Nonmetropolitan Counties Have a Standard Definition: Response to Hughey et al

To the Editor,

We would like to raise a concern about a recent article published in PRiMER: “Risk for Unintended Pregnancy Among Latino Men and Women in a Nonmetropolitan County in Michigan.”¹ Dr Hughey et al studied the risk of unintended pregnancy in Latino men and women in a setting they described as a nonmetropolitan county in Michigan. In the background section, the study was contrasted with previously published studies of metropolitan regions. Their findings showed a higher risk for unintended pregnancy in the study sample than that seen nationally in this population. In the discussion, they hypothesized that this discrepancy may be due to Latinos in nonmetropolitan communities lacking the resources available to those in metropolitan communities, including health-related resources such as language-concordant health care providers and facilities, lower neighborhood social cohesion, or fewer culturally appropriate community resources.

Our major concern with this study is that Washtenaw County is not a nonmetropolitan county.² Washtenaw county is the central county of the Ann Arbor, Michigan metropolitan statistical area, and includes the principal city of Ann Arbor (population 113,934), home of the University of Michigan Health System.³ The Office of Management and Budget considers Washtenaw county metropolitan⁴; the US Department of Agriculture has assigned a Rural-Urban Continuum Code of 2 to Washtenaw County—a metropolitan designation⁵; and the US Census Bureau defines Washtenaw County as a “mostly urban” county, the most urban classification available.⁶

This of course does not mean that every area within Washtenaw county is urban, as on a county basis the definitions above do include some rurality. The US Census Bureau definition of “mostly urban,” for instance, includes all counties where over 50% of the population resides in an urban community. Washtenaw county is 16% rural. However, this small amount of rurality does not make the county nonmetropolitan. And although the authors may have assumed that most of the Latino population resided within the 16% of the county considered rural, this does not bear true when examining census data, which shows the highest numbers of residents of Latino origin in the eastern urban portions of the county, rather than the western rural portions.⁷

Although it is interesting that the population studied here was at higher risk of unintended pregnancy, we do not feel it represents a nonmetropolitan group. Accurate and specific descriptions of the population studied, and the resources available to them, are important. Although the authors do state that the results may not be generalizable, the role of descriptive research is to find information that can be applied to the larger world. Based on the inaccurate nonmetropolitan description, readers may reasonably but inappropriately conclude that the authors have uncovered a phenomenon that may also be true in actual nonmetropolitan communities. Moreover, in describing their population as nonmetropolitan, the authors are losing an opportunity to explore alternative explanations for their findings.

We suggest that the community studied should not be represented as nonmetropolitan, unless the authors are able to more completely justify this assertion in a scholarly manner, either with an alternate county-level designation of nonmetropolitan or on a community level, by documentation of exclusion of those persons of Latino origin that resided within the urban portions of this county.