Health Data and What It Means

• A. 0.2%
• B. 2% ✓
• C. 20%
• D. 50%

Why: Prevalence is the number of cases divided by the total population. 20 divided by 1,000 equals 0.02, which is 2%.

• A. Prevalence
• B. Incidence ✓
• C. Mortality rate
• D. Median

Why: Incidence measures new cases arising during a set time period, while prevalence counts all existing cases at a given point in time.

• A. 50%
• B. 2 percentage points ✓
• C. 2%
• D. 4 percentage points

Why: Absolute risk reduction is the simple difference between the two risks: 4% minus 2% equals 2 percentage points. The 50% figure is the relative risk reduction.

• A. The estimate is definitely correct 95% of the time
• B. The true value is likely within that range, based on the data and method used
• C. 95% of the population falls within that range
• D. The study included 95 participants

• A. Small samples are always biased toward higher numbers
• B. Larger samples reduce random variation, giving more stable and precise estimates
• C. Small samples cannot be analyzed with percentages
• D. Larger samples always show the same result as the population

• A. Secondary data
• B. Primary data ✓
• C. Administrative data
• D. Vital statistics

Why: Primary data is collected firsthand by the researcher for a particular study or purpose, such as a survey designed to measure smoking habits. Secondary data, by contrast, was originally gathered for another purpose and is reused.

• A. Responses from a national household survey on diet
• B. Hospital discharge records and insurance claims ✓
• C. Blood samples collected in a clinical trial
• D. Notes from a focus group on access to care

Why: Administrative data is generated as a byproduct of providing and paying for health services, such as billing records, insurance claims, and hospital discharge records. It was not created primarily for research.

• A. To track individual patients' treatment plans
• B. To measure population-level health events like births, deaths, and marriages ✓
• C. To evaluate the effectiveness of new medications
• D. To record employee health benefits

Why: Vital statistics are collected by government agencies to monitor population-level events such as births, deaths, marriages, and divorces. They support public health planning and tracking of mortality and natality trends.

• A. A single national database containing every citizen's health history
• B. Digital versions of patient charts maintained by healthcare providers ✓
• C. Surveys completed by patients online
• D. Publicly available census data on population health

Why: EHRs are digital records of a patient's clinical information, such as diagnoses, medications, lab results, and visit notes, maintained within a healthcare system. They are not typically a single centralized national database.

• A. Always includes complete medical records
• B. Can be affected by self-report bias, where respondents may inaccurately recall or report behaviors ✓
• C. Cannot be used to estimate population prevalence
• D. Is always collected in real time by machines

Why: Survey data relies on respondents' self-reports, which can be affected by recall errors, social desirability bias, or misunderstanding of questions. Researchers must account for this when interpreting results.

• A. The total number of people who have a disease at a single point in time
• B. The number of new cases of a disease that develop in a population over a specified period ✓
• C. The percentage of deaths caused by a disease in a population
• D. The average age at which people are diagnosed with a disease

Why: Incidence counts new cases arising during a defined time period among people at risk, which distinguishes it from prevalence, the total existing cases at one point in time.

• A. The rate at which new cases appear each year
• B. The proportion of a population that has a condition at a specific point in time ✓
• C. The likelihood that a treatment will succeed
• D. The number of deaths divided by the number of births

Why: Prevalence is a snapshot of how many people have a condition at one moment, including both new and long-standing cases.

• A. 3 per 1,000
• B. 30 per 1,000 ✓
• C. 6 per 1,000 per year
• D. 300 per 1,000

Why: Cumulative incidence is the number of new cases divided by the population at risk: 30/1,000, or 30 per 1,000 over the five-year period. Dividing by five years would give an annual rate, not the cumulative figure.

• A. A guaranteed cause of a disease
• B. Any attribute or exposure associated with an increased likelihood of developing a disease ✓
• C. A test used to confirm a diagnosis
• D. The number of people who recover from a disease

Why: A risk factor is associated with a higher probability of disease. It does not guarantee the disease will occur, and an association does not by itself prove causation.

