This study examined changes in COVID-19-related death certification practices (ie, reporting COVID-19 in Part I of the death certificate) and the quality of cause-of-death (COD) reporting (acceptable and rejected causal sequences) in the United States from 2020 to 2023. In 2020, 2021, 2022, and 2023, 385 293, 463 273, 246 196, and 76 024 death certificates, respectively, mentioned COVID-19. COVID-19 was selected as the underlying COD in 91%, 90%, 76%, and 65% of these certificates, corresponding to reporting COVID-19 in Part I in 93%, 91%, 78%, and 67% of certificates, respectively. Acceptable causal sequences were reported in 51%, 51%, 45%, and 41% of certificates, whereas rejected causal sequences were reported in 21%, 22%, 23%, and 23%, respectively. Using 2020 as the reference year, the adjusted odds ratios (95% confidence intervals) for 2023 were 0.18 (0.18, 0.18) for reporting COVID-19 in Part I, 0.77 (0.76, 0.78) for reporting acceptable causal sequences, and 1.14 (1.12, 1.16) for reporting rejected causal sequences. In conclusion, COVID-19-related death certification practices changed substantially over the study period, paralleling the decline in the virulence of SARS-CoV-2. The quality of COVID-19-related COD reporting declined modestly from 2020 to 2023.

Background:Hemodialysis patients are a priority population for hepatitis C virus (HCV) elimination in Taiwan, where HCV prevalence and end-stage kidney disease burden are high. The introduction of direct-acting antivirals (DAAs) has accelerated elimination efforts. This study evaluated HCV care cascade and identified gaps among Taiwan's hemodialysis patients in the DAA era. Methods:We conducted a retrospective cohort study using National Health Insurance Research Database from 2015 to 2021. Adult patients receiving long-term hemodialysis were included. HCV infection was defined as positive RNA test, and treatment initiation as receipt of ≥7 days of DAA therapy. We assessed the care cascade, annual and cumulative treatment rates, and disparities by demographic and facility characteristics. Results:Among 14 755 HCV antibody-positive patients, 48.2% underwent confirmatory RNA testing. Of 4783 viremic patients, 73.4% initiated DAA therapy. The greatest attrition occurred at the RNA confirmation step. Testing rates were lower among males (47% vs 50% in females), older adults, patients treated in tertiary hospitals (46%), and those in metropolitan areas (48% vs 52% rural). Regional variation was substantial, ranging from 38% to 41% in Eastern, Central, and Taipei regions to 67% in Southern Taiwan. Annual treatment initiation rose from 3.0% in 2017% to 50% in 2019, and then declined to 33% in 2021. Cumulative treatment increased from 10% before 2018% to 73.4% by 2021. Conclusions:HCV care among hemodialysis patients in Taiwan has improved substantially in the DAA era. The major barrier is incomplete RNA testing. Addressing this gap is essential to sustain progress and achieve HCV elimination in this high-risk population.

[This corrects the article DOI: 10.1016/j.ssmph.2025.101884.].

PURPOSE:Suicide rates among adolescents in the United States (US) are rising. We examined intersectional trends in method-specific suicide by race/ethnicity, sex, and age (10-14 vs. 15-19 years) from 2006 to 2023. METHODS:Using US national mortality data, we calculated method-specific suicide rates among adolescents aged 10-19 years across three periods (2006-2011, 2012-2017, and 2018-2023) by race/ethnicity, sex, and age group. Rate ratios with 95% confidence intervals (CIs) were estimated using non-Hispanic (NH) White adolescents as the reference. RESULTS:Across 60 intersectional groups (sex × age × race/ethnicity × period), hanging was the leading suicide method in 51 groups. Suicide rates increased in most groups, but the magnitude varied. Firearm suicide rates rose 10-fold among NH Black girls aged 10-14 years, from 0.06 per 100,000 in 2006-2011 to 0.60 in 2018-2023 versus a three-fold increase among NH White girls (0.15-0.47). The rate ratio for NH Black versus NH White girls increased from 0.40 (95% CI: 0.17-0.93) to 1.28 (95% CI: 0.93-1.76). Poisoning suicide rates increased among NH Black, Hispanic, and Asian and Pacific Islander adolescents aged 15-19 years. Asian and Pacific Islander adolescents aged 15-19 years had the highest rates of suicide by jumping across racial/ethnic groups. DISCUSSION:Method-specific suicide rates are rising among US adolescents, particularly among racial/ethnic minority groups and certain age-sex subgroups. These findings highlight the need for targeted, culturally informed suicide prevention strategies that restrict access to lethal means.

