Transcript Blood pressure and mortality risk in peritoneal dialysis patients

Blood pressure and mortality risk in peritoneal
dialysis patients in England and Wales
Udaya P.Udayaraj , R.Steenkamp, F.Caskey,
D.Ansell, C.Tomson
UK Renal Registry, Bristol, UK
www.renalreg.org
Introduction
• General population – linear or ‘J’ shaped association of BP
with mortality
Lewington et al. Lancet, 2002; 360:1903-13
Introduction
• ‘Reverse J’ or ‘U’ association well described in
hemodialysis (HD) patients
2.5
2
1.5
1
0.5
180+
170-79
160-69
150-59
140-49
130-39
120-29
110-19
0
<110
relative death rate
3
SBP post dialysis mmHg
Zager et al. KI,1998;54:561-69
Kalantar-Zadeh. Hypertension, 2005;45:811-17
Introduction
• Is this ‘reverse’ association related to HD
procedure?
– Creation of AV fistula can predispose to myocardial
ischemia ‡
– Hypotension during HD can precipitate myocardial
ischemia * and is associated with increased mortality †
‡ Savage et al. AJKD 2002; 40: 753-59
* Selby et al. Semin Dial 2007; 20:220-28
† Shoji et al. KI 2004; 66:1212-20
Introduction
• Previous studies in peritoneal dialysis (PD) patients –
inconsistent results due to small sample size † or short follow
up ‡
• Low BP associated with lower mortality in kidney transplant
recipients
• Association of BP and survival amongst dialysis patients
awaiting transplantation is not known
† Lynn et al. KI 2002;62:2281-2287
Ates et al. Kidney Int 60: 767-776, 2001
Jager et al. Kidney Int 55: 1476-1485, 1999
‡ Rumyantazev et al. NDT 2005;20:1693-1701
Rocco et al. Perit Dial Int 22: 371-379, 2002
Aim
To study the association of BP and mortality in
• a larger cohort of peritoneal dialysis patients
• a subgroup of patients who are registered on the
kidney transplant waiting list
Methods (Databases)
• UK Renal Registry (UKRR)
–
–
–
–
Established in 1997
100 % coverage of England and Wales expected by 2007
Fully electronic data extraction from clinical information system
Quarterly biochemical and clinical data on all patients on renal
replacement therapy (RRT)
• UK Transplant (UKT)
– Maintains national organ donor register
– Holds data on kidney transplant waiting list
Study population
Incident RRT patients 1997-2004 in England and Wales
– Inclusion criteria
• From centres with >85% ethnicity data completeness †
• Age >18 years
• On Peritoneal dialysis at day 90
– Exclusion criteria
• Missing data on age/gender/ cause of renal failure/ethnicity
• Transplant prior to day 90
• Missing BP data in 1st or 2nd quarter
† No difference in survival or baseline characteristics of patients in centres with > 85
% and < 85% data
Statistical analysis
• BP measurements
– Mean BP value in the 1st/2nd quarter of RRT
– Systolic, Diastolic, Mean arterial and pulse pressure studied
• Cox regression model
• BP as a continuous variable ( linear and non linear terms)
• Factors adjusted in multivariate model
– Age, gender, ethnicity, cause of renal failure, haemoglobin, corrected
calcium, phosphate, serum albumin
– Change of RRT modality as time dependent variable ( i.e. patients not
censored at modality change)
– Time to inclusion on Transplant waiting list as surrogate for comorbidity
• Outcome
– All cause mortality after day 90
– All patients followed until 31.12.05 or death
• Subgroup analyses
– Diabetics and non diabetics
– Transplant waiting list registration status in 1st year of
RRT ( all cause mortality from 1 year after start of
RRT)
• Waitlisted within 1st year from start of RRT
• Not waitlisted within the 1st year from start of RRT
Results (Study cohort)
Cohort selection
PD at day 90 ( centres >85 % ethnicity data)
With data on ethnicity, cause of renal failure,
previous transplant
With SBP/DBP data*
With data on lab parameters and date of
Transplant waitlisting
no. of patients
included
