healthmedicinet

One hundred and eleven nephrologists across Canada responded to the surveys over the
4-month period, representing 20 % of practicing nephrologists in Canada. To determine
how important the various time frames (1, 3, 5, 10, and 15 years) were perceived by
nephrologists for predicting various key clinical outcomes, respondents were asked
to rate the importance of predicting the risk of each outcome: kidney failure, CV
events, and death in their individual patients with eGFR 15–45 ml/min/1.73 m
²
. Over 80 % of respondents felt that the time frames of 1, 3, and 5 years were most
relevant in predicting risk of each outcome: kidney failure, CV events, and death
in their patients (Fig. 1a–c). In particular, a higher proportion of respondents rated as extremely important
the ability to predict the risk of kidney failure (62 %) as opposed to the risk of
CV events (44 %) or death (51 %). The importance of rating these risks dropped for
the longer time frames: 70 % of respondents rated the time frames of 10 or 15 years
as important. A small subset of respondents expressed that their abilities to predict
the risk of kidney failure (1 % of respondents), CV events (4 %), or death (2 %) did
not matter to them or that they did not think of risks in the described manner.

To assess the needs for the development of risk modeling and tools for informing the
clinical management of CKD, we asked our respondents about the perceived utility and
value of having validated risk scores for predicting the key clinical outcomes. As
shown in Fig. 2a, out of the 5-point scale, most of our survey respondents indicated that they would
“always” or “often” use a validated risk score for predicting specific outcomes in
order to initiate dialysis and transplant planning (76 % of respondents selected one
of those choices), CV risk reduction strategies (66 %), or end-of-life planning (58 %)
with their patients. Since uptake among physicians for a validated risk prediction
score in making decisions about clinical management may be dependent on the satisfaction
levels they might have with the current prediction methods, we asked about satisfaction
with their current certainty to predict specific outcomes for their patients with
eGFR 15–45 ml/min/1.73 m
²
. The majority of the respondents were not satisfied with their ability to predict
the progression to kidney failure, CV events, and death (Fig. 2b). Specifically, the ability of the clinicians to predict CV events and death is most
dissatisfying for them with 82 and 81 %, respectively, rated “not at all satisfied”
or “somewhat satisfied”.

Fig. 2. Predicting risks in CKD management for nephrologists. a Frequency that nephrologists would use a validated risk score for predicting specific
outcomes in order to discuss options in CKD management with their patients. b Levels of satisfaction among nephrologists for their current ability to predict specific
outcomes in their patients

In assessing factors relevant to vascular access planning for kidney failure, we asked
nephrologists what 1-year risk threshold for kidney failure would prompt them to refer
for AVF creation in their patients who had chosen hemodialysis as a treatment option.
In response, 45 % of respondents would refer their patients for AVF if the risk of
kidney failure were 50 %, while 32 and 7 % of respondents would refer at risk thresholds
of 30 and 20 %, respectively (Fig. 3a). The respondents who chose “other” remarked they would only refer if the risk of
kidney failure is much higher and with consideration of local resources available,
i.e., access to surgery and transplantation. When asked at what eGFR they would refer
their patients for AVF, the majority of the responses were split among 15 ml/min/1.73 m
²
(27 % of respondents), 20 ml/min/1.73 m
²
(29 %), and “other” (24 %) (Fig. 3b). As much as 73 % of respondents who chose “other” commented that consideration of
the rate of progression or progressive decline in GFR would also be necessary for
the decision-making.

Fig. 3. Acceptable thresholds for AVF referral. a 1-year risk of kidney failure for AVF referral b eGFR thresholds for AVF referral

Most respondents (39 % “yes, maybe”; 61 % “yes, definitely”) indicated that they would
use validated risk scores accurate to predict specific outcomes, if they were available.
They would more likely use risk scores (29 % “yes, maybe”; 71 % “yes, definitely”)
if the clinical management (i.e., timing of education about ESRD management, planning
for access, referral for transplant, and commencement of medications) of their individual
patients would be altered to improve the outcome.