Acute renal failure: factors influencing nephrology referral and outcome

Transcription

1 QJMed 1997; 90: Acute renal failure: factors influencing nephrology referral and outcome I.H. KHAN, G.R.D. CATTO, N. EDWARD and A.M. MACLEOD From the Department of Medicine and Therapeutics, and Renal Unit, University of Aberdeen, Aberdeen, UK Received 26 June 1997 and in revised form 1 October 1997 Summary We investigated the incidence, factors affecting referral and outcome of acute renal failure (ARF), in an unselected (predominantly Caucasian) population in the Grampian region of Scotland served by a single renal unit. Case-notes were examined for all patients with a serum creatinine ^ 300 uxnol/l. ARF (311 patients) was defined as a temporary rise in serum creatinine ^ 300 imol/l, or, if the patient died during the acute illness, clinical features indicating acute deterioration of previously normal renal function. Advanced ARF at presentation (51 of the 311 with ARF) was defined as a first recorded serum creatinine ^ 500 (imol/l. Patients were classified into low-, medium- and high-risk groups according to presence of comorbidity and age. The annual incidence of ARF was 620/million population (pmp), Introduction Community-based studies have estimated an incidence of advanced acute renal failure in the UK (serum creatinine ^500u.mol/l) of 140 per million of the population 1 and more than 50% of such patients are not referred to a nephrologist. 1 Studies based on data from renal units, therefore, provide an underestimate of the true incidence of ARF. and do not allow factors influencing referral of such patients to nephrologists to be examined. 2 ' 3 Community-based 1 ' 4 studies, although providing true incidence rates, have not studied patterns of referral to nephrology services in detail. We and others have previously shown 5 " 7 that comorbidity and age have a major influence on survival of patients with chronic renal failure and also on the referral of such patients that of advanced ARF 102 pmp. The age-related incidence of ARF ranged from 30 pmp in the age group (0-19 years) to 4266 pmp in the age group >80 years. Overall, 22% were referred to a nephrologist (34% after excluding those with advanced cancer and age >80 years). Referral of patients decreased from 100% in the age group 0-19 to 5% in those >80 years. Referrals in the low-, medium- and high-risk groups were 75%, 30% and 14%, respectively. Patient survival at 2 years was 80%, 42% and 19% for low-, medium- and high-risk groups, respectively (86%, 44% and 32% for referred patients). Referral and outcome in patients with ARF were significantly influenced by age and presence of comorbidity at presentation. to a nephrologist. 8 Using similar methods, we aimed to establish the incidence of ARF (serum creatinine ^300 jimol/l) in an unselected population served by a single renal unit, and the factors affecting referral of these patients to a nephrologist. Methods The renal service for the whole of Grampian region (population in mid-1990) is based in a single renal unit in Aberdeen Royal Infirmary. All serum creatinine assays during the period of the study (01/07/89-30/06/90) were performed in the Department of Clinical Biochemistry Aberdeen Royal Address correspondence to Dr I.H. Khan, Consultant Nephrologist, Aberdeen Royal Hospitals NHS Trust, Foresterhill, Aberdeen AB25 2ZD. i.khan@abdn.ac.uk Oxford University Press 1997

