INTRODUCTION: Preterm birth (PTB) remains a leading cause of perinatal morbidity and mortality worldwide. Understanding Singapore's PTB trends and associated risk factors can inform effective strategies for screening and intervention. This study analyses PTB trends in Singapore from 2017 to 2023, identifies risk factors in this multi-ethnic population and evaluates a predictive model for PTB. METHOD: A retrospective analysis of all PTBs between 22+0 and 36+6 weeks of gestation, from 1 January 2017 to 31 December 2023, was performed by extracting maternal and neonatal data from electronic medical records. These PTBs were taken from the registry of births for Singapore and SingHealth cluster data. Cochran- Armitage trend test and multinomial logistic regression were used. An extreme gradient boosting (XGBoost) model was developed to test and predict the risk of PTB. RESULTS: The PTB rate in Singapore did not show a significant change. However, there was modest downward trend in the SingHealth population from 11.3% to 10.2%, mainly in late spontaneous PTBs (sPTBs). sPTBs accounted for ∼60% of PTBs. Risk factors for very/extreme sPTB included Chinese ethnicity, age ≥35 years, body mass index (BMI) ≥23 kg/m², being unmarried, primiparity, twin pregnancy and maternal blood group AB. The XGBoost model achieved an area under the receiver operating characteristic curve of 0.75, indicating moderate ability to predict PTB. CONCLUSION The overall PTB rate in Singapore has not improved. This study underscores the importance of local factors, particularly advanced maternal age, BMI, primiparity, unmarried, Chinese ethnicity and maternal blood group AB influencing PTB risk. Artificial intelligence methods show promise in improving PTB risk stratification, ultimately supporting personalised care and intervention.
Humans ; Singapore/epidemiology* ; Retrospective Studies ; Female ; Risk Factors ; Premature Birth/ethnology* ; Pregnancy ; Adult ; Infant, Newborn ; Asian People/statistics & numerical data* ; Gestational Age ; Body Mass Index ; Maternal Age ; Logistic Models ; Ethnicity

INTRODUCTIONOver the years, surveys have become powerful tools for assessing a wide range of outcomes among patients. Healthcare managers and professionals now consider patient satisfaction as an outcome by itself. This study aims to determine if results of a patient satisfaction survey are affected by the manner by which the survey instrument is administered.

MATERIALS AND METHODSA patient satisfaction survey was conducted from May 2006 to October 2007 in a tertiary level acute care facility. All patients admitted to the observation unit during the study period were invited to participate. Using a contextualized version of the Consumer Assessment of Healthcare Providers and Systems (CAHPS) Hospital Survey, data was collected through either a phone interview, face to face interview or self-administered questionnaire. Each of these survey modes was administered during 3 different phases within the study period.

RESULTSEight hundred thirty-two (832) patients were included in the survey. Based on results of univariate analysis, out of the 18 questions, responses to 11 (61.1%) were related to survey mode. Face-to-face interview resulted in the greatest proportion of socially desirable responses (72.7%), while phone interview yielded the highest proportion of socially undesirable responses (63.3%). After controlling for possible confounders, logistic regression results showed that responses to 55.6% of the questions were affected by survey mode. Variations in response between phone interview and self-administered questionnaire accounted for 87.5% of the observed differences.

CONCLUSIONSResearchers must be aware that the choice of survey method has serious implications on results of patient satisfaction surveys.

Adult ; Emergency Service, Hospital ; Female ; Health Care Surveys ; methods ; Hospitals ; standards ; Humans ; Interviews as Topic ; Male ; Middle Aged ; Patient Satisfaction ; Quality Assurance, Health Care ; Singapore ; Young Adult