1 / 50

Risks and Benefits of Home-Use HIV Test Kits

Risks and Benefits of Home-Use HIV Test Kits. Richard Forshee, Ph.D. U.S. Food and Drug Administration Center for Biologics Evaluation and Research Office of Biostatistics and Epidemiology BLOOD PRODUCTS ADVISORY COMMITTEE 96th Meeting, November 16-17, 2009 Bethesda Marriott Hotel

cornell
Télécharger la présentation

Risks and Benefits of Home-Use HIV Test Kits

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Risks and Benefits of Home-Use HIV Test Kits Richard Forshee, Ph.D. U.S. Food and Drug Administration Center for Biologics Evaluation and ResearchOffice of Biostatistics and Epidemiology BLOOD PRODUCTS ADVISORY COMMITTEE 96th Meeting, November 16-17, 2009 Bethesda Marriott Hotel 5151 Pooks Hill Rd., Bethesda, MD 20814

  2. Acknowledgments • Mark Walderhaug, Hong Yang, and Arianna SimonettiOffice of Biostatistics and Epidemiology • Elliot Cowan and Hilary HoffmanOffice Of Blood Research & Review • Bernard Branson, Arielle Lasry, and Stephanie SansomCenters for Disease Control and Prevention

  3. Purpose of Model 1 • Design a tool to estimate the public health benefits and risks of a home-use HIV test kit under different test characteristics • If benefits and risks could be quantified, it would be possible to define a region of sensitivity and specificity where the benefits exceed the risks

  4. Purpose of the Model 2 • Benefits and risks of different test outcomes are not comparable, so professional judgment is required • The tool will help decision-makers make informed judgments about the likely tradeoffs • We seek the committee’s input on these issues

  5. Public Health Benefits and Risks Considered by Model • Benefits • True Positive results • True Negative results • Risks • False Negative results • False Positive results • Failed Tests (invalid results or user error) • Benefits and risks are in the context of persons who would not otherwise be tested

  6. Qualitative Summary of Public Health Benefits and Risks of Different Test Outcomes for a Home-Use HIV Test Kit

  7. True Positive • Allows earlier medical intervention and entry into care • Knowledge of individual HIV status allows for behavior modification to prevent HIV transmission, and allows for partner notification • Knowledge of HIV prevalence can allow for better targeting of public health resources as cases come to medical attention

  8. True Negative • Peace of mind • Assistance in appropriate targeting of public health resources

  9. False Positive • Unnecessary personal anxiety • Additional testing required • Will people seek additional testing?

  10. False Negative • False reassurance • Unsuspected transmission of virus and continued high risk behavior • Delayed medical intervention • False negative results may include tests taken during window period

  11. Failed Test, Person is HIV+ • Another test performed • Delay in diagnosis • Additional testing • No further testing • False reassurance • Unsuspected transmission of virus and continued high risk behavior • Delayed medical intervention

  12. Failed Test, Person is HIV- • Another test performed • Possibility for personal anxiety • Delay in determining status • Additional testing • No additional testing • Personal anxiety • Status remains unknown

  13. Subpopulations Modeled • Low Risk Heterosexuals (LRH):0 or 1 partner in previous year • High Risk Heterosexuals (HRH): previous year • Men Who Have Had Sex With Men (MSM): previous year • Injectable Drug Users (IDU):previous year

  14. Time Frame of the Model • Model estimates annual rates for 2nd and succeeding years after introduction by considering currently active at-risk group (with risky behaviors in past 12 months) as regular users of the test kit • Use of test kit might be higher or lower in the 1st year after introduction

  15. Global Inputs • Percent of test kits that fail (5%) • Sensitivity of test kits that did NOT fail • Specificity of test kits that did NOT fail These are treated as hypothetical and will be estimated across a range of plausible values. Sensitivity does not include window period cases.

  16. Sensitivity Values Modeled

  17. Specificity Values Modeled

  18. Sub-population Specific Inputs • Size of sub-population • Percent of sub-population that is untested (previous year) • Percent of untested persons who are HIV+ • Percent of untested persons who would use a home-use HIV test (highly uncertain)

  19. Sub-population Inputs Presenting means only Uncertainty was included in model

  20. Sub-population Size

  21. Percent Untested

  22. Percent Untested HIV+

  23. Percent Untested Who Would Use a Home-Use Test Kit

  24. Size of the Sub-population % Untested Number Untested % HIV + HIV + HIV - Structure of the Model 1

  25. HIV + % Using Test Number Using Test % Failed Failed Test Sensitivity False Negative True Positive Structure of the Model 2

