Unattended Portable Monitors in the Preoperative Screening of Obstructive Sleep Apnoea and Their Role in Predicting Postoperative Cardio-Respiratory Complications

Abstract

Preoperative screening for obstructive sleep apnoea (OSA) is recommended to identify patients at risk of perioperative complications. In addition to screening questionnaires, portable monitors (PM), capable of recording oximetry and nasal airflow signals, are increasingly used to screen for OSA. With their simpler setup than a laboratory study, together with automated scoring, type 3 PM home sleep study (HSS) may be performed without a sleep technician (unattended).

We investigated if an unattended type 3 PM HSS could stratify preoperative anaesthetic clinic (PAC) patients for risk of postoperative complications. Secondary aims were analysis of HSS recording failure rate, factors influencing successful HSS recording, and accuracy of automated scoring for HSS.

Forty-eight patients recruited from a PAC were provided with instructions to self-apply the type 3 PM HSS. Recorded HSS data were manually and automatically scored for Apnoea-Hypopnoea Index (AHI) and Oxygen Desaturation Index (ODI). Continuous oximetry was recorded on the first 3 postoperative nights. Incidence of cardiac and respiratory complications was recorded.

Recording failure rate for unattended type 3 PM HSS was high at 31%, although success rate was improved with use of pictorial instructions. Greater accuracy was noted in the automated scoring of ODI than in the automated scoring of AHI for type 3 PM HSS. Moderate-severe OSA, defined by preoperative AHI>=15 events/hr, was not associated with a higher rate of complications, but ODI>=15 events/hr predicted more postoperative adverse events, including prolonged episodes of undetected nocturnal hypoxaemia.

Our results indicate that unattended type 3 PM HSS may not be well suited for preoperative screening of PAC patients. Rather, oximetry screening alone may be better at identifying PAC patients at risk of postoperative complications.