Robert S. Galen, M.D., M.P.H. Professor and Head, Department of Health Administration, Biostatistics and Epidemiology College of Public Health University.

Slides:



Advertisements
Similar presentations
Nick Curry, MD, MPH Infectious Diseases Prevention Section
Advertisements

Validity and Reliability of Analytical Tests. Analytical Tests include both: Screening Tests Diagnostic Tests.
Diagnosis and Management of Acute HIV Infection HIV Clinical Guidelines from the New York State Department of Health AIDS Institute January 2010 HIV CLINICAL.
Rapid HIV Testing and Its Role in Advancing HIV Prevention: 2004 Update Bernard M. Branson, M.D. Chief, Lab Determinants and Diagnostics Section Centers.
Performance of Bio-Rad Genetic Systems HIV-1/HIV-2 Plus O EIA Followed by Multispot or OraQuick Advance in a Dual Immunoassay HIV Testing Strategy Laura.
Perspectives on Outreach from the NYC Department of Health and Mental Hygiene Benjamin Tsoi, MD, MPH Bureau of HIV/AIDS Prevention and Control NYC Department.
Rapid HIV Testing: 2005 Update Bernard M. Branson, M.D. Associate Director for Laboratory Diagnostics Division of HIV/AIDS Prevention.
Mortality Among a Tuberculosis Outbreak Los Angeles County, 2007–2013 Brian Baker, MD Amit Chitnis, MD MPH Leslie Henry, BSN RN PHN 48th CTCA Educational.
CDC Recommendations for HIV Testing of Adults and Adolescents Christina Price, MPH Delta Region AIDS Education and Training Center.
STD Screening in HIV Clinics: Value and Implications Thomas Farley, MD MPH Tulane University Deborah Cohen, MD MPH RAND Corporation.
HIV Testing in Health-Care Settings
Routine HIV Screening in Health Care Settings David Spach, MD Clinical Director Northwest AIDS Education and Training Center Professor of Medicine, Division.
HIV in the United Kingdom: 2013 HIV and AIDS Reporting Section Centre for Infectious Disease Surveillance and Control (CIDSC) Public Health England London,
Bernard M. Branson, M.D. Associate Director for Laboratory Diagnostics Divisions of HIV/AIDS Prevention National Center for HIV, STD, and TB Prevention.
Rapid HIV Tests Norman Moore, Ph.D. Director of Medical Affairs.
The hidden HIV epidemic: what do mathematical models tell us? The case of France Virginie Supervie, Jacques Ndawinz & Dominique Costagliola U943 Inserm.
Primary HIV Infection: the CDC study Pragna Patel, MD MPH Medical Epidemiologist Behavioral and Clinical Surveillance Branch DHAP, CDC February 28, 2005.
Increasing Our Reach through Rapid HIV Testing Bernard M. Branson, M.D. Chief, Lab Determinants and Diagnostics Section Centers for Disease Control and.
Routine HIV Screening in Health Care Settings David Spach, MD Clinical Director Northwest AIDS Education and Training Center Professor of Medicine, Division.
Is Nucleic Acid Testing for Organ Donors the ‘Right’ Choice? Reference: Humara A, Morrisb M, Blumbergc R, et al. Nucleic acid testing (NAT) of organ donors:
Over The Counter HIV Testing A Technology Whose Time has Come Freya Spielberg, MD MPH
HIV Testing in Health- Care Settings Revised Recommendations for HIV Testing of Adults, Adolescents, and Pregnant Women in Health-Care Settings U.S. Centers.
Journal Club Alcohol, Other Drugs, and Health: Current Evidence January–February 2010.
Enhancing HIV/AIDS Surveillance in California California Department of Public Health Office of AIDS Guide for Health Care Providers.
Bernard M. Branson, M.D. Associate Director for Laboratory Diagnostics Divisions of HIV/AIDS Prevention National Center for HIV, STD, and TB Prevention.
Preventing HIV/AIDS There is no way to tell just by looking whether a person is infected with HIV. Because people are unaware that they are HIV-positive,
HEAPHY 1 & 2 DIAGNOSTIC James HAYES Fri 30 th Aug 2013 Session 2 / Talk 4 11:33 – 12:00 ABSTRACT To estimate population attributable risks for modifiable.
C-Reactive Protein: a Prognosis Factor for Septic Patients Systematic Review and Meta-analysis Introduction to Medicine – 1 st Semester Class 4, First.
