Sunday, February 8, 2009

Lins, et. al, Nephrology Dialysis Transplantation 2009 Study

Lins, et al, has a new study titled: Intermittent versus continuous renal replacement therapy for acute kidney injury patients admitted to the intensive care unit: results of a randomized clinical trial in Nephrology Dialysis Transplant.

The Nephrosphere criticizes it:
This study by Lins, et al. published in Nephrology Dialysis Transplanation addresses the age old question of whether continuous renal replacement therapy (CRRT) is superior to hemodialysis (HD) in acute kidney injury. The authors stratified by severity of illness and conclude that there’s no difference between the groups. In my mind, however, the fatal flaw is the people who were left out of the study. Out of the 650 subjects eligible and in need of RRT, over half (n=344) were excluded, 37% of them for clinical reasons (including coagulation distubrances and hemodynamic instability). In other words, the most unstable patietns, the ones most likely to benefit from CRRT, were excluded from the study! It’s hard to conclude much from this other than CRRT probably isn’t advantageous in more stable patients.
That comment gets right to the heart of the matter, CRRT is for the most unstable patients, ones where the regular hemodialysis is judged to be too much of a strain on the system. As noted above, a number of subjects were excluded for hemodynamic instability, but then the study says: "The main reason for conversion from IRRT to CRRT was haemodynamic instability" (and the other way for coagulation disturbances).

I highlight this point: "An overall mortality of 60.1% was observed." just so people with family members on CRRT in this condition are aware of the most likely outcome.

Looking closer at the outcomes, even though it didn't reach the 10% difference (or patient sample size for that matter) CRRT does coming out slightly "ahead": "Intention-to-treat analysis revealed a mortality of 62.5% in patients treated with IRRT compared to 58.1% in patients treated with CCRT (P = 0.430)." Along with this informative graphic:This is why we keep doing these studies, however, none of them can quite get to the definitive proof needed to claim that CRRT is better in certain patients. Lins, et al. does get closer than other recent attempts, and provides insight into the next steps he feels should be taken. There is certainly opportunity here for a researcher who finds the correct patient group to make a bit of a name for themselves.

The full study is currently available for free at the link.