• A. Relative risk ✓
• B. Mean age
• C. Median household income
• D. Standard deviation

Why: Relative risk divides the risk in the exposed group by the risk in the unexposed group, showing how much more (or less) likely disease is with the exposure.

• A. To ensure the sample size is large enough for statistical significance
• B. To balance known and unknown confounding factors between treatment and control groups ✓
• C. To make the trial faster and less expensive to conduct
• D. To allow participants to choose which group they join

Why: Random assignment distributes both measured and unmeasured confounding variables roughly evenly across groups, so any difference in outcomes can more confidently be attributed to the intervention rather than to pre-existing differences between participants.

• A. 2%
• B. 10 percentage points ✓
• C. 50%
• D. 100%

Why: Absolute risk increase is the simple difference between the two event rates: 20% minus 10% equals 10 percentage points. The relative risk increase would be 100%, which is why absolute and relative measures should be distinguished when interpreting results.

• A. There is a 95% probability the true effect lies within the interval
• B. 95% of patients will respond to the treatment
• C. If the study were repeated many times, about 95% of the intervals constructed would contain the true effect ✓
• D. The treatment is effective 95% of the time

Why: The confidence level describes the long-run performance of the method used to build the interval, not the probability that a specific computed interval contains the true parameter. The true value is fixed; the interval is what varies across repeated samples.

• A. To reduce the cost of manufacturing the placebo
• B. To prevent participants and investigators from having expectations that could bias outcomes or their measurement ✓
• C. To hide the identity of the sponsoring company from regulators
• D. To ensure all participants receive the same dose

Why: When people know who received the active treatment, their expectations can influence reported symptoms, care decisions, and how outcomes are assessed. Blinding reduces these performance and detection biases, making the comparison between groups more trustworthy.

• A. There is a 4% chance the treatment does not work
• B. There is a 96% chance the treatment works
• C. If there were truly no difference, data this extreme or more extreme would occur about 4% of the time ✓
• D. The treatment reduces blood pressure by 4 points

Why: A p-value is the probability of observing data at least as extreme as those obtained, assuming the null hypothesis is true. It does not give the probability that the treatment works, nor does it measure the size of the effect, which requires separate estimation.

• A. How much does the supplement cost per bottle?
• B. What was the study design, and how many participants were in it? ✓
• C. Is the packaging colorful and appealing?
• D. How many years has the company been in business?

Why: Study design and sample size determine how much weight a claim deserves. A small, uncontrolled study or one funded by the seller can produce dramatic-sounding results that don't hold up at scale.

• A. Whether the headline uses the word 'new'
• B. Whether the study was observational or a randomized controlled trial ✓
• C. Whether the reporter is famous
• D. Whether the article includes a photo of coffee

Why: Observational studies can show associations, but they cannot prove cause. Coffee drinkers may differ from non-drinkers in many other ways, so only randomized trials can support a causal claim.

• A. The testimonial is too short
• B. A single person's experience cannot show the product works for everyone, and back pain often improves on its own ✓
• C. Herbal products are never effective
• D. The person should have used a larger cup

Why: Individual anecdotes are unreliable because of natural recovery, placebo effects, and individual variation. Back pain frequently resolves over time regardless of treatment.

• A. What color is the product label?
• B. Compared to what, and what was the actual number of people who improved versus those who didn't? ✓
• C. Who designed the logo?
• D. Is the product available in stores?

Why: Percentages without a baseline are misleading. '90% effective' could mean a drop from 2 cases to 1 case, or it could compare against a meaningful control group. The absolute numbers and comparison group give the real picture.

• A. The site uses a web address ending in .com
• B. It uses emotional urgency and secrecy while promising a dramatic cure without citing credible evidence ✓
• C. It includes a date on the page
• D. It has more than one paragraph of text

Why: Sensational language, claims of hidden knowledge, promises of miraculous cures, and a lack of verifiable sources are common hallmarks of health misinformation.

