MPH-515 Week 4 Sample Paper

Non-Fatal Opioid Overdose Emergency Department Visits in a Midwestern County: Choosing and Reporting an Incidence Rate, 2023-2024

Marcus T. Oyelaran

Chamberlain University, College of Health Professions

MPH-515 Epidemiology and Biostatistics for Public Health Application I

Dr. R. Naganathan, PhD, MPH

Week 4

April 5, 2026

What this page is doingThe title carries the population, the outcome, the measurement decision, and the window, which is what an epidemiology grader reads a title for. Putting the years inside the title tells the reader the denominator is bounded before a single figure appears. Because Chamberlain publishes no deliverable name for this week, the paper is titled by what it measures rather than by an invented assignment name, and the county is a composite, so no real jurisdiction is being described.

Population, Case Definition, and Data Source

The population under study is the resident population of Bellamy County, a composite Midwestern county of roughly 412,000 residents in 2023 and 415,000 in 2024, which yields 827,000 person-years of observation across a window running from January 1, 2023 through December 31, 2024. Denominators are the annual county population estimates the state health department publishes for rate calculation, and each year's estimate is treated as the midyear population for that year. Non-residents treated in county emergency departments are excluded from the numerator, because they never appear in the denominator, and county residents treated across the state line are counted, because they do.

A case is an emergency department visit by a county resident with a discharge diagnosis of opioid poisoning or a chief complaint coded as suspected opioid overdose, following the syndromic definition the Centers for Disease Control and Prevention publishes for overdose surveillance (Centers for Disease Control and Prevention, 2024). Fatal overdoses that never reached an emergency department are excluded and are counted separately in vital records. The chief complaint arm of the definition is deliberately broad, which raises sensitivity and lowers specificity, and that trade is stated here rather than buried later, because it shapes every number that follows it.

Across the window, county emergency departments recorded 1,489 qualifying visits made by 1,241 distinct residents. The gap between those two counts matters more than it looks. Visits are events, and one resident may contribute several of them; residents are the units at risk. Because this analysis reports incidence, the numerator is the 1,241 residents with a first qualifying visit in the window, and the 248 repeat visits are reported separately as a burden measure rather than folded into the incidence numerator. Mixing the two is the commonest arithmetic error in overdose surveillance reporting, and it inflates a rate by roughly 20 percent in these data.

What this page is doingThe denominator arrives before any rate does, which is the entire reason this sheet exists. Person-years are stated, the residency rule is made to match on both sides of the fraction, and the case definition names its own sensitivity and specificity trade instead of hiding it. The strongest move is separating 1,489 visits from 1,241 residents. Graders watch for that distinction, because collapsing events into persons is the error that quietly inflates most submitted overdose rates.

Choosing the Measure

Prevalence is the wrong measure here, and the reason is definitional rather than practical. Prevalence describes the proportion of a population in a given state at a point in time, which requires the outcome to persist long enough to be present when the population is counted. A non-fatal overdose is an acute event with an onset and an end measured in hours, so there is no state for it to be prevalent in. Incidence, which counts new events against the time a population spends at risk, is the measure that matches the outcome (Centers for Disease Control and Prevention, 2012). Reporting prevalence of overdose would not be a conservative choice; it would be a category error.

Within incidence, the choice sits between cumulative incidence, which needs a fixed population followed for a fixed period, and an incidence rate, which uses person-time. A county population over 24 months is dynamic. People move in, move out, are born, and die, and the denominator is an estimate that changes between the two years. Person-time absorbs that movement in a way a fixed denominator cannot, so the result is reported per 100,000 person-years (Lash et al., 2021). The multiplier is not cosmetic either. It is the convention county health departments use for injury and overdose reporting, which keeps this figure directly comparable to published state and national rates.

