Abstract:
The purpose of this research study aimed to examine whether anxiety, perceived sufficient of sleep, perceived self-efficacy, perceived severity of disease, social support, and uncertainty in weaning for patients with assisted ventilators could predict the success of weaning in critically ill patients with assisted ventilators. Ninety-seven in-patients with respiratory failure under invasive mechanical ventilators were on weaning with oxygen T-piece. Males and females aged between 18 and 59 years were recruited from a multistage random sampling in in-patient departments, Chiangrai Prachanukroh and Lampang Hospitals. Questionnaires were composed of 1) demographic information 2) the success of weaning record 3) the assessment of severity of illness 4) anxiety visual analog scale 5) perceived sufficient of sleep 6) perceived self-efficacy for weaning 7) perceived severity of disease 8) social support questionnaire 9) uncertainty in weaning for patients with assisted ventilators and 10) readiness for weaning the mechanical ventilators. The content validity index of questionnaires 1, 2, 4, 5, 6, 7, 8 and 9 were .80 to 1.0 and the reliabilities of questionnaires 4, 5, 6, 7, 8 and 9 were .92 to .98. Descriptive and binary logistic regression statistics were used to analyze data. The findings showed that three factors significantly predicted the success of weaning in critically ill patients with assisted ventilators were anxiety (odds ratio = 0.92, 95% CI 0.84 0.10), perceived sufficient of sleep (odds ratio = 2.05, 95% CI 1.15 3.64), and social support (odds ratio = 1.26, 95% CI 1.02 1.56, p< .05). However, perceived self-efficacy, perceived severity of disease, and uncertainty in weaning for critically ill patients with assisted ventilators did not predict the success of weaning. Binary logistic regression could predict 96.2% of the success of weaning in critically ill patients with assisted ventilators and 82.4% of the non-success of weaning in critically ill patients with assisted ventilators. By average, the regression could predict 93.8% of the correctness.