A new way to measure comorbidity – when a patient suffers from multiple diseases or health conditions – can be used to accurately predict a patient’s risk of hospitalization and future healthcare costs across a population, Cornell researchers report.

Developed by Dr. Mary Charlson, chief of the Division of Clinical Epidemiology and Evaluative Sciences Research and the William Foley Distinguished Professor of Medicine at Weill Cornell Medicine, the Charlson Comorbidity Health Analytics (CCHA) is a measure of 38 chronic health conditions. Using six years of data from more than 27,000 Weill Cornell Medicine employees and dependents, Charlson and colleagues showed that the CCHA surpasses other methods of predicting unplanned hospitalizations – a major source of healthcare costs – and can successfully predict costs over the subsequent five years.

“Comorbidity predicts repeated hospitalizations and costs, because that’s the underlying driver,” Charlson said. “Physicians intrinsically know this, but there haven’t been tools that allow that measurement in routine clinical practice.”

Additionally, the CCHA may offer a way to bring down medical costs. If high-needs patients can be identified and flagged to receive longer primary care visits and additional services that stabilize their health, these interventions may prevent future hospital admissions.

“Patients with a lower comorbidity burden generally require fewer clinical and supportive resources than those with multiple or more complex conditions. However, healthcare systems often rely on standardized models of care that do not adequately account for differences in patient complexity and level of need,” said co-author Martin Wells, the Charles A. Alexander Professor of Statistical Sciences in the Cornell Ann S. Bowers College of Computing and Information Science and the ILR School. “Patients are often allotted the same amount of time with their physician regardless of their clinical complexity. A healthy patient may receive a 15-minute appointment, while a patient with multiple comorbidities and more complex care needs is given the same limited time frame.”

Their new study, “Charlson Comorbidity Health Analytics: A Population Management Strategy to Identify Risk of Hospitalizations, Repeated Hospitalizations, and Resultant High Cost,” published June 29 in PLOS ONE. 

In the CCHA, each chronic condition receives a different point value based on its seriousness. The higher a patient’s total comorbidity score, the more serious their health problems.

The researchers calculated the CCHA for each Weill Cornell Medicine beneficiary – children and adults – using fully anonymized data from claims made from 2016 to 2021. Then they looked at their healthcare costs and risk of being hospitalized each year during the study period. Unlike similar clinical datasets, the Weill Cornell beneficiary data was complete over six years of coverage, giving the whole picture of medical claims and costs for participants.

People with a higher comorbidity score faced a higher risk of hospitalization. Among adults with a score of zero in 2016, only 1.2% ended up in the hospital that year, compared to 61% of adults with a score of eight or higher. 

Higher comorbidity scores also led to increased healthcare costs. In 2016, patients with a score of zero had costs averaging $3,835. Spending increased with every additional point in the CCHA, reaching $53,189 for patients with scores of eight or more. Only 4.3% of patients had a score of five or higher, but these patients accounted for 17.5% of total healthcare spending that year.

Over time, average yearly costs and risk of hospitalization increased in accordance with CCHA scores. Using a lagged analysis, the researchers were able to predict subsequent annual costs and hospital admissions for up to five years, based on patients’ earlier scores.

The team demonstrated the CCHA was a better predictor of hospital admissions than whether a patient had had a previous admission, and also outperformed estimates of future costs and hospital admissions based solely on prior costs.

Giving high-needs patients extra care to keep them out of the hospital is one way to cut spending in a meaningful way, especially at a time when rural hospitals are closing and many people are losing Medicaid coverage, researchers said. 

“One hospitalization is $39,000 to $49,000,” Charlson said. “What’s the cost of an extended primary care visit?”

The researchers have already begun applying the CCHA at scale in a few healthcare systems to see if longer primary care visits can prevent unplanned hospitalizations for this small percentage of the population. If successful, the findings would support new policy recommendations for treating individuals with complicated medical needs.

Dr. James Hollenberg, associate professor of medicine at Weill Cornell Medicine is a co-author on the paper.

The research received support from the Google Cyber NYC Institutional Research Program and the National Science Foundation.

Patricia Waldron is a writer for the Cornell Ann S. Bowers College of Computing and Information Science.