For the geographic comparison, both a rate ratio and a rate difference are reported, because they answer different questions and publishing only one hides half the finding. The rate ratio expresses how much more frequently the outcome occurs in one area than in another and travels well between populations of different sizes. The rate difference expresses excess events per 100,000 person-years and is the figure a health department needs when it decides where to place naloxone distribution and outreach staff. An odds ratio is not used, because full population denominators exist here; the odds ratio approximates a rate ratio for designs that lack them, and using it would trade accuracy for nothing.

What this page is doingThis is the sheet that earns the word analysis. Each measure is chosen against a named alternative rather than asserted: incidence over prevalence on definitional grounds, a rate over cumulative incidence because the population is dynamic, ratio and difference together because they answer different questions, and no odds ratio because real denominators exist. Writing the rejected options down is what separates a paper that ran a calculation from a paper that made a decision.

Results

The crude incidence rate of non-fatal opioid overdose emergency department visits among county residents was 150.1 per 100,000 person-years across the window of January 1, 2023 through December 31, 2024 (1,241 residents over 827,000 person-years; 95 percent confidence interval 141.8 to 158.5). Direct standardization to the year 2000 United States standard population lowered the figure to 143.8 per 100,000 person-years, which reflects a county age structure slightly older than the standard against an outcome concentrated in working-age adults (Anderson & Rosenberg, 1998). Both figures are reported, since the crude rate describes the burden the county's emergency departments actually carried and the standardized rate is the one that supports comparison.

Age-specific rates make that concentration visible. Residents aged 25-44 accounted for 627 of the 1,241 cases against 215,000 person-years, a rate of 291.6 per 100,000 person-years, close to double the crude county figure, which tracks national survey patterns placing opioid misuse mainly among adults under 50 (Substance Abuse and Mental Health Services Administration, 2023). Residents aged 65 and older contributed 41 cases against 148,000 person-years, a rate of 27.7 per 100,000 person-years. One county-wide number averages a group at 291.6 with a group at 27.7 and describes neither of them.

The northeast service area, with 156,500 person-years, recorded 372 cases, a rate of 237.7 per 100,000 person-years. The balance of the county, with 670,500 person-years, recorded 869 cases, a rate of 129.6 per 100,000 person-years. The rate ratio is 1.83 (95 percent confidence interval 1.62 to 2.07), and because that interval excludes the null value of 1.00, sampling variation alone is an unlikely explanation. The rate difference is 108.1 excess cases per 100,000 person-years, which converts to about 169 excess cases in the northeast area over the two years and is the figure that belongs in a resource request.

What this page is doingEvery figure appears with its numerator, its denominator, and its window attached, which is the habit graders are checking for. Crude and standardized rates sit side by side with a stated reason for keeping both. Confidence intervals appear next to their point estimates rather than in a table nobody reads, and the rate difference is converted into a count of excess cases, which is the translation that turns an epidemiological result into a budget argument.

Limitations and What This Rate Cannot Support

Three limitations bound every figure above. The first is numerator undercount. Emergency department data capture only the overdoses that reach an emergency department, and reversals handled by bystanders with naloxone, by emergency medical services without transport, or at home never enter the count, and bystander naloxone is now distributed widely enough that this is not a small residual (World Health Organization, 2023). The direction of the bias is known even where its size is not: the true event rate is higher than 150.1 per 100,000 person-years, not lower. A rate built on care-seeking measures care-seeking as much as it measures disease.

The second is denominator error. Intercensal population estimates carry uncertainty of their own, and that uncertainty is largest for exactly the small geographic areas the comparison depends on, so the northeast rate rests on a shakier denominator than the county rate does. The third is the limit of the geographic comparison itself. A higher rate in one service area is a statement about that area, not about the people who live in it, and reading it as evidence about individual behavior would be an ecological inference these data cannot carry (Lash et al., 2021).

