Nephrol Dial Transplant (2006) 21: 1066–1068
doi:10.1093/ndt/gfk094
Advance Access publication 18 January 2006
Brief Report
Severity scoring and mortality 1 year after acute renal failure
Robert L. Lins1, Monique M. Elseviers2 and Ronny Daelemans1
1
ZNA Stuivenberg, Nephrology-Hypertension, Antwerp and 2Faculty of Medicine,
University of Antwerp, Antwerp, Belgium
Keywords: acute renal failure; long-term survival;
mortality; severity score; intensive care unit;
SHARF score
Introduction
Various scoring systems have been developed to
optimize the use of clinical experience in ICU for
prognosis and to address questions of effectiveness,
efficiency, quality of care and correct allocation of
scarce resources [1,2]. The general severity scoring
systems, however, are inappropriate for a diseasespecific population [3]. We introduced a scoring system,
useful for all patients with acute renal failure (ARF)
admitted to the ICU, whether treated or not with renal
replacement therapy (RRT). The Stuivenberg Hospital
Acute Renal Failure (SHARF) score for hospital
mortality of patients with ARF was developed in
a single centre study, using two scoring moments
(baseline and after 48 h) [4]. In a second phase, the
SHARF score was tested in a multi-centre study in
293 patients. After adaptation, this predictive model
for hospital mortality in ARF proved to be useful in
different settings for comparing groups of patients
and centres [5].
Even more than hospital mortality, long-term
survival and quality of life have become important
outcome parameters studied in other areas of epidemiological research. Also in the treatment of ARF,
it has been recommended that these outcome measures
should be incorporated in future research [6,7]. To date,
however, little is known about the long-term outcome
of ARF.
We used the cohort of the multi-centre SHARF study
to investigate mortality and renal function 1 year after
ARF and to assess the long-term predictive value of the
SHARF score.
Correspondence and offprint requests to: Robert L. Lins, ZNA
Stuivenberg, Nephrology-Hypertension, Antwerp, Belgium.
Email: [email protected];
[email protected]
Subjects and methods
In our prospective cohort study, 8 centres included all
consecutive patients (n ¼ 293) with ARF defined as
a creatinine above 2 mg% without known preexisting renal
disease [5]. Half of the patients died during their hospital
stay. The remaining 145 patients were contacted 1 year after
hospital discharge. After mortality was checked in the National
Registry, a questionnaire was sent to the general practitioner
(GP). Where the GP was not known, the questionnaire was
sent directly to the patient or to the nursing home where the
patient lived. Data on serum creatinine and body weight were
collected. Reminders to GPs, direct mailing to the patient,
consultation of hospital databases and phone calls were used
to complete the missing data.
Renal function at hospital discharge and 1 year later
was investigated using the calculated creatinine clearance
according to the Cockroft and Gault formula [8]. Stages
of chronic kidney disease at hospital discharge and after
1 year were defined according to the NKF K/DOQI guidelines [9].
Data were analysed using the SPSS statistical package. For
descriptive results, differences were tested using Student’s
t and chi-square test with significance level put at P<0.05.
Survival was investigated using the method of life table
analysis.
Results
The population studied consisted of 145 patients that
survived hospitalization after ARF. Patient characteristics at hospital discharge are given in Table 1.
Mortality after hospital discharge could be traced for
all patients: 32 out of the 145 hospital survivors (22.1%)
died within 1 year. Total mortality increased by the
end of hospitalization from 51 to 62% within 1 year
after ARF (Figure 1). At hospital discharge, 46 patients
showed a normal to mild decreased GFR (creatinine
clearance of 60 ml/min and more), 85 patients had a
moderate to severe decreased GFR (creatinine clearance between 15 and 59 ml/min) and 14 patients had
kidney failure (creatinine clearance below 15 ml/min).
In these three groups, 1 year mortality was 33, 18 and
14%, respectively.
