New Insights Into the Alleged Kidney Donor Profile Index Labeling Effect on Kidney Utilization

Abstract

The Kidney Donor Profile Index (KDPI) became a driving factor in deceased donor kidney allocation on December 4, 2014, with the implementation of the kidney allocation system (KAS). On April 20, 2016, the annual recalibration of the Kidney Donor Risk Index into KDPI was incorrectly programmed in DonorNet, resulting in erroneously high KDPI values, by between 1 and 21 percentage points (e.g. actual KDPI of 70% was displayed as 86%). The error was corrected on May 19, 2016, <24 h after being recognized. During this 30-day period, the distribution of recipients largely resembled pre-KAS patterns. The observed discard rate of 22.9% was higher than the post-KAS average of 19.6% (odds ratio [OR]: 1.22) but far lower than the projected rate of 31.4% (OR: 1.96) based on the usual discard rate by KDPI relationship, suggesting clinicians and patients did not rely heavily on this single number (KDPI) in kidney-utilization decisions. Still, risk-adjusted analyses suggest the elevated discard rate was most likely attributable to the erroneously high KDPIs, not a shift in donor characteristics or random chance. The rise in discard rate was sharply higher for kidneys with inflated KDPI that crossed the 85% policy threshold (OR: 1.46; p = 0.049) versus those that did not (OR: 1.06; p = 0.631).

Allocation of deceased donor kidneys: A review of international practices

“…It improves to 0.78 (considered good discriminatory power) when comparing donors with the highest and lowest 20% estimated graft survival, supporting these categories of stratification for allocation . There is, however, a concern that labelling kidneys, ‘high KDPI’ will increase non‐utilization rates and compromise overall utility . Similarly, the c‐statistic for EPTS is 0.69, which has been externally validated in Australia and New Zealand .…”

Insights into the labeling effect of Kidney Donor Performance Index reporting: The Australian experience

“…Examining changes in the kidney nonutilization rate after the introduction of KDPI use in the United States, Bae et al found that although there was no overall increase in the nonutilization rates, among kidneys that were previously labeled as “standard criteria” that were reclassified as “high risk” (KDPI > 85%) there was a significant increase in odds of nonutilization (adjusted OR [aOR]1.42, 95% CI 1.07 to 1.89, P = .02). In an opportunistic study, Stewart et al reported on the impact of KDPI labeling when the index was incorrectly calculated for a 30‐day period in the United States as the result of a programming error, resulting in erroneously high KDPI values being displayed with organ offers. During this period, the kidney nonutilization rate did increase, but not to the degree that would have been predicted based on the usual relationship between KDPI and organ nonutilization.…”

“…Concerns have been raised about the potential labeling effect of KDPI reporting in Australia—that clinicians may be more reluctant to accept offers of kidneys with high KDPI, resulting in resource wastage in donation coordinator time to allocate kidneys and potentially even increased organ nonutilization. Similar issues have been debated in the United States after the introduction of KDPI use in the Kidney Allocation System . It is also possible that increased awareness of kidney quality at the time of organ offer will result in more discerning use of marginal kidneys and directed use of lower‐KDPI kidneys, resulting in improved matching between predicted recipient survival and predicted organ survival.…”

“…Similar issues have been debated in the United States after the introduction of KDPI use in the Kidney Allocation System. [5][6][7] It is also possible that increased awareness of kidney quality at the time of organ offer will result in more discerning use of marginal kidneys and directed use of lower-KDPI kidneys, resulting in improved matching between predicted recipient survival and predicted organ survival.…”

Should UNOS Yield to Competition?

“…The error, which was introduced in 2016 when UNOS updated the KDPI calculator, resulted in a sharp rise in discard rate for deceased‐donor kidneys with an inflated KDPI that crossed the 85% policy threshold. Dr. Segev feels that this costly error should never have occurred …”