Mahdavi Mahdi, Vissers Jan, Elkhuizen Sylvia, van Dijk Mattees, Vanhala Antero, Karampli Eleftheria, Faubel Raquel, Forte Paul, Coroian Elena, van de Klundert Joris
National Institute of Health Research, Tehran University of Medical Sciences, Tehran, Iran.
Institute of Health Policy and Management, Erasmus University Rotterdam, Rotterdam, The Netherlands.
PLoS One. 2018 Feb 15;13(2):e0192599. doi: 10.1371/journal.pone.0192599. eCollection 2018.
While health service provisioning for the chronic condition Type 2 Diabetes (T2D) often involves a network of organisations and professionals, most evidence on the relationships between the structures and processes of service provisioning and the outcomes considers single organisations or solo practitioners. Extending Donabedian's Structure-Process-Outcome (SPO) model, we investigate how differences in quality of life, effective coverage of diabetes, and service satisfaction are associated with differences in the structures, processes, and context of T2D services in six regions in Finland, Germany, Greece, Netherlands, Spain, and UK.
Data collection consisted of: a) systematic modelling of provider network's structures and processes, and b) a cross-sectional survey of patient reported outcomes and other information. The survey resulted in data from 1459 T2D patients, during 2011-2012. Stepwise linear regression models were used to identify how independent cumulative proportion of variance in quality of life and service satisfaction are related to differences in context, structure and process. The selected context, structure and process variables are based on Donabedian's SPO model, a service quality research instrument (SERVQUAL), and previous organization and professional level evidence. Additional analysis deepens the possible bidirectional relation between outcomes and processes.
The regression models explain 44% of variance in service satisfaction, mostly by structure and process variables (such as human resource use and the SERVQUAL dimensions). The models explained 23% of variance in quality of life between the networks, much of which is related to contextual variables. Our results suggest that effectiveness of A1c control is negatively correlated with process variables such as total hours of care provided per year and cost of services per year.
While the selected structure and process variables explain much of the variance in service satisfaction, this is less the case for quality of life. Moreover, it appears that the effect of the clinical outcome A1c control on processes is stronger than the other way around, as poorer control seems to relate to more service use, and higher cost. The standardized operational models used in this research prove to form a basis for expanding the network level evidence base for effective T2D service provisioning.
虽然为2型糖尿病（T2D）这种慢性病提供医疗服务通常涉及一系列组织和专业人员，但关于服务提供的结构与流程和结果之间关系的大多数证据都只考虑单个组织或个体从业者。我们扩展了唐纳贝迪安的结构-过程-结果（SPO）模型，研究芬兰、德国、希腊、荷兰、西班牙和英国六个地区T2D服务在生活质量、糖尿病有效覆盖率和服务满意度方面的差异如何与T2D服务的结构、流程和背景差异相关联。
数据收集包括：a）对医疗服务提供网络的结构和流程进行系统建模，以及b）对患者报告的结果和其他信息进行横断面调查。该调查产生了2011 - 2012年期间1459名T2D患者的数据。使用逐步线性回归模型来确定生活质量和服务满意度中独立的累积方差比例如何与背景、结构和流程的差异相关。所选的背景、结构和流程变量基于唐纳贝迪安的SPO模型、一种服务质量研究工具（SERVQUAL）以及先前的组织和专业层面的证据。进一步的分析深化了结果与流程之间可能的双向关系。
回归模型解释了服务满意度中方差的44%，主要是由结构和流程变量（如人力资源使用和SERVQUAL维度）解释的。这些模型解释了各网络之间生活质量方差的23%，其中大部分与背景变量有关。我们的结果表明，糖化血红蛋白（A1c）控制的有效性与每年提供的护理总时长和每年的服务成本等流程变量呈负相关。
虽然所选的结构和流程变量解释了服务满意度中的大部分方差，但对生活质量的解释程度较低。此外，临床结果A1c控制对流程的影响似乎比反过来的影响更强，因为控制较差似乎与更多的服务使用和更高的成本相关。本研究中使用的标准化运营模型被证明为扩展有效T2D服务提供的网络层面证据基础奠定了基础。