BACKGROUND:Accurate information on where suicides occur; especially in domestic settings, is essential for designing prevention programs. In practice, place of death from death certificates is used as a proxy, though its accuracy is unclear. This study assessed the validity of using death certificate data to determine place of suicide, using linked records from the Suicide Surveillance System in Taiwan. METHODS:We linked mortality data with the Suicide Surveillance System in Kaohsiung City, Taiwan, from 2021 to 2023. Using the place of suicide recorded in the surveillance system as the reference standard, we estimated the proportion of suicide deaths in each place-of-death category (home, hospital, and other) that actually occurred at home. RESULTS:Among 1465 suicide deaths, 965 were successfully linked to the Suicide Surveillance System. The positive predictive value for death at home was high: 99.3% for hanging, 97.8% for charcoal burning, and 100.0% for other poisoning. Many suicides recorded as hospital deaths actually occurred at home, including 79.3% of hanging, 66.7% of charcoal burning, 85.5% of other poisoning, and 71.6% of jumping cases. After adjustment, the estimated proportion of suicides occurring at home was 80.4% for hanging, 68.2% for charcoal burning, 77.9% for other poisoning, 75.0% for jumping from a height, and 60.2% for other methods. CONCLUSIONS:Using death certificates alone underestimates suicides at home. Improved data integration is needed to support more accurate surveillance and better-targeted suicide prevention efforts.

Studies have indicated that the risk of death on the first day of life (day 0) was higher than risk of death during other periods (days 1 to 6 and 7 to 27). However, little is known about whether the pattern of mortality trends on day 0 differs from those on days 1 to 6 and 7 to 27. We aimed in this study to examine NMRs trends by age at death in Japan, Korea, and Taiwan. In this cross-sectional study, we calculated NMRs (deaths per 1000 live births) by age at death from 2005 to 2021 in Japan, 2005 to 2022 in Korea, and 2005 to 2023 in Taiwan. Joinpoint regression model was used to estimate the annual percent change (APC) for each segment of the trend in NMRs to examine whether the trend changed significantly. A slowdown of decreasing trend on days 0 to 27 was observed from 2015 to 2021 with APC of − 4.3

This study assessed the quality of cause-of-death reporting in the United States before and during the COVID-19 pandemic. We used the selection rate and the adjusted odds ratio (aOR) to analyze each cause identified by the National Center for Health Statistics as unsuitable for the underlying cause of death (UCOD). The selection rate was defined as the proportion of deaths with mention of a particular unsuitable UCOD on the death certificate where that cause was ultimately selected as the UCOD. Out of 36 unsuitable UCODs, 33 exhibited a significant decline in selection rates from 2019 to 2021. However, when deaths with mention of COVID-19 on the death certificate were excluded, only 19 causes revealed a significant decline. In analyses that controlled for the age of decedents, aORs in 2021 were significantly lower compared with 2019 for 26 causes, and this number decreased to 17 causes in analyses that excluded COVID-19-related deaths. In conclusion, the overall quality of COD reporting improved during the COVID-19 pandemic, attributable mainly to the fact that over one-tenth of the deaths were related to COVID-19. Yet, for deaths that did not involve COVID-19, improvements in the quality of COD reporting were less prominent for certain causes.

The World Health Organization (WHO) classification scheme for leading causes of death (LCODs) provides greater granularity than the National Center for Health Statistics (NCHS) scheme by subdividing broad categories such as cancer, heart disease, and accidents into 17, 8, and 7 specific groups, respectively. This study compared LCOD rankings among U.S. adolescents aged 10-19 years across racial and ethnic groups using CDC WONDER mortality data. Across 48 sex-age-race/ethnicity-period strata, top-ranked causes differed between schemes in 31 strata. For example, policy priorities for non-Hispanic White boys aged 15-19 years would differ by classification: accidents ranked first under the NCHS scheme, whereas suicide or transportation accidents ranked highest under the WHO scheme. Substantial changes in ranks of top five LCODs across years were observed in some racial/ethnic groups. Among boys aged 10-14 years, transportation accidents rose from third in 2018-2019 to first in 2020-2021 and remained first in 2022-2023 among non-Hispanic Asians, while remaining consistently second among non-Hispanic Black and White boys and first among Hispanic boys. The WHO and NCHS schemes offer complementary LCOD profiles that can guide adolescent mortality surveillance and health policy decision-making.

