第一部分 全球先心病死亡率的变化趋势及其年龄-时期-队列效应

背景

先天性心脏病（先心病）是全球范围内出生缺陷疾病负担和死亡的首要病因。本研究旨在分析过去30年间全球204个国家和地区先心病死亡率的变化趋势以及与年龄、时期和出生队列的关系。

方法

采用2019年全球疾病负担研究（Global Burden of Disease study, GBD 2019）的先心病死亡数据进行分析，运用年龄-时期-队列模型（Age-Period-Cohort model， APC model）估计先心病死亡率的净变化值（年均百分比变化值，%/年）以及不同年龄段（0-4岁、5-9岁、…60-64岁、65-69岁）死亡率的局部变化值（%/年），并计算1990-2019年时期和队列的相对死亡风险。这种建模方法可以区分死亡率趋势中的年龄、时期和队列效应，有助于比较先心病医疗质量的差异和差距。

结果

先心病是20岁以下人群因非慢性传染性疾病死亡的主要原因。2019年，全球范围内先心病死亡人数为21.7万（95%UI: 17.7-26.2万）。印度、中国、巴基斯坦和尼日利亚四国的先心病死亡负担最高，占全球死亡人数的39.7%。在1990-2019年期间，高社会人口指数（SDI）国家先心病死亡率的净变化值为-2.41%/年（95% CI: [-2.55, -2.67]），低SDI国家仅为-0.62%/年（95%CI: [-0.82, -0.42]）。全球范围内，先心病死亡的年龄分布逐渐从儿童转变为成年人群，而墨西哥和巴基斯坦等国10-34岁人群的死亡率呈现上升趋势。过去30年间大多数高SDI国家比如韩国（净变化值：-4.0%/年, 95%CI: [-4.8, -3.1]）和美国（-2.3%/年[-2.5, -2.0]）以及许多中等SDI国家例如巴西（-2.7%/年[-3.1, 2.4]）和南非（-2.5%/年[-3.2, -1.8]）的死亡率呈现稳定下降趋势。然而，在129个死亡人数50以上的国家中，52个国家的死亡率呈现增长趋势（净变化值≥0.0%/年）或下降趋势停滞（≥-0.5%/年）。除沙特阿拉伯和哈萨克斯坦外，中高SDI和高SDI国家随着时间的推移和出生队列相对死亡风险普遍呈现改善的趋势。厄瓜多尔和墨西哥等14个中等SDI国家、包括印度在内的16个中低SDI国家以及包括巴基斯坦在内的20个低SDI国家，时期和出生队列的相对死亡风险出现不断恶化的趋势。

结论

先心病死亡率是评估儿童和成年先心病医疗服务质量的有用和可获得的指标。先心病医疗服务体系的改善将在人群层面降低出生队列的死亡风险，并随着时间的推移降低所有年龄段患者的死亡风险。尽管过去30年全球先心病死亡率持续降低，但许多国家地区的不良时期和队列效应表明所有年龄段的先心病人群可能未获得充分治疗。先心病医疗服务的有效性对于实现联合国关于5岁以下儿童死亡率和非传染性疾病死亡率的可持续发展目标具有重要意义。

第二部分 我国北方地区先心病外科现状调查和医疗资源可及性分析

背景

先心病是我国排名首位的先天缺陷，许多先心病患者需要接受先心病手术才能获得最佳的预后。然而，目前关于我国先心病外科医疗资源的分布和人群地理可及性的现状依然未知。

方法

从2018-2019年中国心外科和体外循环白皮书获得我国北方地区开展先心病外科手术的医院相关信息，并采用美国儿科学会制定的先心病诊治能力标准将中心分为两组（诊治能力完备或诊治能力受限）。基于我国人口普查人口数据和道路网络数据均来自官方数据库。运用地理信息系统技术（Geographic Information System, GIS technology）进行先心病外科医疗中心的服务区分析并描绘其地理可及性。

