Skip to main content

Table 1 Labels grouped by DTA-specificity

From: Use of test accuracy study design labels in NICE’s diagnostic guidance

Label Interpretation Similar/synonym labels
1. DTA-specific labels (n = 19)
 1.1 DTA-informative (n = 6)
  Comparative performance Comparison of minimum two index test/strategies [3]. Comparative (diagnostic) accuracy studies
  Direct comparisons1 Comparison of  two index test/strategies within the same group/population (i.e. with the same set of eligibility criteria). Head to head comparisons; within-study comparisons
  (Prospective cohort) randomized to either the index test or comparator2 Randomization to either the index test or reference standard/comparator. Randomized head-to-head comparisons
  (Prospective cohort) receiving the comparator and at least one index test with follow up3 All participants receive both an index test and the reference standard/comparator and are followed up for a final outcome measure/status.  
  Reference standard positive studies Only participants with a positive reference standard test receive the index test.  
  Single/two-gate diagnostic studies Including one (single set of eligibility criteria) or two/multiple groups (separate sets of eligibility criteria) for comparison [9]. Two-study cohorts
 1.2 DTA-descriptors (n = 9):
  Single/multi-centre study Participants are sampled from a single or from multiple centres. Large, multi-centre prospective (UK) study; two-centre
  Primary care study Participants are sampled from settings in the primary (vs. secondary or tertiary) care.  
  Consecutive recruitment Participants are sampled consecutively (vs. random or convenient sampling) [2].  
  Symptomatic study Including only symptomatic participants.  
  (Single-gate studies) recruiting populations at high risk Including only high risk participants.  
  Population (-based) studies Including/sampling participants that represent the target/study population [9]. (Multi-centre) community-based study
  Pro- or retrospective studies Data collection (or recruitment) was planned before (prospective study) the index test and reference standard were performed or after (retrospective study) (STARD) [7]. Prospective cohort studies; prospective, consecutive cohort study; prospective, international multicentre
  Retrospective analysis (of prospective database) Post-hoc or non-pre-specified statistical analysis.  
  Discordant case analysis4 A study  that excludes discordant cases in the analysis.  
 1.3 General DTA labels (n = 4)
  (Diagnostic) cohort studies A diagnostic cohort study in which “individuals are enrolled before the final outcome (presence or absence of the target condition) is known” [9] (i.e. a single cohort/group of suspected participants). (Retrospective) cohort studies
  (Diagnostic) case-control studies/design “(…) studies in which the disease status is already known before the index test is performed”.
“(…) the reference test is applied only to a subsample of the participants with or without the target condition”.
  (Diagnostic) cross-sectional studies Comparison of the result of the index test with that of the reference standard in the same participant at the same time.
  (Diagnostic) observational studies A DTA-study that only make use of observational data e.g. re-interpretation/review of existing tests results.  
2. Umbrella terms (n = 4):
 Diagnostic accuracy studies A study reporting diagnostic accuracy performance measures. (Large prospective cohort) studies collecting diagnostic accuracy data
End-to-end studies A diagnostic accuracy study that directly links test results to patient outcomes (such as RCTs) [8].  
Long-term studies (following patients for several years) A study  that includes follow-up of any given outcome measure/status.  
 Pilot studies A study that is testing/piloting a setting/feasibility, often with a small sample size.  
3. Non-DTA-specific labels (n = 9)
 Test-treat trials A study in which a test guides the treatment strategy.  
 Controlled clinical trials A clinical trial including a control (comparator) group to the intervention group.  
 Cluster-randomized controlled trial A study that randomizes clusters of participants to either intervention or control group.  
 Derivation study A study that focuses on the development of a prediction model (contra validation of a model).  
 Observational studies A study that only includes observational data (non-intervention data).  
 (Observational) cohort studies A study using data from all cohort-members (in contrast to case-control study that uses data from an outcome-selected subset of the source population) [5, 10].  
 Prospective, two cohorts (feasibility, validation) A study with two separate cohorts recruited prospectively.  
Randomized controlled trials (RCTs) A study with randomization of participants to either intervention group(s) or a control (comparator) group (vs. non-randomized). (Large multi-centre) RCT
 Validation study A study that focuses on validation (of an existing/developed model).  
4. Unclear labels (n = 3)
 Diagnostic studies with a control group
 (Multi-centre) tracker study
 Mixed design (of within-study comparisons)
  1. The table shows the classification of the labels identified and their most likely interpretations. The labels in italic are labels (also) used in the guidance
  2. 1Elsewhere, other possible interpretation could be comparison of index test and reference test within the same patient (i.e. fully paired) [11]
  3. 2In DTA terminology, the term “comparator” usually refers to one of two index tests (in a comparative study). However, in this case, the “comparator” seems to refer to other tests that are used for the comparison of the index tests (which is used when a commonly accepted/implemented reference standard is absent) [12]. Other possible interpretation could be randomization to one of several index tests or randomization to two different test-sequence strategies
  4. 3In other contexts of DTA-studies, “follow-up” can imply the clinical follow-up compose the reference standard itself (i.e. the final outcome, diagnosis, is established by follow-up). This is also referred to as a “delayed type cross-sectional study” [13]. Yet, it can also mean that any given reference standard is performed at a later time point (delayed) than the index test
  5. 4This could also mean studies that identify discordant cases