Wearable Monitors Have Not Shown Earlier Warning of Decline

Not shown so far. A review pooled 7 studies of wearable continuous monitoring against standard care in hospitalized patients. Across those 7 studies, ICU transfers did not fall significantly (risk ratio 0.87, 95% CI 0.66 to 1.15), and neither did rapid response or cardiac arrest calls (0.84, 0.69 to 1.01). In one COVID-19 ward study, the sensor detected none of the 4 ICU admissions. A 400-patient trial found continuous alerts did not significantly cut severe vital sign deviations.

Does continuous wearable monitoring catch decline earlier?

It has not been shown to. A 2021 systematic review searched for studies comparing wearable monitoring systems with standard care in hospitalized patients. Ten studies qualified and seven were pooled.

The pooled results were not significant. ICU transfers had a risk ratio of 0.87 (95% CI 0.66 to 1.15). Rapid response or cardiac arrest team calls were 0.84 (0.69 to 1.01). Total complications were 0.77 (0.44 to 1.32) and mortality 0.48 (0.18 to 1.29).

The authors conclude there is no current evidence that these systems affect early deterioration detection or related outcomes. They cite differing designs and quality across studies.

What happened in a ward that tested it directly?

The sensor missed the ICU admissions. A COVID-19 ward gave patients a wireless wearable that continuously measured heart rate and respiratory rate. Two ICU physicians independently reviewed the data to mark when an intervention might be needed. A third extracted clinical events from the medical record.

Forty-eight clinical events occurred. None of the 4 ICU admissions was detected by the sensor. Of 62 sensor events, 13 were true alarms, and two of those related to rapid response team calls. True alarms came on average 39 minutes (SD 113) before the record events.

Agreement between the two physicians was 10%. The 38 undetected events were as severe as the 10 detected ones. The authors conclude that continuous monitoring of heart rate and respiratory rate does not reliably detect deterioration in COVID-19 patients.

Did continuous alerts change outcomes in a randomized trial?

Partly. A 2025 trial randomized 400 adults after major noncardiac surgery to usual intermittent checks or usual care plus continuous wireless monitoring with real-time alerts to staff phones. The primary outcome was total time in severe vital sign deviations.

Median time was 60 minutes a day with monitoring and 76 without (P = .19). That is not a significant difference. Time with oxygen saturation under 88% fell by 47 minutes a day (95% CI 18 to 80). Adverse events differed between groups, at 42.5% and 31.5% (P = .02). The authors report significant reductions in desaturations and adverse events.

The authors say the results support further studies. The primary result was null.

Are the sensors' readings accurate enough to trust an alarm?

Sometimes. In 20 patients in a post-anesthesia unit, wireless heart rate had a bias of -0.85 beats per minute (limits -6.2 to 4.5). Wireless respiratory rate against direct observation had a bias of 0.0 breaths per minute (limits -2.6 to 2.6). Eighty percent of all values were within the study's clinical limits.

The Healthdot, an accelerometer patch on the lower rib, was tested in 30 bariatric surgery patients across 22 hours. Heart rate met a 5 beats per minute limit 87.5% of the time and respiration 92.3% of the time.

The RespiraSense monitor did worse on respiratory wards. In 31 patients, automated readings averaged 2.5 breaths per minute higher than manual ones. The 95% limits of agreement were -7.9 to 12.9, wider than the plus or minus 3 that the authors call acceptable. Trends agreed in only 56% of cases.

Manual counts also clustered at 20 and 24, while automated readings followed a smooth curve. See whether a wearable measures respiratory rate accurately and wrist versus chest placement.

What does this not settle?

The evidence covers hospital-grade wireless sensors in wards and recovery units. It does not cover consumer watches or rings, or use at home.

One ward study of a specific disease and sensor cannot show what a different device does in another ward. A 90-patient trial of ViSi Mobile and HealthPatch found most patients and staff positive. Acceptance is not effectiveness.

Common questions