Assessing the impact of acute kidney injury in secondary care and developing strategies to improve outcomes

Access to files

•   FULL-TEXT.PDF (pdf)

Abstract

Abstract: Assessing the impact of acute kidney injury in secondary care and developing strategies to improve outcomes (Dr Lynne Sykes) Introduction: Acute kidney injury (AKI) is associated with up to one in five emergency admissions to hospital and over 300,000 deaths per year in the UK. This thesis, presented in the alternative format, examines work undertaken to better describe the etiology of AKI in secondary care and then strategies to reduce AKI incidence, progression and complications. Methods: Selected anonymised data from the hospital’s ‘data warehouse’ was analysed using SPSS to calculate risk for mortality and critical care admission, analyse background user data, or calculate precision and bias of different point of care tests. The International Health Institute’s Breakthrough Series Model was used for our quality improvement methodology. Results: The literature review suggested education, an e-alert to trigger an AKI bundle and an in-built redundancy in the system were key to reducing mortality and critical care admission. The literature also demonstrates a high event rate of AKI and significant heterogeneity in cause and patient phenotype. The first three results chapters describe the epidemiology of our cohort of secondary care AKI patients in more detail. Chapters 3 and 4 examine the risks of different stages of AKI and the impacts they have on mortality, depending on admission diagnosis. Chapter 3 shows stark differences between patient mortality in those admitted with acute coronary syndrome and AKI 3 compared to those without AKI (OR 12.8 [4.8-33.8] p<0.001) and those admitted with fractured neck of femur and AKI 3 compared to those without (OR 24.6 [8.9-67.9]). In Chapter 4, the percentage of patients admitted with heart failure dying is similar in AKI 2 and AKI 3 (50% versus 47% respectively), demonstrating that escalating AKI stage does not always equate to escalating risk of mortality. Chapter 5 shows a specific ‘at-risk’ AKI population: patients with existing chronic kidney disease. Here, after a first AKI, subsequent episodes of AKI are more likely to be severe. Also, the risk for needing renal replacement therapy increases fourteenfold if a second AKI is stage 2 or 3, or twenty-eightfold if there are three or more episodes of AKI. The second three results chapters, Chapters 6, 7 and 8, describe the quality improvement work undertaken to reduce the incidence and progression of AKI, and also its complications. The large quality improvement programme is described in Chapter 6: it reduced hospital-acquired AKI by 22% and AKI progression by 48% on participating wards. The AKI collaborative group developed an AKI app to support education and signpost to references. Its use is detailed in Chapter 7. We compared results from point of care (POC) analysers with laboratory values in Chapter 8 and found that performance in the normal range showed excellent precision, and that in several scenarios POC tests could be used (with clinical judgment) to alter management. Conclusion: This thesis uses big data to better describe the granularity of cases of AKI. Both the cause and effect of AKI can be heterogeneous and it should be seen as an ‘illness barometer’. With early recognition, education and a set of actions within an AKI bundle, we have shown that AKI incidence and progression can be reduced.

Keyword(s)

AKI; acute kidney injury; acute renal failure; quality improvement

Bibliographic metadata

Institutional metadata

Record metadata