IHP 515 Module 2 Measures Paper Example

From Counts to Rates: Measuring Overdose Deaths in Harlan County

[Student Name]

Southern New Hampshire University

IHP 515: Population-Based Epidemiology

Module Two Paper

[Instructor Name]

[Date]

From Counts to Rates: Measuring Overdose Deaths in Harlan County

At a county board meeting, a commissioner said overdose deaths had nearly doubled and asked whether Harlan County was worse off than its neighbor, Pike County, which recorded 54 deaths in 2023. The honest answer required more than counts. This paper explains how epidemiologists turn counts into rates, why age-specific rates and age adjustment matter, how incidence differs from prevalence and where the data come from, using Harlan County's composite overdose data.

The National Context

Rudd et al. (2016) reported that drug overdose deaths in the United States rose sharply from 2000 to 2014, with the rate of opioid-involved deaths increasing about threefold and the steepest recent increases involving heroin and synthetic opioids other than methadone, largely illicitly made fentanyl. They noted that most overdose deaths involved opioids by the end of the period. Harlan County's rise fits this national pattern, but local measures are needed to guide local action.

Counts and Crude Rates

A count says how many events happened; a rate ties the count to the population at risk over a period. Harlan County's population was about 210,000 in 2019 and 212,000 in 2023. The crude death rate for 2019 was 38 divided by 210,000, multiplied by 100,000, or 18.1 deaths per 100,000 residents per year. For 2023 it was 71 divided by 212,000, multiplied by 100,000, or 33.5 per 100,000. That is an increase of roughly 85%, slightly less than the 87% rise in counts because the population grew.

Pike County, with a population of about 145,000, recorded 54 deaths in 2023, a crude rate of 37.2 per 100,000. On crude rates alone, Pike looks somewhat worse than Harlan despite fewer deaths, which shows why counts mislead when populations differ in size.

Age-Specific Rates

Crude rates can hide patterns across age groups. In 2023, Harlan recorded 8 deaths among residents aged 15 to 24, out of about 28,000, a rate of 28.6 per 100,000; 39 deaths among those aged 25 to 44, out of about 52,000, a rate of 75.0; 21 among those aged 45 to 64, out of about 58,000, a rate of 36.2; and 3 among those 65 and older, out of about 38,000, a rate of 7.9. Adults aged 25 to 44 face a rate more than twice the county's crude rate, which points prevention toward that group.

Table 1. Harlan County Overdose Deaths by Age, 2023

Note. Figures are illustrative for the composite county; children under 15 had no deaths.

Why Age Adjustment Matters

Because overdose rates differ sharply by age, a county with more young adults will tend to have a higher crude rate even if the risk within each age group is the same. Pike County's population is younger than Harlan's, with a larger share aged 25 to 44. To compare fairly, epidemiologists age-adjust rates by applying each county's age-specific rates to a common standard population, a method called direct standardization.

A Simplified Age Adjustment

Using a simplified standard population for residents aged 15 to 64 with weights of 20% for ages 15 to 24, 40% for 25 to 44 and 40% for 45 to 64, Harlan's adjusted rate is 0.20 times 28.6 plus 0.40 times 75.0 plus 0.40 times 36.2, or 50.2 per 100,000. Pike's age-specific rates for the same groups were 24.0, 68.5 and 30.1, giving an adjusted rate of 0.20 times 24.0 plus 0.40 times 68.5 plus 0.40 times 30.1, or 44.2 per 100,000. After adjustment, Harlan's rate is higher than Pike's, reversing the crude comparison. Pike's crude rate looked higher mainly because more of its residents are in the highest-risk age group. In practice, health departments use the full U.S. 2000 standard population across all age groups, but the logic is the same.

Incidence and Prevalence

Grimes and Schulz (2002) describe how research designs relate to measures: cross-sectional surveys estimate prevalence, while cohort designs follow people over time to measure incidence. Incidence tallies fresh events arising in a population across a stretch of time, whereas prevalence tallies all current sufferers, whether measured on one day or across an interval. In 2023, Harlan's emergency departments recorded 412 visits for nonfatal overdoses among residents, an incidence of about 194 visits per 100,000 per year. A county survey estimated that about 2.1% of adults had opioid use disorder, a prevalence of roughly 2,100 per 100,000. Prevalence is shaped both by the flow of new cases and by how long each case persists, so rising prevalence can mean more new cases, longer survival or both.

Where the Data Come From

In the account of Thacker and Berkelman (1988), surveillance means gathering, analyzing and interpreting health data without pause and getting the results quickly to people who can act on them. Harlan's overdose measures draw on death certificates and medical examiner reports for deaths, emergency department data for nonfatal overdoses and periodic surveys for prevalence. Each source has limits: death certificates may not specify drugs involved, emergency data miss overdoses reversed outside hospitals and surveys undercount stigmatized behaviors.

Small Numbers and Uncertainty

County-level rates rest on small numbers, which makes them unstable. A change of five deaths in a year can move Harlan's rate by more than two points per 100,000, and age-specific rates built on single-digit counts, such as the three deaths among residents 65 and older, can swing widely from year to year by chance alone. Epidemiologists handle this by reporting confidence intervals, combining several years of data for small groups and avoiding strong conclusions from one year's change in a small subgroup. For the board, the rise from 18.1 to 33.5 per 100,000 is large enough and consistent enough across four years to be meaningful, but the rate for older adults should be treated as uncertain.

Answering the Commissioner

The commissioner's question has a clearer answer now. Harlan's overdose death rate rose about 85% in four years, deaths are concentrated among adults aged 25 to 44 and, after accounting for age, Harlan's rate is higher than Pike's, not lower. Those findings point prevention toward young and middle-aged adults and justify regional cooperation.

Conclusion

Counts start the conversation; rates, age-specific rates and age adjustment make comparisons fair, and incidence and prevalence answer different questions. Showing each calculation lets decision makers see what the numbers mean and trust the recommendations built on them.

References

Grimes, D. A., & Schulz, K. F. (2002). An overview of clinical research: The lay of the land. The Lancet, 359(9300), 57-61. https://doi.org/10.1016/S0140-6736(02)07283-5

Rudd, R. A., Aleshire, N., Zibbell, J. E., & Gladden, R. M. (2016). Increases in drug and opioid overdose deaths: United States, 2000-2014. MMWR. Morbidity and Mortality Weekly Report, 64(50-51), 1378-1382. https://doi.org/10.15585/mmwr.mm6450a3

Thacker, S. B., & Berkelman, R. L. (1988). Public health surveillance in the United States. Epidemiologic Reviews, 10(1), 164-190. https://doi.org/10.1093/oxfordjournals.epirev.a036021