14.4
感度、特異度、及び予測値
医療診断では、臨床検定が様々な病状の特定と診断に重要な役割を果たします。ただし、検定結果の解釈は必ずしも簡単ではありません。異常な検定結果が必ずしも病気の存在を裏付けるとは限らず、正常な結果が病気の不在を保証するとも限りません。これらの診断ツールの信頼性を評価するために、医療従事者は 2 つの重要な…
健康科学では、感度とは、病気が存在するときに診断テストで陽性の結果を示す確率を指します。
一方、特異度は、病気が存在しないときに検査が陰性の結果を返す確率を測定します。
陽性の検査結果が与えられた場合、病気にかかる確率は予測値陽性です。
一方、予測値陰性は、検査結果が陰性のときに病気にかからない確率です。
これらの指標は通常、疾患の実際の有無に基づいており、臨床現場で実施された広範な検証研究から導き出されます。
ウイルス感染スクリーニングの例を考えてみましょう。これは、最初に予備検査に基づいて行われ、後で精巧な血液パラメータを使用して行われます。
したがって、感度は a/(a + b)、特異度は d/(c + d)、予測値は正の a/(a + c)、予測値は負の d/(b + d) です。
View the full transcript and gain access to JoVE Core videos
Q1: What is the difference between sensitivity and specificity in diagnostic testing?
Sensitivity measures the probability that a test correctly identifies individuals with the disease, producing a positive result when disease is present. Specificity measures the probability that a test correctly identifies individuals without the disease, yielding a negative result when disease is absent. High sensitivity minimizes false negatives in screening, while high specificity reduces false positives in confirmatory testing.
Q2: How do you calculate sensitivity and specificity from a 2x2 contingency table?
Sensitivity is calculated as a/(a + b), where a represents true positives and b represents false negatives. Specificity is calculated as d/(c + d), where d represents true negatives and c represents false positives. These formulas derive from validation studies conducted in clinical settings to assess diagnostic test accuracy.
Q3: What is positive predictive value and how does disease prevalence affect it?
Positive predictive value (PPV) is the probability that an individual with a positive test result truly has the disease, calculated as a/(a + c). PPV depends on test sensitivity, specificity, and disease prevalence in the population. Higher disease prevalence increases PPV, making a positive result more likely to indicate a true case in populations with higher baseline risk.
Q4: Why is negative predictive value important in clinical decision-making?
Negative predictive value (NPV) is the probability that an individual with a negative test result truly does not have the disease, calculated as d/(b + d). NPV is influenced by disease prevalence and is particularly valuable in populations where the disease is rare, providing clinicians with confidence that a negative result accurately confirms disease absence for informed patient care decisions.
Q5: How do sensitivity and specificity differ from predictive values in clinical practice?
Sensitivity and specificity assess test accuracy under controlled conditions based on disease presence or absence. Predictive values provide practical insights for real-world clinical decision-making by indicating the likelihood of actually having or not having disease given a specific test result. Together, these measures help clinicians balance risks of false positives and negatives using statistical methods for analyzing epidemiological data.
Q6: Why is high sensitivity essential for screening tests?
High sensitivity is critical for screening tests because it minimizes false negatives, reducing the chance of missing disease cases. A sensitive test produces a positive result when disease is present, ensuring that individuals with the condition are identified during initial screening. This is particularly important in preliminary examinations before more elaborate diagnostic procedures are performed.
Q7: When is high specificity more important than high sensitivity?
High specificity is crucial for confirmatory diagnostic tests because it reduces false positives, preventing healthy individuals from being incorrectly labeled as having disease. A specific test yields a negative result when disease is absent, providing confidence in diagnosis confirmation. This is essential after initial screening to avoid unnecessary treatment and psychological burden on patients.