Polypharmacy Predicts Onset and Transition of Frailty, Malnutrition, and Adverse Outcomes in Peritoneal Dialysis Patient.

BACKGROUND: Polypharmacy, frailty and malnutrition are known predictors of adverse outcomes in dialysis patients. Little has reported about their interaction and composite prognostic values. We aimed to describe the interaction between polypharmacy, frailty, nutrition, hospitalization, and survival in peritoneal dialysis patients. METHODS: In this prospective cohort study, we recruited 573 peritoneal dialysis patients. Drug burden was measured by medication number and daily pill load. Frailty and nutrition were assessed by the validated Frailty Score (FQ) and Subjective Global Assessment (SGA) respectively. All patients were followed for two years. Primary outcome was all-cause mortality. Secondary outcomes were fall and fracture episodes, hospitalization, change in FQ and SGA. RESULTS: At baseline, each patient took 7.5 ± 2.6 medications with 15.5 ± 8.5 tablets per day. Medication number, but not daily pill load predicted baseline FQ (p = 0.004) and SGA (p = 0.03). Over 2 years, there were 69 fall and 1,606 hospitalization episodes. In addition, 148 (25.8%) patients died, while FQ and SGA changed by 0.73 ± 4.23 and -0.07 ± 1.06 respectively in survivors. Medication number (hospitalization: p = 0.02, survival: p = 0.005), FQ (hospitalization: p < 0.001; survival: p = 0.01) predicted hospitalization and survival. Medication number also predicted fall episodes (p = 0.02) and frailty progression (p = 0.002). Daily pill load did not predict any of these outcomes. CONCLUSIONS: Drug burden is high in peritoneal dialysis patients, and it carries important prognostic implication. Medication number but not pill load significantly predicted onset and progression of frailty, malnutrition, fall, hospitalization, and mortality.

Characteristics and clinical outcomes of living renal donors in Hong Kong.

INTRODUCTION: In Asia, few reports are available on the outcomes for living renal donors. We report the short- and long-term clinical outcomes of individuals following living donor nephrectomy in Hong Kong. METHODS: We retrospectively reviewed the characteristics and clinical outcomes of all living renal donors who underwent surgery from January 1990 to December 2015 at a teaching hospital in Hong Kong. Information was obtained from hospital records and territory-wide electronic patient records. RESULTS: During the study period, 83 individuals underwent donor nephrectomy. The mean (± standard deviation) follow-up time was 12.0 ± 8.3 years, and the mean age at nephrectomy was 37.3 ± 10.0 years. A total of 44 (53.0%), four (4.8%), and 35 (42.2%) donors underwent living donor nephrectomy via an open, hand-port assisted laparoscopic, and laparoscopic approach, respectively. The overall incidence of complications was 36.6%, with most being grade 1 or 2. There were three (9.4%) grade 3a complications; all were related to open donor nephrectomy. The mean glomerular filtration rate was 96.0 ± 17.5 mL/min/1.73 m2 at baseline and significantly lower at 66.8 ± 13.5 mL/min/1.73 m2 at first annual follow-up (P<0.01). The latest mean glomerular filtration rate was 75.6% ± 15.1% of baseline. No donor died or developed renal failure. Of the donors, 14 (18.2%) developed hypertension, two (2.6%) had diabetes mellitus, and three (4.0%) had experienced proteinuria. CONCLUSION: The overall perioperative outcomes are good, with very few serious complications. The introduction of a laparoscopic approach has decreased perioperative blood loss and also shortened hospital stay. Long-term kidney function is satisfactory and no patients developed end-stage renal disease. The incidences of new-onset medical diseases and pregnancy-related complications were also low.

Urinary sediment ICAM-1 level in lupus nephritis.

