Kadir and colleagues have outlined some of the challenges that affect the ability to accurately quantify the effect of conflict on mortality and health, and associated risk factors. However, they inaccurately suggest that mortality estimates “are based primarily on reports from combating parties, news media, and non-governmental organisations.” This might be the case for the Uppsala Conflict Data Program to which they refer, which draws from the media and reports from non-governmental organisations, as well as a smaller sample of other sources. However, in addition to systematic body count analyses, many other approaches exist to estimate mortality, including the commonly used retrospective mortality survey, key informant interviews for acute recall, and, where feasible, prospective surveillance. To counteract limitations intrinsic to each approach, triangulation of sources is key. Capture-recapture analyses have proven helpful in this regard, as shown by a recent analysis of deaths during civil demonstrations in Sudan in 2019.

中文翻译：

---

人道主义危机中的知识生产：提防创新陷阱。

Kadir及其同事概述了一些挑战，这些挑战影响准确量化冲突对死亡率和健康的影响以及相关风险因素的能力。但是，他们不准确地表明死亡率估计“主要基于战斗党，新闻媒体和非政府组织的报道。” 他们所引用的《乌普萨拉冲突数据计划》就是这种情况，该计划取自媒体和非政府组织的报告，以及其他来源的较小样本。但是，除了系统的身体计数分析外，还有许多其他方法可以估算死亡率，包括常用的回顾性死亡率调查，急性召回的关键知情人访谈以及在可行的情况下进行的前瞻性监测。为了抵消每种方法固有的局限性，对源进行三角测量是关键。捕获-捕获分析在这方面被证明是有帮助的，最近对2019年苏丹民事示威期间死亡的分析表明了这一点。