Identifying the patient at risk of acute kidney injury: a predictive scoring system for the development of acute kidney injury in acute medical patients

Background: Acute kidney injury (AKI) in hospitalized patients has significant implications in terms of morbidity and mortality, length of hospital stay and associated costs. To date, no interventions are proven to prevent the development of AKI but this is hampered in part by the lack of early recognition of patients at risk. We aimed to determine whether a simple system could be devised from both physiological and demographic data in order to identify individuals at increased risk from the development of inpatient AKI.

Method: Our observational, population-based single-centred study took place in an 870-bed associated university hospital. All patients admitted to the acute medical admissions unit on the Worthing site of the Western Sussex Hospitals Trust during the study period were included.

Results: Multivariate logistic regression analysis demonstrated that age, respiratory rate and disturbed consciousness together with a history of chronic kidney disease, diabetes mellitus, congestive cardiac failure and liver disease were associated with an increased risk of developing AKI within 7 days of admission. We derived a simple scoring system to identify acute medical patients at greater risk of developing AKI.

Conclusions: The incidence of AKI complicating inpatient admissions remains high, however with the application of the derived AKI prediction score it is hoped that early recognition will translate to improved outcomes.