Broderick Kathryn, Vaidyanathan Arthi, Ponticiello Matthew, Hooda Misha, Kulkarni Vaishali, Chalem Andrea, Chebrolu Puja, Onawale Ashlesha, Baumann Ana, Mathad Jyoti, Sundararajan Radhika
Department of Family Medicine, Sydney Kimmel Medical College, Philadelphia, PA, USA.
School of Medicine, Duke University, Durham, NC, USA.
Implement Sci. 2024 Dec 24;19(1):81. doi: 10.1186/s13012-024-01407-2.
A central goal of implementation science is to generate insights that allow evidence-based practices to be successfully applied across diverse settings. However, challenges often arise in preserving programs' effectiveness outside the context of their intervention development. We propose that qualitative data can inform generalizability via elucidating mechanisms of an intervention. Critical realist thematic analysis provides a framework for applying qualitative data to identify causal relationships. This approach can be used to develop mechanism maps, a tool rooted in policy that has been used in health systems interventions, to explain how and why interventions work. We illustrate use of these approaches through a case example of a community health worker (CHW)-delivered gestational diabetes (GDM) screening intervention in Pune, India. CHWs successfully improved uptake of oral glucose tolerance tests (OGTT) among pregnant women, however clinical management of GDM was suboptimal.
Qualitative interviews were conducted with 53 purposively sampled participants (pregnant women, CHWs, maternal health clinicians). Interview transcripts were reviewed using a critical realist thematic analysis approach to develop a coding scheme pertinent to our research questions: "What caused high uptake of GDM screening?" and "Why did most women with GDM referred to clinics did not receive evidence-based management?". Mechanism maps were retrospectively generated using short- and long-term outcomes as fenceposts to illustrate causal pathways of the CHW-delivered program and subsequent clinical GDM management.
Critical realist thematic analysis generated mechanism maps showed that CHWs facilitated GDM screening uptake through affective, cognitive and logistic pathways of influence. Lack of evidence-based treatment of GDM at clinics was caused by 1) clinicians lacking time or initiative to provide GDM counseling and 2) low perceived pre-test probability of GDM in this population of women without traditional risk factors. Mechanism mapping identified areas for adaptation to improve the intervention for future iterations.
Mechanism maps created by repeated engagement following the critical realist thematic analysis method can provide a retrospective framework to understand causal relationships between factors driving intervention successes or failures. This process, in turn, can inform the generalizability of health programs by identifying constituent factors and their interrelationships that are central to implementation.
实施科学的一个核心目标是产生见解，使基于证据的实践能够在不同环境中成功应用。然而，在干预措施开发背景之外保持项目有效性时，往往会出现挑战。我们认为定性数据可以通过阐明干预机制来为普遍性提供信息。批判实在主义主题分析提供了一个应用定性数据来识别因果关系的框架。这种方法可用于开发机制图，这是一种源于政策的工具，已用于卫生系统干预措施，以解释干预措施如何以及为何起作用。我们通过印度浦那社区卫生工作者（CHW）提供的妊娠期糖尿病（GDM）筛查干预措施的案例来说明这些方法的使用。社区卫生工作者成功提高了孕妇口服葡萄糖耐量试验（OGTT）的接受率，然而GDM的临床管理并不理想。
对53名有目的抽样的参与者（孕妇、社区卫生工作者、孕产妇健康临床医生）进行了定性访谈。使用批判实在主义主题分析方法对访谈记录进行审查，以制定与我们的研究问题相关的编码方案：“是什么导致了GDM筛查的高接受率？”以及“为什么转诊到诊所的大多数GDM女性没有接受循证管理？”。使用短期和长期结果作为标杆，回顾性地生成机制图，以说明社区卫生工作者提供的项目以及随后的GDM临床管理的因果途径。
批判实在主义主题分析生成的机制图显示，社区卫生工作者通过情感、认知和后勤影响途径促进了GDM筛查的接受。诊所缺乏GDM的循证治疗是由以下原因导致的：1）临床医生缺乏提供GDM咨询的时间或主动性；2）在没有传统风险因素的这群女性中，GDM的预测试概率较低。机制图确定了需要调整的领域，以改进未来迭代的干预措施。
通过遵循批判实在主义主题分析方法反复参与创建的机制图可以提供一个回顾性框架，以了解驱动干预成功或失败的因素之间的因果关系。反过来，这个过程可以通过识别对实施至关重要的构成因素及其相互关系，为卫生项目的普遍性提供信息。
