An examination of the most expensive adult hospitalizations in America

Link to article at PubMed

Hosp Pract (1995). 2022 Oct;50(4):340-345. doi: 10.1080/21548331.2022.2121572. Epub 2022 Sep 8.


BACKGROUND: While no hospitalization is inexpensive, some are extremely costly. Learning from these exceptions is critical. The patients and conditions that ultimately translate into the most exorbitant adult hospitalizations have not been characterized.

OBJECTIVE: To analyze and detail characteristics of extreme high-cost adult hospitalizations in the United States using the most recently available Nationwide Inpatient Sample (NIS) data.

DESIGN/SETTING/PARTICIPANTS: The NIS 2018 database was queried for all adult hospitalizations with hospital charges greater than $333,000. Multivariable linear regression was used in the analyses.

MEASURES: The main outcome measures were total charges, mortality, length of stay, admitting diagnosis, and procedures.

RESULTS: There were 538,121 adults age ≥18 years with total hospital charges ≥$333,333. Among these patients 481,856 (89.5%) survived their hospitalization and 56,265 (10.4%) died. Males, older patients, being insured by Medicare, having more comorbid illness, and those who were transferred from another hospital were significantly more likely to die during the incident hospitalization (all p < 0.01). Patients who died had even more costly hospitalizations with more procedures (mean [SD]: 10.7 [±6.4] versus 7.0 [± 5.9], p < 0.01), and longer lengths of stay after adjustment for confounders (p = 0.01).

CONCLUSIONS: Hundreds of thousands of adult patients are hospitalized in the US each year at extremely high costs. For both those who survive and the 10% who die, there may be opportunities for reducing the expense. Interventions, such as predictive modeling and systematic goals of care discussions with all patients, deserve further study.

PMID:36062489 | DOI:10.1080/21548331.2022.2121572

Leave a Reply

Your email address will not be published. Required fields are marked *