Corrigendum to 'Continuous Vital Sign Monitoring on Surgical Wards: The COSMOS Pilot

Alerts for vital sign abnormalities seek to identify meaningful patient instability while limiting alarm fatigue. Optimal vital sign alarm settings for postoperative patients remain unknown, as is whether alerts lead to effective clinical responses reducing vital sign disturbances. We conducted a 2-phase pilot study to identify thresholds and delays and test the hypothesis that alerts from continuous monitoring reduce the duration of vital sign abnormalities.

Two-phase pilot.

250 adults having major non-cardiac surgery.

Surgical wards.

All patients had routine vital sign monitoring by nurses at 4-h intervals. We initially continuously recorded clinician-blinded saturation, heart rate, and respiratory rate in 100 patients. In the second phase, we randomized 150 patients to blinded versus unblinded continuous vital sign monitoring. In unblinded patients, nurses were verbally alerted to abnormal vital signs.

In the first phase, we modeled expected alarm counts using 6082 h of continuous oxygen saturation, heart rate, and respiratory rate data. Thresholds and delays targeting a limited number of meaningful alerts were selected for phase two. The primary analysis in phase 2 assessed the effect of unblinded monitoring across a 5-component composite of cumulative durations of vital sign abnormalities. Secondary outcomes included a fraction of alerts deemed meaningful by nurses and number of clinical interventions.

In phase one, we identified alarm settings that yielded an average of 1.0 alerts per patient per day. In phase two, there were an average of 0.45 alerts per patient per day, with fewer alerts presumably resulting because nurses were allowed to alter the initial notification thresholds. The median [Q1, Q3] duration of SpO₂ <85 % was 6.7 [1.2, 18] minutes in unblinded patients vs. 4.6 [0.83, 40] in the blinded group. For RR <4 breaths/min, durations were 0 [0,0] vs. 0 [0, 0.15] (unblinded vs. blinded). RR >30 was 9.0 [2.4, 23] vs. 11 [2.6, 22] minutes. HR <45 bpm lasted 0.02 [0, 1.2] vs. 0.44 [0, 2.2] minutes. HR >130 bpm was 0 [0, 1.3] vs. 0 [0, 0.9] minutes. The average relative effect ratio of geometric means for duration of vital signs exceeding thresholds was 0.72 [95 % CI: 0.47, 1.1], P = 0.11. Among the 73 alarms, 60 (82 %) were considered useful in unblinded patients, leading to 49 interventions in the unblinded group, compared to 30 interventions in the blinded group.

Using the continuous saturation, heart rate, and respiratory rate thresholds established in Phase 1, we generated approximately 0.45 alerts per patient per day during Phase 2, nearly all of which were considered useful by nurses Although the difference in vital sign abnormalities between the unblinded and blinded groups was not statistically significant, unblinded monitoring and nursing alerts led to more interventions (mostly increasing oxygen delivery).

govregistration:NCT05280574.