Abstract Background Opioid use disorder (OUD) contributes to a heavy burden of disease across the globe, including excessive early mortality and poor quality of life. Methadone and buprenorphine are classified by the World Health Organization (WHO) as Essential Medicines as opioid agonist treatment (OAT) of OUD1. Although evidence showed OAT substantially reduces all-cause mortality among people with OUD2, the cost- effectiveness of OAT should be further evaluated. For the evaluation of health utility difference, one optimal measure is health benefit in terms of quality-adjusted life year (QALY), of which both survival and quality of life (QoL) are taken into consideration. Also, it will be more suitable to take a lifetime perspectives accounting for both life expectancy and QoL when capturing the health impact of OAT on opioid-dependent individuals. Aims & Objectives We first estimated the life expectancy (LE) and loss of life expectancy (loss-of-LE) from an OUD cohort stratified by OAT for comparison, then attempted to quantify the difference in loss of quality-adjusted life expectancy (QALE) for demonstration the health utility gain by OAT. The lifetime medical costs were then calculated using the National Health Insurance (NHI) database in Taiwan. We evaluated the cost-effectiveness of OAT versus non-OAT treatment by integrating a survival function with mean-cost over a lifelong horizon. Methods A total of 1283 participants with OUD recruited in the beginning of OAT implementation were linked to the National Mortality Registry for 8 years with stratification by OAT. Kaplan-Meier estimation for survival was performed, and it was extrapolated to 70 years to obtain the LE using a semi-parametric method. We further estimated the loss-of-LE for both cohorts by subtracting their life expectancies from the age- and sex- matched referents of the general population. Cause-specific standardized mortality ratios (SMRs) were calculated and compared with the national cohort to validate the representativeness of this sample. Then, a subsample of subjects' quality of life with EQ-5D was collected for utility values and multiplied by survival functions to estimate the QALE, while that for the age- and sex-matched referents were adjusted to the value collected from the 2009 National Health Interview Survey, and the difference between them was the loss-of-QALE. Finally, we linked the medical claims of NHI datasets with the survival function to estimate the mean cumulative costs since the first date of their diagnosis then extrapolate to lifetime healthcare expenditure of both the OAT/non-OAT cohorts. Results Both survival and QOL of the OAT group were significantly better than those of the non-OAT group. After lifetime extrapolation, the estimated QALE and loss-of-QALE were 17.8 and 18.2 QALY for OAT subjects, respectively, while those of the non-OAT group were 9.2 and 27.9 QALY. The incremental cost- effectiveness ratio (ICER) of OAT versus non-OAT group were US$1,808 per life-year saved and US$1,454 per QALY saved, respectively. Discussion & Conclusion Take cost-effectiveness into consideration, OAT should be further promoted for opioid-dependent population. References 1.World Health Organization. 2005. WHO Model List of Essential Medicines 14th Edition. http://whqlibdoc.who.int/hq/2005/a87017_eng.pdf. Accessed February 2009. 2.Sordo L, Barrio G, Bravo MJ, Indave BI, Degenhardt L, Wiessing L, Ferri M, Pastor-Barriuso R. 2017. Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ. DOI: 10.1136/bmj.j1550.

Cesarean section (CS) rate has become increasingly prevalent worldwide, which has raised concerns about the possible risks as they often result in frequently longer recovery periods for mothers and possible complications for both the mother and the child. The World Health Organization (WHO) recommends a 10–15

