Disease, Diagnostic test and screening
Diagnosis
disease v disease free making a diagnosis symptoms fasting blood test to decide diabetes. is a result of a process each step of process to test if person has the disease or injury if it is present or absent. likelihood increases with each step.
A Valid Diagnostic study needs
Assembly of appropriate spectrum of patients/persons Applies diagnostic test and reference standard to them all Interprets blind to each other Repeats itself in a 2nd independent test set to all patients
Specificity
is probability that a person without the disease(Noncase) will be correctly identified as nondisease by the test. it is thus the probability of Correctly Identifying a non diseased person with a test( syn: true-negative probability) Truth for non-diseased persons.
Screening
presumptive identification of unrecognized disease or defect by application of tests. screening tests sort out well persons secondary prevention diagnostic tests to verify main purpose and assumption early detection leads to favourable outcome
Diagnosis definition
process of determining health status and factors responsible for producing it. eg individual, family, group community. the term is applied both the process of determination and to its findings.
Combine tests sequentially
when available tests are not highly specific then combine tests sequentially increasing specificity and PPV lowering sensitivity and NPV
Take home message
Validity V Reliability SpPin SnNout relationship between prevelence and predictive value Diagnostic tests and screening test share similar study design screening programs vs diagnosis good ideas, doesn't mean an effective screening program.
Screening considerations
cost effectiveness ethics Biases
Negative Predictive Value
if the test result is negative for a patient, what is the probability they have the disease the probability that a person with a negative test result is a true negative. the proportion of patients with negative results who are correctly diagnosed NPV= No of diseased who screen-ve/ total no of people who screened -ve
Sequentially or Simultaneously
often use multiple tests sequential :use less invasive less expensive first results increase the post test probability through each test.
When to screen if the disease
1. High morbidity, mortality and costs 2. high prevelence and incidence , high prevalence of preclinical stage 3. Known natural history and biology 4. Effective treatment of the early stages 5. a long period between pre clinical and clinical stages
When to Screen if the test
1. can detect disease in preclinical stages 2. Accurate and reliable low false negative and false positive results 3. Safe, acceptable and painless 4. Simple and inexpensive.
Predictive Value
the probability of the disease given the results of the test Asks, how good is the test at identifying people with the disease and people without the disease. Positive predictive Value : the probability that person with a positive test is a true positive PPV= no of diseased who screen +ve/ total no of people who screen +ve TP/tp +FP= a/a+b Extent to which a test can differentiate between presence or absence of a person's condition
Validity of Test
Able to distinguish between with and without the target disorder Must be applied to a group of people with the disease and a group without Results of new diagnostic test under investigation compared with known true result.
Influencing Accuracy of Screening tests
1. change threshold value for a positive test 2. combine tests in parallel sequentially
Specificity testing
specificity =True negative / all non diseased (TN+FP) d/b+d x100% Among healthy people how specific is the test to tell the truth(negative) Rules out the condition.
SpPin and SnNout
can be problem if sensitivity or specificity is low SnNout : sensitivity with a negative result, rules out SpPin specificity with a positive result, rules in
Reliability
consistency, repeatability, precision or reproducibility) Reliability implies the diagnostic test were repeated within the same participants by the same or different testers the Results of the repeated test would be similar or identical to the first findings.
Biases in evaluating screening
Lead time bias. the apparent lengthening of survival achieved by early diagnosis, rather than actual benefit from early intervention to avoid , must compare people screened with a population that isn't screened. Length Bias the detection at screening of cases with the disease that is of slower progression or longer duration. those not detected by screening may have more aggressive disease,
Assessing Reliability
Needs to apply to test at least twice Compares variation between repeated measures If 1st and 2nd readings identical then there is perfect reliability Kappa statistic
Screening Test Results
Positive = abnormal= disease Negative=normal= no disease Then verified by diagnosis( a gold standard) Some results are true/correct Some are false/incorrect.
Quality of a Diagnostic Test
Reliability ( consistence, repeatability, precision or reproducibility) Validity ( conformity, accuracy, or correctness
Screening Test Sensitivity
Sensitivity = True Positive/ All diseased TP/ TP+FN a/ a+c X100% How sensitive is the test to pick up all or most cases of disease in the tested population.
specificity
TN/ TN+FP
NPV negative predicative value
TN/TN+FN
Sensitivity
TN/TP+FN
PPV positive predictive value
TP/TP+FP
WHO principles of Screening
1. Disease a public health problem 2. Natural History of disease understood 3. Recognizable latent or early symptomatic stage 4. Acceptable treatment available 5. Consensus on whom to treat 6. Facilities for screening and treatment available 7. Economic balanced between case finding and subsequent medical care program sustainable.
Over view of tests
Tests can be problematic if sensitivity and specificity are low low specificity generate numerous false positives if low prevelence SnNout senstivity with a negative result rules out SpPin specificity with positive result rules in.
Likelihood ratio
The probability that a test result will be obtained in a patient with the condition of interest as compared to a patient without the condition of interest LR+ = sensitivity/(1-specificity) LR- = (1-sensitivity)/specificity
Screening test results
True Positive : a positive test result in a pt. True Negative: a negative test result in a healthy person False Positive : a positive result for a healthy person False Negative : a negative result for a patient.
Measuring outcomes of Screening programs
UPTAKE the proportion of people offered screening who were eventually screened DETECTION RATE the prevalence of preclinical cancers detected ACCURACY the proportion of test results that are confirmed by definitive diagnosis. CAUSE SPECIFIC MORTALITY RATE among all screened people ( those with or without positive test results COSTS PER CANCER DETECTED mainly economic factors.
Screening
Unrecognised disease - applying tests exams, procedures Secondary prevention Often requires diagnostic tests to verify Main purpose and basic assumption: early detection of disease will lead to a more favorably outcome Some time during subclincial stage
Likelihood ratio Application.
Useful to assess how good a diagnostic test is and to help in selecting appropriate diagnostic test or sequence of tests if pretest odds or pretest probability you can work out post test odds or post test probability Tells you what the diagnostic test has done for you getting closer to the diagnosis.
Validity
Validity ( conformity, accuracy, or correctness) Validity implies the result of the diagnostic test is correct;the diagnostic test is able to differentiate correctly between people with a disease and people free from disease
Effectiveness of screening
do benefits outweigh the harm to health and well being do the net benefits merit the economic costs
Sensitivity
is probability that a diseased person (case) in the population test will be identified as diseased by the test (syn:positive probability). Sensitivity is thus the probability of correctly diagnosing a case or the probability that any given case will be identified, "Pick up" for cases
Combine Tests in parallel
need a very sensitive test only have 2-3 relatively insensitive tests Positive if test result positive in any test Negative if test result negative in both tests. Increasing sensitivity, referral centres diagnose disease that local centres do not. Consequence of a false positive : further evaluation and treatment.
Screening and Diagnostic test
Evaluated using same quality measures Sensitivity and Specificity Definitions and calculations apply to both. Systematic reviews useful for consideration