Materials and Methods

Stool samples from 173 participants were analyzed via FIT. Each participant underwent a colonoscopy and biopsy for cancer diagnosis. Three independent pathologists evaluated sample results without prior knowledge of the biopsy outcomes.

Results

Colon cancer was confirmed in 145 participants. The FIT was positive for 120 of these patients, and among the 28 without cancer, 21 had a negative FIT. Of those with a positive FIT, 94% were diagnosed with colon cancer.

Table of Diagnostic Results

Abbreviations: TP = true positive; FP = false positive; FN = false negative; TN = true negative.

Conclusion

The FIT shows promising diagnostic accuracy for colon cancer in individuals over 50. Future research is needed on its efficacy in broader populations.

Sensitivity Case Study

To find sensitivity, use the formula:

$$ \text{Sensitivity} = \frac{TP}{TP + FN} $$ Among all patients with colon cancer, the sensitivity calculated is 83%, indicating the proportion of patients with cancer who tested positive.

Specificity Case Study

The specificity formula is:

$$ \text{Specificity} = \frac{TN}{TN + FP} $$ Thus, the specificity calculated is 75%.

Positive Predictive Value Case Study

The PPV is determined as follows:

$$ \text{PPV} = \frac{TP}{TP + FP} $$ The calculated PPV is 94%, indicating how many patients with a positive FIT actually have cancer.

Negative Predictive Value Case Study

Utilizing the NPV formula:

$$ \text{NPV} = \frac{TN}{TN + FN} $$ The calculated NPV is 45%.

Prevalence and Incidence Case Study

Prevalence is defined as:

$$ \text{Prevalence} = \frac{TP + FN}{ ext{Total Population}} $$ The prevalence computed from this study is 83%.

For incidence:

$$ \text{Incidence Rate} = \frac{# \, \text{new cases}}{ ext{# \, people at risk}} $$ Wherein in this study, 5% of 500 adults diagnosed with new colon cancer equates to 25 new cases.

Factors for Justifying a Test

To determine if testing is warranted, consider: - Diagnostic characteristics - Cost - Harms - Pre-test probability of disease - Likelihood ratios

Case Vignette: Pre-Test/Post-Test Probability

A pregnancy test may not be warranted for a male due to a pre-test probability of 0%. This indicates a positive result would be a false positive. Pre-test probability determines the likelihood of having a disease prior to testing.

Study on CT vs. US for Appendicitis

Study compared CT and US for acute appendicitis. CT showed: - Sensitivity: 95% - Specificity: 90% Hence, CT is more accurate than US for diagnosing appendicitis.

Likelihood Ratios

LRs indicate how a test modifies the probability of disease. Formula for positive LR is:

And negative LR:

Cutoff Values

A test cutoff is a predefined point that determines positive or negative results, maximizing diagnostic value.

ROC Curve

The ROC curve plots sensitivity against (1-specificity). AUC shows the model's accuracy—closer to 1 indicates a better test.

Risk Quantification

Relative risk and absolute risk reduction help assess treatment efficacy alongside NNT (number needed to treat). Higher NNT is preferable to avoid harm.

Beers Criteria

Beers Criteria aim to reduce potentially inappropriate prescriptions in geriatrics—common medications to be cautious about include: - Anticholinergics - Benzodiazepines - Chronic NSAIDs