Background Population aging has increased the prevalence of multimorbidity, jeopardizing the sustainability and efficiency of healthcare systems. This study aimed to evaluate the effects of an integrated ambulatory care program (IACP) on healthcare utilization and costs among older patients with multimorbidity while accounting for the confounding effects of frailty. Methods A retrospective cohort study using propensity matching including patients aged 65 or older with two or more chronic conditions attending the outpatient clinic at our hospital between June 1 and December 31, 2019, was conducted. Exposure was defined as receipt of IACP care. Patients not undergoing the IACP comprised the unexposed group and were matched at a ratio of 1:4 to patients undergoing the IACP group according to sex, age, Charlson Comorbidity Index score, multimorbidity frailty index score, and number of outpatient visits within 6 months before the index date. Outcomes were changes in healthcare utilization and related costs between 6 months before and after receiving IACP care. Multivariate regression analyses were used for data analysis and the Generalized Estimation Equation method was used to fit the regression models. Results A total of 166 (IACP) and 664 (non-exposed) patients were analyzed. The mean participant baseline ages were 77.15 ± 7.77 (IACP) and 77.28 ± 7.90 years (unexposed). In univariate analyses, the IACP group demonstrated greater reductions than the unexposed group in the frequency of outpatient visits (-3.16 vs. -1.36, p < 0.001), number of physicians visited (-0.99 vs. -0.17, p < 0.001), diagnostic fees (-1300 New Taiwan Dollar [NTD] vs. -520 NTD, p < 0.001), drug prescription fees (-250 NTD vs. -70 NTD, p < 0.001), and examination fees (-1620 NTD vs. -700 NTD, p = 0.014). Multivariate analyses demonstrated that patients in the IACP group experienced significant reduction in the frequency of outpatient visits (95% CI: -0.357 to -0.181, p < 0.001), number of physicians visited (95% CI: -0.334 to -0.199, p < 0.001), and overall outpatient costs (95% CI: -0.082 to -0.011, p = 0.01). However, emergency department utilization, hospitalization, and costs did not differ significantly. Conclusions Expanding IACPs may help patients with multimorbidity reduce their use of outpatient clinics at the 6-month follow-up, reduce care fragmentation, and promote sustainability of the healthcare system.

Background Dengue is a significant mosquito-borne disease. Several studies have utilized estimates from the Global Burden of Disease (GBD) study to assess the global, regional or national burden of dengue over time. However, our recent investigation suggests that GBD's estimates for dengue cases in Taiwan are unrealistically high. The current study extends the scope to compare reported dengue cases with GBD estimates across 30 high-burden countries and territories, aiming to assess the accuracy and interpretability of the GBD's dengue estimates.Methods Data for this study were sourced from the GBD 2019 study and various national and international databases documenting reported dengue cases. The analysis targeted the top 30 countries and territories with the highest 10-year average of reported cases from 2010 to 2019. Discrepancies were quantified by computing absolute differences and ratios between the 10-year average of reported cases and GBD estimates. Coefficients of variation (CV) and estimated annual percentage changes (EAPCs) were calculated to assess variations and trends in the two data sources.Results Significant discrepancies were noted between reported data and GBD estimates in the number of dengue cases, incidence rates, and EAPCs. GBD estimates were substantially higher than reported cases for many entities, with the most notable differences found in China (570.0-fold), India (303.0-fold), Bangladesh (115.4-fold), Taiwan (85.5-fold) and Indonesia (23.2-fold). Furthermore, the GBD's estimates did not accurately reflect the extensive yearly fluctuations in dengue outbreaks, particularly in non-endemic regions such as Taiwan, China and Argentina, as evidenced by high CVs.Conclusions This study reveals substantial discrepancies between GBD estimates and reported dengue cases, underscoring the imperative for comprehensive analysis in areas with pronounced disparities. The failure of GBD estimates to represent the considerable annual fluctuations in dengue outbreaks highlights the critical need for improvement in disease burden estimation methodologies for dengue.

