Public and patient involvement in quantitative health research: A statistical perspective
Public and patient involvement in quantitative health research: A statistical perspective
复制标题
DOI：
10.1111/hex.12800
复制
发表时间：
2018-12-01
影响因子：
3.2
通讯作者：
Hannigan, Ailish
中科院分区：
文献类型：
--
作者：
Hannigan, Ailish
背景资料：最近关于公众和患者参与（PPI）对健康研究的影响的综述中纳入的大多数研究都采用了定性设计。PPI在单纯的定量设计中的探索不足，特别是它对统计分析的影响。统计学家在实践中有着悠久的历史，他们在健康研究中的作用（间接）和合作（直接），但他们对PPI在定量健康研究中的观点从未被明确检查过。目的：从统计角度探讨生产者价格指数在定量研究的不同阶段，即抽样、测量和统计分析的潜力和挑战，结论：统计分析的基础是具有代表性的样本，对生产者价格指数采取合作或直接的办法可能有助于实现这一目标，办法是支持接触人口中代表性不足的群体并增加他们的参与。承认和重视非专业人员在统计分析和决定衡量何种变量方面的背景知识的作用，可以支持集体学习和促进科学理解，其他学科使用参与性建模就是证明。定量研究人员经常遇到的问题反映了定量抽样方法，即PPI贡献者的选择和所需数量，这需要进一步的方法发展。在定量卫生研究中直接使用生产者价格指数可能会增加其影响，但所需的促进和伙伴关系技能可能需要对包括统计人员在内的所有利益攸关方进行进一步培训。
Background: The majority of studies included in recent reviews of impact for public and patient involvement (PPI) in health research had a qualitative design. PPI in solely quantitative designs is underexplored, particularly its impact on statistical analysis. Statisticians in practice have a long history of working in both consultative (indirect) and collaborative (direct) roles in health research, yet their perspective on PPI in quantitative health research has never been explicitly examined.Objective: To explore the potential and challenges of PPI from a statistical perspective at distinct stages of quantitative research, that is sampling, measurement and statistical analysis, distinguishing between indirect and direct PPI.Conclusions: Statistical analysis is underpinned by having a representative sample, and a collaborative or direct approach to PPI may help achieve that by supporting access to and increasing participation of under-represented groups in the population. Acknowledging and valuing the role of lay knowledge of the context in statistical analysis and in deciding what variables to measure may support collective learning and advance scientific understanding, as evidenced by the use of participatory modelling in other disciplines. A recurring issue for quantitative researchers, which reflects quantitative sampling methods, is the selection and required number of PPI contributors, and this requires further methodological development. Direct approaches to PPI in quantitative health research may potentially increase its impact, but the facilitation and partnership skills required may require further training for all stakeholders, including statisticians.