AKI with VAD Placement Increased from 2006 to 2015

An estimated 6.5 million adults in
the United States are affected by heart failure and the prevalence is expected
to increase by nearly 50% from 2012 to 2030. The lifetime risk for heart
failure is estimated at 20% to 45%. Increasingly, ventricular assist devices
(VADs) are used for treatment of patients with advanced heart failure that is
not managed with more conservative therapies. Outcomes in patients with VADs
and reduced kidney function are poor both pre- and postoperatively. Estimates
of the incidence and outcomes of acute kidney injury (AKI) in the setting of
VAD placement are hampered by the wide variation in the definition of AKI.

In conjunction with an increase in
the use of VADs, containment of healthcare costs becomes an issue. Among patients
covered by Medicare who received VADs from 2006 to 2011, the mean Medicare
payment for the VAD implantation hospitalization was $210,000; the costs
remained stable during that 5-year period.

Carl P. Walther, MD, MS, and colleagues conducted a cohort
study to compare outcomes by the absence versus presence of diagnosed AKI and
AKI requiring dialysis (AKI-D) during hospitalizations with implantable VAS
placement in the United States. The researchers sought to examine whether
recent trends in relevant outcomes differed in the two AKI categories. Results
of the study were reported in the American Journal of Kidney Diseases
[2019;74(5):650-658].

The study extracted data from the
National Inpatient Sample (NIS) database from January 1, 2006, to December 31,
2015. The NIS, a 20% stratified sample of discharges from US community
hospitals, includes data on all payers of inpatient healthcare in the United
States from more than 7 million hospital stays annually.

Patients ≥18 years of age who
received an implanted VAD and had a diagnosis code indicating heart failure
and/or shock or had a separately coded cardiac surgery during the
hospitalization were included in the study. In 2005, the International
Classification of Diseases, Ninth Revision, Clinical Modification, refined
procedure codes to distinguish implantable VADs from nonimplantable VADs. Exclusion
criteria were codes for end-stage renal disease (ESRD) but none for AKI, and
receiving dialysis but not having codes for either AKI or ESRD. Individuals
with codes for AKI and ESRD were included.

The primary outcome of interest was
in-hospital mortality; length of stay was also evaluated.

Hospital costs were estimated using
total hospital charges, reported for each discharge in the NIS, multiplied by
the applicable cost-to-charge ratios. The Consumer Price Index for Hospital and
Related Services was used to adjust costs to reflect changes in
hospital-related costs, indexed to 2015.

During the study period, there were
an estimated 24,140 hospitalizations with VAD implantation. The number
increased from 853 in 2006 to 3945 in 2015.

Of the 24,140 hospitalizations with
VAD implantation, AKI was among the first 15 discharge codes in 56.1%: 49.6%
had a code for AKI but no dialysis procedure and 6.5% had AKI and dialysis
procedure coded. There was an increase from 2006 to 2015 in the proportion of
hospitalizations with coded AKI, both with and without dialysis, from 44.0% in
2006 to 2007 to 61.7% in 2014 to 2015. During the same period, the proportions
of hospitalizations with AKI-D declined from 9.3% to 5.2%. The proportion of
AKI would increase from 56.1% to 57.8% if AKI diagnosis ascertainment were
extended to all codes in each record.

Overall, in-hospital mortality was
14.5% (95% confidence interval [CI], 13.3%-15.8%). During the study period,
mortality declined from 36.7% (95% CI, 29.7%-43.7%) in 2006 to 2007 to 10.2%
(95% CI, 8.7%-11.7%) in 2014 to 2015. In patients without a diagnosis of AKI
and among those with an AKI diagnosis without dialysis, there was a marked
decline in unadjusted in-hospital mortality during the study period; the
decline was more prominent in the earlier years. Among those with AKI-D, the
decline in mortality was much less pronounced.

The median length of stay was 29
days. Among the patients who survived the index hospitalization, patients with
AKI-D had the longest length of stay (48 days) compared with those without AKI
(25 days) or those with AKI not requiring dialysis (33 days).

In 2006 to 2007, mean hospitalization
cost (in 2015 terms) was $319,096 (95% CI, $288,145-$350,047); costs decreased
to $245,943 (95% CI, $234,697-$257,191) in 2014 to 2015. Among patients without
AKI, mean hospitalization costs were $232,492 (95% CI, $223,905-$241,078),
compared with $288,448 (95% CI, $278,321-$298,575) among patients with AKI not
requiring dialysis and $365,080 (95% CI, $337,662-$392,497) among patients with
AKI-D.

Compared with hospitalizations
without AKI, costs were 19.1% (95% CI, 15.4%-22.9%) higher in patients with AKI
without dialysis and 39.6% (95% CI, 30.6%-49.3%) higher in patients with AKI-D,
following adjustment for demographics, comorbid conditions, and hospitalization
circumstance.

At discharge, 19.6% of patients
without AKI were transferred to another facility (another acute-care hospital,
skilled nursing facility, or rehabilitation facility). Among patients with AKI
not requiring dialysis, the proportion was 31.5%; among those with AKI-D, the
proportion was 47.7%.

The inability to determine the timing
of AKI with respect to VAD implantation was cited by the authors as a
limitation to the study. Also cited were limitations in determining the
prevalence of pre-existing chronic kidney disease and discharge weights could not
be determined for a subpopulation of interest.

“In conclusion, as VADs are
increasingly used in the management of end-stage heart failure refractory to
medical management, understanding and ameliorating pre- and postoperative
decreases in kidney function is necessary. Diagnosis of AKI has increased
during the study period, likely due to more appreciation of the importance of
AKI and increasing sensitivity of AKI definition, but dialysis-requiring AKI
has decreased. Mortality risk among VAD recipients with AKI not requiring
dialysis is improving, but among persons with AKI-D, excess mortality remains
high. This study highlights the need for further investigations into
understanding and reducing the severity of AKI related to end-stage heart
failure and VAD implantation,” the researchers said.

Takeaway Points

1. Patients undergoing implantation of
   ventricular assist devices (VADs) to manage end-stage heart failure are at risk
   for acute kidney injury (AKI) that may or may not require dialysis.
2. Of patients who underwent implantation
   of a VAD from 2006 to 2015, 56.1% developed AKI; of those, 6.5% required
   dialysis (AKI-D). AKI diagnosis increased during the study period; the
   prevalence of AKI-D declined. Mortality declined in all AKI categories; the
   smallest decline was among patients with AKI-D.
3. Compared with patients with no AKI,
   adjusted hospitalization costs were 19.1% higher in patients with AKI and 39.6%
   higher among patients with AKI-D.