3,991
3,822
2,061
2,035
* No difference in baseline characteristics and survival between those
included and excluded except % males was more in excluded cohort (62.8% vs
58.2%)
Baseline Characteristics
Number of patients
median age years (range)
Males %
Cause of renal failure %
Diabetes
Glomerulonephritis
Hypertension
Others
Polycystic kidney disease
Pyelonephritis
Renovascular disease
Uncertain etiology
Ethnicity %
Asian
Black
Other
White
Median duration of followup years (range)
2035
59 (18 -90)
58
20.2
13.4
6.4
12.2
8.7
8.8
6.2
24.0
6.6
2.6
1.1
89.7
2.98 (0.27 -9.0)
Cox regression models
Relative Hazard
2.0
P< 0.0002
1.5
P= 0.01
1.0
0.5
unadjusted
adjusted
†
0.0
100 110 120 130 140 150 160 170 180 190
Systolic blood pressure mmHg
†Adjusted for age, gender, cause of renal failure, ethnicity, lab parameters, treatment modality, time to transplant waiting list
Cox Regression models
Relative Hazard
2.5
unadjusted
2.0
adjusted
P= 0.0007
1.5
†
P=0.05
1.0
0.5
0.0
70
80
90
100
110
120
130
Mean arterial pressure mmHg
†Adjusted for age, gender, cause of renal failure, ethnicity, lab parameters, treatment modality, time to transplant waiting list
Subgroup analyses – Time to transplant
waiting list (WL)
WL < 1year
4.0
2.0
Hazard ratio
P =0.07
WL > 1 year / not WL
3.0
P= ns
1.0
0.0
90
110
130
150
170
190
WL < 1 year
WL > 1 year / Not WL
3.0
p =0.01
2.0
P = ns
1.0
0.0
210
70
Systolic Blood pressure mmHg
80
90
100 110 120 130 140
Mean arterial pressure mmHg
WL < 1 year (n=675) HR per 10 mmHg 1.14, 95 % CI 0.99-1.31
HR per 10 mmHg 1.29, 95 % CI 1.05-1.59
4.0
WL < 1 year
Hazard ratio
Hazard ratio
4.0
3.0
WL > 1 year / not WL
p =0.01
2.0
HR per 10 mmHg 1.33, 95 % CI 1.07-1.67
1.0
P = ns
0.0
60
70
80
90
100
110
Diastolic Blood pressure mmHg
† Survival after 1 year, Adjusted for age, gender, cause of renal failure, ethnicity, lab parameters, treatment modality
Subgroup analyses - Diabetes
• Diabetic – no association of any BP component with
mortality
• Non diabetic – U shaped for SBP and reverse J for
MAP (as for entire study cohort)
† Adjusted for age, gender, ethnicity, lab parameters,treatment modality, time to transplant waiting list
Summary
adjusted
Pattern of association of BP and mortality was
– For the entire study population
• U shaped for SBP
• Reverse J for MAP
• No association for DBP and PP
2.0
2.0
1.5
1.5
1.0
0.5
1.0
0.0
0.5
– For those registered on the transplant waiting list within first year
0.0
• Linear for SBP, DBP, MAP
3.0
• No association for PP
2.0
– For patients with diabetes
• No association with any BP components
1.0
0.0
• Strengths
– Largest study to date on PD patients
– First to describe the association of BP and survival in
patients on the transplant waiting list
– Novel analytic strategy of adjusting for modality change
allowed extended follow up of patients without censoring
• Limitations
– BP measurement not standardised
– Lack of data on co-morbidity, residual renal function
dialysis dose
– No data on cardiac function or anti HT medications
Conclusions
• High and low SBP are risk factors for increased
mortality in incident PD patients
• Low SBP was not associated with increased mortality
in a subgroup of patients activated early on the
transplant waiting list
• Association of low SBP and mortality seen in entire
study cohort could reflect poor cardiac function/
health status
• The BP associated with best survival may vary for
different patient subgroups
Baseline Characteristics
SBP mean + SD mmHg
DBP mean + SD mmHg
MAP mean + SD mmHg
PP mean + SD mmHg
143 + 22
81 + 12
102 + 14
62 + 17
Haemoglobin mean + SD g/dl
Corr. Calcium mean + SD mmol/L
Serum phosphate mean + SD mmol/L
11.2 + 1.5
2.4 + 0.2
1.6 + 0.4
Time to Transplant waitling list from start of RRT %
< 6months
> 6 months
not waitlisted
22.2
25
52.8