Pavel Roshanov
@PavelRoshanov
Followers
471
Following
922
Statuses
986
Nephrologist-scientist @WesternU and @LHSCCanada, Scientist @LHSCRI; Investigator @PHRIresearch
London, Ontario
Joined May 2010
Landmark analysis of 2 prospective studies (N=45K) reveals preoperative kidney function as a dominant predictor of cardiac complications in non-cardiac surgery. eGFR contributes the most predictive information - far exceeding established risk factors like coronary artery disease, diabetes, and previous stroke. A 50-year-old with eGFR of 30 approximates the cardiac risk of an 80-year-old with eGFR of 90. These findings challenge current risk calculators that rely on binary kidney function thresholds of serum creatinine. Full paper: Editorial:
1
2
6
The analyses demonstrate the value of modeling continuous variables like eGFR as continuous with flexible functions that allow for departures from linearity. Much thanks to the excellent work of @f2harrell in the rms package and his book on Regression Modeling Strategies that made this possible.
1
0
1
RT @f2harrell: The R rms package, now 34 years old, has a massive update. A key new #Statistics capability is profile likelihood confidenc…
0
39
0
RT @AmitXGarg: I'm excited to present the late-breaking study "Effect of a Perioperative Hypotension Avoidance Strategy vs. a Hypertension…
0
3
0
RT @Richard_D_Riley: *NEW pre-print* "Extended sample size calculations for evaluation of prediction models using a threshold for classifi…
0
16
0
RT @mvaduganathan: Wow, game-changing #SELECT data out at #HeartFailure2024! Sema 2.4mg ⬇️ CVd+HF events by 21% in HF subgroup of n=4,286…
0
105
0
0
0
0
@hswapnil @CSNSCN @turbo_dc Congrats to brilliant @turbo_dc who I get to work with closely thanks to @lastwalsh
1
0
4
RT @Nefrontera: 👏🏽Nice timeline of major trials of uremic solute removal by @PavelRoshanov at #CSNAGM
@CSNSCN
0
11
0
RT @turbo_dc: Swapnil putting on a show at the Top 10 Trials session CSN AGM @CSNSCN @CNTN_RCEN complete with amazing @NephJC visual abstra…
0
5
0
@docSandraO @PHRIresearch I can think of few scenarios--focused squarely around quality of life--where it seems appropriate to use mortality as a non-inferiority outcome.
1
0
1
RT @docSandraO: When one gets an opportunity to run with an idea from a great scholar & be guided by great scholars…
0
3
0
RT @CaroAlba05: Risk #models make accurate #predictions in #heartfailure. Physicians significantly overestimate risk. When absolute risk is…
0
10
0
RT @LGHemkens: Saying trials dont measure "real world effects" because they exclude patients with comorbidities is not evidence based: "We…
0
19
0