A Bayesian network model to predict the role of hospital noise, annoyance, and sensitivity in quality of patient care

Abstract

Background Hospital noise can threaten nurses' health, their cognitive function and emotion and in turn, influence the quality of patient care and patient safety. The aim of this study was to predict the role hospital noise, annoyance, and sensitivity in quality of patient care. Methods This descriptive and cross-sectional study was carried out among nurses in an Iranian hospital. To determine nurses' noise exposure level, the noise was measured in the 1510 points in the hospital in accordance with ISO 9612 using KIMO DB 300/2 sound level meter and analyzer. Then, an online survey was conducted to collect individual data. The applied questionnaires were a demographic questionnaire, Weinstein noise sensitivity scale, noise annoyance scale, and quality patient care scale. Finally, to analyze data, Bayesian Networks (BNs), as probabilistic and graphical models were used. Results For the high noise exposure, high noise sensitivity, and high annoyance with the probability of 100%, the probability of desirable quality of patient care decreased by 21, 14, and 23 percent, respectively. Moreover, at the concurrently high noise exposure and high noise sensitivity with the probability of 100%, the desirable quality of patient care decreased by 26%. The most effective mean values belonged to the association of noise exposure and annoyance (0.636). Moreover, the annoyance had highest association with the physical aspect (0.400) and the sensitivity had greatest association with the communication aspect (0.283). Conclusion Annoyance induced from noise and sensitivity affected the quality of patient care adversely. Moreover, noise and sensitivity had a separate direct adverse effect upon the quality of patient care and their coincidence occurrence reduced the quality of patient care potentially.