectiveness of a new diagnostic test, the universal rule is, that your question needs to be as specific as possible.
The question that you want to ask, needs to specify the patient group, the diagnostic test (s), the reference standard (golden standard) and the outcomes. This is abbreviated into PIRO (Population, Index test, Reference standard and Outcome).
Population (patient group)
This should include a detailed description of the population, including the specific disease (e.g severity, symptoms) and relevant demographic factors (e.g age and gender). If relevant, you should include in which context, the test should be used (e.g private praksis, hospital). The precision and balance of positive and adverse effects of the test may variate, depending on the context in which the test is being applied (e.g primary, secondary or tertiary sector). You may also include an estimate of disease prevalence when assessed in different clinical situations.
Which test (s) do you want to evaluate? Here you should include the diagnostic thresholds (when is someone considered sick versus healthy?), who should apply and interpreate the test and under which conditions is the test being be applied (in the laboratory or clinic?)
What is the reference standard (golden standard)? The reference standard is the test (s), which are currently used to determine the absence or presence of the specific disease. It is considered the most reliable procedure for the given disease. It is a MUST that the reference standard is established, since this is crucial when wanting to assess the sensitivity and specificity of the new test being investigated.
If there is no reference standard or if this is uncertain, then the assessment of the test can only be done based on patient relevant outcome. In this case, you need to apply the PICO format.
Diagnostic precision: You need to consider which impact the diagnostic precision (fake positive/negative) has on psychological and prognostic factors as well as further treatment. If the reference standard is uncertain, then assessing the precision of a new test may be deceptive and has to be interpreted with great caution, since you are lacking a test that you can compare the precision of your new test to.
What is essential to know in order to evaluate the positive and adverse effects of the new diagnostic test? The outcome investigated should be determined on two levels:
- Patient relevant outcome
- The diagnostic precision
The must optimal would be to assess patient relevant outcome, however this is rarely possible, thus the value of the diagnostic test may be evaluated using surrogate outcomes, including sensitivity and specificity. A tests sensitivity is the ability of a test to correctly identify those with the disease (true positive), whereas test specificity is the ability of the test to correctly identify those without the disease (true negative). This does not necessarily require randomized studies, but can be quantified using non-randomized studies, in which the patients initially are evaluated using the index test (new test) followed by the already exsisting test (reference standard). If you find that the exsisting test increases patient relevant outcome, then applying this test would be the most ethical.
What is the role of the new test?
You need to consider the role of the new test as well as the risk of overdiagnosing. Should the new test replace an existing test, be a part of a triage or rather an add-on.
- Replacement: The new test may be cheeper, easier or quicker to perform, have less adverse events ect.
- Triage: This includes a test that determines which patients should be offered follow-up tests. It may be relevant, if the follow-up test has adverse effects, be expensive or less accessible.
- Add-on: An additional test that may improve precision of exsisting methods.
When looking into a new diagnostic test, you need to assess the possibility of overdiagnosis. Have a look at this blog post that talks about diagnostic precision.