Week 3 Seminar - Quantitative Methods PDF

Summary

This document is a seminar on COPHP Quantitative Methods, which covered topics on screening, reliability, validity, and predictive values. It includes practice questions on those topics. The target audience appears to be undergraduate students.

Full Transcript

COPHP: QUANTITATIVE METHODS Winter 2023 Week 3 Agenda Review Q&A Practice When is it appropriate to screen? Important public health problem High prevalence (common) Natural history (disease course/progression) understood Long latency period Early detection improves outcomes A good screening tool exi...

COPHP: QUANTITATIVE METHODS Winter 2023 Week 3 Agenda Review Q&A Practice When is it appropriate to screen? Important public health problem High prevalence (common) Natural history (disease course/progression) understood Long latency period Early detection improves outcomes A good screening tool exists Reliability and Validity Valid Reliable Unreliable Invalid 2x2 Tables for Screening Gold Standard / True Disease Status Diseased Test/Screening Result Positive True Positive Negative False Negative Not Diseased False Positive True Negative Sensitivity and Specificity SENSITIVITY SPECIFICITY Of all people who truly have the disease, what is the probability that our screening tool returns a positive result? Of all people who are truly healthy, what is the probability that our screening tool returns a negative result? 𝑻𝑷 Sensitivity = 𝑻𝑷 + 𝑭𝑵 𝑻𝑵 Specificity = 𝑻𝑵 + 𝑭𝑷 Predictive Values POSITIVE PREDICTIVE VALUE (PPV) NEGATIVE PREDICTIVE VALUE (NPV) Of all people who test positive, what is the probability that they actually have the disease? Of all people who test negative, what is the probability that they are truly healthy? 𝑻𝑷 PPV = 𝑻𝑷 + 𝑭𝑷 𝑻𝑵 NPV = 𝑻𝑵 + 𝑭𝑵 Determinants of Predictive Values Performance of the test Increasing specificity causes biggest increase in positive predictive value Prevalence of the disease in the population With higher prevalence, the same test will have a higher PPV 9 Potential costs of screening Financial Expense Implications of incorrect results False Positive? False Negative? Morbidity Initial Screening Subsequent Procedures Questions? Prepare for Practice: pollev.com/quantblock or Text QUANTBLOCK to 22333 Topics Review Q&A Practice Practice Chlamydia is a sexually transmitted infection. The long-term sequelae of untreated chlamydia can include pelvic inflammatory disease, which can ultimately lead to infertility. Because chlamydia is frequently asymptomatic, especially in people assigned female at birth, it is recommended that sexually active, AFAB individuals under the age of 25 be routinely screened for chlamydia. This is especially true if they have a new sexual partner with whom they are not using a barrier. What type of screening strategy is this? 1. Universal 2. Targeted/Indicated 3. Not enough information Practice Chlamydia is a sexually transmitted infection. The long-term sequelae of untreated chlamydia can include pelvic inflammatory disease, which can ultimately lead to infertility. Because chlamydia is frequently asymptomatic, especially in people assigned female at birth, it is recommended that sexually active, AFAB individuals under the age of 25 be routinely screened for chlamydia. This is especially true if they have a new sexual partner with whom they are not using a barrier. What type of screening strategy is this? 1. Universal 2. Targeted/Indicated 3. Not enough information Practice Seattle Public Schools perform annual vision and hearing checks for all students in 3rd, 6th, and 9th grades.* 1. Universal 2. Targeted/Indicated 3. Not enough information What type of screening strategy is this? * This is a hypothetical example. Please consult your classmates’ LOs for the actual frequency/types of screening offered by SPS. Practice Seattle Public Schools perform annual vision and hearing checks for all students in 3rd, 6th, and 9th grades.* 1. Universal 2. Targeted/Indicated 3. Not enough information What type of screening strategy is this? * This is a hypothetical example. Please consult your classmates’ LOs for the actual frequency/types of screening offered by SPS. Practice Syphilis is a bacterial STI that can also be passed from parent to child during birth (congenital syphilis). Syphilis during pregnancy/at the time of birth is associated with increased risk of stillbirth, neonatal death, and severe (potentially life threatening) bone and neurological deformities. Because syphilis is frequently asymptomatic, especially in people assigned female at birth, the USPSTF recommends that all pregnant persons be screened for syphilis during their pregnancy. What type of screening strategy is this? 1. Universal 2. Targeted/Indicated 3. Not enough information Practice Syphilis is a bacterial STI that can also be passed from parent to child during birth (congenital syphilis). Syphilis during pregnancy/at the time of