Identification of valid and multidimensional quality indicators in diabetic foot care, useful to study quality of care in diabetic foot clinics

Diabetic foot ulceration (DFU) is a common late-stage complication, experienced by up to 25% of diabetics in their lifetime.1,2 Its impact, both on quality of life3 and on resource utilization4, drives diabetic foot clinics (DFCs) and national healthcare systems to improve foot care organization. Monitoring of performance and providing feedback to care providers (audit-feedback), which includes benchmarking, is one strategy to improve quality of care.5 The Initiative for Quality Improvement and Epidemiology in Multidisciplinary Diabetic Foot Clinics (IQED-Foot), implemented in Belgium since 2005, is one application of this principle.6 The specific aim of this project is to identify valid and multidimensional quality indicators useful for quality monitoring in specialized diabetic foot care. The existing IQED-Foot project has a number of shortcomings that need to be addressed to yield an improved audit methodology with regard to validity, acceptability and usability. In multiple European countries efforts are made to optimize diabetic foot care, notably by improving referral of patients to multidisciplinary DFCs and by creating accreditation programs; however valid and practical quality monitoring tool are still sorely needed. Analyses of IQED-Foot data have shown that DFUs are about 2.5 times more likely to heal in the “best” DFC compared to the “worst” DFC. Large gains in resources and quality-of-life can be expected if quality differences are identified and addressed. The first research question focuses on the three dimensions of care usually used to assess quality of care; two phases are planned. Based on the current definitions the indicators will be refined and their validity checked in phase 1. Structure indicators will be investigated, e.g. pattern and delay of referral, efforts to educate first-line health care professionals and patients, organization of in-hospital care (delay until diagnostic procedure or treatments). Process indicators include delivery of podiatric care, DFU off-loading, revascularization, DFU debridement, orthopedic surgery or composites thereof. DFU healing, major amputation, death and their risk-adjusted counterparts will be the considered outcome indicators. Identification of valid risk-adjustment models for the above outcomes will be the first step in this phase, because the availability of risk-adjusted outcomes is essential to study the correlation with structure and process indicators in phase 2. These analyses will be published: a first paper will describe the determinants of these outcomes, i.e. the patient and ulcer characteristics identified in the risk-adjustment models; in a potential second paper we will derive a prognostic model from the risk-adjustment model. Face validity and acceptability of the resulting indicators will be checked among experts. Based on the Delphi process, the indicator definitions and risk adjustment models will be revised and a final set of indicators will be used for Phase 2. During phase 2, based on the set of validated indicators, we will correlate structure and process indicators to risk-adjusted outcome indicators, at the patient-level, the level of individual care teams and the national level, both cross-sectionally and longitudinally. The second research question provides a critical review of the use of patients-reported measures (PROMs/PREMs) in the area of diabetic foot care. We will identify the specific instruments that exist to measure limitations in social functioning (including work) and activities of daily living (ADL). We intend then to score instruments on their utility and feasibility for inclusion in the existing audit system. These findings will be published in a systematic review (paper 3). The third research question consists of determining the final scope of quality monitoring. Based on the findings from research questions 1 and 2, we will put forward recommendations about the future scope and methodology of quality monitoring of DFU care and their anticipated impact on data collection burden. In case valid outcome indicators are not associated with structure or process indicators, we will focus on outcomes to evaluate care quality. In this scenario, another QI approach will be used to identify and address differences between DFCs in terms of structure and process of care. If indicators are associated, the focus should shift to structure or process indicators, because they are more actionable. We are going to submit recommendations (using the Delphi method) to stakeholders (including experts, the DFCs and representatives of relevant patient organizations) to assess acceptability of the chosen set of indicators. In order to get to a leaner, more valid QI initiative, allowing better decision-making, their views will be captured and incorporated. Applicability in an international context will be also considered. The final 6 months of this phase will be spent on implementing the findings of this project, which should be complete in time for the 2020 cycle of audit-feedback.The final months will be spent on the redaction of the text of the thesis and its defense.