• A. Any health-related information that is publicly available online
• B. Individually identifiable health information held or transmitted by covered entities or their business associates ✓
• C. Only diagnostic codes used by insurance companies
• D. Aggregated statistics about population health trends

Why: PHI refers to individually identifiable health information that is created, received, stored, or transmitted by covered entities or their business associates. Aggregated, de-identified statistics generally fall outside this definition.

• A. To make the data easier to store on hard drives
• B. To remove or obscure information that could reasonably identify a specific individual ✓
• C. To increase the number of diagnoses recorded in the file
• D. To ensure the data can be sold to advertisers

Why: De-identification removes direct identifiers (such as names and Social Security numbers) and reduces indirect identifiers so that individuals cannot reasonably be linked to their health information, protecting privacy while allowing research to proceed.

• A. The right to delete any entry a doctor has written
• B. The right to access and obtain a copy of their health records ✓
• C. The right to control which hospital employees can see their lab results
• D. The right to prevent providers from recording any diagnosis

Why: HIPAA gives patients the right to inspect and receive copies of their health records held by covered entities. They can also request amendments to inaccurate information, but they cannot unilaterally delete entries written by providers.

• A. Health data should always be stored in the smallest possible file format
• B. Only the least amount of information needed to accomplish a specific purpose should be used or disclosed ✓
• C. Patients must provide the minimum amount of information during every visit
• D. Hospitals must keep records for at least the minimum number of years set by law

Why: The minimum necessary principle requires that uses and disclosures of PHI be limited to what is reasonably necessary for the intended purpose, reducing unnecessary exposure of sensitive information.

• A. Health records are always stored on paper, making them easy to lose
• B. Health information is sensitive and can lead to identity theft, discrimination, and lasting personal harm that cannot be easily reversed ✓
• C. Breaches of health data only affect hospital finances
• D. Health records expire quickly, so breaches have little long-term impact

Why: Health information is highly sensitive and permanent. Exposure can enable medical identity theft, fraud, discrimination in employment or insurance, and embarrassment, and a person cannot change their medical history the way they can change a password or credit card.

• A. Your heart is failing and you should stop exercising immediately
• B. A possible improvement in cardiovascular fitness, though one reading alone should not be treated as proof ✓
• C. The device is definitely broken and the data is worthless
• D. Your blood pressure has doubled

Why: A lower resting heart rate over time is often associated with improved cardiovascular fitness. However, wearable data can be affected by many factors, so it is a trend to monitor rather than a definitive diagnosis.

• A. The mean, because it is always the most accurate average
• B. The median, because it is less affected by extreme values ✓
• C. The range, because it shows the highest number
• D. The mode, because it always represents the middle of the data

Why: The median is the middle value and is resistant to outliers. Extreme values can pull the mean upward, making the median a better summary of typical values in skewed data.

• A. Whether the study was published in a famous journal
• B. Whether the result is absolute risk or relative risk, and how large the baseline risk was ✓
• C. Whether coffee tastes good to most people
• D. Whether the researchers were paid by a coffee company

Why: A 20% relative reduction can mean very different things depending on the baseline risk. If the original risk was 1 in 1,000, a 20% reduction is a small absolute change. Understanding absolute versus relative risk helps you judge real-world significance.

• A. Ignore it because weekend data is always unreliable
• B. Look for a consistent pattern over several weeks and set a specific, realistic goal to address it, such as a short walk each weekend day ✓
• C. Immediately buy a more expensive tracker
• D. Assume you are unhealthy and stop walking altogether

Why: Good data-informed decisions rely on patterns observed over time and lead to specific, achievable actions. Tracking a consistent pattern and setting a concrete goal turns raw data into a useful behavior change.

• A. Because the test is only valid if the person is wearing a watch
• B. Because fasting status changes what counts as a normal range, so the same number can mean different things in different situations ✓
• C. Because glucose numbers never change during the day
• D. Because context only matters for cholesterol, not glucose

Why: Reference ranges depend on conditions such as fasting or after a meal. Without knowing the context, a reading cannot be compared properly to guidelines, which is why health data must be interpreted alongside its circumstances.