Screening and Diagnostic Testing

Information about Screening and Diagnostic Testing

Published on July 14, 2014

Author: amitakashyap1

Source: authorstream.com

Content

Screening and Diagnostic Testing: Screening and Diagnostic Testing Amita kashyap Assessing the Validity and Reliability of Diagnostic and Screening Tests: Assessing the Validity and Reliability of Diagnostic and Screening Tests S creening and diagnostic tests - To distinguish between people who have the disease and those who do not. Hence quality of screening and diagnostic tests is a critical issue . In using a test to so distinguish, it is important to understand how characteristics are distributed in human populations. PowerPoint Presentation: Variability and Bias: Variability and Bias A 45 yr old man’s BP was 140/86 during routine check up for job, he was obese. His father died of MI at 65 yrs of age Total S Cholesterol (non fasting) was 242 mg/ dl No other abnormality Physician asked him to come after 2 weeks; fasting, for further testing Repeat total S Cholesterol (fasting) was 198 mg/ dl Physician’s decision to treat by drugs changed!! PowerPoint Presentation: Random Variability Systematic Variability - Bias PowerPoint Presentation: Levels features Individual Individual variability Measurement variability Population Genetic variability btw individuals Environmental variability Measurement variability Sample Manner of Sampling Size of Sample Measurement variability Levels of Variability PowerPoint Presentation: Sources of variability features Individual characteristics Diurnal variation Factors like Age, diet and exercise Environmental like season and temperature Measurement characteristics Poor calibration of instrument Inherent lack of precision of the instrument Observers misreading or recording Potential Sources of Variability Validity : Validity The degree to which a measurement or study reaches a correct conclusion Internal Validity – the extent to which the results of an accurately reflect the true situation. To improve on it we decrease the impact of factors extraneous to the study question by restricting the type of subjects and the environment in which the study is performed External Validity - generalizability Bias – a threat to validity: Bias – a threat to validity The systematic error in a study that leads to a distortion of the results Randomization reduces the chance difference between the groups Selection Bias Information Bias Confounding : (can be quantified) otherwise evaluation of bias is subjective Likelihood of 1) the presence of bias and 2) its potential magnitude of effect PowerPoint Presentation: Total Population Sample Frame Sampling scheme Eligible Subjects Inclusion Criteria Exclusions Subjects asked to participate Informed consent Non Participants Participants Lost to Follow Up Participants complete the study PowerPoint Presentation: Information Bias Unacceptability Bias Recall and interviewer Bias PowerPoint Presentation: Oob Obesity Mayocardial Infarction Total Cholesterol Confounding A potential confounder must satisfy two conditions: Association with the disease of interest in the absence of exposure Association with the exposure but not as a result of being exposed PowerPoint Presentation: Confounding = high total S Cholesterol Practice of clinical Medicine is the artful application of Science. Variability is the law of life. No two individuals react or behave alike – probability is the guide of Life!: Practice of clinical Medicine is the artful application of Science. Variability is the law of life. No two individuals react or behave alike – probability is the guide of Life! Diagnostic Testing: Diagnostic Testing Patient Profile : A 54 year old school teacher got her physical examination for insurance. She had no complaints; ( she had hot flashes a year ago but had resolved without treatment ). Physical examination, including breast, pelvic (PAP smear), and rectal examination; NAD. Physician recommended mammogram. ( ? ) Mammogram was not normal, hence she was referred to a surgeon who also found Breast normal but Based on mammographic abnormality however; both