Remote patient monitoring for managing acute COVID-19, and mortality and hospital use in Sydney, New South Wales, 2021–22: a retrospective observational cohort study

Objectives
To evaluate the influence of remote patient monitoring (RPM) for managing people with acute coronavirus disease 2019 (COVID-19) on 28-day mortality and hospital use in Australia.

Study design
Retrospective observational cohort study; analysis of deterministically linked NSW Notifiable Conditions Information Management System and hospital, emergency department, and non-admitted patient data.

Setting, participants
South Eastern Sydney Local Health District catchment area residents aged 15 years or older for whom positive severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) test results (polymerase chain reaction or rapid antigen testing) during 26 November 2021 – 30 June 2022 were recorded.

Main outcome measures
Primary outcome: All-cause mortality within 28 days of positive SARS-CoV-2 test result. Secondary outcomes: Hospital length of stay, and numbers of emergency department presentations, hospital admissions, and intensive care unit admissions within 14 days of positive test results. All analyses were undertaken for the unadjusted data (original cohort analysis) and after propensity score matching and inverse probability treatment weighting.

Results
Of 276 236 people aged 15 years or older with positive SARS-CoV-2 test results and complete demographic information, 4399 (1.6%) participated in RPM. Twenty-eight-day mortality was lower for the RPM group than the usual care group (propensity score-matched: adjusted odds ratio [aOR], 0.19; 95% confidence interval [CI], 0.08–0.43; inverse probability treatment-weighted: aOR, 0.21; 95% CI, 0.10–0.46). The 14-day likelihood of intensive care unit admission and emergency department presentation was similar for both groups; the likelihood of hospital admission was higher for the RPM group (propensity score-matched: aOR, 1.42; 95% CI, 1.12–1.78; inverse probability treatment-weighted: aOR, 1.51; 95% CI, 1.28–1.78), but the mean hospital length of stay was shorter (adjusted mean difference, original cohort: –2.01 [95% CI, –2.81 to –1.21] days; propensity score-matched: –3.54 [95% CI, –6.39 to –0.69] days; inverse probability treatment-weighted: –3.26 [95% CI, –6.01 to –0.50] days).

Conclusion
RPM was associated with greater 14-day likelihood of hospital admission, but also with shorter mean length of stay and lower 28-day mortality, which may indicate that clinical deterioration was detected and treated earlier than with usual care. The benefit of RPM for managing other acute health conditions in the community, particularly infectious diseases, should be examined.