Quality Campaign Cuts Dialysis Deaths

Published: Sep 17, 2013

By Crystal Phend, Senior Staff Writer, MedPage Today

save

|

A

A

A national quality improvement initiative cut 1-year mortality among new dialysis patients by 13.6%, an independent analysis showed.

While that reduction from 2009 to 2012 didn't reach the 20% target, further declines are expected based on the bigger 25% reduction in 90-day mortality, Vincent Mor, PhD, of Brown University School of Public Health, and colleagues wrote in a report released from the PEAK campaign.

"It did not achieve that bold goal, but it was a fairly ambitious goal to begin with," Mor told MedPage Today. "We actually do anticipate a continued drop in 1-year mortality over the next 5 or 6 months because of the 90-day mortality."

He served as project director for the independent data analysis commissioned by Kidney Care Partners, the amalgam of professional, industry, and advocacy groups that organized the initiative.

Overall end-stage renal disease mortality has been on the decline for a decade, whereas deaths in the first year have remained stubbornly high at an annualized rate of about 30%, the PEAK (Performance Excellence and Accountability in Kidney Care) websiteexplained as the reason for its focus.

The campaign included a number of voluntary "best practices" aimed at reducing catheter use in favor of arteriovenous fistulas, improving cardiovascular outcomes, managing nutrition and inflammation, handling anemia, and optimizing dialysis, along with depression screening, patient education, and social support.

The analysis couldn't pinpoint the reason for the decline seen during the campaign, but did hint at longer dialysis sessions as a contributing factor.

Case mix didn't change over the study period for albumin levels at dialysis initiation or other clinical and demographic characteristics, except for a small but steady decline in baselinehemoglobin levels and a possible increase in baseline creatinine among older age groups, suggesting slightly worse kidney function over time when starting dialysis.

Use of a catheter instead of an arteriovenous fistula when starting dialysis changed little over the study period, though patients may have gone on to get a fistula later in treatment, the researchers pointed out.

The one striking trend was more patients initially prescribed dialysis for at least 4 hours' duration.

That proportion rose 10 percentage points from before the initiative in 2008 through 2011, from just under 65% to nearly 75%, and then spiked up to almost 80% by the end of 2012.

While that shift was associated with a drop in mortality, Mor cautioned that the association could not be determined as causal based on the available data and that there was no data on what happened to prescribed dialysis duration later in the course of treatment.

His group pulled national data from the same Centers for Medicare and Medicaid Services (CMS)'s Renal Management Information System used by the United States Renal Data System (USRDS).

The decline in first-year mortality from 0.25 deaths per person-year when the campaign launched in June 2009 to less than 0.22 per person-year by the end of 2012 came out almost identical whether calculated by USRDS methods or by the PEAK method.

The proportion of patients dying within 1 year of starting dialysis fell by 13.7%, from about 21% to less than 19% over the same period.

Early deaths in the first 90 days declined from around 8% to about 6% by the end of 2012. That was a 22.5% improvement since the beginning of the program after adjustment for demographic, clinical, and treatment characteristics at baseline.

Notably, more than 60% of that decline occurred in those starting dialysis in 2012, "suggesting that the impact of PEAK may not have completed its impact on reductions in 1-year mortality," as early survival typically predicts 1-year survival, the report pointed out.

While the report was considered final, there are plans to follow mortality rates for another 6 months to see if that is the case, Mor said.

"Given the findings, now the important thing is to try to find out why," he suggested.

The USRDS has additional, more detailed data available, which should clarify the factors most closely related to the declining mortality so that those components can be applied more uniformly across practices, Mor noted.

Also, there were variations within regions that could be informative, he pointed out.