NR 441 Week 8 Global Health Course Reflection Example

Someone Always Writes Down the Cause: A Reflection on a Belief That Turned Out to Be Local

Student Name

College of Nursing, Chamberlain University

NR 441: Community Health Nursing - Global

Instructor Name

Month Day, Year

Someone Always Writes Down the Cause: A Reflection on a Belief That Turned Out to Be Local

The Belief

At the start of this course I believed that when a person dies, someone qualified records why. I did not think of it as a belief; it was simply how death works in my practice. On my medical unit in the United States, when a patient dies, a physician completes a death certificate within a day or two, listing an immediate cause and the conditions that led to it, and the certificate goes to the state vital records office. I had filled out the nursing parts of that process many times. When I read a statistic that heart disease was the leading cause of death in some country, I assumed that behind the number were certificates like the ones on my unit, written by physicians who had cared for the patient.

What Made It Local

The fact that made the belief untrue elsewhere came in the third week, when I prepared a burden brief on India. Most deaths in India occur at home, and only a minority of registered deaths carry a medically certified cause; India's own certification reports put the certified share at about twenty percent of registered deaths toward the end of that decade (Pandey et al., 2021). For most other deaths, the cause is assigned through verbal autopsy, an interview with a relative conducted by a trained nonmedical worker, or it is estimated through statistical models. The World Health Organization's country page for India goes further, warning readers that its registration data cannot be used as they stand and that its cause-of-death figures deserve caution. This is not unusual. Mikkelsen et al. (2015) found that a large share of the world's deaths occur in countries where civil registration and cause-of-death data are incomplete or of low quality.

Once I understood this, the tables I had been reading looked different. The leading cause of death for India in the Global Health Estimates is not a count of certificates; it is a modeled estimate with an uncertainty range, and for some causes, such as diarrheal disease, the range runs from about a quarter of a million to nearly a million deaths (World Health Organization, 2024). The belief I had brought from my unit was not a professional truth about death. It was a feature of one country's registration system.

Why the Belief Mattered

The belief was not harmless. In the first week, I wrote a discussion post comparing heart disease deaths in the United States with those in a low-income country and concluded that heart disease was less of a problem there because the rate was lower. I did not ask how many deaths in that country had a certified cause at all, or how many heart attacks at home would have been recorded as sudden death, old age or an ill-defined cause. A rate built on incomplete certification can be lower because the disease is rarer or because the deaths were never attributed to it, and my post had no way of telling which. If a program officer had used reasoning like mine to decide where to spend money on cardiovascular care, the money could have gone to the wrong problem.

The same belief shaped how I read improvements. When a country reports a falling death rate from a cause, I assumed fewer people were dying of it. But if the share of deaths that are certified is rising at the same time, a cause can appear to rise or fall because of better recording rather than a change in disease. Learning to ask how the cause was assigned changed how I read every trend in the course.

Where I Went Wrong Along the Way

I made a related mistake before I understood this. In my country health profile of Rwanda, my first draft stated that maternal mortality had fallen from 229 to 203 per 100,000 live births, because I had placed the World Bank figure beside the survey figure and read the lower number as the more recent. In fact, the 229 was a modeled estimate for 2023 and the 203 was a survey measure covering the years before 2020, produced by a different method with a wide confidence interval. The two numbers did not describe a change at all. My instructor's comment, asking what each figure was made of, was the first time I understood that two numbers in a table can look comparable and not be. I corrected the draft, and the final profile reports each figure with its method, but I would have submitted a false trend if I had not been asked.

What I Check Now

I now ask three questions before I use a statistic from another country, and I ask them of statistics from my own country too. First, was this counted or estimated? A registered count, a survey measure and a model output are different kinds of evidence, and the source usually says which, if I look. Second, who recorded the underlying event, and with what training? A cause of death written by a physician, one assigned from a relative's interview and one inferred by a model carry different weight. Third, does the figure come with an uncertainty range, and how wide is it? A narrow range and a wide one tell me how much to trust the rank order in a table.

The change has already reached my work at home. Our hospital's quality report lists the leading causes of inpatient death on my unit, and I had always read it as settled. Last month I asked our quality nurse how the causes were coded, and learned that they come from discharge diagnosis codes rather than from the death certificate, which means a patient admitted for pneumonia who died of a heart attack may be counted under pneumonia. The question I learned to ask about India turned out to apply on my own unit.

I also carry a fourth question into any conversation about priorities, which is who is missing from the number altogether. A death that is never registered does not appear in a rate, and the people least likely to be registered are often the people a community health nurse most needs to reach. Some of what this course asked of me was technical, learning the difference between a survey and a model. The larger change is that I no longer assume the absence of a figure means the absence of a problem.

References

Mikkelsen, L., Phillips, D. E., AbouZahr, C., Setel, P. W., de Savigny, D., Lozano, R., & Lopez, A. D. (2015). A global assessment of civil registration and vital statistics systems: Monitoring data quality and progress. The Lancet, 386(10001), 1395-1406. https://doi.org/10.1016/S0140-6736(15)60171-4

Pandey, A. K., Gautam, D., Thomas M, B., & Kharakwal, Y. (2021). Registration and medical certification of deaths in the Indian states: A comparative analysis of data of CRS and MCCD reports (2010-2019) [Preprint]. medRxiv. https://doi.org/10.1101/2021.12.09.21267291

World Health Organization. (2024). Global health estimates 2021: Deaths by cause, age, sex, by country and by region, 2000-2021 [Data set]. https://www.who.int/data/gho/data/themes/mortality-and-global-health-estimates