surgeon and radiologist agreed for FNA under radiologic guidance for abnormal breast. FNA specimen revealed cancer cells and patient was scheduled for further surgery next week. PowerPoint Presentation:  0.3 13 20 40 60 80 100 64 After positive FNA result 54 yr old women Before Mamogram After positive mammogram Probability of Breast Cancer (Percent) palpable lump Prior to mamogram Tests are performed to detect the disease, assess its severity, predict outcome, or to monitor response to therapy Schematic Diagram of the estimated Probability of Breast Cancer in a 54 yr old women without palpable Breast Mass, after A positive mammogram and following a positive FNA test result 1% H/o Br Ca In mother PowerPoint Presentation: Total a + c b + d PowerPoint Presentation: Sensitivity and Specificity Surgical Biopsy (Gold Standard) FNA results positive Disease No Disease Total 14 8 22 negative 1 91 92 TOTAL 15 99 114 Sensitivity = 14 15 ( 14+1 ) X 100 = 93% Specificity = 91 99 ( 91+8 ) X 100 = 92% PV+ = 14 14 + 8 X 100 = 64% PV – = 91 91 + 1 X 100 = 99% Accuracy ( validity )- determining the ‘True Status’ of the disease Descriptors of test accuracy - Sensitivity and Specificity - the validity of the test assessed relative to gold standard Pre FNA P(Br Ca) =15/114 =0.13 Pre FNA P(No Br Ca) =99/114 =0.87 PowerPoint Presentation: Post FNA probability of disease for + ve or – ve test result guide further action Whether the probability of Br. Ca is 13% or 64%; further workup is required, A – ve test result would reduce the probability that Br. Ca is present to 1% (100% minus PV- ve ) So! Now no Biopsy…but keep watch The greater the sensitivity, the more likely the test will detect the persons with the disease Predictive value (+ ve and - ve )-estimation of the probability of the presence or absence of disease if test is positive or negative Predictive value of a test is affected by the prevalence of the disease. PowerPoint Presentation: Surgical biopsy FNA results Positive Cancer No Cancer Total 14 8 22 Negative 1 91 92 Total 15 99 114 Surgical biopsy FNA results Positive Cancer No Cancer Total 113 15 128 Negative 8 181 189 Total 121 196 317 Effect of Prevalence on Predictive value of a test: For Patients without palpable masses For Patients With palpable masses Prevalence= 13% Sensitivity = 14/15=93% Specificity = 91/99 =92% PV + =14/22= 64% PV - = 91/92= 99% Prevalence= 38% Sensitivity = 93% Specificity = 92% PV + = 88% PV - = 96% Specific Example: Specific Example Test Result Pts with disease Pts without the disease Threshold: Test Result Call these patients “negative” Call these patients “positive” Threshold Some definitions ...: Test Result Call these patients “negative” Call these patients “positive” without the disease with the disease True Positives Some definitions ... PowerPoint Presentation: Test Result Call these patients “negative” Call these patients “positive” without the disease with the disease False Positives PowerPoint Presentation: Test Result Call these patients “negative” Call these patients “positive ” without the disease with the disease True negatives PowerPoint Presentation: Test Result Call these patients “negative” Call these patients “positive” without the disease with the disease False negatives Moving the Threshold: left: Test Result without the disease with the disease ‘‘-’’ ‘‘+’’ Moving the Threshold: left e.g. Suspicious FNA results considered positive Moving the Threshold: right: Test Result without the disease with the disease ‘‘-’’ ‘‘+’’ Moving the Threshold: right e.g. Suspicious FNA results considered negative PowerPoint Presentation: Surgical biopsy FNA results positive Cancer No Cancer Total 113 15 128 negative 8 181 189 Total 121 196 317 Effect of cut off value: Suspicious FNA results considered positive Prevalence= 38% Sensitivity = 93% Specificity = 92% PV + = 88% PV - = 96% Surgical biopsy FNA results positive Cancer No Cancer Total 91 0 91 negative 30 196 226 Total 121 196 317 Suspicious FNA results considered negative Prevalence= 38% Sensitivity = 75% Specificity = 100% PV + = 100% PV - = 87% Likelihood Ratios (LR) – in interpretation