Health data poverty refers to the lack of representative, high-quality health data for certain populations, leading to gaps in research, diagnosis, treatment and public health decision-making.
The term encompasses the systematic underrepresentation of racial and ethnic minorities, rural communities, low-income populations, people with disabilities, Indigenous communities and others in clinical trials, electronic health records, genomic databases and the artificial intelligence systems used in health care.
Deeper dive
Health data poverty can occur for many reasons. Some communities have limited access to health care, resulting in fewer documented health encounters. Others may distrust health systems because of historical discrimination, making participation in research less likely. In some cases, data collection methods themselves fail to capture meaningful demographic information or exclude people without reliable internet access or digital health tools.
Health data poverty is an important concept to recognize when evaluating new research. Some questions to consider when evaluating health data: Were participants representative of the broader population? Did investigators report outcomes by race, ethnicity, sex, age, disability status or geography? If AI or predictive models were used, were they validated across diverse populations? Understanding these limitations can help reporters avoid overstating findings and identify potential inequities hidden within the data.
As precision medicine, digital health and AI become more central to health care, reducing health data poverty is increasingly viewed as an essential step toward improving health equity and ensuring that innovations benefit all populations.