Diagnostic Reasoning and Evidence-Based Medicine Flashcards

The likelihood ratio positive (LR+) for a diagnostic test is calculated as:

Answer: Sensitivity / (1 - Specificity)

LR+ = sensitivity / (1 – specificity), quantifying how much a positive test result increases the odds of disease relative to the pre-test probability.

When interpreting diagnostic tests, which situation should prompt concern about verification bias?

Answer: Disease status confirmed only in patients with positive screening tests

Verification bias occurs when gold standard testing is applied selectively (e.g., only to positive screens), overestimating sensitivity and underestimating specificity.

In clinical practice, a decision threshold (test threshold) represents the probability at which:

Answer: Testing is warranted based on pre-test probability and test risk/benefit

The test threshold is the pre-test probability below which testing is unnecessary, while the treatment threshold is the probability above which treatment proceeds without testing.

Which statistical test is most appropriate for comparing means of a continuous variable between two independent groups with normally distributed data?

Answer: Independent samples t-test

The independent samples t-test compares means of normally distributed continuous variables between two independent groups to test for statistically significant differences.

Premature closure in diagnostic reasoning is best described as:

Answer: Accepting a diagnosis before all relevant information has been gathered

Premature closure stops diagnostic inquiry once a plausible diagnosis is found, missing alternative or comorbid diagnoses that would have been revealed by further evaluation.

Forest plots in meta-analyses display each study's result as a square with a horizontal line. What does the horizontal line represent?

Answer: The 95% confidence interval for that study's effect estimate

In a forest plot, the horizontal line through each study's square extends to the upper and lower bounds of the 95% confidence interval for that study's effect size.