1. CPH Domain 1: Data Sources and Epidemiologic Measures
After this lesson you can
- Differentiate primary from secondary data and match sources such as BRFSS, NHANES and vital statistics to a question
- Distinguish validity from reliability and name content, criterion and construct validity
- Spot item-writing flaws such as double-barreled and leading questions
- Apply range checks, logic checks, duplicate checks and missing-data codes when cleaning data
- Compute and interpret prevalence, incidence, case fatality and age-adjusted rates
- Calculate relative risk, odds ratio, attributable risk and number needed to treat from a two-by-two table
The first of two lessons on Domain 1, Data, Analytics, and Informatics. It covers primary and secondary data sources, validity and reliability, survey design, data cleaning, and the epidemiologic measures of morbidity, mortality and risk, worked through Harbor County examples.
Questions and answers
Questions a learner would ask about this lesson. Think through your own answer before you open each one.
- What is the difference between primary and secondary data?
Show answer
Primary data are data you collect yourself for your specific question, such as surveys, interviews, focus groups or clinical measurement, so you control the instrument, timing and population. Secondary data were collected by someone else for another purpose, such as vital records, registries, claims or national surveys. Secondary data are cheaper and larger, but you inherit their definitions, so fit to your question is the risk.
- How do I tell validity from reliability on the exam?
Show answer
Validity is whether an instrument measures what it intends to measure, which is accuracy. Reliability is whether it gives consistent results, which is consistency. An instrument can be reliable and invalid, which is the signature of systematic bias, but it cannot be valid and unreliable. Both properties belong to a tool used with a specific population for a specific purpose.
- Which national data source should I pick for which question?
Show answer
Use BRFSS, the ongoing telephone survey of adults, for self-reported behaviors and conditions at the state level. Use NHANES, which adds a physical exam and lab tests, when the condition may be undiagnosed. Use vital statistics for deaths. Remember that hospital discharge and claims data count encounters, not people, so one patient admitted four times appears four times.
- What is the difference between prevalence and incidence?
Show answer
Prevalence counts all existing cases, old and new, over the total population at a point or period in time. Incidence counts only new cases among the population at risk. Cumulative incidence divides new cases by people at risk, and the incidence rate divides new cases by person-time. Prevalence rises when people survive longer with a disease.
- What do the Harbor County two-by-two numbers show?
Show answer
Of 1,000 adults with prediabetes followed for a year, 60 of 400 not enrolled developed diabetes (15 per 100) and 30 of 600 enrolled did (5 per 100). The relative risk is 15 over 5, or 3, and the attributable risk is 10 per 100. The number needed to treat is 1 divided by 10 per 100, which is 10.
Practice quiz
- Question 1A county wants to estimate how many adults have high blood pressure that has never been diagnosed. Which source fits best?
- Question 2Which statement about the denominator of cumulative incidence is correct?
- Question 3In the Harbor County cohort, the one-year risk of diabetes was 15 per 100 among adults not enrolled in the program and 5 per 100 among those enrolled. What is the relative risk for not enrolling?
- Question 4Using the same Harbor County numbers, how many adults would need to enroll to prevent one case of diabetes in a year?
- Question 5A study begins by selecting people who already have a disease and comparable people who do not, then looks back at exposure. Which measure of association does it report?
- Question 6In an invented Harbor County outbreak there are 200 confirmed cases and four deaths. What is the case fatality rate?
- Question 7A clinic scale gives the same reading every time a person steps on it, but the reading is always several pounds lower than the measured weight on a calibrated scale. The scale is best described as:
- Question 8A depression screener is judged by how well it forecasts a diagnosis made months later. Which type of validity is being assessed?
- Question 9During cleaning, an analyst flags a record in which the respondent reports never having smoked but also reports smoking twenty cigarettes a day. This is an example of a:
- Question 10Harbor County's crude death rate is about 288 per 100,000 and Bay County's is about 145, yet their age-specific rates are identical and both age-adjusted rates are 240 per 100,000. What explains the gap in crude rates?
Saved in your browser only: no account, no server.
CPH Exam Pathway is an independent course produced by Subthesis for educational purposes. It is not affiliated with or endorsed by the National Board of Public Health Examiners (NBPHE), and completing it does not confer certification. Confirm fees, eligibility, and exam rules against the current NBPHE candidate handbook.