The Fracture Risk Assessment Tool (FRAX®) is a widely utilized country-specific calculator for identifying individuals with high fracture risk; its score is calculated from 12 variables, but its formulation is not publicly disclosed. We aimed to decompose and simplify the FRAX® by utilizing a nationwide community survey database as a reference module for creating a local assessment tool for osteoporotic fracture community screening in any country. Participants (n = 16384; predominantly women (75%); mean age = 64.8 years) were enrolled from the Taiwan OsteoPorosis Survey, a nationwide cross-sectional community survey collected from 2008 to 2011. We identified 11 clinical risk factors from the health questionnaires. BMD was assessed via dual-energy X-ray absorptiometry in a mobile DXA vehicle, and 10-year fracture risk scores, including major osteoporotic fracture (MOF) and hip fracture (HF) risk scores, were calculated using the FRAX®. The mean femoral neck BMD was 0.7 ± 0.1 g/cm2, the T-score was −1.9 ± 1.2, the MOF was 8.9 ± 7.1%, and the HF was 3.2 ± 4.7%. Following FRAX® decomposition with multiple linear regression, the adjusted R2 values were 0.9206 for MOF and 0.9376 for HF when BMD was included and 0.9538 for MOF and 0.9554 for HF when BMD was excluded. The FRAX® demonstrated better prediction for women and younger individuals than for men and elderly individuals after sex and age stratification analysis. Excluding femoral neck BMD, age, sex, and previous fractures emerged as 3 primary clinical risk factors for simplified FRAX® according to the decision tree analysis in this study population. The adjusted R2 values for the simplified country-specific FRAX® incorporating 3 premier clinical risk factors were 0.8210 for MOF and 0.8528 for HF. After decomposition, the newly simplified module provides a straightforward formulation for estimating 10-year fracture risk, even without femoral neck BMD, making it suitable for community or clinical osteoporotic fracture risk screening.

The Fracture Risk Assessment Tool (FRAX (R)) is a widely utilized country-specific calculator for identifying individuals with high fracture risk; its score is calculated from 12 variables, but its formulation is not publicly disclosed. We aimed to decompose and simplify the FRAX (R) by utilizing a nationwide community survey database as a reference module for creating a local assessment tool for osteoporotic fracture community screening in any country. Participants (n = 16384; predominantly women (75%); mean age = 64.8 years) were enrolled from the Taiwan OsteoPorosis Survey, a nationwide cross-sectional community survey collected from 2008 to 2011. We identified 11 clinical risk factors from the health questionnaires. BMD was assessed via dual-energy X-ray absorptiometry in a mobile DXA vehicle, and 10-year fracture risk scores, including major osteoporotic fracture (MOF) and hip fracture (HF) risk scores, were calculated using the FRAX (R). The mean femoral neck BMD was 0.7 +/- 0.1 g/cm(2), the T-score was -1.9 +/- 1.2, the MOF was 8.9 +/- 7.1%, and the HF was 3.2 +/- 4.7%. Following FRAX (R) decomposition with multiple linear regression, the adjusted R-2 values were 0.9206 for MOF and 0.9376 for HF when BMD was included and 0.9538 for MOF and 0.9554 for HF when BMD was excluded. The FRAX (R) demonstrated better prediction for women and younger individuals than for men and elderly individuals after sex and age stratification analysis. Excluding femoral neck BMD, age, sex, and previous fractures emerged as 3 primary clinical risk factors for simplified FRAX (R) according to the decision tree analysis in this study population. The adjusted R-2 values for the simplified country-specific FRAX (R) incorporating 3 premier clinical risk factors were 0.8210 for MOF and 0.8528 for HF. After decomposition, the newly simplified module provides a straightforward formulation for estimating 10-year fracture risk, even without femoral neck BMD, making it suitable for community or clinical osteoporotic fracture risk screening.

This cross-sectional study examines international variations in neonatal mortality rates with and without minimum thresholds for gestational age (22 weeks) and birth weight (500 g) in 2010 and 2020.

Objective: Whether pregnancy is associated with severe injuries from motor vehicle crashes (MVCs) remains unclear. This study aimed to investigate the potential relationship between pregnancy and severity of injuries from MVCs. Methods: We identified a total of 23,559 pregnant women victims who encountered MVCs during pregnancy as well as 94,236 age- and calendar year-at MVC matched non-pregnant women victims that are also involved in MVCs. Injury severity was assessed using the Maximum Abbreviated Injury Scale (MAIS) based on the diagnosis of medical claims after MVCs. Multinomial logistic regression models were used to estimate the odds ratio and corresponding 95 % CI of injury severity levels associated with pregnancy. Results: Pregnant women had a significantly higher risk of both severe (adjusted odds ratio, aOR = 1.79, 95 % CI = 1.54-2.08) and mild injuries (aOR = 8.63, 95 % CI = 8.21-9.07) following MVCs as compared to non-pregnant women victims. Particularly, pregnant women who were riding scooters had an increased risk of severe injury (aOR = 4.25, 95 % CI = 3.58-5.04). In addition, pregnant women who experienced MVC but without any injury were more likely to visit a clinic than non-pregnant MVC victims. Conclusion: Pregnant women victims, particularly those who were riding scooters involved in MVCs suffered from a higher risk of severe injury as compared to their non-pregnant counterparts. Our findings suggest that women should consider avoiding riding a scooter and must use restrictive devices during pregnancy, which would help reduce the severity of injuries sustained following an MVC.

