Slide 1

Thank you Doctor Port and thank you Doctor Cannata.

Slide 2

In the DOPPS III, 340 randomly selected hemodialysis units were stratified by unit type and geographic region in Japan, Northern America, Australia, New Zealand and in Western Europe. All together detailed data for more than 40.000 hemodialysis patients has been collected since the first DOPPS in 1996.

Slide 3

Now prior DOPPS work has indicated that for many practices facility-level analyses are superior to patient-level analyses in reducing treatment-by-indication bias.
When many measured confounders do not differ across levels of a facility practice, such balance suggests that practice is provided independently of both measured and likely of most unmeasured confounders. In these cases, the facility practice has the potential to be substantially free from both measured and unmeasured confoundings.
So the question can then be tested, is there a survival benefit for hemodialysis patients treated in facilities having a greater use of a particular practice?

Slide 4

So, let’s start by looking at dialysis dose.

Slide 5

This is the facility percent of patients with a single pool Kt/V of less than 1.2 in different geographic regions. The median facility percent of patients with a Kt/V of less than 1.2 is about 5% in U.K. and in US and in France, Sweden and Spain less than 10%. However, the median facility percent of patients with a spKt/V of less than 1.2 is more than 20% in Japan and in Belgium with quite a wide range.

Slide 6

This is the relative risk of all-cause mortality by facility percent of patients with a Kt/V of less than 1.2 adjusted for age, male, black, time on dialysis, BMI, comorbid conditions for patients included in DOPPS I, II and III. The reference was set to less than 25% with a relative risk of 1. As you can see, with increasing facility percent of patients with a Kt/V of less than 1.2 there is an increase in the relative risk of all-cause mortality being 73% higher with 75% more patients with a Kt/V of less than 1.2. The relative risk is 1.07 per 10% more patients with a spKt/V of less than 1.2.

Slide 7

So, how about treatment time?

Slide 8

This is the relative risk of mortality adjusted for Kt/V by facility average treatment time in minutes. The reference was set to 1 for a facility average treatment time of more than 4 hours. You can see that with shorter facility average mean treatment time the higher the relative risk of mortality being 19% higher with shorter treatment times than 3.5 hours. The relative risk is 0.96 per 15 minutes longer average treatment time. Treatment time also has an impact on phosphorous control.

Slide 9

This is the adjusted odds ratio of having a phosphorous of more than 5.5 mg/dl or 1.8 mmol/l versus less than 5.5 mg/dl by facility mean treatment time. Again, the reference was set to 1 for treatment time above 4 hours. The shorter the treatment time, the higher the risk of having a high serum phosphorous being 44% higher with a treatment time shorter than 3.5 hours. This is a 16% lower risk per 30 minutes longer facility mean treatment time for having a high phosphorous level.

Slide 10

The quality of life also seems to be better with a longer facility mean treatment time. These are different components of the physical quality of life scale. These are different components of the mental physical quality of life scale. These are the facility mean treatment times of less than 3.5 hours. This is between 3.5-4 hours. This is the reference and this is treatment time of more than 4 hours. You can see that the summary of physical component and the mental component summary is significantly higher with a facility mean treatment time of more than 4 hours compared to the facility mean treatment time of 3.5-4 hours.

Slide 11

Now, serum phosphorous also has an impact on mortality and morbidity in hemodialysis patients.

Slide 12

This is the risk of all-cause mortality and cardiovascular mortality by serum phosphorous from 1mg/dl up to 10 mg/dl for patients included in DOPPS I, II and II. We can see that the curve is U formed and that the all-cause mortality is significantly higher with low serum phosphorous probably reflecting the presence of malnutrition. The all-cause mortality and cardiovascular mortality is increased in patients with high serum phosphorous levels maybe reflecting the presence of vascular calcification and secondary hyperparathyroidism.

Slide 13

This is at the facility level. The risk of all-cause and cardiovascular mortality associated with 10% more patients in the phosphorous category between 6-7 and above 7. You can see that the relative risk is increased in such facilities.

Slide 14

Now, the type of vascular access also has an impact on mortality and morbidity, especially the risk of infections and septicemia.

