TL;DR
This study compares the prevalence of chronic conditions (CCs) and multiple chronic conditions (MCCs) in older adults in Florida versus the entire United States using two large datasets: the OneFlorida Data Trust and the Healthcare Cost and Utilization Project (HCUP) National Inpatient Sample (NIS). The findings suggest that these data networks, like OneFlorida, are valuable for large-scale secondary data analysis, as their results closely mirror national trends.
What Problem Does This Paper Address?
As the population ages, the prevalence of chronic conditions, especially multiple chronic conditions (MCCs), among older adults is a growing concern. MCCs lead to increased healthcare needs, higher medical costs, and a significant burden on patients, caregivers, and the healthcare system. To effectively manage and prevent these conditions, researchers need reliable data to understand their prevalence and trends. While national surveys and claims data have been used, the increasing availability of electronic health records (EHRs) and the development of clinical data networks offer new opportunities. However, it's crucial to validate whether data from these networks, like the OneFlorida Data Trust, accurately represent national populations. This paper addresses this by comparing prevalence estimates from OneFlorida with those from the established HCUP NIS.
Methods
The study employed a comparative analysis of two large datasets. They first identified patients with chronic conditions (CCs) using the Centers for Medicare and Medicaid Services (CMS) Chronic Condition Warehouse (CCW) algorithm, which utilizes validated ICD-9-CM and ICD-10 diagnosis codes, number of occurrences, and consecutive years of diagnosis. For the purpose of this study, they focused on 25 specific CCs and defined MCCs as having two or more CCs, based on the CMS algorithm.
The analysis involved:
- Descriptive statistics: To characterize the older adult populations in both datasets.
- Prevalence analysis: To determine the frequency of each of the 25 CCs and MCCs.
- Stratified analysis: Examining hypertension (the most prevalent CC) and MCCs by gender, age, and race/ethnicity.
- Trend analysis: Investigating the longitudinal trends in MCC prevalence over time for both datasets.
Chi-square tests were used to compare group differences, and Pearson correlation coefficients were calculated for trend analysis.
Datasets
- OneFlorida Data Trust: A secure, centralized data repository integrating data from contributing organizations within the OneFlorida research consortium in Florida. This study utilized inpatient discharge records from 2012-2016.
- Healthcare Cost and Utilization Project (HCUP) National Inpatient Sample (NIS): A comprehensive, all-payer inpatient healthcare database in the United States. The study used NIS data for 2013 for cross-sectional analysis and 2003-2013 for longitudinal analysis.
Key Insights and Ideas
- EHR Data Networks as Valid Proxies: The study strongly suggests that clinical data networks like OneFlorida, built from heterogeneous EHR and claims data, can provide reliable estimates comparable to national datasets like the NIS. This is crucial for future research leveraging these growing data resources.
- Ubiquity of Chronic Conditions in Older Adults: The findings underscore that chronic conditions and multiple chronic conditions are highly prevalent in the older adult population, both nationally and within Florida.
- Consistent Top Conditions: The top 10 most prevalent chronic conditions were consistent across both the OneFlorida and NIS datasets, indicating a shared pattern of disease burden in older adults.
- Similar MCC Trends: The longitudinal analysis revealed similar increasing trends in the prevalence of MCCs over time in both datasets, suggesting a consistent societal shift towards more complex chronic disease profiles in aging populations.
- Demographic Similarities in MCC Prevalence: Notably, the prevalence of MCCs did not show significant differences between genders or across major racial/ethnic groups in the OneFlorida and NIS data, contrasting with findings in younger populations where gender differences are more pronounced.
Key Results
- Prevalence of Chronic Conditions: The rankings of the top 10 prevalent chronic conditions were identical across the OneFlorida and NIS datasets. Hypertension was the most prevalent condition in all datasets.
- Prevalence of Multiple Chronic Conditions: More than 18% of older adults had 4 or more chronic conditions, and over 65% had 4 or more conditions in the analyzed datasets.
- Hypertension Differences: While generally similar, the largest differences in hypertension prevalence between the 2013 OneFlorida and NIS datasets were around 3% for specific subgroups (e.g., non-Hispanic white females aged 85+).
- MCC Trends: Both the NIS (2003-2013) and OneFlorida (2012-2016) datasets showed a steady, monotonic increasing trend in the prevalence of MCCs over time.
- Demographic Comparisons: The study found no notable differences in MCC prevalence by gender or race/ethnicity between the OneFlorida and NIS datasets, although women in younger age groups (18-64) tend to have a higher prevalence of MCCs.
Limitations and Future Work
- Data Source Limitations: Both datasets primarily capture conditions confirmed by healthcare professionals in inpatient settings. This may lead to an underrepresentation of conditions managed in outpatient settings or those that remain undiagnosed (e.g., arthritis).
- Data Quality and Duplication: The study acknowledges potential for false positives in case identification and the challenge of data linkage and deduplication across different healthcare organizations within the OneFlorida consortium.
- Geographic Representation: While OneFlorida covers a significant portion of Florida, it may not fully represent all key healthcare markets within the state.
- Future Work: The authors suggest enhancing case assignment by incorporating laboratory test results and medication data. They also propose comparing outpatient data from OneFlorida with national outpatient databases to gain a more comprehensive understanding of chronic conditions.
Summary updated