结果

我国北方地区13个省市有93家医院开展先心病手术，其中只有30.1%具有完备的先心病诊治能力；缺乏小儿心内科医师和机械循环支持是其主要限制因素。在2018年，61.3%的中心开展先心病手术少于200例。93家医院共完成30921例先心病外科手术，其中31.6%的手术由先心病诊治能力受限的中心完成，这些中心的年手术量远低于诊治能力完备的中心（中位手术量：85 vs 368）。北京地区的先心病手术密度为每百万人口367例，远远高于北方地区其他省市地区（山东：35例/每百万人口；河南：69例/每百万人口，陕西：124例/每百万人口）。在我国北方地区5岁以下儿童人群中，只有12.9%居住在先心病外科医疗中心服务区30公里的范围之内（通勤时间通常为半天）。与东部地区的儿童（31.4%）相比，71.8%中部地区和70.2%西部地区的儿童需要通勤至少180公里（通常需要在就医城市过夜）才能在诊治能力完备的先心病外科中心获得治疗机会。

结论

我国北方地区大部分先心病手术集中在大中心开展，北京地区的人均手术密度与其人口不成比例。许多开展先心病外科手术的中心不具有完备的先心病诊治能力，相当一部分先心病患者可能接受质量欠佳的手术治疗。公共卫生政策应当从增加医疗资源供给和优化医疗资源配置两方面解决我国偏远地区先心病人群获得高质量先心病医疗服务的地域差异。

第三部分 医疗资源可及性和复杂先心病手术预后的观察性队列研究

背景

基于地理位置的医疗资源可及性与先心病患者的求医行为和医疗服务利用密切相关，可能对先心病的预后有重要影响。然而，目前我国缺乏关于先心病医疗资源可及性和先心病预后的研究。

方法

本研究设计为单中心观察性队列研究，纳入2012年至2015年期间在中国医学科学院阜外医院接受心脏手术的复杂先心病患儿。所有纳入分析的患儿在术后至少完成一次门诊或电话随访。通过电子病历数据库和电话随访收集患儿的社会人口学、社会经济学以及临床变量数据。采用美国胸外科医师协会先心病手术死亡评分（STAT score）进行手术风险分级。根据患儿家庭住址与其最近先心病外科医疗中心的地理位置关系，分为三个医疗资源可及性分区（I区: 0-90 km；II区: 90-180 km；III区: >180 km）。主要研究终点为全因死亡和计划外再入院事件，次要研究终点为门诊随访频率。应用Cox比例风险回归模型分析医疗资源可及性与全因死亡和非计划再入院的关系，应用logistic回归模型分析术后回阜外医院门诊随访的影响因素。

结果

在1994名纳入分析的复杂先心病患儿中，中位手术年龄为10.45月（IQR: 6.64-21.65）。法洛四联症（42.3%）、单心室（14.8%）、肺动脉闭锁（9.7%）以及大动脉转位（9.9%）是最常见的复杂先心病类型。大部分手术的风险等级在STAT-3级以下（71.2%）。在中位随访时间为33.1月（21.3-45.1）、总随访时间为5506.2人年期间，共发生153例死亡事件，全因死亡率为2.78 (95CI: 2.36-3.25)/100人年；共发生971例非计划再入院事件，发生率为1.97 (1.85-2.10)次/10人年。全因死亡率和非计划再入院发生率在I区~ III区中均呈现递增的趋势。在校正性别、手术年龄、父母教育水平、STAT手术风险等级以及其他院内变量后，医疗资源可及性和全因死亡风险有关（HR: II区:1.13 [0.74-1.72], P=0.579; III区:1.47 [1.01-2.17], P=0.048）；医疗资源可及性和非计划再入院的风险呈现增加的趋势，但未达到统计学显著性（HR: II区:1.17 [0.93-1.47], P=0.176; III区:1.18 [0.95-1.47], P=0.140）。随访期间，患儿在阜外医院门诊的平均就医频率为1.60次/人年（1.57-1.63），共有251人出院后从未回院随访。父母受教育水平较低、STAT手术风险等级>3级是复杂先心病患儿术后从未回门诊随访的独立危险因素。

结论

基于地理位置的先心病医疗资源可及性和先心病术后死亡风险独立相关，医疗资源可及性较低的患者术后门诊失访以及非计划再入院的风险呈现增加的趋势。应当从个体层面采取必要的措施减少先心病医疗资源可及性对于弱势人群的不良影响，从而提高先心病患者的手术预后。

Part 1. Global time trends in congenital heart disease mortality: an Age-Period-Cohort analysis

Background

Congenital heart disease (CHD) is the leading cause of morbidity and mortality from birth defects worldwide. We report an overview of trends in CHD mortality in 204 countries and territories over the past 30 years and associations with age, period, and birth cohort.