Bernard M. Branson, M.D. Associate Director for Laboratory Diagnostics Divisions of HIV/AIDS Prevention National Center for HIV, STD, and TB Prevention.
HIV Screening and Women’s Health Health Care Education & Training, Inc. Originally developed by: Section 5: Test Options.
Rapid HIV Testing: 2003 Update Bernard M. Branson, M.D. Chief, Lab Determinants and Diagnostics Section Centers for Disease Control and Prevention.
Toward Elimination of Perinatal HIV Transmission in the U.S. Margaret A. Lampe, RN, MPH Division of HIV/AIDS Prevention Centers for Disease Control & Prevention.
Implementing Rapid HIV Testing in the United States Bernard M. Branson, M.D. Centers for Disease Control and Prevention Overview and Background.
Performance of rapid HIV tests used singly and in combination: Moving toward a point of care diagnosis. Kevin Delaney, MPH HIV Diagnostics: New Developments.
Statistics in Screening/Diagnosis
Multiple Choice Questions for discussion
*p
HIV/AIDS is preventable and treatable, but is incurable.
HIV Testing In Vermont Update 2007 Cathleen Harris, MD Fletcher Allen Health Care.
Epidemiology The Basics Only… Adapted with permission from a class presentation developed by Dr. Charles Lynch – University of Iowa, Iowa City.
Implementing a Rapid HIV Testing Guideline for L&D NNEPQIN April 30, 2007.
Welcome Enhanced Perinatal Surveillance (EPS) Meeting.
John W. Hogan, M.D Howard University College of Medicine.
Alliance Discussion with Office of AIDS: November HIV/AIDS Surveillance Surveillance overview HIV Incidence Surveillance Second Surveillance Stakeholder.
HIV and STI Department, Health Protection Agency - Colindale HIV and AIDS Reporting System HIV in the United Kingdom: 2012 Overview.
1 Counseling and HIV Testing HAIVN Harvard Medical School AIDS Initiatives in Vietnam.
Introduction to OraQuick Rapid HIV Testing William F. Ryan Community Health Center School Based Health Program.
Evidence-Based Medicine Diagnosis Component 2 / Unit 5 1 Health IT Workforce Curriculum Version 1.0 /Fall 2010.
Rapid HIV Testing In Labor and Delivery Unit Presented by Danielle Joseph-White Public Health Investigator Specialist Houston Department of Health and.
HIV DISEASE IN PREGNANCY
Screening of diseases Dr Zhian S Ramzi Screening 1 Dr. Zhian S Ramzi.
This material was developed by Oregon Health & Science University, funded by the Department of Health and Human Services, Office of the National Coordinator.
HIV/AIDS is preventable and treatable, but is incurable.
HIV Testing in Medical Settings Mark Thrun, MD Denver Public Health
SCH Journal Club Use of time from fever onset improves the diagnostic accuracy of C-reactive protein in identifying bacterial infections Wednesday 13 th.
Screening.  “...the identification of unrecognized disease or defect by the application of tests, examinations or other procedures...”  “...sort out.
Session: 3 The four pronged approach to comprehensive prevention of HIV in infants and young children Dr.Pushpalatha, Assistant Professor, Dept of Pediatrics,
HIV Testing in Acute Care Settings Rich Rothman, MD, PhD, FACEP CDC, DHHS, OraSure Technologies, Abbott  Historical.
Timothy Wiemken, PhD MPH Assistant Professor Division of Infectious Diseases Diagnostic Tests.
SCREENING FOR DISEASE. Learning Objectives Definition of screening; Principles of Screening.
BioPlex 2200 HIV Ag-Ab Assay
Jan 2002 EDMA The central role of the Medical Laboratory in a World of Managed Health An EDMA presentation of the benefits of in vitro testing as a basis.
Universal Opt-Out Screening for HIV in Health Care Settings, Cost Effectiveness in Action Douglas K. Owens, MD, MS VA Palo Alto Health Care System and.
© 2010 Jones and Bartlett Publishers, LLC. Chapter 12 Clinical Epidemiology.
1 Module 2: HIV Counseling and Testing for PMTCT Ministry of Health/HAPCO, Ethiopia.
DEPARTMENT OF PAEDIATRICS, THE QUEEN ELIZABETH HOSPITAL,
How do we delay disease progress once it has started?
AIDS in the United States
1985: First HIV-1 ELISA Approved by FDA
Presentation transcript:

Robert S. Galen, M.D., M.P.H. Professor and Head, Department of Health Administration, Biostatistics and Epidemiology College of Public Health University of Georgia Athens, Georgia, USA The Predictive Value of Laboratory Tests: Past, Present and Future

United States Atlanta Los Angeles Chicago Dallas New York LOCATION Athens Atlanta State of Georgia 60 miles northeast of Atlanta

HISTORY UGA The University of Georgia America’s oldest state chartered university January 27, 1785 Chartered by Georgia General Assembly The Arch, Symbol of UGA

Aerial photo of Sanford Stadium, holds over 92,000 cheering fans UGA Mascot, UGA VI UGA Quarterback, D.J. Shockley SANFORD STADIUM

Evaluating the Usefulness of Screening Tests Sensitivity, Specificity, Predictive Value Effects of pre-test probability on predictive value Trade-offs in choosing different cut-off values Trade-offs in choosing test combinations Series and Parallel Testing When to test ? Which test is best ? What does the result mean?

Predictive Value Table Number with positive test result Number with negative test result Totals Number with disease TPFNTP + FN Number without disease FPTNFP + TN TotalsTP + FPFN + TNTP + FP + TN + FN

Measures of the Validity of Tests Sensitivity Specificity Predictive value (+) Predictive value (-) Accuracy (efficiency) Sensitivity Specificity Predictive value (+) Predictive value (-) Accuracy (efficiency)

Sensitivität = Positivität bei Krankheit = x 100 Spezifität = Negativität bei Gesundheit = x 100 Prädiktiver Wert eines positiven Tests = x 100 TP TP+FN TN TN+FP TP TP+FP

Effect of prevalence on predictive value when sensitivity and specificity equal 95% Prevalence of disease (%) Predictive value of a positive test (%)

Effect of prevalence on predictive value when sensitivity and specificity equal 99% Prevalence (%) Predictive value of a positive test (%)

Sensitivity and Specificity of Anti-HIV Tests in One Study ReagentSensitivity (%)Specificity (%)Sum (%) Abbott, Routine Abbott, p Abbott, p Abbott, p24 & p Abbott, p24 or p41 or both duPont Electro-Nucleonics, Routine Electro-Nucleonics, Sorin Litton Organon Pasteur Wellcome

Coin flip data arranged in tabular format Number with positive coin flip Number with negative coin flip Totals Renovascular hypertension ,000 No renovascular Hypertension 45,000 90,000 Totals50, ,000

Combination testing for hypothetical data Test(s)Sensitivity (%)Specificity (%) Single Series Parallel A B A and B A or B

ROC Curves

Area under the ROC curve = Standard error = % Confidence interval = to P (Area=0.5) =

Predictive Value Number with positive test result Number with negative test result Totals Number with disease pap(1 – a)p Number without disease (1 – p)(1 – b)(1 – p)b1 – p Totalspa + (1 – p)(1 – b)p(1 – a) + (1 – p)b1

IN MEMORIUM Bayes, Reverend Thomas. An essay toward solving a problem in the doctrine of chance. Philo. Trans. Roy. Soc. 53: , 1763.

Laboratory advances in the predictive value of tests can drive clinical applications and improve quality of care: a) protecting the nation’s blood supply b) screening high risk patients c) screening all pregnant women d) screening everyone ? There are some challenges, however, as tests improve. The best test may not always be the most useful in a particular clinical situation. The HIV Epidemic

Number HIV infected 850, ,000 Number unaware of their HIV infection 180, ,000 Awareness of Serostatus among Persons with HIV, United States

HIV Testing Challenges “In the United States, 32 % of the people who test positive don’t come back for their results.” Dr Bernard Branson, CDC

Four priorities: 1. Make voluntary HIV testing a routine part of medical care 2. Implement new models for diagnosing HIV infections outside medical settings 3. Prevent new infections by working with persons diagnosed with HIV and their partners 4. Further decrease perinatal HIV transmission Advancing HIV Prevention: New Strategies for a Changing Epidemic MMWR April 18, 2003