Slide 15

There is an international trend toward a greater use of dialysis catheters. These are the countries included in both DOPPS I and in DOPPS III. You can see that there’s an increase in the use of dialysis catheters during this period of time in all countries except in Japan and in the UK and you can see that in DOPPS III about 20-26% of the patients had a dialysis catheter in Spain and in the US and in the UK. The panel to the right shows patients included in DOPPS II and DOPPS III and also here there’s an increase in the use of dialysis catheters and you can see that in countries like Sweden, Canada and Belgium 32-40% of the patients have a dialysis catheter.

Slide 16

This is the relative risk of all-cause mortality by facility catheter use. The reference was set from 0-9% with the relative risk of all-cause mortality of 1. You can see that the greater facility percent use of catheter, the higher the relative risk of mortality being 30-45% higher with greater facility catheter use. These are patients included in all DOPPS, DOPPS I, II and III. The relative risk is 21% higher per 20% greater facility catheter use.

Slide 17

This also corresponds to septicemia hospitalization events. These are data from DOPPS II showing septicemia events per 100 patient years in relation to country catheter use. You can see that countries that use a lot of catheters like Sweden, US, Canada and Belgium have a high septicemia event rate and this also corresponds to significant costs.

Slide 18

The DOPPS practice related risk maybe a potential quality improvement tool.

Slide 19

This practice related risk score includes the percent patients with a Kt/V of more or equal to 1.2. The percent of patients with a hemoglobin of more or equal to 11g/dl. The percent of patients with a serum albumin of more or equal to 4. The percent of patients with catheters. A facility practice related risk score is calculated by multiplying the relative risk from a COX model for each of the 4 factors based upon the facility level of practice based on prevalent cross-section of DOPPS I patients. The possible PRS range is between 1 which is the lowest risk up to 2.4. This is now in press if you want to read more about this practice related risk. I will show you some pictures from this.

Slide 20

This is the relative risk of death in relation to quartiles of the practice-related risk score. Now, for the lowest quartile was set for a relative risk of death of 1. With increasing practice related risk score there’s an increased risk of mortality being 40% higher in both the highest quartiles.

Slide 21

This is also important that the change in the practice related risk score from DOPPS I to DOPPS II correlate significantly to a change in the standardized mortality ratio. This is a 20% decrease in the practice related risk score which corresponds significantly to a 19% decrease in the standardized mortality ratio which means, indicates that the practice related risk score maybe an important quality tool in the future.

Slide 22

Now, many studies have shown that inflammation is associated with morbidity and mortality in hemodialysis patients.

Slide 23

This is the percent of patients with a reported baseline CRP in DOPPS II and in DOPPS III. You can see that in DOPPS II 55% of the patients from Japan had a reported baseline CRP well this was much more uncommon in the other countries included in the DOPPS. However, in DOPPS III there was an increase in the percent of patients with reported baseline CRP in all countries except maybe for the US. Above 60% of the patients have a reported baseline CRP in Japan, Belgium, Italy, Spain, Germany, France and in Sweden.

Slide 24

This is data from Japan only. 2000 patients. This is the relative risk of death in relation to baseline CRP in mg/l. The reference was set to less than 1.0 mg/dl. Relative risk of 1. You can see that with an increase in CRP and already with a CRP of 3-5 there’s a significant increase in the relative risk of death being 64% higher in that range and this increase is with higher CRP levels adjusted for a number of other parameters. 

Slide 25

This is the mortality risk in facilities measuring CRP in more than 50% of their patients versus units measuring CRP in less than 50% of their patients. You can see that the all-cause mortality is lower in facilities measuring CRP in more than 50% of the patients versus less than 50% of their patients but the analysis was not statistically significant.
However, for cardiovascular and cerebrovascular mortality there’s a 28% decrease in the relative risk of death in facilities measuring CRP in more than 50% of the patients compared to units measuring CRP in less than 50% of the patients and this is also statistically significant.

Slide 26

So to summarise, DOPPS has applied the strength of its facility-based practice pattern study designed to help reduce treatment-by-indication biases inherent in patient-level observational data. When outcomes are tested in relationship to a facility practice rather than the treatment given to an individual patient, DOPPS findings provide strong support for guidelines and practices that avoid low dialysis dose a Kt/V of less than 1.2, that provide longer treatment times, that avoid high serum phosphorous levels that promote the use of AV fistulas and discourage the use of catheters and that utilize CRP as an aspect of ongoing care maybe with more attention to the treatment of infections and so on. If you want to read more, find more information you can go to the website www.dopps.org. Thank you very much for your attention.