BACKGROUND: Urinary intercellular adhesion molecule-1 (ICAM-1) level is potentially a valuable biomarker of lupus nephritis (LN), but because ICAM-1 is a cell-surface molecule, soluble ICAM-1 level in urinary supernatant measured by ELISA may not be biologically relevant. METHODS: The ICAM-1 level in urine sediment of 12 LN patients, 10 patients with pauci-immune necrotizing glomerulonephritis (NecGN), and six healthy controls were determined with a polymerase chain reaction (PCR)-based assay. The urinary sediment levels of miR-221, miR-222, miR-339-3P and miR-339-5P, which are involved in the regulation of ICAM-1 production, were also quantified. RESULTS: LN patients had lower urinary sediment ICAM-1 levels than the other two groups (overall p = 0.034). In addition, urinary sediment ICAM-1 level inversely correlated with the estimated glomerular filtration rate (GFR) (r = -0.474, p = 0.026) but not other markers of lupus activity, or urinary sediment levels of miR-221, miR-222, miR-339-3P, or miR-339-5P. However, serum anti-dsDNA level inversely correlated with urinary sediment levels of miR-221 (r = -0.591, p = 0.043) and miR-222 (r = -0.689, p = 0.013), while urinary sediment miR-221 level also correlated with serum C3 level (r = 0.658, p = 0.02). CONCLUSIONS: We conclude that urinary sediment ICAM-1 level was significantly reduced in LN, and the level inversely correlated with renal function. Urinary sediment miR-221 and miR-222 levels correlate with lupus disease activity and may serve as biomarkers of LN.

Fracture risk after thiazide-associated hyponatraemia.

BACKGROUND/AIM: Although thiazide-type diuretics can promote a positive calcium balance, thiazide can be associated with hyponatraemia, which is recently linked with heightened fracture risk. We examine the chance of developing fracture in patients with and without hyponatraemia after taking thiazide diuretics. METHODS: In this single-centre retrospective study, we followed up a previously published cohort of patients with (n= 223) and without (n= 216) thiazide-induced hyponatraemia. RESULTS: A total of 61 osteoporotic fractures was recorded during a mean follow-up period of 82 months. Using univariate regression analysis, the hazard ratio of thiazide-induced hyponatraemia was 1.78 (95% confidence interval (CI), 1.05-3.03; P= 0.033). Cox proportional hazards regression analysis, however, showed that age, body mass index and diabetes mellitus were the only independent predictors of osteoporotic fractures. No association of a history of thiazide-induced hyponatraemia and risk of fracture was evident in the final model. CONCLUSION: Since a history of thiazide-induced hyponatraemia is associated with osteoporotic fracture in univariate but not multivariate analyses, an alternative explanation is that confounding factors of older age and low body mass index accounted for the apparently increased risk of osteoporotic fracture in patients with thiazide-induced hyponatraemia.

Measurements on the routine chest radiograph as prognostic markers in Chinese peritoneal dialysis patients.

BACKGROUND: Fluid overload is a common problem in peritoneal dialysis (PD) patients. Cardiothoracic ratio (CTR) and vascular pedicle width (VPW) in routine chest radiograph are useful indicators of intravascular volume status and may represent important prognostic factors of PD patients. METHODS: We measured VPW and CTR in 286 unselected prevalent PD patients. VPW was further adjusted for the thoracic diameter (VPWR). One-year actuarial survival, technique survival, and duration of hospitalization were analyzed. RESULTS: The mean values of VPW, CTR, VPWR were 47.31 ± 4.73 mm, 0.542 ± 0.074, 0.170 ± 0.024, respectively. VPW correlated with age (r = 0.143; p = 0.016), body weight (r = 0.371; p < 0.001), body height (r = 0.271; p < 0.001), and Charlson's index score (r = 0.153; p = 0.01). One-year patient survival was 87.8%, and technique survival was 82.2%. None of the radiological measurements had an independent effect on one-year actuarial or technique survival by multivariate analysis. Both CTR and VPWR correlated with the duration of hospitalization (r = 0.192 and 0.186, respectively (p = 0.001 and 0.002). Multivariate regression analysis by log-linear modeling showed that independent predictors of one-year hospitalization were VPWR, serum albumin, and SGA overall score. CONCLUSIONS: In Chinese PD patients, VPW was significantly correlated with age, body weight, body height and Charlson's index score. VPWR was an independent predictor of the duration of hospitalization. Further studies are needed to confirm the prognostic value of these radiographic measurements in PD patients.

Relationship between carnosinase gene CNDP1 leucine repeat polymorphism and the clinical outcome of Chinese PD patients.