Methods

Cause-specific CHD mortality estimates were derived from the Global Burden of Disease 2019 study. We utilised an age-period-cohort model to estimate overall annual percentage changes in mortality (net drifts), annual percentage changes from 0–4 to 65–69 years (local drifts), period and cohort relative risks (period/cohort effects) between 1990 and 2019. This approach allows for the examination and differentiation of age, period, and cohort effects in the mortality trends, with the potential to identify disparities and treatment gaps in cardiac care.

Results

CHD is the leading cause of deaths from non-communicable diseases (NCDs) in those under 20 years. Global CHD deaths in 2019 were 217,000 (95% uncertainty interval 177,000–262,000). India, China, Pakistan, and Nigeria had the highest mortality, accounting for 39.7% of deaths globally. Between 1990 and 2019, the net drift of CHD mortality ranged from –2.41% per year (95% confidence interval [CI] –2.55, –2.67) in high Socio-demographic Index (SDI) countries to –0.62% per year (95% CI: –0.82, –0.42) in low-SDI countries. Globally, there was an emerging transition in the age distribution of deaths from paediatric to adult populations, except for an increasing trend of mortality in those aged 10–34 years in Mexico and Pakistan. During the past 30 years, favourable mortality reductions were generally found in most high-SDI countries like South Korea (net drift = –4.0% [95% CI –4.8 to –3.1] per year) and the United States (–2.3% [–2.5 to –2.0]), and also in many middle-SDI countries like Brazil (–2.7% [–3.1 to 2.4]) and South Africa (–2.5% [–3.2 to –1.8]). However, 52 of 129 countries had either increasing trends (net drifts ≥0.0%) or stagnated reductions (≥–0.5%) in mortality. The relative risk of mortality generally showed improving trends over time and in successively younger birth cohorts among high- and high-middle-SDI countries, with the exceptions of Saudi Arabia and Kazakhstan. 14 middle-SDI countries such as Ecuador and Mexico, and 16 low-middle-SDI countries including India and 20 low-SDI countries including Pakistan, had unfavourable or worsening risks for recent periods and birth cohorts.

Conclusions

CHD mortality is a useful and accessible indicator of trends in the provision of congenital cardiac care both in early childhood and across later life. Improvements in the treatment of CHD should reduce the risk for successively younger cohorts and shift the risk for all age groups over time. Although there were gains in CHD mortality globally over the past three decades, unfavourable period and cohort effects were found in many countries, raising questions about adequacy of their health care for CHD patients across all age groups. These failings carry significant implications for the likelihood of achieving the Sustainable Development Goal targets for under-5 and NCD mortality.

Part 2. An investigation of congenital heart surgery and geographic access to congenital cardiac care in north China

Background

Congenital heart disease (CHD) is the leading birth defect in China and many patients require congenital heart surgery (CHS) to achieve optimal outcomes. However, the current landscape and population accessibility to specialist resources for CHS in China are unclear.

Methods

Care facilities performing CHS were identified from the 2018 white book of Chinese cardiovascular surgery and were categorized as full or limited facilities based on specialist capacity according to standards set by American Academy of Pediatrics. Census-based population data and road data were obtained from administrative sources. Service area analysis of all facilities was performed to estimate geographic accessibility.

Results

Of 93 facilities in 14 provinces and municipalities in north China, only 30.1% had full specialist capacity. The shortage of pediatric cardiologists and mechanical circulatory support was the primary limiting factor. In 2018, 61.3% of facilities performed < 200 cases of CHS, and 31.6% of all CHS (N = 30,921) were performed in limited facilities with substantially lower volume than full-capacity facilities (Median: 85.0  versus 368.0). Beijing had a disproportionately higher CHS volume (367 cases per million population) than other provinces (Shandong: 35 cases per million population; Henan: 69 cases per million population, Shaanxi: 124 cases per million population). Of all children under 5 in north China, only 12.9% live within 30 kilometers (typical half‐day visit) of the service areas of all facilities. Compared with children from the eastern region (31.4%), 71.8% and 70.2% of children from central and western regions needed to travel >180 kilometers (typical overnight visit) to receive care in full-capacity facilities.

