Related Experiment Video
Updated: Aug 6, 2025
Identification of Mycobacterium Species by DNA Microarray Chip Method
Published on: June 24, 2025
Over-interpretation of findings in diagnostic accuracy studies of infectious diseases
Solange Bramer1, Ho Yee Cheung2, Wesley Do2
1Oxford University Hospitals, Oxford University, Oxford, United Kingdom.
Objectives:
To assess the prevalence of overly positive interpretation, also called 'spin',-of results in diagnostic accuracy studies of infectious diseases and to identify suggestions for improvement.
Methods:
A PubMed search was performed to identify diagnostic accuracy studies of infectious diseases published between January and March 2019. Each article was assessed by two authors independently to identify study characteristics and forms of actual and potential over-interpretation. 'Actual over-interpretation' was defined as conclusions that were not on the basis of study aims or conclusions that were more favourable than justified by the study findings. There are other practices that may result in the over-interpretation of study findings and these have been described as 'potential over-interpretation'.
Results:
The final analysis included 120 studies. Favourable or promising recommendations were made in the main text of 101 (84%) of the included studies. Evidence of actual over-interpretation (spin) was found in 30 (25%) articles, with 22 (18%) studies reporting a conclusion that did not match the study aims and 56 (47%) studies with a more positive conclusion in the abstract than the main text. All analysed studies exhibited at least one form of potential over-interpretation, with was most commonly a lack of sample size calculation (n = 109, 91%) and not reporting a null hypothesis (n = 115, 96%).
Discussion:
Evidence of over-interpretation of results was found in one-third of the included studies. We have proposed possible interventions to prevent overly positive interpretations of results in diagnostic accuracy studies.
Insights
Overly positive interpretations, or spin, were found in one-third of infectious disease diagnostic accuracy studies. This research highlights the need for improved reporting to prevent biased results in scientific literature.
Area of Science:
- Medical research methodology
- Infectious diseases diagnostics
Background:
- Overly positive interpretation, termed 'spin', can bias the perceived accuracy of diagnostic tests.
- Assessing the prevalence of spin in infectious disease diagnostic accuracy studies is crucial for reliable evidence synthesis.
Purpose of the Study:
- To determine the frequency of spin in diagnostic accuracy studies for infectious diseases.
- To identify potential interventions to mitigate overly positive interpretations in research findings.
Main Methods:
- A systematic PubMed search identified diagnostic accuracy studies for infectious diseases published in early 2019.
- Two independent reviewers assessed 120 studies for characteristics and forms of actual and potential over-interpretation.
- Actual over-interpretation included conclusions not aligned with study aims or findings; potential over-interpretation encompassed reporting deficiencies.
Main Results:
- Overly positive recommendations appeared in 84% of studies.
- Actual spin was present in 25% of articles, with 18% showing mismatched conclusions and 47% having more positive abstracts than main texts.
- All studies exhibited potential over-interpretation, commonly due to lack of sample size calculation (91%) and unreported null hypotheses (96%).
Conclusions:
- One-third of diagnostic accuracy studies for infectious diseases demonstrated evidence of result over-interpretation.
- Proposed interventions aim to prevent overly positive interpretations and enhance the integrity of diagnostic accuracy research.
Related Concept Videos
01:13Sensitivity, Specificity, and Predicted Value
Sensitivity is the...
01:15Receiver Operating Characteristic Plot
01:26Urine Studies II: Urine Culture and Sensitivity Test
01:29Bias in Epidemiological Studies
01:13Accuracy and Errors in Hypothesis Testing
In hypothesis testing, the probability of making a Type I error, denoted as α, is commonly set at 0.05. This significance level indicates a 5%...
01:10Documentation of Nursing Diagnosis
In some settings, data-driven computerized decision support systems are in place, allowing for more accurate nursing diagnoses. The database within one of these systems includes diagnostic labels defining characteristics, activities, and indicators for nursing. A nurse enters...