Roberton Timothy, Litvin Kate, Self Andrew, Stegmuller Angela R
Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA.
BMC Public Health. 2017 Nov 7;17(Suppl 4):785. doi: 10.1186/s12889-017-4751-4.
Modeling tools have potential to aid decision making for program planning and evaluation at all levels, but are still largely the domain of technical experts, consultants, and global-level staff. One model that can improve decision making for maternal and child health is the Lives Saved Tool (LiST). We examined respondents' perceptions of LiST's strengths and weaknesses, to identify ways in which LiST - and similar modeling tools - can adapt to be more accessible and helpful to policy makers.
We interviewed 21 purposefully sampled LiST users. First, we identified the characteristics that respondents explicitly stated, or implicitly implied, were important in a modeling tool, and then used these results to create a framework for reviewing a modeling tool. Second, we used this framework to categorize the strengths and weaknesses of LiST that respondents articulated.
Two overarching qualities were important to respondents: usability and accuracy. For some users, LiST already meets these criteria: it allows for customized input parameters to increase specificity; the interface is intuitive; the assumptions and calculations are scientifically sound; and the standard metric of "additional lives saved" is understood and comparable across settings. Other respondents had different views, although their complaints were typically not that the tool is unusable or inaccurate, but that aspects of the tool could be better explained or easier to understand.
Government and agency staff at all levels should be empowered to use the data available to them, including the use of models to make full use of these data. For this, we need tools that meet a threshold of both accuracy, so results clarify rather than mislead, and usability, so tools can be used readily and widely, not just by select experts. With these ideals in mind, there are ways in which LiST might continue to be improved or adapted to further advance its uptake and impact.
建模工具具有辅助各级项目规划和评估决策的潜力，但在很大程度上仍主要由技术专家、顾问和全球层面的工作人员使用。一种可改善孕产妇和儿童健康决策的模型是挽救生命工具（LiST）。我们研究了受访者对LiST优缺点的看法，以确定LiST以及类似建模工具可如何进行调整，从而对政策制定者更具实用性和帮助。
我们对21名经过有目的抽样的LiST用户进行了访谈。首先，我们确定了受访者明确提及或隐含暗示的、在建模工具中很重要的特征，然后利用这些结果创建了一个用于评估建模工具的框架。其次，我们使用该框架对受访者阐述的LiST的优缺点进行分类。
对受访者来说，两个总体特征很重要：可用性和准确性。对一些用户而言，LiST已符合这些标准：它允许定制输入参数以提高特异性；界面直观；假设和计算在科学上合理；“额外挽救的生命”这一标准指标易于理解且在不同环境中具有可比性。其他受访者则有不同看法，尽管他们通常并非抱怨该工具不可用或不准确，而是认为该工具的某些方面可以得到更好的解释或更易于理解。
各级政府和机构工作人员应被赋予权力，以便利用他们可获取的数据，包括使用模型来充分利用这些数据。为此，我们需要既符合准确性门槛（以便结果能阐明而非误导）又具备可用性（以便工具能被广泛且便捷地使用，而不仅仅是特定专家使用）的工具。牢记这些理想目标，LiST仍有一些方法可以继续改进或调整，以进一步提高其应用和影响力。