birth is associated with increased risk of stillbirth, neonatal death, and severe (potentially life threatening) bone and neurological deformities. Because syphilis is frequently asymptomatic, especially in people assigned female at birth, the USPSTF recommends that all pregnant persons be screened for syphilis during their pregnancy. What type of screening strategy is this? 1. Universal 2. Targeted/Indicated 3. Not enough information Practice Which of the following statements about sensitivity is true? The sensitivity of a test… 1. …is influenced by the prevalence of the disease 2. …has a huge impact on a test’s positive predictive value 3. …tells us what proportion of truly sick people will get a positive test result 4. …does none of the above Practice Which of the following statements about sensitivity is true? The sensitivity of a test… 1. …is influenced by the prevalence of the disease 2. …has a huge impact on a test’s positive predictive value 3. …tells us what proportion of truly sick people will get a positive test result 4. …does none of the above Practice Which of the following statements about negative predictive value is true? The negative predictive value of a test… 1. …tells us how likely it is that a person who is truly negative will test negative 2. …tells us how likely it is that a person who tests negative is truly negative 3. …is a characteristic of a test that does not change. 4. …does none of the above Practice Which of the following statements about negative predictive value is true? The negative predictive value of a test… 1. …tells us how likely it is that a person who is truly negative will test negative 2. …tells us how likely it is that a person who tests negative is truly negative 3. …is a characteristic of a test that does not change. 4. …does none of the above Practice The prevalence of hearing loss in people 70+ is 43%. What is the appropriate value of X? Diseased Not Diseased Total Positive Negative 1. 2. 3. 4. 4,300 5,700 430 Not enough information Total X 10,000 Hearing Loss in Older Adults | USPSTF Practice The prevalence of hearing loss in people 70+ is 43%. What is the appropriate value of X? Diseased Not Diseased Total Positive Negative 1. 2. 3. 4. 4,300 5,700 430 Not enough information Total X 10,000 Hearing Loss in Older Adults | USPSTF Practice To detect moderate hearing loss, an 8-question survey has a sensitivity of 68%. What value belongs in box A? Diseased Positive 1. 2. 3. 4. Not Diseased Total 5,700 10,000 A Negative Total 4,300 1,376 3,876 2,924 Not enough information Practice 𝑻𝑷 Sensitivity = 𝑻𝑷 + 𝑭𝑵 To detect moderate hearing loss, an 8-question survey has a sensitivity of 68%. What value belongs in box A? 𝑨 0.68 = 𝟒𝟑𝟎𝟎 Diseased Positive Not Diseased Total A 0.68 ∗ 4300 = A Negative Total 4,300 5,700 10,000 Practice To detect moderate hearing loss, an 8-question survey has a sensitivity of 68%. What value belongs in box A? Diseased Positive Not Diseased 1. 2. 3. 4. Total A 0.68 ∗ 4300 = A Negative Total 4,300 1,376 3,876 2,924 Not enough information 5,700 10,000 Practice The questionnaire has a specificity of 79% (documented in the table below). What is the negative predictive value in this population? 1. 2. 3. 4. Diseased Not Diseased Total Positive 2,924 1,197 4,121 Negative 1,376 4,503 5,879 Total 4,300 5,700 10,000 76.6% 79.0% 70.9% Not enough information Practice 𝑻𝑵 NPV = 𝑻𝑵 + 𝑭𝑵 The questionnaire has a specificity of 79% (documented in the table below). What is the negative predictive value in this population? Diseased Not Diseased Total Positive 2,924 1,197 4,121 Negative 1,376 4,503 5,879 Total 4,300 5,700 10,000 𝟒, 𝟓𝟎𝟑 NPV = 𝟓, 𝟖𝟕𝟗 Practice The questionnaire has a specificity of 79% (documented in the table below). What is the negative predictive value in this population? 1. 2. 3. 4. Diseased Not Diseased Total Positive 2,924 1,197 4,121 Negative 1,376 4,503 5,879 Total 4,300 5,700 10,000 76.6% 79.0% 70.9% Not enough information 𝟒, 𝟓𝟎𝟑 NPV = 𝟓, 𝟖𝟕𝟗 Discussion Questions Practice Physicians and researchers have developed a screening tool for ovarian cancer. The screening consists of a blood draw and a transvaginal ultrasound. Based on this information alone, do you think this is a good screening tool in the general population? Practice Physicians and researchers have developed a screening tool for ovarian cancer. The screening consists of a blood draw and a transvaginal ultrasound. The follow-up diagnostics for a positive result on the test require surgery for confirmation. Do you think this is a good screening tool for use in the general population? Practice The screening consists of a blood draw and a transvaginal ultrasound. Follow-up diagnostics for a positive result on the test require surgery for confirmation. After testing the screening protocol, researchers find that the screening has a low positive predictive value. Do you think this is a good screening tool for use in the general population?