Conclusions

Most CHS in north China were performed in large centers, and Beijing had concentrated CHS volume disproportionate to its local population. Many facilities for CHS in north China had limited specialist capacity and many CHD patients received suboptimal surgical care. Policy measures should increase resource availability and optimize resource distribution to address the significant geographic disparities to receive high-quality surgical care among disadvantaged patients.

Part 3. The association of geographic access to specialist care with outcomes after complex congenital heart surgery: an observational cohort study

Background

Geographic accessibility to specialist medical resources is closely related to healthcare seeking behavior of patients with CHD and may have an important impact on the prognosis of CHD. However, at present, there is a lack of studies on the geographic accessibility of specialist medical resources and prognosis of CHD in China.

Method

This study was designed as a single center observational cohort study and included children with complex CHD who underwent cardiac surgery at Fuwai Hospital between 2012 and 2015. All included patients completed at least one outpatient visit or telephone follow-up postoperatively. Data on sociodemographic, socioeconomic, and clinical variables were collected from electronic medical record database and follow-up. Surgical risk was categorized using Society of Thoracic Surgeons‑European Association for cardiothoracic surgery congenital heart surgery mortality score (STAT risk score). Based on the geographical relationship between home address and the most proximal congenital heart surgical center, each patient was classified as one of three groups (group I, II & III were within 0-90 km, 90-180 km, and > 180 km of the most proximal center). The primary study endpoint was all-cause death and unplanned readmission events, and the secondary study endpoint was frequency of specialist outpatient visit at Fuwai hospital. Cox proportional hazards regression model was applied to analyze the association between geographic accessibility to CHD resources and study endpoints. Logistic regression model was applied to analyze the predictors for outpatient follow-up at Fuwai Hospital after surgery.

Results

Among 1994 children with complex CHD included in this analysis, the median age at surgery was 10.45 months (IQR: 6.64-21.65). The most common diagnosis was Tetralogy of Fallot (42.3%), single ventricle (14.8%), pulmonary atresia (9.7%), and transposition of the great arteries (9.9%). Most surgeries were below STAT-3 (71.2%). During median 33.1 months (IQR: 21.3-45.1) and 5506.2 person years of follow-up, 153 deaths had occurred, resulting in an all-cause mortality rate of 2.78 (95CI: 2.36-3.25) per 100 person-years, and there were 971 unplanned readmissions with an incidence rate of 1.97 (1.85-2.10) events per 10 person-years. All-cause mortality and unplanned readmission rates showed increasing trends across group I to III. After adjusting for sex, age at operation, parental education level, STAT category and other in-hospital variables, geographic accessibility to CHD resources was associated with all-cause mortality (HR: group II: 1.13 [0.74-1.72], P = 0.579; group III: 1.47 [1.01-2.17], P = 0.048); The association of geographic accessibility to CHD resources and unplanned readmission showed an increasing trend but did not reach statistical significance (HR: group II: 1.17 [0.93-1.47], P = 0.176; group III: 1.18 [0.95-1.47], P = 0.140). During follow-up, the number of specialist clinic visit at Fuwai Hospital was 1.60 (1.57-1.63) per person-year, and a total of 251 patients never returned for follow-up after discharge from the hospital. Lower parental education level and STAT category > 3 were independent risk factors for children with complex CHD who never returned for outpatient follow-up after surgery.

Conclusions

Geographic accessibility to medical resources for CHD were independently associated with the risk of postoperative mortality, with patients of lower accessibility to CHD resources having a higher risk of missed clinic visit and unplanned readmissions. Public health policy should target at the individual level to reduce the adverse impact of geographic inaccessibility to cardiac surgical resources for vulnerable populations, thereby improving surgical outcomes for patients with CHD.