Using Metrics and Measurement to Improve Oral Health
Preventistry is about changing four interconnected systems to improve oral health for everyone. With the Preventistry approach, we must address policy, finance, care and community aspects that together determine people’s ability to be healthy. But we need supporting data and research to affect that change.
Data can tell powerful stories. It’s an effective tool to uncover information that might otherwise remain unseen. But it’s challenging to track improvement without comparable data, making it critical to consider exactly what is being measured for accuracy and apples-to-apples comparisons. This is why one of DentaQuest’s key goals is helping to establish a single national oral health measurement system.
Let’s outline an example: Many things fall under the umbrella of care, including how we define “provider,” the rules that providers must follow, how providers engage with patients, and even how providers engage with other providers. Successful analyses of data to evaluate these rules and levels of engagement can influence policymakers and guide future research – if the research questions are approached with clear definition.
This requires considering:
- the patient level,
- the claims level, and
- the hospital/health system level.
At every level, it’s important to understand how the individual variables within a dataset are defined. An example of this is race/ethnicity. If you’re attempting to determine the influence of race/ethnicity on utilization of dental services in a specific area, but the dataset only contains information about one race, then it’s not going to answer your research question.
When reviewing research, it’s also critical to review how that particular study defined race/ethnicity, given the different ways to label, identify, and categorize people. For instance, if data shows the highest incidence of caries among seniors who identify as African American, you cannot accurately say black seniors have more cavities than the other races. “Black” as a label encompasses more than people who identify as “African American” – the terms are not interchangeable when it comes to datasets.
Given the detailed nature of how researchers collect and use data, storytelling to influence change may take more than just looking at outcomes. Understanding data may require reviewing a codebook, speaking to other researchers who have utilized the same dataset, or running smaller subsets and reviewing those results. This process may also require returning to the owner or creator of the dataset to verify results.
In the end, metrics and measurement can affect change, but only when accurately used. In the future, universal metrics will ensure effective systems transformation and thus a healthier nation overall.