  26. Model Results

  27. Absolute Numbers:Model Results Totals for all sub-populations Persons who would not otherwise be tested

  28. Test Sensitivity

  29. Test Specificity

  30. Using Ratios to Explore Tradeoffs • As a surrogate for quantitative estimates of benefit and risk, ratios of benefits to risks can provide some insight into public health tradeoffs • We will explore the number of beneficial test outcomes to the number of adverse test outcomes • There is no accepted threshold value for any of these ratios • Professional judgment is required to determine what ratio is acceptable

  31. True Positive to False Negative • The ratio indicates how many newly identified True Positive results are obtained for each False Negative result • True Positive and False Negative results are related solely to the sensitivity of the test and the number of HIV+ persons taking the test • Failed tests for HIV+ persons are counted as False Negative results in this calculation

  32. Tradeoff between True Positive Test Results and False Negative Test Results

  33. True Negative to False Positive • The ratio indicates how many True Negative results are obtained for each False Positive • True Negative and False Positive results are related solely to the specificity of the test and the number of HIV- persons taking the test • Failed tests for HIV- persons are counted as False Positive results in this calculation

  34. Tradeoff between True Negative Test Results and False Positive Test Results

  35. True Positive to False Positive • True Positive and False Positive results are related to both the sensitivity and the specificity of the test as well as the number of HIV+ and HIV- persons taking the test • The ratio indicates how many True Positive results are obtained for each False Positive

  36. True Positive to False Positive Ratio Sensitivity Specificity Under current assumptions, the ratio of true positive to false positive test results is more strongly related to specificity than to sensitivity.

  37. Current BPAC Recommended Performance • The lower bound of the 95% confidence interval for sensitivity and specificity must be greater than or equal to 95% • What is the mean sensitivity and specificity needed to meet this? • Assume a 2% prevalence and a study population consisting of 100 positive subjects and 4,900 negative subjects

  38. Test Sensitivity to Meet BPAC Recommended Lower Bound of the 95% CI

  39. Test Sensitivity to Meet BPAC Recommended Lower Bound of the 95% CI

  40. High Sensitivity, High Specificity Typical Currently Approved Rapid Tests 7.6 Low Sensitivity, Low Specificity BPAC Minimum (see assumptions) 0.34 Ratio of True+ to False+ by Characteristics Estimated ratio of true + to false + test results as a function of sensitivity and specificity of tests. Colors represent a temperature map. Assumes 111,000 HIV+ test users and 7,200,000 HIV- test users (mean values from the simulation). Specificity has a strong relationship with the ratio.

  41. More HIV+ and Fewer HIV- Users Typical Currently Approved Rapid Tests 10.4 BPAC Minimum (see assumptions) 0.47 Ratio of True+ and False+ by Test Characteristics. Estimated ratio of true + to false + test results as a function of sensitivity and specificity of tests. Colors represent a temperature map. Assumes 126,000 HIV+ test users and 6,000,000 HIV- test users (75 and 25 percentile values from the simulation). Specificity has a strong relationship with the ratio.

  42. Summary • We have presented a tool to explore the public health benefits and risks of a home-use HIV test kit with different test characteristics • Tool provides estimates of the absolute numbers of each type of test result and three ratios to assist in making informed judgments about public health tradeoffs • Each model result is important for determining the overall public health impact

  43. Discussion 1 • Several of the input parameters have significant uncertainty because of a lack of data • Our estimates that a large number of HIV- persons will use the test implies that high specificity will be required to minimize False+ results • Ratio of True + to False + is strongly related to specificity of the test under current assumptions • Ratio of True+ to False- is related to sensitivity of the test

  44. Discussion 2 • Assessing the public health impact of each test outcome requires a value judgment • We seek the committee’s input regarding the appropriate balance between different test outcomes, e.g. how many False- test results would be acceptable for each newly identified True+ result?

  45. Thank you! Richard Forshee, Ph.D. FDA/CBER Richard.Forshee@fda.hhs.gov

  46. Reserve Slides

  47. Sub-population Size National Survey of Family Growth National Survey of Family Growth, various criteria Brady 2008 Anderson et al. 2009

  48. Percent Untested National Health Interview Survey Jenness 2009 Heimer 2007 Assumption

  49. Percent Untested HIV+ Testing of military applicants, CDC website STD clinics, CDC website Entering drug treatment, CDC website Anderson et al. 2009

  50. Percent Untested Who Would Use a Home-Use Test Kit Assumption Assumption Assumption Assumption

More Related