Predictors of pulse oximetry data failure.

Background: Pulse oximeters have been reported to fail to record data in 1.12-2.50% of cases in which anesthesia records were handwritten. There is reason to believe that these may be underestimates. Computerized anesthesia records may provide insight into the true incidence of pulse oximetry data failures and factors that are associated with such failures.
Methods: The current study reviewed case files of 9,203 computerized anesthesia records. Pulse oximetry data failure was defined as the presence of at least one continuous gap in data > or = 10 min in duration in a case. A multivariate logistic regression model was used to identify predictors of pulse oximetry data failure, and a modified case-control method was used to determine whether extremes of blood pressure and hypothermia during the procedure were associated with pulse oximetry data failure.
Results: The overall incidence of cases that had at least one continuous gap of > or = 10 min in pulse oximetry data was 9.18%. The independent preoperative predictors of pulse oximetry data failure were ASA physical status 3,4, or 5 and orthopedic, vascular, and cardiac surgery. Intraoperative hypothermia, hypotension, hypertension, and duration of procedure were also independent risk factors for pulse oximetry data failure.
Conclusions: Pulse oximetry data failure rates based on review of computerized records were markedly greater than those previously reported. Physical status, type of surgery, and intraoperative variables were risk factors for pulse oximetry data failure. Regulations and expectations regarding pulse oximetry monitoring should reflect the limitations of the technology.