Four FDA-approved Rapid HIV Tests Sensitivity (95% C.I.) Specificity (95% C.I.) OraQuick Advance - whole blood - oral fluid - plasma 99.6 ( ) 99.3 ( ) 99.6 ( ) 100 ( ) 99.8 (99.6 – 99.9) 99.9 (99.6 – 99.9) Uni-Gold Recombigen - whole blood - serum/plasma 100 (99.5 – 100) 99.7 (99.0 – 100) 99.8 (99.3 – 100)

OraQuick Advance HIV-1/2 CLIA-waived for finger stick, whole blood, oral fluid; moderate complexity with plasma Store at room temperature Screens for HIV-1 and 2 Results in 20 minutes

Obtain finger stick specimen…

Insert loop into vial and stir

Collect oral fluid specimens by swabbing gums with test device. Gloves optional; waste not biohazardous

Insert device; test develops in 20 minutes

Positive Negative Positive HIV-1/2 Reactive Control Read results in 20 – 40 minutes

Remember the tradeoffs… Good News: More HIV-positive people receive their test results. Bad News: Some people will receive a false-positive result before confirmatory testing.

Positive Predictive Value of a Single Test Depends on Specificity & Varies with Prevalence Test Specificity HIV Prevalence Predictive Value, Positive Test 10% 99% 98% 92% 5% 98% 96% 85% 2% 95% 91% 69% 1% 91% 83% 53% 0.5% 83% 71%36% 0.3% 75%60% 25% 0.1% 50% 33% 10% OraQuickSingle EIAReveal 99.9%99.8%99.1% 97% 95% 87% 77% 63% 50% 25% Uni-Gold 99.7%

HIV Screening with OraQuick in Labor and Delivery: the MIRIAD Study Testing of pregnant women in labor for whom no HIV test results are available; 16 hospitals in 6 cities: Atlanta, Baton Rouge, Chicago, Miami, New Orleans, New York Results:  4849 women screened  34 (0.7%) new HIV infections identified  Sensitivity 100 % Specificity 99.9 %  Positive Predictive value: OraQuick 90%; EIA 76% Bulterys et al, JAMA 2004; 292:

USPSTF Recommends Screening All Pregnant Women for HIV “The United States Preventive Services Task Force continues to recommend screening all adolescents and adults at high risk for HIV and now also recommends screening all pregnant women.”

USPSTF Recommends Screening All Pregnant Women for HIV Evidence is good that pregnant women find recommended regimens of highly antiretroviral therapy (HAART) to be acceptable, and that HAART significantly lowers rates of mother-to-child transmission. Early diagnosis of maternal HIV infection also facilitates discussion of elective cesarean section and avoidance of breast-feeding, which may lower HIV transmission rates.

HIV Screening with OraQuick in Labor and Delivery: the MIRIAD Study The CDC now recommends routine rapid HIV testing using an opt-out approach (ie, a woman is informed that HIV testing will be routinely done during labor if her HIV status is unknown but she may decline testing). Rationale: there is a brief window of opportunity for interventions to decrease HIV transmission to the newborn Bulterys et al, JAMA 2004; 292:

HIV Screening with OraQuick in Labor and Delivery: the MIRIAD Study “In many settings, including in the developing world, pregnant women with unknown HIV status are often seen by clinicians for the first time during labor. Rapid testing during labor can enable pregnant women with undocumented HIV status to learn their HIV infection status so they can receive antiretroviral prophylaxis and be referred for comprehensive medical care and follow-up.” Bulterys et al, JAMA 2004; 292:

Nucleic acid amplification testing Two steps forward and one step back: Do antibody tests miss cases that could otherwise be diagnosed? We know there have been false-negatives during the acute infection period. How bad is the problem ?

Nucleic acid amplification testing “ In this study, we found that antibody tests alone detected only 96 percent of HIV infections, as compared with an algorithm that included nucleic acid amplification tests to detect acute HIV infections.” Pilcher et.al. NEJM 2005; 352:

Nucleic acid amplification testing “The addition of nucleic amplification testing to an HIV testing algorithm significantly increases the identification of cases of infection without impairing the performance of diagnostic testing. The detection of highly contagious, acutely infected persons creates new opportunities for HIV surveillance and prevention.” Pilcher et.al. NEJM 2005; 352:

Nucleic acid amplification testing “ We believe that the work that has been done to date…is now sufficient to allow us to conclude that this form of testing should be a standard tool for the prevention and surveillance of HIV infection and for the care of infected persons.” Pilcher et.al. NEJM 2005; 352:

Nucleic acid amplification testing What is the value of a test, if patients don’t get the result ? In some clinical situations the best test may not solve the problem! Tests must be selected in the clincial context they will be used, not in a vacuum!

