DHA 731 Week 2 Applying Epidemiologic Measures Example

From Rates to Numbers: Applying Epidemiologic Measures to Diabetes Across 39 Census Tracts in Northeastern North Carolina

[Student Name]

University of Phoenix

DHA/731: Population Health and Epidemiology

Week 2 Assignment

[Instructor Name]

[Date]

The health network and its planning decisions are composites written for a model paper. Tract and county figures are real CDC PLACES 2025 release estimates for Bertie, Halifax, Hertford and Northampton counties, North Carolina, with counts calculated from them; the state figure is a BRFSS 2023 estimate; research findings come from the sources cited.

At a planning meeting for the rural hospitals' diabetes program, the finance director asked a simple question that no one could answer: how many people in our counties have diabetes? The public health data the team had presented reported percentages. Clinic staffing, supply budgets and outreach plans require numbers. This paper applies epidemiologic measures to diabetes in the four counties the hospitals serve, Bertie, Halifax, Hertford and Northampton, to answer her and to show what each measure can and cannot tell a leader.

Prevalence and Incidence

Prevalence is a snapshot: the share of the population that has the condition at one point in time. Incidence answers a different question: among people who did not have the condition, how many develop it over a set period. Diabetes is a chronic condition, so prevalence builds up over the years as new diagnoses are added and people survive longer with the disease. For planning, prevalence sizes today's caseload, while incidence shows how quickly tomorrow's caseload will grow.

The Data

The CDC PLACES project publishes model-based estimates of diagnosed diabetes among adults for every census tract. For the four counties, the 2025 release provides crude prevalence estimates for 39 tracts, based on 2023 data (Centers for Disease Control and Prevention [CDC], 2025b). Crude prevalence, the share of all adults with diabetes, is the right measure for estimating how many people need care, because it reflects the population's actual age structure.

Variation Across Tracts

Tract estimates ranged from 11.9% to 26.6%, with a median of 19.6% (CDC, 2025b). The highest tract, in Bertie County, had more than twice the prevalence of the lowest, in Hertford County, a prevalence ratio of 26.6 divided by 11.9, or about 2.2.

Why One Tract Differs

The lowest tract is not necessarily the healthiest. It includes a small college, and many of its adults are young students, who rarely have diabetes. A leader who read that tract's rate as a success would misunderstand it. Outliers should prompt questions about who lives there before conclusions are drawn. Tracts containing prisons, nursing homes or military housing can mislead in the same way, and a quick look at census data on group quarters usually explains them.

From Prevalence to Counts

Multiplying each tract's prevalence by its adult population and adding the results gives the number of adults with diagnosed diabetes. Across the 39 tracts, about 16,600 of roughly 85,900 adults have diagnosed diabetes, an overall prevalence of about 19.4% (CDC, 2025b). By county, the estimates are about 7,600 adults in Halifax, 3,200 in Hertford, 3,000 in Northampton and 2,800 in Bertie. Nineteen percent is a statistic; sixteen thousand six hundred is a patient list.

Comparing With the State: Relative Measures

For the state as a whole, the 2023 survey estimate of diagnosed diabetes is 12.4% of adults (CDC, 2025c). The prevalence ratio for the four counties compared with the state is 19.4 divided by 12.4, or about 1.56: an adult in these counties is roughly 56% more likely than the average North Carolina adult to report diagnosed diabetes.

Comparing With the State: Absolute Measures

The prevalence difference is 19.4 minus 12.4, or 7.0 percentage points. Multiplied by 85,900 adults, that difference represents about 6,000 more adults with diabetes than there would be if the counties matched the state rate. Relative measures describe how strongly a population differs; absolute measures describe how many people are affected, which is what budgets require.

A Caution About Age

The comparison uses crude rates, and the counties have a larger share of residents over sixty-five than North Carolina as a whole. Age-adjusted county estimates, which remove the effect of age structure, range from 14.2% to 15.2% (CDC, 2025a). Part of the excess reflects an older population, which affects what kinds of services are needed but does not reduce the number of people who need them.

Diagnosed Is Not All

PLACES estimates diagnosed diabetes only. Some adults have diabetes that has not been diagnosed, and more have prediabetes. National surveys that test blood samples consistently find a meaningful share of adults with diabetes who have never been told, and the share is likely higher in rural areas where fewer people have regular checkups. A screening campaign would therefore raise the number of known cases, which the region should expect and plan for rather than read as a worsening epidemic. The true burden is therefore larger than the counts suggest, and outreach to find undiagnosed cases is part of the task.

How the Estimates Are Made

Survey samples in rural tracts are too small to estimate prevalence directly. PLACES relies on a method developed by CDC researchers: Zhang and colleagues described multilevel regression and poststratification, which fits a model to national survey data with individual and area-level predictors and then applies it to census population counts for small areas; when they checked the modeled figures for chronic lung disease against what the survey measured directly, agreement was nearly perfect for states and strong for counties (Zhang et al., 2014).

Limits of Model-Based Estimates

Because the estimates are modeled, they reflect the characteristics of each tract's population, such as age, sex and race, more than local conditions that the model does not include. They are good for planning and for comparing areas, but they cannot detect the effect of a local program, since a program's success would not change the model's inputs.

Using the Measures for Planning

The counts guide capacity, and they change conversations with finance staff. If each adult with diabetes needs at least two primary care visits a year, the region needs capacity for about 33,000 diabetes visits. Tract estimates guide where to send community health workers, starting with the seven tracts at or above 24%, four in Halifax County, two in Northampton and one in Bertie. Those tracts also tend to have high rates of food insecurity and transportation barriers, so outreach there must include help with food and rides, not only clinical visits. Counts by county will also guide how the region divides a limited budget for diabetes educators among its clinics. The absolute excess guides the scale of prevention programs. Closing even a fifth of the gap with the state would mean about 1,200 fewer adults living with diabetes, a concrete target for a long-term plan.

Using the Measures for Evaluation

Because PLACES cannot show local program effects, the region will evaluate its diabetes program with its own clinical data: the number of patients in registries, the share with blood sugar under control and incidence of complications such as amputations and dialysis.

Conclusion

Epidemiologic measures turn data into decisions. Prevalence estimates from 39 tracts become about 16,600 adults with diagnosed diabetes. Ratios show the counties' adults are roughly 56% more likely to have diabetes than adults statewide, and differences translate that into about 6,000 extra people. Understanding how the estimates were made keeps leaders from misreading them.

References

Centers for Disease Control and Prevention. (2025a). PLACES: Local data for better health, county data, 2025 release [Data set]. data.cdc.gov. https://data.cdc.gov/d/swc5-untb

Centers for Disease Control and Prevention. (2025b). PLACES: Local data for better health, census tract data, 2025 release [Data set]. data.cdc.gov. https://data.cdc.gov/d/cwsq-ngmh

Centers for Disease Control and Prevention. (2025c). Behavioral Risk Factor Surveillance System (BRFSS) prevalence data (2011 to present) [Data set]. data.cdc.gov. https://data.cdc.gov/d/dttw-5yxu

Zhang, X., Holt, J. B., Lu, H., Wheaton, A. G., Ford, E. S., Greenlund, K. J., & Croft, J. B. (2014). Multilevel regression and poststratification for small-area estimation of population health outcomes: A case study of chronic obstructive pulmonary disease prevalence using the Behavioral Risk Factor Surveillance System. American Journal of Epidemiology, 179(8), 1025-1033. https://doi.org/10.1093/aje/kwu018