Automated, electronic alerts for acute kidney injury: a single-blind, parallel-group, randomised controlled trial.
Wilson FP, Shashaty M, Testani J, Aqeel I, Borovskiy Y, Ellenberg SS, Feldman HI, Fernandez H, Gitelman Y, Lin J, Negoianu D, Parikh CR,Reese PP, Urbani R, Fuchs B.
Lancet. 2015 Feb 25. pii: S0140-6736(15)60266-5. doi: 10.1016/S0140-6736(15)60266-5. [Epub ahead of print]
Join us for the #NephJC chat on
Tuesday 9 PM EST on April 28 (Americas)
Wednesday 8 PM GMT April 29 (EU/Africa)
For the next NephJC (#25!) we return to a study presented first at ASN Kidney Week and at #NephJCLive. This was F. Perry Wilson (aka @methodsmanmd) and the AKI alert study, which has been subsequently published in the Lancet.
Summary
Background: Acute kidney injury (AKI) is common in hospitalized patients and carries a higher risk of mortality. Even small increases in serum creatinine translate into higher mortality. However, the trials examining various interventions to prevent or treat AKI (e.g. prophylactic nesiritide, fenoldopam, antioxidants) have been unfortunately negative. One of the arguments explaining the negative results has been the delay in the diagnosis of AKI and the loss of window of opportunity. The KDIGO guidelines suggest early treatment including drug dose adjustment, nephrotoxin avoidance, and attention to volume status and perfusion pressure. It is indeed possible that AKI is missed in early stages. However, it is not known if diagnosing it early actually changes outcomes. In a prospective, but uncontrolled study, Colpaert et al showed that a real-time electronic alert system or "AKI sniffer," alerting of every worsening RIFLE class, increased the number and timeliness of early therapeutic interventions with improvement in the short term renal outcomes. The study we will discuss at the next NephJC is a randomized controlled trial, in which the authors studied if an automated electronic alert for acute kidney injury would reduce the severity of acute kidney injury and improve clinical outcomes in patients in hospital.
Methods:
Design: Single-blind, parallel-group, randomized controlled trial.
Setting: The hospital of the University of Pennsylvania in Philadelphia, PA, USA.
Participants: Adults with acute kidney injury as defined by the KDIGO creatinine based criteria. (the current serum creatinine value at least 0·3 mg/dL greater than the lowest value that occurred in the previous 48 h or 50% greater than the lowest value that occurred in the previous 7 days.)
Exclusion criteria:
- Initial hospital creatinine 4·0 mg/dL or greater, f
- ewer than two creatinine values measured,
- inability to determine the covering provider, a
- dmission to hospice or the observation unit,
- previous randomization,
- end-stage renal disease.
Intervention:
- In the alert group, the covering provider and unit pharmacist received a text page within 1 hour of the patient meeting AKI criteria, informing them of the presence of AKI (alert). The text message provided a link to the study website, containing study information and a link to the KDIGO acute kidney injury practice guidelines
- The usual care group did not receive any alerts.
Randomization: 1:1 allocation to AKI alert group or usual care group. Block randomization for four mutually exclusive strata, medical vs surgical service and ICU vs non-ICU.
Blinding: Clinicians were aware of the allocation but the study personnel and outcome assessors were kept masked to the allocation.
Primary outcome: composite of relative maximum change in creatinine, dialysis, and death at 7 days.
Results:
In about seven months, 1201 were assigned to the acute kidney injury alert group and 1192 were assigned to the usual care group. 30% were in ICUs and 42% had been admitted to surgical services.
The population had a mean age 60 years. Baseline characteristics of both the groups were strikingly similar.
Many patients in both groups never achieved a creatinine higher than the one initially meeting the criteria for AKI.
Overall, the primary outcome of the composite relative maximum change in creatinine, dialysis, and death at 7 days was no different between the groups.
- At 7 days after randomization, median maximum relative change in creatinine concentrations was 0·0% in the alert group and 0·6% in the usual care group (p=0·81).
- 87 (7·2%) patients in the alert group and 70 (5·9%) patients in usual care group had received dialysis (OR 1·25 [95% CI 0·90–1·74] p=0·18)
- 71 (5·9%) patients in the alert group and 61 (5·1%) patients in the usual care group had died (1·16 [0·81–1·68]; p=0·40).
There was no difference between the number of patients getting a Nephrology consultation, or a nephrotoxic agent like IV contrast, aminoglycosides or NSAIDs between the two groups. More frequent creatinine testing occurred in the usual care group compared with the alert group within 48 h of randomization (p=0·05), but not within 7 days.
There was no difference in pre-dialysis serum concentrations of creatinine, potassium, bicarbonate, or blood urea nitrogen between the groups in any strata.
Less than 10% of the study population received a nephrology consult. Less than half of the population had this AKI documented in the charts. About one fourth of the patients had a urinalysis done in 48 hours.
Although the rate of death did not differ between any subgroups, the alert was significantly associated with an increased rate of renal consult (p=0·001) and dialysis (p=0·07) among surgical ward patients.
To examine the effect of study itself on the heightened awareness about the AKI among the caregivers, changes in the effect of the alert on the primary outcome were assessed as the study progressed. Primary endpoint did not change over the course of the study but as the study progressed, authors noted a diminution of the effect of the alert on maximum relative change in creatinine (p=0·03) and a decrease in the effect of the alert increasing dialysis rates (p=0·02).
Discussion:
Alerts made no differences in two groups in the primary composite endpoint at 7 days. They did not make any difference in the interventions as well. Several explanations are put forth by the authors:
- The alert might not be a useful clinical intervention and the providers were already aware of AKI.
- Early identification of the disease does not lead to meaningful changes in management.
- Changes in management strategy are not enough to improve clinical outcomes across the spectrum of causes of AKI. Because of heterogeneous study population, these changes which may be useful in one group, may not affect outcome in other groups.
- Alert effectiveness is likely to be lower where the quality of usual care is better.
Surgical ward patients in alert group were more likely to get nephrology consultation, to receive dialysis, thus utilizing more resources without changing overall mortality. (In fact, in-hospital mortality was higher!) This needs to be studied systematically.
Strengths:
· Relevant research question for the electronic alerting systems in future.
· Well designed, large, adequately powered study.
· Efficient delivery of alerts
· Explicit description of population, setting, interventions, and outcome measures.
Limitations:
· Single center
· Outcome measures studied at 7 days.
· Controls are not actual controls as the caregivers overlap, and are ‘educated’ by the alerts in intervention group.
· Not all the causes of AKI have meaningful preventive interventions; clubbing all of them together dilutes the effects of truly useful timely intervention. Possible effects of alerts ought to be through the interventions done thereafter; however the interventions (e.g. fluid management, medications changes) were not different in two groups.
· Possibility of alert fatigue,
· Urine output criteria were not used
Authors conclusions:
There was no short term meaningful benefit of an electronic alert system for acute kidney injury in patients in hospital. They do not change the pattern of nephrology consultations, interventions or outcomes.
Increased use of dialysis and higher in-hospital mortality in surgical ward subgroup in intervention group caution against routine use of electronic alerts for AKI.