Taiwan's dengue cases vary annually, peaking in infrequent epidemics, which differ substantially from the Global Burden of Disease Study's (GBD's) projections. Although the GBD study provides invaluable insights into global health trends, its modelling approach fails to capture the dynamic change of dengue transmission.

BackgroundTo evaluate the association between maternal MVCs during pregnancy and neurodevelopmental disorders (NDDs, including intellectual disability, ADHD, ASD, and infantile cerebral palsy) in children.MethodsThis population-based cohort of live births in Taiwan was analyzed, comparing children born to mothers involved in MVCs during pregnancy with those without such exposure. Children were linked to the insurance database to identify the possible diagnosis of NDDs. The Cox proportional hazards regression model was used to estimate the relative hazards.ResultsA total of 19,277 children with maternal MVCs and 76,015 children without exposure were included. Children exposed to maternal MVCs during the first two trimesters or whose mothers sustained mild to severe injuries showed a higher risk of intellectual disability. Severe maternal injuries also increased the risk of infantile cerebral palsy (aHR = 3.86; 1.27-11.78). MVCs in the third trimester, or mild maternal injuries, were associated with a higher risk of ASD (third trimester: aHR = 1.40; 1.04-1.87; mild injuries: aHR = 1.38; 1.09-1.74).ConclusionChildren exposed to maternal MVCs with severe injuries had a higher risk of intellectual disability and cerebral palsy. Third-trimester exposure may increase the risk of ASD. However, these findings should be interpreted cautiously as genetic factors may contribute to the observed association.ImpactThere is some evidence linking maternal MVCs during pregnancy to the development of neurodevelopmental disorders in children.Children of mothers with severely injured were more likely to suffer from infantile cerebral palsy and intellectual disability.The risk of autism spectrum disorder is higher in children whose mothers are involved in MVCs during the late stage of pregnancy, and there is also an increased risk of intellectual disability during the first two trimesters.

Objective: Late initiation of prenatal care (LIPC; i.e., initiation of the first prenatal visit after the first trimester) is a major public health concern for pregnant mothers and fetuses. This study examined the social determinants of LIPC in Taiwan, where universal health coverage has been provided since 1995. Methods: The study population comprised 185,855 babies born in 2017 to mothers who were Taiwanese nationals. Data from several databases, namely the Birth Registration Database, Birth Certificate Application Database, Low- and Middle-Low-Income Households Database, Persons with Disabilities Database, and National Health Insurance Research Database, were retrieved to link data on social determinants to medical claims. The gradient effect of each investigated social determinant was assessed using the Cochran–Armitage trend test in SAS 9.4. Results: The overall incidence rate of LIPC in Taiwan was 8.6% in 2017. Notably, the disparities in the initiation of prenatal care were closely associated with social disadvantages. In particular, extraordinarily high LIPC rates were identified among teenagers (40.0%), mothers with drug addiction problems (31.9%), unmarried mothers (27.6%), mothers from low income (23.0%), mothers with a high school or lower level of education (23.0%), migrants (19.2%), and mothers with disabilities (18.9%). Several social determinants had significant adverse gradient effects on LIPC, namely employment (P < .0001), education (P < .0001), socioeconomic status (P < .0001), and urbanization (P < .0001). Logistic regression revealed higher odds of LIPC among the mothers in the poorest 20% (adjusted odds ratio [aOR], 1.52; 95% confidence interval [CI], 1.43–1.61), those with a parity of >3 (aOR, 1.99; 95% CI, 1.89–2.09), unmarried mothers (aOR, 2.88; 95% CI, 2.73–3.04), mothers with drug addiction problems (aOR, 5.24; 95% CI, 4.78–5.75), mothers aged <19 years (aOR, 5.30; 95% CI, 4.84–5.81) or 20–24 years (aOR, 1.77; 95% CI, 1.67–1.89), and mothers with disabilities (aOR, 1.64; 95% CI, 1.41–1.92). Conclusions: The findings of this study highlight the social inequalities in the initiation of prenatal care in Taiwan, indicating that the maternal and child health-care system in Taiwan should be further improved.