Geospatial suicide risk prediction in the State of Maryland and the limits of discernible intent

Every year, thousands of people in the United States take their own lives, a tragedy that leaves communities searching for patterns to prevent future loss. For decades, public health experts have known that these deaths do not happen randomly; they cluster in specific places and are linked to factors like poverty, lack of medical care, and social isolation. However, a significant portion of these deaths presents a difficult puzzle for officials. When a person dies, a medical examiner must determine the cause and the manner of death, deciding if it was an accident, a homicide, a suicide, or something else. Sometimes, the evidence is simply not clear enough to make that call. These cases are labeled as having an "undetermined" manner of death. It is possible that some of these uncertain cases are actually suicides, but without a clear diagnosis or a history of treatment, the intent remains hidden. Understanding where these hidden risks might be hiding is crucial for saving lives, yet the map of these uncertain deaths has remained largely uncharted.

A team of researchers at Johns Hopkins University set out to draw that map for the state of Maryland. They wanted to see if the places where confirmed suicides occurred were the same as the places where deaths were left undetermined. To do this, they gathered data on nearly 5,000 confirmed suicides and over 10,000 undetermined deaths that occurred between 2012 and 2020. They broke the state down into small geographic neighborhoods, known as census tracts, and looked at the conditions in each one. They examined factors such as how many people lived there, the age of the population, how much money people earned, whether the area was rural or urban, and how many residents suffered from conditions like depression or alcohol misuse. They used two different methods to find patterns: one method looked for statistical links between the death rates and the local conditions, while the other method scanned the map to find specific "hot spots" where deaths happened more often than expected by chance.

The researchers discovered that the two types of deaths were telling very different stories about the state. Confirmed suicides were most common in rural areas and smaller towns, particularly in the western part of Maryland and on the Eastern Shore. These areas tended to have older populations and fewer people from minority racial backgrounds. In contrast, deaths labeled as undetermined were concentrated in the heart of the state's cities, especially around Baltimore. These urban hot spots were characterized by higher poverty, greater social deprivation, and a higher proportion of minority residents. The data suggested that in these crowded, disadvantaged city neighborhoods, medical examiners faced a much harder time figuring out what happened. The lack of clear medical records, treatment history, or clinical context in these areas may have made it difficult to confirm suicide, leading to a higher number of cases being left in the "undetermined" category.

When the team looked closer at the specific causes of death within these clusters, a striking connection emerged. The areas with the highest rates of firearm deaths largely matched the areas with the highest rates of confirmed suicide. This makes sense, as a gunshot wound often leaves little doubt about the intent. However, the areas with the highest rates of undetermined deaths were the same places where overdose and poisoning deaths were most common. It is often harder for an examiner to tell if an overdose was a deliberate act or an accident, especially if the person had no known history of mental health struggles. This overlap suggests that many of the undetermined deaths in these urban, deprived areas might actually be suicides that went unrecognized. The study implies that the difficulty in identifying intent is not just a random error, but a pattern linked to the social and economic conditions of the community.

The findings offer a new way to look at public health surveillance. By recognizing that "undetermined" deaths often cluster in the same places as confirmed suicides, health officials can identify communities that are at risk even when the official numbers are incomplete. The study suggests that in rural areas, the focus might be on the specific challenges of isolation and access to care, while in urban centers, the priority should be addressing the deep social and economic burdens that obscure the true nature of these tragedies. The researchers noted that their work covers only one state and a specific time period, so the patterns might look different elsewhere. However, the approach of looking at both confirmed and uncertain deaths together provides a clearer picture of where help is needed most. It highlights that to truly understand the scope of suicide, we must look not just at the deaths we can clearly identify, but also at the ones that remain a mystery, for they may hold the key to understanding the full scale of the crisis.