Flooding Interrupted Care for Home Health Care Recipients

Areas beset by hurricane-related flooding can prove detrimental to live in for residents, as patients requiring home health care (HHC) had worse outcomes in care compared with HHC patients in other areas, according to new research published in JAMA Health Forum.¹

More than 3.3 million Medicare beneficiaries use HHC consistently,² totaling $133.4 billion in health care expenses and making it a key part of health care in the US.¹ Patients using HHC are often more frail, older, and have difficulty coming into a doctor’s office, which would make them more vulnerable to natural disasters impeding their care. This study aimed to assess how the outcomes for HHC recipients were affected by the flooding related to Hurricane Sandy for patients living in New York City, New York, New Jersey, and Connecticut.

The study included all beneficiaries who were admitted to HHC in October 2012, when Hurricane Sandy was most impactful. These patients were identified through claims data from the Outcome and Assessment Information Set. This included a 20% national sample of Medicare fee-for-service beneficiaries who were enrolled in Parts A/B of Medicare continuously. Zip codes were used to identify areas of any flooding, and the comparison group resided in areas within a 10-mile radius of flooded zip codes.

Race and ethnicity were self-reported in both groups. Timely start of care, mean HHC length of stay, improved medication management, discharge to the community, and discharge to a skilled nursing facility were all outcomes evaluated during the analysis.

There were 761 beneficiaries who were included in the study, where the mean (SD) age was 77.3 (12.7) years, and 66.6% were women. A total of 67.4% identified as non-Hispanic White, 15.1% identified as Black, and 17.5% identified as other. Flooding was experienced by 432 patients in the analysis, with no significant differences in the demographics or clinical characteristics between the flooded and nonflooded groups.

Successful discharge to the community was more common in beneficiaries in the nonflooded zip codes (44.7% vs 39.6%) compared with the flooded zones. The mean HHC length of stay was also significantly different between the nonflooded and flooded groups (70.6 [221.6] days vs 83.5 [212.3] days). Timely start of care was more common in the nonflooded group compared with the flooded group (63.8% vs 55.5%). Discharge to a skilled nursing facility and improved medication management were not statistically different between the 2 groups.

There were some limitations to this study. The results may not be generalizable to the Medicare Advantage population. HHC agency characteristics could not be controlled for, as the data were not available in 2012. Differences in other health outcomes could not be revealed due to limited power.

“These findings underscore the vulnerability of HHC recipients to flooding and invite further questions about the incorporation of HHC into broader health care emergency planning efforts at local and regional levels,” the authors concluded.

References

1. Ghosh AK, Unruh MA, Jung HY, Yun H, Sterling MR. Exposure to hurricane-related flooding and outcomes of home health care patients. JAMA Health Forum. 2026;7(6):e261758. doi:10.1001/jamahealthforum.2026.1758
2. Kim S, Qi M, Konetzka RT, Werner RM. Home health care use among Medicare beneficiaries from 2010 to 2020. Med Care Res Rev. 2025;82(3):260-268. doi:10.1177/10775587251318407