Healthcare costs of foodborne diseases in the United States: an analysis using electronic health records.

This study used electronic health records from The Ohio State University Healthcare System to estimate hospital costs (payrolls, supplies, utilities) for foodborne diseases (FBD) across a range of clinical diagnoses. We considered the relationship between FBD and comorbidities since comorbidities may increase the risk of FBD or FBD sequelae. Our regression analysis advances prior work by estimating incremental FBD costs while adjusting for chronic conditions, comorbidities, and other confounders (e.g., demographics, payor and patient types) using large-scale EHR data, filling a critical gap in FBD cost estimation. We analyzed 8,147,264 encounter episodes from 2017 to 2021, totaling 697,732 patient years, with 930 being FBD-associated. Poisson regression yielded baseline annual FBD hospital costs of $8,258 per patient (cost-to-charge adjusted). When adjusting and comparing different types of comorbidities, the Clogg Z-test showed that cost estimates without comorbidity controls were $3,735 higher than baseline costs per patient-year (p = 0.098). However, there was little difference between the baseline and the estimates that took comorbidities into account. Instead of using foodborne pathogen-specific codes, costs were marginally lower by $1,492 (p = 0.058) when using gastroenteritis codes (such as infectious diarrhea). In contrast, gastrointestinal symptom codes such as diarrhea and others were associated with $3,440 (p < 0.001) higher costs than codes specific to foodborne pathogens. We discussed the implications of FBD cost variations through sensitivity analyses of comorbidity adjustments and diagnostic coding in methodology, and how these variations mimic liability assessments in the consumer and food industry.