Predictive Modeling of Preoperative Sleep Disorder Risk in Older Adults by Using Data From Wearable Monitoring...

Area of Science:

• Geriatric Medicine
• Sleep Medicine
• Surgical Risk Assessment

Background:

• Sleep disorders are prevalent in older surgical patients, increasing complications and healthcare costs.
• Existing surgical risk tools overlook specific factors affecting sleep in older adults.
• Wearable devices offer objective sleep data, but models often lack surgical context.

Purpose of the Study:

• To develop and validate a risk prediction model for preoperative sleep disorders in older surgical patients.
• To integrate smart wearable device data with clinical assessments for comprehensive risk evaluation.
• To enable early identification of influencing factors for personalized perioperative care planning.

Main Methods:

• Prospective study of 242 older surgical patients using smart rings pre-surgery.
• Collected sociodemographic, cognitive, and psychological data alongside objective sleep metrics.
• Developed and validated a risk model using logistic regression and bootstrap sampling, assessing performance via ROC curves, calibration, and decision analysis.

Main Results:

• Key predictors identified: Hospital Anxiety and Depression Scale score, number of awakenings, and durations of REM and light sleep.
• The model demonstrated strong discrimination (AUC=0.92) and good calibration.
• Decision curve analysis confirmed high clinical utility across various risk thresholds.

Conclusions:

• A smart ring-based risk model effectively identifies older surgical patients with preoperative sleep disturbances.
• This facilitates timely, individualized interventions to improve perioperative care.
• The model offers a scientific basis for enhancing surgical outcomes and reducing healthcare burden in older adults.