2 782 I.H. Khan etal. Hospitals NHS Trust and the results were stored on computer. Case-notes of all patients with a serum creatinine concentration ^300(j.mol/l were examined 2 years after the date of the assay. Acute renal failure was defined as a rise in serum creatinine concentration for the first time above 300 jamol/l, following which either (i) the serum creatinine returned to below that level and remained at a lower level, or returned to normal or (ii) the patient died during the acute illness and the clinical features were suggestive of acute deterioration (the patient previously having had normal renal function). Patients who had only one serum creatinine measurement were considered to have had acute renal failure if the clinical features suggested an acute illness, and they had died or were discharged without a further serum creatinine estimation. Such patients could include those with late presentation of end-stage renal disease and are therefore difficult to classify; they were included only if the clinicians regarded them as having had acute renal failure. Patients with an acute deterioration in renal function during the course of known chronic renal failure (creatinine persistently ^200 umol/l) and those who remained dialysis-dependent for more than 90 days after commencing renal replacement therapy were excluded from this study. The so called '90 day rule' is arbitrary and was used as a cut-off for defining acute renal failure. Patients who subsequently recovered renal function were regarded as having had acute renal failure. Patients whose first measured serum creatinine was ^500 (imol/l were classified as presenting with advanced ARF. The following information was extracted from case-notes: age and sex, dates of creatinine assays, date of referral to a nephrologist, cause of ARF, antecedent comorbidity at presentation, date of death and date of first dialysis if applicable. Each patient was assigned a risk category of low, medium or high, based on age and comorbidity at presentation (Table 1). The age-related annual incidence of ARF was calculated using data from the report of the Registrar General of Scotland, 1990 which provided information on population in different age bands. Statistics The x 2 test w ' tn Yates' correction was used to compare discreet variables. Continuous data were compared using Mann-Whitney U and Student's t test. The level of significance was set at 5% for all comparisons. Results During the 12-month period 311 patients were identified who fulfilled the criteria for ARF as defined above. Complete data were available for 310 patients (malerfemale ratio 1.23). Of these, 51 patients (male: female ratio 2.4) presented with advanced ARF (creatinine at presentation ^500 uxnol/l). Overall 70/310 (22%) patients were referred to a nephrologist (35% of those with advanced ARF); when those with malignant disease and age above 80 years were excluded, the referral rate for the remainder rose to 34.2%. The mean age (95% Cl) was 72.3 ( ) years for patients who were not referred to a nephrologist and 56.0 ( ) for those who were referred (p<0.0001). The mean (95% Cl) ages in the low-, medium- and high-risk groups were 38.6 ( ), 62.0 ( ) and 74.6 ( ) years, respectively. All 20 patients in the low-risk group were <70 years and 45% of those in the high-risk group were >80 years. Patients with volume depletion and cardiogenic shock were referred relatively infrequently (Table 2). Referral and survival according to risk group Significantly more patients in the low-risk group were referred (75%) compared with those in the medium- and high-risk groups (30% and 14%, respectively, Table 3). Overall survival at 2 years was 80%, 46% and 19% in low-, medium- and high-risk groups, respectively (80%, 51% and 32% at 3 months) (Table 4). Survival for those referred was 86% in the low-risk, 44% in the medium-risk and 32% in the high-risk group, and for those not referred Table 1 Risk groups for ARF patients Risk group Low Medium High Definition Age <70 years and no comorbid illness. Age or age <80 years with any one of the following: angina, previous myocardial infarction, cardiac failure, chronic obstructive airways disease, pulmonary fibrosis, or liver disease (cirrhosis, chronic hepatitis), [peripheral vascular and cerebrovascular disease] or <70 with diabetes mellitus, in addition to ARF. Age >80 or any age with two or more organ dysfunction in addition to end-stage renal disease, or any age with diabetes and cardiopulmonary disease, or any age with visceral malignancy in addition to ARF.

3 Referral in ARF 783 Table 2 Causes ot renal failure Table 5 Annual incidence of ARF according to age Cause Volume depletion Cardiogenic Sepsis and infection Surgical cases Haematological Urinary obstruction (not prostatic) Intrinsic renal Uncertain Prostatic Drugs and poisoning Hepatorenal Cardiac surgery Miscellaneous Numbers (no. referred) All patients 49(3) 43(2) 33(8) 26(11) 23(5) 23(3) 22(14) 22(0) 18(2) 16(6) 14(6) 12(7) 9(3) 310 Advanced ARF* 6(2) 3(0) 6(1) 2(1) 3(2) 9(1) 7(5) 1(0) 7(1) 2(2) 3(1) KD KD 51 *Creatinine ^500 u,mol/l. Numbers in parentheses are patients referred to a nephrologist. Table 3 Referral according to risk group Risk group Referred Non-referred Low Medium High 15(75) 27(30) 28(14) 70(23) 5(25) 63(70) 1 72(86) 240(77) %, 48% and 16% in the three risk groups, respectively. Survival was statistically significant (p<0.0001) only in the high-risk group patients who were referred. Dialysis treatment A total of 25 patients (50 per million/year) received dialysis, seven, eight and 10 in the low-, mediumand high-risk groups, respectively. Two-year survival in these patients was 71%, 25% and 20%, in the low-, medium- and high-risk groups, respectively. Age group (years) ^ Referred (%) Per million population* 30 (30) 157 (117) 834 (606) 2694 (1604) 5188 (4266) * Parentheses indicate the incidence when those with malignancy are excluded. Advanced ARF Fifty-one patients presented with a serum creatinine ^500uxnol/l. Intrinsic renal disease and urinary obstruction (both prostatic and non-prostatic) were the two main causes of renal failure in these patients (Table 2). Referral was 71%, 38% and 27%, in the low-, medium- and high-risk groups, respectively. Overall, 47% survived at 2 years (85% of the low-, 64% of the medium- and 30% of the high-risk group). Annual incidence of ARF After excluding those with a history of visceral malignant disease and those over 80 years, the incidence was 318/million. The annual incidence of ARF (creatinine ^300umol/l) was 620/million population and that of advanced ARF (creatinine ^500 u.mol/1) was 102/million. The incidence of ARF increased with advancing age from 30/mi11 ion/year for the age group 0-19 years to 4266/mil lion/year for those over the age of 80 years (excluding those with malignant disease) (Table 5). Referral of patients decreased from 100% in the age group 0-19 years to only 5% in those ^80 years. Discussion In this study, age and comorbidity had a major influence on referral to a nephrologist of patients with ARF. There is no agreed definition of ARF, Table 4 Survival at 2 years according to risk group Risk group Referred Not referred Alive Dead Alive Dead Alive Dead Low Medium High