ß The Author [2006]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please email: [email protected]
Severity scoring and mortality 1-year after acute renal failure
1067
Table 1. Characteristics of survivors of ARF at hospital discharge
Number of survivors
145
Age in years: mean (range)
Sex: male/female
Cause of ARF
Medical (%)
Surgical (%)
Type of ARF
Prerenal (%)
Renal (%)
Patients receiving RRT for ARF treatment (%)
Hospital length of stay in days: mean (SD)
Renal status at hospital discharge
GFR 60 ml/min (%)
GFR ¼ 15–59 ml/min (%)
GFR < 15 ml/min (%)
68 (21–89)
92/53
50.3
49.7
67.2
32.8
30.3
24.9 (22.8)
31.7
58.6
9.7
Fig. 2. One-year survival in patients treated with and without RRT
during their ICU stay.
Discussion
Fig. 1. Survival, hospital mortality and 1-year mortality in patients
admitted to the ICU with ARF.
One-year non-survivors were older than survivors
[73.5 (SD ¼ 10.1) vs 65.5 (SD ¼ 14.3) year; P ¼ 0.004]
and had more heart failure (59 vs 31%; P ¼ 0.003) and
multi-organ failure (38 vs 21%; P ¼ 0.052) during
ICU stay. At hospital discharge, non-survivors had a
mean serum creatinine of 1.7 mg/dl (SD ¼ 0.9) compared
with 2.1 mg/dl (SD ¼ 1.9) in survivors (P ¼ 0.282). No
difference in 1-year survival could be observed in
patients treated or not treated with RRT (Figure 2).
In contrast to APACHE II and Liano, the SHARF
scores calculated during ICU stay differed significantly
between the 1-year survivors and non-survivors (Table 2).
Renal function 1 year after hospital discharge could
be documented for 105 out of 113 survivors. Mean serum
creatinine of these survivors was 1.96 mg/dl (SD ¼ 1.74)
at hospital discharge and 1.94 mg/dl (SD ¼ 1.50) after
1 year (P ¼ 0.897). Mean creatinine clearances were
51.7 (SD ¼ 29.5) ml/min and 48.9 (SD ¼ 29.3) ml/min
(P ¼ 0.340), respectively. At hospital discharge, ESKD
was diagnosed in 14 patients. In four of them, renal
function still improved and dialysis treatment could be
stopped afterwards. Three additional patients developed ESKD during the 1-year observation period.
In the original cohort of 293 ARF patients, a hospital
mortality of 51% was observed. When we contacted the
survivors again 1 year after hospital discharge, a high
additional mortality of 11% was found.
It is difficult to compare this observation of longterm high mortality with published data. Information
on long-term survival after ARF is limited and
published data differed in population studied and
observation period used. Kresse and colleagues [10]
studied retrospectively long-term mortality and
found overall mortality rates of 60 and 79% over
a period of 18 and 36 months, respectively. Korkeila
and colleagues [11] reported on long-term mortality
in 62 ARF patients and found an additional
mortality of 10% at 6 months (from 45% at hospital
discharge to 55%) and a total mortality of 65% at
5 years.
The observation, that the highest mortality was
observed in patients discharged with normal to mild
decreased GFR and lowest mortality in ESRD patients,
cannot be explained. It is possibly due to the low
numbers and the rather short observation period for
mortality. It can be indicative for ‘survival of the fittest’
in the ESRD group. Overall, it is clearly demonstrated
that renal failure strongly increases cardiovascular risk
[12], albeit that a 1-year period is rather short to result
in an increased mortality. It is still possible that renal
failure of patients at hospital discharge was an indirect
sign of their bad cardiovascular situation. This was
suggested by the finding that heart failure during
ICU stay was more prevalent in non-survivors than
in survivors.
The bad prognosis of ARF is generally expressed
as hospital mortality. The SHARF score and other
1068
R. L. Lins et al.
Table 2. Severity of illness scores of long-term survivors and non-survivors
APACHE II
Liano
SHARF 0
SHARF 48
a
One-year survivorsa
(n ¼ 113)
One-year non-survivorsa
(n ¼ 32)
P-value of
difference
Predictive value
of mortality area
under ROC curve
18.9
0.53
34.2
27.8
20.0
0.59
42.9
39.2
0.315
0.054
0.028
0.011
0.623
0.587
0.791
0.833
(5.7)
(0.16)
(18.6)
(19.4)
(5.0)
(0.14)
(21.4)
(24.9)
Mean (SD).
severity scores were developed using hospital mortality
as an outcome variable. One can question, however,
whether this kind of score remains partly predictive
for long-term survival. We were able to demonstrate
that the SHARF score showed significant differences
between survivors and non-survivors while the
APACHE II and Liano scores did not.
