BACKGROUND In all health-care systems, decisions need to be made regarding allocation of available resources. Evidence is needed for these decisions, especially in low-income countries. We aimed to estimate how health-care resources provided by the public sector were used in Malawi during 2015-19 and to estimate the effects of strengthening health-care services. METHODS For this modelling study, we used the Thanzi La Onse model, an individual-based simulation model. The scope of the model was health care provided by the public sector in Malawi during 2015-19. Health-care services were delivered during health-care system interaction (HSI) events, which we characterised as occurring at a particular facility level and requiring a particular number of appointments. We developed mechanistic models for the causes of death and disability that were estimated to account for approximately 81% of deaths and approximately 72% of disability-adjusted life-years (DALYs) in Malawi during 2015-19, according to the Global Burden of Disease (GBD) estimates; we computed DALYs incurred in the population as the sum of years of life lost and years lived with disability. The disease models could interact with one another and with the underlying properties of each person. Each person in the Thanzi La Onse model had specific properties (eg, sex, district of residence, wealth percentile, smoking status, and BMI, among others), for which we measured distribution and evolution over time using demographic and health survey data. We also estimated the effect of different types of health-care system improvement. FINDINGS We estimated that the public-sector health-care system in Malawi averted 41·2 million DALYs (95% UI 38·6-43·8) during 2015-19, approximately half of the 84·3 million DALYs (81·5-86·9) that the population would otherwise have incurred. DALYs averted were heavily skewed to children aged 0-4 years due to services averting DALYs that would be caused by acute lower respiratory tract infection, HIV or AIDS, malaria, or neonatal disorders. DALYs averted among adults were mostly attributed to HIV or AIDS and tuberculosis. Under a scenario whereby each appointment took the time expected and health-care workers did not work for longer than contracted, the health-care system in Malawi during 2015-19 would have averted only 19·1 million DALYs (95% UI 17·1-22·4), suggesting that approximately 21·3 million DALYS (20·0-23·6) of total effect were derived through overwork of health-care workers. If people becoming ill immediately accessed care, all referrals were successfully completed, diagnostic accuracy of health-care workers was as good as possible, and consumables (ie, medicines) were always available, 28·2% (95% UI 25·7-30·9) more DALYS (ie, 12·2 million DALYs [95% UI 10·9-13·8]) could be averted. INTERPRETATION The health-care system in Malawi provides substantial health gains with scarce resources. Strengthening interventions could potentially increase these gains, so should be a priority for investigation and investment. An individual-based simulation model of health-care service delivery is valuable for health-care system planning and strengthening. FUNDING The Wellcome Trust, UK Research and Innovation, the UK Medical Research Council, and Community Jameel.

中文翻译：

---

2015-19 年马拉维公共部门医疗保健系统资源使用情况的估计和加强医疗保健服务的效果：一项建模研究 （Thanzi La Onse）。


背景 在所有医疗保健系统中，都需要就可用资源的分配做出决定。这些决定需要证据，尤其是在低收入国家。我们旨在估计 2015-19 年马拉维公共部门提供的医疗保健资源的使用情况，并估计加强医疗保健服务的效果。方法 对于这项建模研究，我们使用了 Thanzi La Onse 模型，这是一种基于个人的模拟模型。该模型的范围是 2015-19 年马拉维公共部门提供的医疗保健。医疗保健服务是在医疗保健系统互动 （HSI） 事件期间提供的，我们将其描述为发生在特定的设施级别并需要特定数量的预约。我们开发了死亡和残疾原因的机制模型，根据全球疾病负担 （GBD） 估计，该模型约占 2015-19 年马拉维死亡人数的 81% 和残疾调整生命年 （DALY） 的约 72%;我们计算了人口中发生的 DALY，即损失的生命年数和残疾寿命年数的总和。疾病模型可以相互交互，并与每个人的基本特性相互作用。Thanzi La Onse 模型中的每个人都具有特定的属性（例如，性别、居住地区、财富百分位数、吸烟状况和 BMI 等），为此，我们使用人口和健康调查数据测量了随时间推移的分布和演变。我们还评估了不同类型的医疗保健系统改善的效果。 调查结果 我们估计，马拉维的公共部门医疗保健系统在 2015-19 年期间避免了 41·200 万残疾调整生命年 （95% UI 38·6-43·8），大约是人口本来会产生的 84·30 万残疾调整年 （81·5-86·9） 的一半。避免的 DALY 严重偏向 0-4 岁的儿童，因为服务避免了由急性下呼吸道感染、HIV 或 AIDS、疟疾或新生儿疾病引起的 DALY。在成年人中避免的 DALY 主要归因于 HIV 或 AIDS 和结核病。在每次预约都花费了预期的时间，并且医护人员的工作时间不超过合同规定的时间的情况下，2015-19 年马拉维的卫生保健系统只能避免 19·100 万残疾调整生命年（95% 的 UI 17·1-22·4），这表明大约 21·30 万残疾调整生命年（20·0-23·6）的总效果是通过医护人员的过度劳累产生的。如果生病的人立即获得护理，所有转诊都成功完成，卫生保健工作者的诊断准确性尽可能高，消耗品（即药物）始终可用，可以避免 28·2%（95% UI 25·7-30·9）的 DALYS（即 12·20 万 DALY [95% UI 10·9-13·8]）。解释 马拉维的医疗保健系统以稀缺的资源提供了可观的健康收益。加强干预措施可能会增加这些收益，因此应成为调查和投资的优先事项。基于个体的医疗保健服务提供模拟模型对于医疗保健系统的规划和加强很有价值。资助 The Wellcome Trust、UK Research and Innovation、UK Medical Research Council 和 Community Jameel。