4 784 I.H. Khan et al. which has made it difficult to compare results from different studies. We chose a definition based on clinical evidence of acute deterioration of renal function in patients with no evidence of renal failure prior to presentation and, in order accurately to define the incidence of ARF, we have excluded patients with acute deterioration of renal function during the course of known chronic renal insufficiency. We have also excluded patients with acute renal failure who remained dialysis-dependent for longer than 90 days, who were thus considered as having end-stage renal disease. Such patients, for whom the term 'acute irreversible renal failure' has been coined, may comprise almost 17% of those presenting with acute renal failure. 9 Our study, however, underestimated the true incidence of ARF in the community, as only those patients whose serum biochemistry had been measured could be studied. Previous studies of the incidence and outcome of ARF, both renal-unit-based and community-based, have shown a high mortality for this condition. 10 ' 11 It is difficult to compare the incidence of ARF in our study with those previously published, as we have aimed to define the incidence of patients with a serum creatinine at presentation of 300 u.mol/1. Feest et al. studied the incidence of advanced ARF (creatinine >500 u.mol/1) and found an incidence of 140 per million. 1 The incidence of advanced ARF (creatinine at presentation ^500jamol/l) in our study was 102 per million. This study has shown the major influence of antecedent comorbidity and advancing age on nephrology referral and survival. These two factors had an overriding influence on survival, irrespective of referral to a nephrologist. Survival was significantly better in the referred patients in the high-risk group. It can be speculated that non-referral of high-risk patients was appropriate in view of the presence of severe comorbidity. Whether improved survival in the referred high-risk patients was related to nephrology referral, cannot be determined, due to the retrospective and uncontrolled nature of the study. The 25 patients who received dialysis treatment (50/million of the population) had an overall survival of 64%. In these patients survival was best in low-risk (71%) and worst in the high-risk patients (20%). Thus overall crude mortality figures mask the influence of comorbidity on survival. We did not record the reasons for not offering dialysis to patients, as these data are difficult to obtain reliably, working retrospectively from case-notes. There are no agreed guidelines for referral of patients to a nephrologist. Formulating such referral guidelines might save considerable time and resources if 'inappropriate' referrals, such as patients with transient rises of serum creatinine due to readily reversible disorders (e.g. volume depletion) or those with terminal renal failure in the course of advanced malignant disease, are avoided. Secondly those patients with intrinsic renal disease, or those with multi-organ failure, who might benefit most from nephrology referral, could be referred early in the course of ARF, which might improve outcome. Guidelines for the initial management of ARF (such as the need for ultrasound, serum and urinary osmolality, urine microscopy) may also help physicians determine the need for nephrology referral. The incidence of ARF increases dramatically in the older age groups, and in this study the annual incidence increased from 30 per million in the age group 0-19 to 5188 per million in those over 80 years of age. There was a marked reduction in referral rates for older patients, decreasing from 44% for the age group years to 19% for those aged 70-79, and only 5% for those > 80 years of age Only 22% of patients with ARF (as defined in this study) were referred to a nephrologist. The survival of patients referred in the high-risk group was better than those not referred and for all patients, comorbidity and age had a major influence on referral and survival. Guidelines for nephrology referral in this condition might help identify patients who might benefit from renal services. Increased referral of patients, however, is likely to require increased resources. Acknowledgements This study was supported by a grant from the Clinical Resource and Audit Group of the SOHHD. We thank Dr I. Broom of the Dept of Biochemistry Aberdeen Royal Infirmary, all the Physicians and Surgeons in Grampian who allowed us to study their patients, and Miss C. Ritchie for technical assistance. References 1. Feest TC, Round A, Hamad S. Incidence of severe acute renal failure in adults: results of a community based study. 8r /Vied; 1993; 306: Kennedy AC, Burton JA, Luke RC, Briggs JD, Lindsay RM, Allison MEM, era/. Factors affecting the prognosis in acute renal failure. QJ Med 1973; 47: Turney JH, Marshall DH, Brownjohn AM, Ellis CM, Parsons FM. The evolution of acute renal failure Quart JMed 1990; 73: Kaufman J, Dhakal M, Patel B, Hownburger R. Community acquired acute renal failure. Am J Kidney Dis 1991; 17: Khan IH, Catto GRD, Edward N, Fleming LW, Henderson IS, Macleod AM. Influence of coexisting disease on survival on renal replacement therapy. Lancet 1993; 341: Khan IH, Catto GRD, Edward N, Macleod AM. Death during the first 90 days of dialysis: a case-control study. Am J Kidney Dis 1995; 25:

5 Referral in ARF 785 Wright LF. Survival in patients with end-stage renal disease. Am) Kidney D/s 1991; 17:25-8. Khan IH, Catto GRD, Edward N, Macleod AM. Chronic renal failure: factors influencing nephrology referral. Q J Med 1994; 87: Bhandari S, Turney JM. Survivors of acute renal failure who do not recover renal function. QJ Med. 1996; 89: Turney JH. Outcome in acute renal failure: In Raine AEG, ed. Advanced Renal Medicine. Oxford, Oxford University Press, 1992: Wheeler DC, Feehally J, Walls J. High risk acute renal failure. QJ Med 1986; 61:

6 Tiny Pulse {Saffron on Japanese paper)

7 Hot Tongue {Chinese ginseng in a bowl)