We conclude that a large number of patients
surviving ARF in ICU died during the first year after
hospital discharge. Further investigation should focus
on long-time survival and look for preventive measures
in this field, such as a better follow-up of ARF patients
after hospital discharge.
Acknowledgements. The study was made possible by the commitment and cooperation of the investigators of the Stuivenberg
Hospital Acute Renal Failure study group. They collected the data
in the participating centres.
Stuivenberg Hospital Acute Renal Failure study group: P. Arnouts,
St Jozef Hospital Turnhout; J. Berwaerts, General Hospital
Middelheim; J. Bierens, General Hospital Stuivenberg; J.-M.
Billiouw, O.L. Vrouw Hospital; C. Claessens, St Jozef Hospital
Turnhout; M. Couttenye, University Hospital Antwerp;
R. Daelemans, General Hospital Stuivenberg, M.E. De Broe,
University Hospital Antwerp; H. Demey, University Hospital
Antwerp; I. Demeyer, O.L. Vrouw Hospital; M. Elseviers,
University Hospital Antwerp; E. Gheuens, General Hospital
Stuivenberg; J. Huygens, Free University Brussels; J. Nagler,
General Hospital Middelheim; G. Nollet, O.L. Vrouw Hospital;
P. Peeters, Free University Brussels; P. Rogiers, General Hospital
Middelheim; R. Rutsaert, University Hospital Antwerp; D.
van Caesbroeck, St Jozef Hospital Turnhout; P. van der Niepen,
Free University Brussels and P. Zachée, General Hospital
Stuivenberg.
Conflict of interest statement. None declared.
References
1. Knauss WA. Apache II: a severity of disease classification
system. Crit Care Med 1985; 13: 818–829
2. Liano F, Pascual J, Garcia-Martin F et al. Prognosis of acute
tubular necrosis: an extended prospectively contrasted study.
Nephron 1993; 63: 21-31
3. Douma CE, Redekop WK, van der Meulen JH et al. Predicting
mortality in intensive care patients with acute renal failure
treated with dialysis. J Am Soc Nephrol 1997; 8: 111–117
4. Lins R, Elseviers M, Daelemans R et al. Prognostic value of
a new scoring system for hospital mortality in acute renal
failure. Clin Nephrol 2000; 53: 10–17L
5. Lins RL, Elseviers MM, Daelemans R et al. Re-evaluation and
modification of the Stuivenberg Hospital Acute Renal Failure
(SHARF) scoring system for the prognosis of acute renal
failure: an independent multicentre, prospective study. Nephrol
Dial Transplant 2004; 19: 2282–2288
6. Suter P, Armaganidis A, Beaufils F et al. Predicting outcome in
ICU patients. Intensive Care Med 1994; 20: 390–397
7. Tonelli M, Manns B, Feller-Kopman D. Acute renal failure in
the intensive care unit: a systematic review of the impact of
dialytic modality on mortality and renal recovery. Am J Kidney
Dis 2002; 40: 875–885
8. Cockcroft DW, Gault MH. Prediction of creatinine clearance
from serum creatinine. Nephron 1976; 16: 31–41
9. K/DOQI Clinical Practice Guidelines for Chronic Kidney
Disease: Evaluation, Classification, and Stratification. Part 4.
Definition and classification of stages of chronic kidney disease.
http://www.kidney.org/professionals/kdoqi/guidelines_ckd
10. Kresse S, Schlee H, Deuber HJ, Koall W, Osten B. Influence of
renal replacement therapy on outcome of patients with acute
renal failure. Kidney Int Suppl 1999; 72: S75–S83
11. Korkeila M, Ruokonen E, Takala J. Costs of care, long-term
prognosis and quality of life in patients requiring renal
replacement therapy during intensive care. Intensive Care Med
2000; 26: 1824–1831
12. Go AS, Cherton GM, McCulloch CE, Hsu CY. Chronic
kidney disease and the risk of death, cardiovascular events and
hospitalization. N Engl J Med 2004; 351: 1296–1305
Received for publication: 25.2.05
Accepted in revised form: 4.1.06