138 / 2019-11-05 18:46:12

Spatial-temporal distribution of preterm birth in China: a systematic review and modeling analysis

Preterm birth;,Estimate;,Mainland China;,Linear mixed regression model

Background
Preterm birth is the leading cause of neonatal death worldwide. China has the second highest number of preterm births, which remains a crucial barrier in control of child mortality and improvement of maternal and newborn care. It is important to understand the temporal and spatial patterns of preterm birth in China, which provides useful information for the control and intervention.
Methods
We systematically reviewed data on preterm birth for 31 provinces and municipalities in mainland China from 1 Jan 1990 to 15 September 2019 in databases of PubMed, Embase, Web of Science, CNKI, WANFANG, and VIP. Data of potential influencing factors were obtained through other open-access databases. We applied a Bayesian mixed linear regression model to estimate preterm birth rates in three time periods with spatial random effects at the provincial level. The population-adjusted national estimates was obtained based on the provincial level estimates and the corresponding population.
Results
A total of 27,641 literatures had been reviewed, among which 271 met the inclusion criteria, resulting in 471 survey data in the final modeling analysis. There was an increasing trend of preterm birth rates in China, and the national preterm birth rates were estimated to 2.67% (95% BCI 2.26%-3.16%), 3.84% (95% BCI 3.14%-4.49%) and 4.50% (95% BCI 3.78%-5.28%) in the period 1990-1999, 2000-2009 and from 2010 onwards, corresponding to 0.42 million (95% BCI 0.35-0.49), 0.56 million (95% BCI 0.46-0.66), 0.71 million (95% BCI 0.60-0.84) of live births, respectively. Maps of preterm birth rates at provincial level shows geographical diversity across the country, with Tibet estimated the highest (7.92%, 95% BCI 4.24%-13.91%), followed by Heilongjiang and Shanghai, while Yunnan had the lowest (3.34%, 95% BCI 1.79%-6.32%) preterm birth rate from 2010 onwards. Positive associations were identified between preterm birth risk with GDP per capita, carbon emissions from fire, NO2 level, number of maternal and child health hospitals per capita, while very low or very high temperature may increase the risk. Prospective studies or literatures with high quality tend to report lower rates, whereas studies conducted in tertiary hospitals may reported higher rates compared to lower level hospitals and population-based studies.
Conclusion
This study is the first effort to estimate the spatial-temporal distribution of preterm birth risk in China during the past 30 years, which contributes to a better understanding of the epidemiology of preterm birth in China and assisting of the planning of relevant policies and intervention strategies.