Reliability & Diagnostic Utility of the Clinical Exam

Question 1

Assessment of diagnostic tests

Answer 1

• Reliability: produces precise, accurate, reproducible information
• Diagnostic accuracy: discriminate between patients with & without the disorder

Question 2

Describe reliability

Answer 2

• degree of consistency that a test provides the same result between
• Random errors: deviation from the true measurement that occurs due to chance (patient liability, instrument error, & human error/tester error)
• Percent agreement: in general a 75% agreement is considered acceptable however it is rarely used (alone) due to limitations (does not take chance into account & overestimates percent agreement)
• Interclass correlation (CC) or Kappa’s coefficient: incorporates chance, results range form -1 to +1, & in healthcare 0.75 is acceptable

Question 3

Reliability interpretation ICC or Kappa

Answer 3

0.81-1.0: substantial agreement
0.61-0.80: moderate agreement
0.41-0.60: fair agreement
0.11-0.40: slight agreement
0.0-(-)1.0: no agreement

Question 4

Define diagnostic accuracy

Answer 4

• measure of agreement between the clinical test & a reference standard

Question 5

Describe the positive predictive value (PPV)

Answer 5

• likelihood a patient with a positive test result has the disease/disorder
• high PPV means the positive result is a strong indicator the patient has the disease

Question 6

Describe the negative predictive value (NPV)

Answer 6

• likelihood a patient with a negative test result does not have the disorder
• high NPV means the negative result is a strong indicator the patient does not have the disease

Question 7

Describe sensitivity (SnNout)

Answer 7

• indicates the tests ability to detect those patients that have the disorder
• a negative result is good for ruling out a disorder

Question 8

Describe specificity (SpPin)

Answer 8

• indicates the tests ability to detect patients who actually do not have the disorder
• a positive result is good for ruling in the disorder

Question 9

Difference between pretest and posttest probability

Answer 9

• Pretest: likelihood a patient exhibits a specific disorder before the clinical exam is performed
• Posttest: likelihood a patient has a diagnosis after the clinical exam has been performed

Question 10

Describe the positive likelihood ratio

Answer 10

• a shift in probability that favors the existence of the disorder
• positive ratios greater than 10 represent large & conclusive shifts in probability
• a likelihood ratio of 1 (either pos. or neg.) is of little clinical value

Question 11

Describe negative likelihood ratio

Answer 11

• a shift in probability that favors the absence of a disorder
• negative LR close to 0 indicate large & conclusive shifts in probability
• a likelihood ratio of 1 (either pos. or neg.) is of little clinical value

Question 12

Likelihood ratios chart

Answer 12

≥10 (+LR) or <0.1 (-LR): large interpretation
5.0-10.0 (+LR) or 0.1-0.2 (-LR): moderate interpretation
2.0-5.0 (+LR) or 0.2-0.5 (-LR): small interpretation
1.0-2.0 (+LR) or 0.5-1.0 (-LR): rarely important interpretation

Question 13

Describe the confidence interval (CI)

Answer 13

• a 95% CI is typical for test results
  -a probability (p) value of <0.05 infers there is 95% probability that a true relationship exists between the results of the outcome measures

Question 14

Describe clinical prediction rules (CPRs)

Answer 14

• identification of variables that can be used to predict the probability of if a disorder will occur or what may predict a treatment strategy that my be useful
• usually multiple components of an exam: Hx & special tests that increase the likelihood of a disorder or treatment category