Routine Population-Wide HIV Screening May Be Cost-Effective “The findings of Paltiel et al. and Sanders et al. show that, given the availability of effective therapy and preventive measures, it is possible to improve care and perhaps influence the course of the epidemic through widespread, effective, and cost-effective screening.” Bozzette S.A. N.E.J.M. 2005; 352:

Routine Population-Wide HIV Screening May Be Cost-Effective “Failure to implement widespread routine screening for HIV infection represents a critical disservice to patients who are currently infected, those at risk for infection, and the future health of the nation.” Bozzette S.A. N.E.J.M. 2005; 352:

What is a risk factor? The traits, factors, and characteristics that predispose to the development of atherosclerosis have been collectively termed “risk factors.” Not all risk factors are useful laboratory tests

Not all risk factors are useful lab tests Predicting risk is a different problem than classification. We classify patients to decide to treat or not. If we want to use risk factors this way, they need to be evaluated by the test characteristics of sensitivity, specificity, predictive value and ROC curve analysis. This was done and published in July, 2006.

ROC Curves

July, 2006: A challenging month for novel risk factors Is routine screening of CRP warranted? Folsom et al. Arch Intern Med 166: Lloyd-Jones & Tian Lloyd-Jones et al. Annals: 145:35-42 Cook et al. Annals: 145: Smith et al. Annals: 145: 70-72

ARIC Study shows little benefit of measuring 19 novel risk factors 15,792 adults followed since Participants had physical exam, assessment of major risk factors and novel markers Inflammation, endothelial function, fibrin formation, fibrinolysis, B vitamins, antibodies to infectious agents Folsom et al. Arch Intern Med. 2006; 166:

ARIC Study shows little benefit of measuring 19 novel risk factors Change in AUC was used to assess the additional contribution of novel risk markers to CHD prediction beyond that of traditional risk factors Traditional risk factor model predicted CHD well: AUC 0.8 CRP did not add significantly to the AUC and neither did most other novel risk factors Folsom et al. Arch Intern Med. 2006; 166:

ARIC Study shows little benefit of measuring 19 novel risk factors “C-reactive protein level does not emerge as a clinically useful addition to basic risk factor assessment for identifying patients at risk of a first CHD event.” “Routine measurement of these novel markers is not warranted for risk assessment.” Folsom et al. Arch Intern Med. 2006; 166:

Predicting Cardiovascular Risk “On their face, these results appear astonishing. Can it really be that 19 of the most exciting new markers for CVD, about which there are thousands of published articles and on which entire careers have been based, do not add anything substantial to risk prediction for CVD?” Lloyd-Jones, Tian. Arch Intern Med. 2006; 166:

Predicting Cardiovascular Risk Age adjusted hazard ratios may be similar to those of traditional risk factors Traditional risk factors have been proven to have a causal role in CHD and are targets of therapy They must form the basic risk prediction model Lloyd-Jones, Tian. Arch Intern Med. 2006; 166:

Predicting Cardiovascular Risk A new marker is useful only if it corrects a substantial portion of misclassification by the existing risk score, which the 19 novel markers in the ARIC study did not do. Lloyd-Jones, Tian. Arch Intern Med. 2006; 166:

Predicting Cardiovascular Risk The established risk factors should remain the focus of CHD risk estimation and prevention for now and routine measurement of any of these 19 novel markers for the entire population cannot be recommended. Lloyd-Jones, Tian. Arch Intern Med. 2006; 166:

Predicting Cardiovascular Risk “There is no definitive evidence (in the literature) that, for most individuals, CRP adds substantial predictive value above that provided by risk estimation using traditional risk factors for CVD.” Lloyd-Jones et al. Ann Intern Med. 2006; 145: 35-42

Quality Laboratory Testing Doing the right test (precisely and accurately) in the right patient at the right time to effectuate a positive clinical outcome in a cost-effective way. Much of the testing we engage in does not meet these requirements.

Quality Laboratory Testing In the future, laboratory tests will be evaluated in the context of the clinical decision making they contribute to, not just in the laboratory! Quality testing will be integrated into cost effective medical care, based on sound evidence and clinical trials.