Clinical Test Performance
Overview
Clinical test performance describes the quantitative measures used to evaluate how well a genetic test performs: specifically, how accurately it identifies individuals who do and do not have a condition. This chapter covers sensitivity and specificity (properties of the test itself) and positive and negative predictive values (which depend on both the test and the population being tested). These concepts are fundamental to evidence-based practice and come up routinely in clinical genetics.
Understanding test performance is essential every time you order a screening test, interpret a result, or counsel a patient about what a positive or negative result actually means. A common clinical task is calculating the positive predictive value (PPV) of a screening test given its sensitivity, specificity, and the prevalence of the condition in the screened population. The key insight, that PPV depends heavily on disease prevalence, explains why the same test can be highly informative in a high-risk population but generate many false positives in a low-risk population.
These concepts apply directly to clinical scenarios across genetics: prenatal cfDNA screening, expanded carrier screening, newborn screening, and predictive testing for hereditary cancer. The ability to explain test performance in plain language, helping patients understand what a positive screen does and does not mean, is a core genetic counseling competency.
Key Concepts
- Sensitivity: the probability that a test correctly identifies affected individuals (true positive rate)
- Specificity: the probability that a test correctly identifies unaffected individuals (true negative rate)
- Positive predictive value (PPV): the probability that a person with a positive result truly has the condition
- Negative predictive value (NPV): the probability that a person with a negative result truly does not have the condition
- 2x2 table construction: setting up the classic table with disease status vs. test result to calculate all four measures
- Prevalence and PPV: understanding why PPV drops dramatically when prevalence is low, even with a highly sensitive and specific test
- Screening vs. diagnostic: relating test performance measures to the clinical distinction between screening and confirmation
Predictive Values
Positive and negative predictive value answer the question patients actually ask: "My test was positive, what are the chances I actually have this condition?" PPV and NPV depend on disease prevalence in addition to sensitivity and specificity. The calculations use 2x2 tables and the critical insight is that PPV drops sharply when prevalence is low, even when the test itself is highly sensitive and specific. This explains why the PPV of cfDNA screening for trisomy 21 is much higher than for rarer conditions like trisomy 13 despite similar test sensitivity.
Screening metrics define how the test performs; predictive values translate that performance into clinically actionable information for a specific patient.
Screening Metrics
Sensitivity and specificity are the intrinsic performance characteristics of a test. Sensitivity (detection rate, true positive rate) measures how well the test catches true cases; specificity (true negative rate) measures how well it excludes non-cases. There is an inherent tradeoff between the two when adjusting cutoff thresholds, and both relate directly to the false-negative and false-positive rates. A test with 99% sensitivity still misses 1 in 100 affected individuals; that residual risk is essential to communicate clearly during counseling.