Two decades of primary health care expansion in Latin America: evaluation and forecasting study for health-related SDGs

最近的《阿斯塔纳宣言》强调了建立在强有力的初级卫生保健基础上的卫生系统的重要性，初级卫生保健应成为解决健康的经济和社会决定因素的多部门行动的一部分并加以协调。初级卫生保健应成为构成福利国家的干预网络的主要枢纽之一，将医疗保健系统与社会和扶贫干预措施协同联系起来。几个拉丁美洲国家在过去二十年中实施并扩大了不同程度强度和联系的初级保健、卫生保健系统组成部分和社会援助干预措施。然而，最近的经济危机和紧缩措施正在威胁到大多数拉丁美洲和加勒比地区公共医疗保健系统和福利国家的巩固。在少数国家对初级保健的一些评估发现，初级保健对特定的健康结果产生了积极影响，但没有研究进行了多国比较，也没有研究采用全面的方法来估计初级保健项目对广泛的健康结果（应通报的疾病、住院和死亡率——总体上以及针对特定原因和年龄组）的影响——包括长期影响。衡量其与其他医疗系统组成部分和社会援助，特别是有条件（CCT）和社会养老金（SP）的协同影响。同样，也没有研究衡量初级保健覆盖时间和环境因素（如地方治理指数或人类发展指数）的影响。初级保健与有条件现金援助的协同效应是可以预期的，因为初级保健单位应该关注健康条件，而初级保健与初级保健的相互作用是可信的，因为减贫的生理效应，特别是在预防和治疗途径上。该项目的目的是系统地处理所有这些问题，以便全面评价初级保健在巴西、哥伦比亚、厄瓜多尔和墨西哥（巴西、哥伦比亚、厄瓜多尔和墨西哥）尽可能广泛的发病率和死亡率结果方面的有效性。之所以能够进行比较评价，是因为BCEM在过去二十年中实施了不同的初级保健、医疗保健和社会援助。每个BCEM将合并来自广泛来源的总体和个人层面、纵向和横断面、社会经济和健康数据，并将使用最可靠的准实验设计来评估这些全面的初级保健对如此广泛的结果的影响。利用来自所有回顾性影响评估的大量数据和参数，并使用微观模拟建模方法（可以说是目前最准确的预测技术之一），我们将预测由于预算限制或紧缩措施而导致的不同初级保健覆盖减少的影响，并结合上述福利国家组成部分，与覆盖稳定或覆盖扩大的情况相比较。评估到2030年BCEM在广泛健康结果方面的差异。使用所有粒度的数据和效果估计，我们将评估未来几年在特定亚群上最有效的PHC， CCT和SP扩展覆盖动态。我们还将校准和选择最佳政策方案，以减少卫生不平等现象，并在每个BCME中实现每个与卫生有关的可持续发展目标。该项目还将巩固拉加经委会的研究人员网络，并将开发算法和方法，利用生态和个人一级的数据综合公共政策的事后和事前影响评价。研究中创建的所有数据集将在一个开放的动态平台上提供，该平台包含来自医疗保健系统、社会干预和健康结果的综合LAC数据，研究网络将使用该平台和开发的算法来刺激监测、评估和预测的连续循环。

The recent Astana Declaration emphasizes the importance of health systems based on a strong Primary Health Care (PHC), which should be part and coordinate multisectoral actions addressing economic and social determinants of health. PHC should be one of the main hubs of the network of interventions composing the welfare state, synergistically connecting the healthcare system with social and poverty-relief interventions. Several Latin American countries (LAC) have implemented and expanded over the last two decades PHC, healthcare system components, and social assistance interventions with different degrees of intensity and connection. However, recent economic crisis and austerity measures are threatening the consolidation of public healthcare systems and of the welfare state in the majority of LAC. Some evaluations of PHC in a limited number of countries have found positive effects on specific health outcomes, but no study has performed multi-country comparisons or adopted a comprehensive approach for the estimation of the effects -including long-term effects- of PHC programs on a broad range of health outcomes (notifiable diseases, hospitalizations and mortality - overall and for specific causes and age-groups), measuring its synergistic impact with the other healthcare system components and with social assistance, in particular conditional (CCT) and social pensions (SP). Similarly, no research has measured the influence of PHC coverage duration and of contextual factors such as local governance indexes or human development index. While the synergistic impact of PHC with CCT is expected because the health conditionalities should be attended in the PHC units, the interaction effects of SP with PHC are plausible because of the physiological effects of the relief from poverty, in particular in the pathways of prevention and cure. The aim of this project is to systematically address all these issues for a comprehensive evaluation of the effectiveness of PHC on the broadest possible range of morbidity and mortality outcomes in Brazil, Colombia, Ecuador and Mexico (BCEM). The comparative evaluation will be possible because BCEM have undertaken different PHC, healthcare and social assistance implementations during the last two decades. Aggregate and individual-level, longitudinal and cross-sectional, socioeconomic and health data from a wide range of sources in each BCEM will be merged and the most robust quasi-experimental designs will be used to evaluate these comprehensive PHC impacts on such a wide number of outcomes. With the large amount of data and parameters from all the retrospective impact evaluations, and using a microsimulation modelling approach -arguably one of the most accurate forecasting techniques up-to-date - we will forecast the effects of different PHC coverage reductions - due to budget constraints or austerity measures - in conjunction with the above mentioned welfare state components, versus scenarios of coverage stabilization or coverage expansion, evaluating differences in a broad range of health outcomes in BCEM up to 2030. Using all the granularity of data and effect estimates, we will evaluate the most effective PHC, CCT and SP expansion coverage dynamics -on specific subpopulations- in the next years. We will also calibrate and select the best policy scenarios for the reduction of health inequalities and the achievement of each health-related SDG in each BCME. This project will also consolidate a network of researchers in LAC and will develop algorithms and methods to integrate ex-post and ex-ante impact evaluations of public policies using ecologic and individual-level data. All the datasets created in the study will be made available in an open dynamic platform of comprehensive LAC data from healthcare systems, social interventions and health outcome, and the research network will use the platform and the developed algorithms to stimulate continuous cycles of monitoring, evaluation and forecasting.