of Dx tests: Likelihood Ratios (LR) – in interpretation of Dx tests Definition: An LR is the probability of a particular test result for a persons with the disease divided by the probability of that test result in non-diseased persons LR+ - Probability of + ve test result for a person with disease (true positive/ total diseased) Probability of + ve test result for a person without disease (false positive/ total Non-diseased) Sensitivity / 1-specificity = (14/15)/(8/99)=.93/.08= 11.63 Sensitivity and specificity are expressed as proportion An LR+ve of 1 indicates? PowerPoint Presentation: LR¯ - Probability of - ve test result for a person with the disease (false positive/ total diseased) Probability of - ve test result for a person without disease (true negatives/ total Non-diseased) i.e. 1-Sensitivity)/Specificity Surgical Biopsy (Gold Standard) FNA results positive Disease No Disease Total 14 8 22 negative 1 91 92 TOTAL 15 99 114 LR+ = Sensitivity / 1-specificity = 0.93/1-0.92 =0.93/0.08=11.63 LR¯ - 1-Sensitivity)/Specificity = 1-0.93/0.92 = 0.07/0.92=0.08 In contrast to PV, LR does not vary as a function of Prevalence Receiver Operating Characteristic (ROC) Curve: Receiver Operating Characteristic (ROC) Curve Diagnostic tests giving quantitative outcome e.g. serum levels of enzymes, there are many options about where to set a cut off point – as the cut off point rises (from 200 to 250mg/dl for total cholesterol) the sensitivity will increase with a corresponding decrease in specificity. At each cutoff point, sensitivity and (1- specificity) is calculated and plotted on ‘y’ and ‘x’ axis respectively along the full range of cutoff points PowerPoint Presentation: 1 0 %1s %0s Say 1 Say 0 PowerPoint Presentation: Say 1 Say 0 PowerPoint Presentation: Say 1 Say 0 PowerPoint Presentation: Say 1 Say 0 PowerPoint Presentation: Say 0 Say 1 PowerPoint Presentation: Say 1 Say 0 PowerPoint Presentation: Say 1 Say 0 PowerPoint Presentation: Say 0 Say 1 ROC curve: True Positive Rate (sensitivity) 0% 100% False Positive Rate (1-specificity) 0% 100% ROC curve LR+ = 1, + ve test is equally likely in persons with or without the disease Signal Noise Substantial gain in sensitivity with only modest reduction in specificity AUC - summary Index Highest possible value = 1 Area under diagonal line=0.5 ROC curve comparison: True Positive Rate 0% 100% False Positive Rate 0% 100% True Positive Rate 0% 100% False Positive Rate 0% 100% A good test: A poor test: ROC curve comparison ROC curve extremes: Best Test: Worst test: True Positive Rate 0% 100% False Positive Rate 0% 100% True Positive Rate 0% 100% False Positive Rate 0% 100% The distributions don’t overlap at all The distributions overlap completely ROC curve extremes Screening Test: Screening Test Identify individuals with a disease before it is detected by routine diagnosis (survival may remain same but appear more- lead time bias ) Treatment initiated after screening (early than routine) will improve chance of survival Length biased sampling occurs when a screening program detects a less aggressive (…slow progressing) disease only To overcome these biases – age specific mortality rates are calculated in entire population (screened and not screened). It is important to identify false negative results PowerPoint Presentation: High FP rate and low PV+ is due to low prevalence of the disease in general population Criteria for Screening Test – morbidity & motality must be sufficient concern A high risk population must exist Test should be sensitive and specific with minimal risk & acceptable Effective intervention known Disease Status Mammography positive Cancer No Cancer Total 132 985 1117 negative 47 62,295 62,342 Total 179 63,280 63,459 Prevalence= 0.3% Sensitivity = 73.7% Specificity = 98.4% PV + = 11.8% PV - = 99.9% Usefulness of Mammography PowerPoint Presentation: The process of making an objective and systematic analysis of information from all the randomized controlled trials

Related presentations


Other presentations created by amitakashyap1

IDSP
05. 07. 2014
0 views

IDSP

epidemiology UG
05. 07. 2014
0 views

epidemiology UG