The study was designed to develop and validate risk models to define risk of AKI or of worsening AKI during hospital admission. We determined three time points during the period of admission when significant clinical decision-making takes place at which the use of risk models would have the greatest impact on clinical care and patient management. These time points are described below.
The point of admission to hospital
The model applied in this case (referred to as risk model 1) uses all electronic data up until the point of admission (and the reason for admission to hospital) to determine risk of patients already having AKI on admission. In this way, the model is guiding the testing of kidney function to ensure that patients who are likely to have AKI have their kidney function tested to unmask the condition and allow efficient appropriate clinical intervention to treat the AKI.
After 24 hours of admission
These models use all electronic data both prior to admission and up to 24 hours after admission to determine the risk of developing AKI (model 2) or of worsening AKI (model 3) in the first 72 hours after admission. At the 24-hour point, patients will be admitted to the ward and are likely to then be reviewed by the admitting clinical team on the post-take ward round, or by the ward team on their ward round of new patients. The models guides clinical management at this point to define patients at risk of AKI who require:
l Management changes to include the stopping of nephrotoxic medication, fluid assessment and ensuring fluid replete, monitoring of blood pressure and ensuring adequate blood pressure and appropriate use of antihypertensive drugs.
l Daily renal function testing to observe for development of AKI. If there is consideration at this point of discharge from hospital, the models at this point may inform the discharge decision or, if discharge is still intended, guide follow-up in primary care to observe for AKI (and also management of medications that may have been stopped temporarily while the risk of AKI exists).
After 72 hours of admission
The purpose of this model was to predict patients who will develop AKI, or worsening AKI if already present, during the rest of the hospital admission. During the progression of this study, the clinical experts on the board (informed by the results of the risk modelling) determined that this point of risk assessment would not add clinical benefit for a number of reasons:
l The risk models at this point were not sufficiently accurate to determine risk and guide clinical management, which are also related to and a consequence of the other reasons.
l Most patients remain in hospital for< 3 days (72 hours). For those who do stay in hospital over 72 hours, the period of admission may range from 3 days up to as high as 90–365 days. It is very difficult to determine risk of developing AKI in a widely varying time period, most importantly because these patients will develop new conditions and changes in blood results (variables in the risk models) after 72 hours, which change their risk and cannot be accounted for.
Clinical alerting
The purpose of the risk models is to guide clinical management. This will be achieved through alerting clinicians at the point of care. These systems have been developed as part of this project, but are awaiting formal assessment in a trial setting. The clinical alerting system will not solely alert to risk of AKI; it will also alert to patients with established AKI and provide clinical guidance for improving the management of these patients and reducing both the progression of AKI and the development of resultant sequelae. It is therefore important that both alerting to AKI and alerting to AKI risk are brought together in a clinical practice algorithm to guide both the alerting and the clinical management pathway following the alert.
Alerting at the point of admission
Initial risk model development for the point of admission included pre-admission AKI as a variable in the models. As would be expected, if a patient has AKI just prior to admission then there is a high risk that they will still have AKI at the point of admission. The initial models developed confirmed that the variable ‘pre-admission AKI’ was by far the strongest predictor of AKI on admission and diluted other variables in the model. As part of the clinical practice algorithm, the decision was made by the clinical experts on the project team that all patients with pre-admission AKI should be alerted by the alerting system as ‘pre-admission AKI’, irrespective of risk assessment. In the cases of these patients, renal function testing should always occur on admission and the management changes above should be implemented. Patients with pre-admission AKI were, therefore, removed from the admission models for predicting AKI, and hence the population of assessment for the point of admission models included only patients without pre-admission AKI or patients who did not have pre-admission AKI status determined because they did not undergo renal function testing pre admission.
Alerting after 24 hours of admission
For patients who have AKI on admission, the clinical alerting system will suggest appropriate management interventions as above listed (see Points of decision-making), including daily renal function testing. Again, as part of the clinical practice algorithm, patients with AKI on admission were removed from the risk models at 24 hours to predict AKI at 72 hours, as these patients have already been highlighted by the system. Moreover, including AKI on admission as a variable in a model to predict AKI at 72 hours would be highly predictive, dilute the other variables and reduce the clinical utility of the system, especially as these patients will already be flagged up by the system anyway.
This led to the development of the clinical practice algorithm (Figure 5) to guide:
l AKI alerting
l patients to be assessed by the risk models at the point of admission and at 24 hours after admission
l appropriate renal function testing
Risk of having AKI on admission Patient admission/A&E attendance
Pre-admission AKI?
High risk of AKI?
AKI on admission?
Usual management
At 24 hours
At 72 hours
Alert Admission risk model (1) Y N Y N Alert • Check renal function daily • Medication and fluid review
• Check renal function • Medication and fluid review
Usual management
Y N
Alert
• Check renal function daily • Medication and fluid review
High risk of AKI?
AKI at 72 hours?
Usual management Alert 24-hour risk
model (2)
Y N
Y N
Risk of having AKI in next 48 hours
(i.e. up to 72 hours into admission)? Alert • Check renal function daily • Medication and fluid review
• Check renal function daily • Medication and fluid review Point of admission