Remote Monitoring After Sepsis Did Not Improve Home Days

Key Takeaways

• Remote therapeutic monitoring did not improve postdischarge home days compared with usual care.
• Readmissions, mortality, and other secondary outcomes were not meaningfully different across groups overall.
• Less favorable subgroup patterns were observed in older adults and patients leaving skilled nursing facilities, while participants often described the technology as workable but somewhat impersonal.

A randomized comparison published in JAMA Network Open found no clear advantage for postdischarge remote therapeutic monitoring over usual care after hospitalization for serious infection. Among adults discharged after sepsis or lower respiratory tract infection, median postdischarge home days were 90 in every study arm over 90 days. The trial compared four monitoring strategies with structured telephone-based usual care after discharge, and none changed the main recovery outcome over follow-up.

This open-label, mixed-methods randomized clinical trial used intention-to-treat analysis across 19 hospitals, with recruitment from March 25, 2021, through September 3, 2024. Eligible patients were at least 21 years old, lived in western Pennsylvania, had qualifying insurance, had internet-connected technology, lacked cognitive impairment, and had moderate or high readmission risk. The analyzed cohort included 1286 patients, who were randomized on the first weekday at home after discharge. The four strategies combined low- or high-intensity questionnaires with standard or enhanced response teams and were compared with structured telephone support. The primary endpoint was postdischarge home days at 90 days, with death assigned a value of minus 1.

In the primary analysis, all superiority probabilities remained below 55%, and cumulative odds ratios versus usual care were 0.96, 0.86, 1.01, and 0.96. Secondary outcomes, including mortality, readmission, functional status, quality of life, emergency department use, and hospice services, also did not differ significantly. At least one readmission occurred in 37.8% of usual care patients and in 36.3% to 44.2% across intervention groups. Mortality and other postdischarge measures were likewise broadly similar between monitored and usual-care strategies.

Among patients aged 65 years or older, standard-response and enhanced-response approaches were associated with fewer home days than usual care. Their cumulative odds ratios were 0.56 and 0.67, with inferiority probabilities of 99.6% and 97.9%, respectively. Among patients discharged to a skilled nursing facility before randomization, the corresponding cumulative odds ratios were 0.41 for standard response and 0.29 for enhanced response. No subgroup differences were observed among patients discharged home, leaving these signals within an otherwise neutral overall trial.

Of 887 patients assigned to remote monitoring, 529 enrolled in the program, yielding a 59.6% engagement rate. Monitoring started once patients returned home and was paused during any readmission over follow-up. Qualitative interviews suggested that many participants found the technology easy to use, while others were frustrated by having to remember questionnaires, the standardized nature of the monitoring and call center responses, and limited interpersonal connection. Some participants also questioned whether the program could resolve problems after discharge. The trial was limited to an insurance-defined western Pennsylvania cohort with required internet-connected devices, evaluated symptom-based rather than physiologic monitoring, home visits, or hospital-at-home care, and did not improve the main recovery measure overall.