School of Public Health, MRC Centre for Global Infectious Disease Analysis, Imperial College London, London, England.
Division of Population Health, Health Services Research & Primary Care, University of Manchester, Manchester, England.
Health Qual Life Outcomes. 2020 Mar 18;18(1):78. doi: 10.1186/s12955-020-01323-1.
The Short Form Survey 12-item (SF12) mental and physical health version has been applied in several studies on populations from Sub-Saharan Africa. However, the SF12 has not been computed and validated for these populations. We address in this paper these gaps in the literature and use a health intervention example in Malawi to show the importance of our analysis for health policy.
We firstly compute the weights of the SF12 physical and mental health measure for the Malawian population using principal component analysis on a sample of 2838 adults from wave four (2006) of Malawian Longitudinal Study of Aging (MLSFH). We secondly test the construct validity of our computed and the US-population weighted SF12 measures using regression analysis and Fixed Effect estimation on waves four, seven (2012) and eight (2013) of the MLSFH. Finally, we use a Malawian cash transfer programme to exemplify the implications of using US- and Malawi-weighted SF12 mental health measures in policy evaluation.
We find that the Malawian SF12 health measure weighted by our computed Malawian population weights is strongly associated with other mental health measures (Depression:-0.501, p = < 0.001; Anxiety:-1.755; p = < 0.001) and shows better construct validity in comparison to the US-weighted SF12 mental health component (rs = 0.675 versus rs = 0.495). None of the SF12 measures shows strong associations with other measures of physical health. The estimated average effect of the cash transfer is significant when using the Malawi-weighted SF12 mental health measure (treatment effect: 1.124; p = < 0.1), but not when using the US-weighted counterpart (treatment effect: 1.129; p > 0.1). The weightings affect the size of the impacts across mental health quantiles suggesting that the weighting scheme matters for empirical health policy analysis.
Mental health shows more pronounced associations with the physical health dimension in a Low-Income Country like Malawi compared to the US. This is important for the construct validity of the SF12 health measures and has strong implications in health policy analysis. Further analysis is required for the physical health dimension of the SF12.
短表单 12 项（SF12）心理健康和身体健康版本已应用于撒哈拉以南非洲人群的多项研究中。然而，SF12 尚未针对这些人群进行计算和验证。我们在本文中弥补了这一文献空白，并以马拉维的一项健康干预为例，说明了我们的分析对健康政策的重要性。
我们首先使用主成分分析（principal component analysis）在马拉维老龄化纵向研究（MLSFH）第四波（2006 年）的 2838 名成年人样本中计算 SF12 身体健康和心理健康测量的权重。其次，我们使用回归分析和 MLSFH 的第四波、第七波（2012 年）和第八波（2013 年）的固定效应估计来测试我们计算的和美国人群加权的 SF12 测量的结构有效性。最后，我们使用马拉维的现金转移计划来说明在政策评估中使用美国和马拉维加权的 SF12 心理健康测量的影响。
我们发现，由我们计算的马拉维人群权重加权的马拉维 SF12 健康测量与其他心理健康测量（抑郁：-0.501，p<0.001；焦虑：-1.755；p<0.001）有很强的关联，并与美国加权的 SF12 心理健康成分相比具有更好的结构有效性（rs=0.675 与 rs=0.495）。SF12 测量均与其他身体健康测量无明显关联。当使用马拉维加权的 SF12 心理健康测量时，现金转移的估计平均效应显著（治疗效果：1.124；p<0.1），但当使用美国加权的对应物时则不显著（治疗效果：1.129；p>0.1）。权重会影响心理健康分位数的影响大小，这表明权重方案对实证健康政策分析很重要。
与美国相比，心理健康在马拉维这样的低收入国家与身体健康维度的关联更为显著。这对 SF12 健康测量的结构有效性很重要，并对健康政策分析有重要影响。SF12 的身体健康维度还需要进一步分析。