What the rate does support is a defensible statement of burden and a defensible allocation argument. The county can say that residents experienced non-fatal overdose at 150.1 per 100,000 person-years over 2023 and 2024, that working-age adults carried the concentration, and that one service area carried 108.1 excess cases per 100,000 person-years against the rest of the county. What it cannot support is a claim about cause. Establishing why the northeast area differs would need a design with individual-level exposure data, which is the reasonable next step rather than a stronger reading of these numbers.

What this page is doingLimitations here are specific to these data, and each one carries a direction. Numerator undercount is named along with the way the bias runs, denominator error is tied to the small-area comparison that depends on it, and the ecological limit is stated before a reader can misread the geographic finding. The sheet then says plainly what the rate does support, so the paper does not end in a general disclaimer that surrenders the findings it just defended.

References

Anderson, R. N., & Rosenberg, H. M. (1998). Age standardization of death rates: Implementation of the year 2000 standard (National Vital Statistics Reports, Vol. 47, No. 3). National Center for Health Statistics. https://www.cdc.gov/nchs/products/nvsr.htm

Centers for Disease Control and Prevention. (2012). Principles of epidemiology in public health practice (3rd ed.). U.S. Department of Health and Human Services. https://www.cdc.gov/csels/dsepd/ss1978/index.html

Centers for Disease Control and Prevention. (2024). Drug Overdose Surveillance and Epidemiology (DOSE) system. U.S. Department of Health and Human Services. https://www.cdc.gov/overdose-prevention/data-research/

Lash, T. L., VanderWeele, T. J., Haneuse, S., & Rothman, K. J. (2021). Modern epidemiology (4th ed.). Wolters Kluwer.

Substance Abuse and Mental Health Services Administration. (2023). Key substance use and mental health indicators in the United States: Results from the 2022 National Survey on Drug Use and Health. U.S. Department of Health and Human Services. https://www.samhsa.gov/data/

World Health Organization. (2023). Opioid overdose [Fact sheet]. World Health Organization. https://www.who.int/news-room/fact-sheets/detail/opioid-overdose

How this MPH 515 Week 4 example is structured

Chamberlain publishes no weekly deliverable names for Epidemiology and Biostatistics for Public Health, so this MPH-515 Week 4 example is written to the genre a fourth week of applied epidemiology typically calls for: take surveillance data for a defined population, calculate and interpret a measure of disease frequency, and defend the measure you chose. Your classroom instructions decide the exact form. The paper is ordered the way an epidemiologist has to think. The denominator sheet comes first, because a rate with an undefined population is not a rate. Measure selection comes second and is argued rather than assumed. Results come third, so every number the reader meets is already anchored to a population and a window. Limitations come last and are specific to these data rather than generic hedging.

MPH-515 Week 4 questions, answered

What does the Week 4 work in MPH-515 usually ask for?

In many sections the fourth week of Epidemiology and Biostatistics for Public Health asks for an applied analysis of surveillance data: define a population, calculate a measure of disease frequency, defend the measure against its alternatives, and state the limitations. Chamberlain publishes no weekly deliverable names, so treat that as the likely genre and follow your classroom instructions on form and length.

Is a case count enough, or do I have to report a rate?

A count with no denominator cannot be compared to anything, so it is not yet a finding. Report the numerator, the population that produced it, and the window over which it accumulated, then express the result per 100,000 person-years or another stated multiplier. Applied epidemiology graders read for that triple, and a paper missing any one of the three loses ground fast.

Should I use real county data or build a scenario?

Either works if the arithmetic is honest. This example uses a composite county so that no real jurisdiction is described, but every rate in it is computed from the counts and person-years actually stated, so a reader can check the math. Whichever you use, name your data source and your case definition; a rate whose case definition is unstated cannot be evaluated.

Write yours, or have the desk draft it

This paper is an original model document written by our desk, not a submitted student paper and not an official Chamberlain University document. Read it for the moves, then write your own to the instructions in your classroom. If you want one built to your exact prompt and rubric, the first custom sample is free and arrives in 24 to 48 hours.

MPH-515 Week 4: sample paper, in real form