PU570 Unit 10 Epidemiological Report Example

Stroke Is Falling in the Old and Rising in the Young: A Unit 10 Epidemiological Report Measuring Age-Specific Stroke Rates in One Composite State, With Annual Percent Change, Severity at Admission and What the Report Will Not Claim

[Student Name]

Purdue University Global

PU570: Chronic Disease Epidemiology

Unit 10 Epidemiological Report

[Instructor Name]

[Date]

The composite state and its rates are composite but follow published national patterns; the trend evidence, the risk factor findings and the research cited are real.

Purpose and Question

The state health department asked whether stroke in this state follows the national pattern of falling rates overall and rising rates in younger adults, and what the state's own data say about why. The report covers hospital discharge data for 2004 through 2023, with first-ever ischemic stroke identified by principal diagnosis and residence in the state.

The Headline the Report Does Not Lead With

Age-adjusted first-ever ischemic stroke incidence fell from 198 to 161 per 100,000 over the period, a 19 percent decline. Reported alone, this figure would be read as unambiguous progress and would be used to justify holding stroke prevention funding flat.

Age-Specific Rates, Which Is the Finding

1. Age 75 and over: 1,142 per 100,000 in 2004, 812 in 2023. A 29 percent decline.

2. Age 65 to 74: 584 to 441. A 24 percent decline.

3. Age 55 to 64: 268 to 236. A 12 percent decline.

4. Age 45 to 54: 86 to 104. A 21 percent increase.

5. Age 25 to 44: 17 to 26. A 53 percent increase.

The pattern matches what a population-based surveillance study first established nationally, where the mean age at stroke fell and incidence rose in adults under 55 while falling in older groups (Kissela et al., 2012).

Annual Percent Change, and Where the Trend Turns

Fitted log-linear models give an annual percent change of minus 1.8 percent for adults 65 and over across the whole period. For adults 25 to 44 the trend is not linear: minus 0.4 percent from 2004 to 2010, then plus 3.1 percent from 2011 to 2023, with the joinpoint in 2010 or 2011 depending on the model. The report gives both the single-slope and the two-segment estimates, because the policy question is whether something changed around 2010 and a single slope cannot answer it. The state's overall rate fell because the age groups that contribute most of the cases are shrinking their rates faster than the youngest groups are raising theirs. Nothing in the combined figure is false, and nothing in it is useful.

Severity at Admission

Median admission stroke severity score fell in every age group, consistent with earlier presentation and with milder strokes being diagnosed as imaging improved. In the two youngest bands, the share of admissions with a severity score below 5 rose from 31 to 48 percent, which raises the detection question this course has handled before: part of the rise in young-adult stroke may be milder events now imaged and coded as stroke that would previously have been recorded as a transient episode or missed. The report cannot separate that component and says so.

Risk Factor Prevalence Among Young Adult Cases

Among the state's stroke admissions aged 25 to 44, documented hypertension rose from 41 to 58 percent of cases, diabetes from 17 to 26 percent, obesity from 19 to 34 percent, and tobacco use stayed near 40 percent. This mirrors the national finding that traditional vascular risk factors are increasingly present among younger stroke patients rather than absent from them (George et al., 2017). The report treats this as the most plausible driver of a real increase and is explicit that case-level risk factor documentation cannot establish it.

What the Report Recommends

1. Report stroke incidence by age band in every future surveillance product, and retire the single age-adjusted figure from press material.

2. Ask the state's hospitals to supply severity scores for all stroke admissions, which three of twelve currently omit, so the detection question can be addressed rather than acknowledged.

3. Direct hypertension control effort at adults aged 35 to 54, where prevalence is rising and treatment rates are lowest, rather than at the age groups whose rates are already falling.

4. Repeat this analysis in three years with a joinpoint model rather than a single slope.

How the Rates Were Built, and the Three Decisions Inside Them

A report of this kind is only as good as three choices a reader cannot see in the figures. First, case definition: first-ever ischemic stroke by principal diagnosis excludes hemorrhagic stroke and excludes cases where stroke was a secondary diagnosis, which undercounts strokes occurring during admission for something else. Second, the first-ever rule is enforced by looking back five years within the state's own data, so a resident whose earlier stroke happened out of state is counted as incident here, and in-migration at older ages makes that error larger in the oldest bands. Third, coding changed mid-series, and the transition to the tenth revision in 2015 shifted which cases carried a stroke principal diagnosis; the report fits its trend models with a term for the transition year and reports the estimates with and without it. None of these three decisions reverses the age divergence, which is large and consistent across specifications, and all three affect the absolute rates enough that they should not be quoted to the nearest whole number. Causal reasoning from a trend of this kind is a judgment about competing explanations rather than a measurement of an effect, which is the standing caution in this literature (Rothman & Greenland, 2005).

What This Report Will Not Claim

It will not claim that stroke is becoming a disease of the young: the absolute rates in older adults remain an order of magnitude higher, and the youngest bands contribute a small share of the state's cases. It will not claim the rise in young adults is entirely real, since milder events and better imaging account for part of it and the report cannot say how much. It will not attribute the decline in older adults to any specific program, since blood pressure treatment, smoking decline, anticoagulation for atrial fibrillation and statin use all changed over the same twenty years and this design cannot apportion them. And it will not report rates by race or ethnicity from this data source without a denominator correction, because the discharge data's race field is incomplete for 14 percent of admissions, and a rate built on an undercounted numerator and a census denominator would be wrong in a direction that matters.

References

George, M. G., Tong, X., & Bowman, B. A. (2017). Prevalence of cardiovascular risk factors and strokes in younger adults. JAMA Neurology, 74(6), 695-703. https://doi.org/10.1001/jamaneurol.2017.0020

Kissela, B. M., Khoury, J. C., Alwell, K., Moomaw, C. J., Woo, D., Adeoye, O., Flaherty, M. L., Khatri, P., Ferioli, S., De Los Rios La Rosa, F., Broderick, J. P., & Kleindorfer, D. O. (2012). Age at stroke: Temporal trends in stroke incidence in a large, biracial population. Neurology, 79(17), 1781-1787. https://doi.org/10.1212/WNL.0b013e318270401d

Rothman, K. J., & Greenland, S. (2005). Causation and causal inference in epidemiology. American Journal of Public Health, 95(S1), S144-S150. https://doi.org/10.2105/AJPH.2004.059204