Pearl has released the Pearl Oral Health Index, a report based on an analysis of more than 26 million dental x-rays representing 15 million patients and 737 million teeth. According to the company, the report uses data generated by its FDA-cleared artificial intelligence platform to examine oral health patterns across age, location, gender, and other demographic factors.
According to Pearl, the analysis found that the prevalence of untreated dental decay is substantially higher than estimates reported by the Centers for Disease Control and Prevention's National Health and Nutrition Examination Survey (NHANES). The company reported an average of 6.07 decayed teeth per patient compared with NHANES' reported average of 0.7. At the same time, Pearl said its findings for missing teeth (2.16 versus 2.0 per patient) and restored teeth (7.01 versus 6.0 per patient) closely aligned with NHANES data, supporting its methodology. The company noted that everyone included in its dataset had already sought dental care, meaning the findings may not represent the entire population.
“For the first time, we can see what's actually happening inside the mouths of America, not through a survey, not through a sample, but through the raw radiographic truth of tens of millions of visits,” said Ophir Tanz, founder and CEO, Pearl. “The picture is dramatically different from what the public data has told us for decades. There is more disease, more inequality, and more variability in care than anyone realized. AI didn't create this problem, but it finally let us measure it.”
The report also examined oral health by age group. According to Pearl, adults aged 18 to 24 had the highest proportion of untreated disease, with 52% of dental disease remaining untreated. The company reported that fillings began to outnumber untreated decay among adults aged 35 to 44, while adults aged 75 years and older averaged 19.14 teeth showing signs of decay, loss, or restoration, with 28% of that damage remaining untreated.
Geographic comparisons identified Hawaii as having the lowest overall disease burden among states included in the analysis, while Delaware had the highest. According to the report, New Jersey had the highest rate of calculus, and Kansas had the highest proportion of affected teeth extracted rather than restored.
The report also evaluated variation among dental practices. Across 282 practices that each treated more than 200 adult patients, Pearl reported that practices at the 10th percentile identified an average of 3.1 affected teeth per patient compared with 12.4 teeth per patient at the 90th percentile. According to the company, differences reflected variation in diagnosis, training, awareness, and diagnostic consistency rather than disease prevalence because the AI analyzed every radiograph using the same criteria.
Pearl also examined access to dental care by comparing ZIP codes with National Provider Identifier registry data. According to the report, 20.8% of affected teeth were extracted in ZIP codes with no dentists compared with 14.9% in ZIP codes with 51 or more dentists, representing a 40% higher extraction rate in communities the report describes as "dental deserts."
The report found differences in dental utilization by sex. According to Pearl, women represented 56.2% of patients in the U.S. dataset compared with 43.8% for men. Clinical findings were similar once patients received care, although men demonstrated higher rates of calculus while women had more restored teeth.
Pearl also compared oral health findings between the United States and the United Kingdom. According to the report, patients in the United Kingdom averaged 6.36 missing teeth compared with 2.16 in the United States, and approximately 31% of affected teeth were extracted in the United Kingdom versus 14.5% in the United States. The company attributed these differences to variations in the two countries' dental care systems.
According to Pearl, the Pearl Oral Health Index draws on two datasets collected between April 2024 and March 2026. The FMX Series Dataset includes 1.49 million patients with complete full-mouth radiographic series and was used to calculate decayed, missing, and filled tooth measurements. The Expanded Encounters Dataset includes 14.8 million patients and 26 million tooth-level dental encounters used for state comparisons, gender analyses, tooth-level timelines, and longitudinal evaluations.
Pearl, founded in 2019, develops artificial intelligence software for dentistry. The company said its FDA-cleared AI technology is authorized in 120 countries to assist dentists with interpreting dental radiographs.
The full Pearl Oral Health Index is available at https://hellopearl.com/oral-health-index.
