Professional quality of life and its associated factors among Vietnamese doctors and nurses.

BACKGROUND: Prevalence of health workers with occupational health issues ranked fourth among all careers resulting in a reduction in quality of life. However, tools to measure professional quality of life (ProQoL) are unavailable in Vietnamese. This study aims to develop a Vietnamese version of the ProQoL, and examine ProQoL and its associated factors among doctors and nurses. METHODS: The ProQoL is comprised of 30 items measures compassion satisfaction (CS), burnout (BO), and secondary traumatic stress (STS). The tool was translated into Vietnamese following the Guideline by Guillemin et. al (1993), reviewed by expert panels, and validated for internal consistency and test-retest reliability among 38 health workers working at hospitals in HCMC. The validated tool was then used in a cross-sectional study to measure the ProQoL of full-time doctors and nurses working in clinical departments at the University Medical Center, University of Medicine and Pharmacy at Ho Chi Minh City, Vietnam. In addition to the ProQoL, self-reported data about demographic and occupational characteristics were collected. RESULTS: The Vietnamese version of ProQoL achieved high internal consistency (alphas between 0.85 and 0.91) and Intra-class Correlation Coefficients (ICCs between 0.71 and 0.89) for all subscales. Among 316 health workers, mean scores of CS, BO, STS were 36.4 (SD = 5.4), 24.9 (SD = 5.1), 25.9 (SD = 5.3), respectively, indicating moderate levels of CS, BO and STS. Participants who were older (b = 0.17, 95%CI = 0.08, 0.26), had sufficient perceived income (b = 2.59, 95%CI = 0.93, 4.24), and > 10 years of working experience (b = 2.15, 95%CI = 0.68, 3.62), had higher CS scores. Those who were older (b=-0.15, 95%CI=-0.23, -0.07), had sufficient perceived income (b=-2.64, 95%CI=-4.18, -1.09), > 10 years of experience (b=-1.38, 95%CI=-2.76, -0.01), worked in surgical department (b=-1.46, 95%CI=-2.54, -0.38) and 8 hours/day (b=-1.52, 95%CI=-2.61, -0.44), had lower BO scores. Moreover, those in a relationship (b=-2.27, 95%CI=-3.53, -1.01) and had sufficient perceived income (b=-1.98, 95%CI=-3.64, -0.32) had lower STS scores. CONCLUSIONS: The Vietnamese version of ProQoL is valid and reliable for use among Vietnamese health workers. Age, marital status, perceived income status, years of working experience, daily working hours, and specialty was associated with at least one component of ProQoL but gender, religion, education level, and monthly income were not.

Applying Machine Learning to Identify Anti-Vaccination Tweets during the COVID-19 Pandemic.

Anti-vaccination attitudes have been an issue since the development of the first vaccines. The increasing use of social media as a source of health information may contribute to vaccine hesitancy due to anti-vaccination content widely available on social media, including Twitter. Being able to identify anti-vaccination tweets could provide useful information for formulating strategies to reduce anti-vaccination sentiments among different groups. This study aims to evaluate the performance of different natural language processing models to identify anti-vaccination tweets that were published during the COVID-19 pandemic. We compared the performance of the bidirectional encoder representations from transformers (BERT) and the bidirectional long short-term memory networks with pre-trained GLoVe embeddings (Bi-LSTM) with classic machine learning methods including support vector machine (SVM) and naïve Bayes (NB). The results show that performance on the test set of the BERT model was: accuracy = 91.6%, precision = 93.4%, recall = 97.6%, F1 score = 95.5%, and AUC = 84.7%. Bi-LSTM model performance showed: accuracy = 89.8%, precision = 44.0%, recall = 47.2%, F1 score = 45.5%, and AUC = 85.8%. SVM with linear kernel performed at: accuracy = 92.3%, Precision = 19.5%, Recall = 78.6%, F1 score = 31.2%, and AUC = 85.6%. Complement NB demonstrated: accuracy = 88.8%, precision = 23.0%, recall = 32.8%, F1 score = 27.1%, and AUC = 62.7%. In conclusion, the BERT models outperformed the Bi-LSTM, SVM, and NB models in this task. Moreover, the BERT model achieved excellent performance and can be used to identify anti-vaccination tweets in future studies.