BACKGROUND: pre-clinical studies showed that carnosine may have a beneficial cardiovascular effect. We studied the effect of tri-nucleotide repeat (CTGs) polymorphisms in exon 2 of the CNDP1 gene, which codes for carnosinase and is responsible for the degradation of carnosine, on the clinical outcome of Chinese peritoneal dialysis (PD) patients. METHODS: we studied 442 PD subjects. Genotyping was done by direct sequencing of genomic DNA. Patients were followed for 43.5 ± 16.2 months. RESULTS: the prevalence of 6-6, 5-6, 5-5 and 4-6 CTGs genotypes was 80.3%, 18.6%, 0.9% and 0.2%, respectively. A total of 270 patients (61.1%) developed the primary composite end point during follow-up. The 5-year event-free survival of the 6-6 CTGs and non 6-6 group was 37.1% and 21.3%, respectively (log rank test, p = 0.3). CONCLUSION: the CTGs polymorphism of the CNDP1 gene does not affect survival of Chinese PD subjects. The role of carnosine and CNDP1 gene polymorphism in the pathogenesis of cardiovascular disease requires further study.

Relation between polymorphisms of receptor for advanced glycation end products (RAGE) and cardiovascular diseases in Chinese patients with diabetic nephropathy.

BACKGROUND: Interaction of receptor for advanced glycation end products (RAGE) with advanced glycation end products (AGEs) is an important pathogenic mechanism of diabetic complications. Three mutations in the promoter region of the RAGE gene (T-429C, T-374A and a 63 bp deletion spanning from -407 to -345 nucleotides) were known to have increased transcriptional activities. We investigated the relationship between these polymorphisms and the risk of cardiovascular diseases in Chinese subjects with overt diabetic nephropathy. METHODS: A total of 219 Type 2 diabetic subjects with nephropathy were recruited. Genotyping of the three polymorphisms in the genomic DNA was done by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). Patients were followed for 8 years for the development of cardiovascular events and survival. RESULTS: The T-429 C and T-374 A polymorphism had no effect on the event-free survival of the subjects. For the 63 bp deletion polymorphism, the event-free survival was 37.0% and 63.2% at 96 months for del-/- and del-/+ genotypes, respectively (log-rank test, p = 0.034). After adjusting for confounders, the 63 bp deletion polymorphism had a marginal effect on event-free survival (adjusted hazard ratio: 3.517, 95% CI: 0.852 - 14.521, p = 0.082). Subjects without any mutation of the three polymorphisms have significantly higher risk of first ischemic heart disease than those with any of the three mutations (adjusted hazard ratio: 0.218, 95% CI: 0.062 - 0.764, p = 0.017). CONCLUSION: The 63 bp del-/+ genotype of the RAGE gene has a marginal benefit on the cardiovascular event-free survival in subjects with diabetic nephropathy. Subjects with any of the three mutations have a lower risk of ischemic heart disease. The role of RAGE in the pathogenesis of cardiovascular disease in diabetic patients requires further study.

Tacrolimus for the treatment of systemic lupus erythematosus with pure class V nephritis.

OBJECTIVES: The treatment of pure membranous (class V) lupus nephropathy remains unsatisfactory. We studied the efficacy and safety of tacrolimus in the treatment of membranous nephritis secondary to SLE. METHODS: We recruited 18 consecutive SLE patients (tacrolimus group) with recently confirmed biopsy-proven class V lupus nephritis. They were treated with a tailing dose of oral prednisolone and tacrolimus 0.1-0.2 mg/kg/day for 6 months, followed by maintenance prednisolone and AZA. The rate of resolution of proteinuria and SLEDAI were compared with 19 historical controls treated with oral cyclophosphamide or AZA (control group). All patients were followed for 12 months. RESULTS: Baseline clinical characteristics were comparable between the groups. For the tacrolimus group, the complete and partial remission rates were 27.8 and 50.0%, respectively at 12 weeks; for the control group, they were 15.8 and 47.4%, respectively (overall chi-square test, P = 0.5). However, tacrolimus group had faster resolution of proteinuria than the control group by the general linear model with repeated measures (P = 0.032). At 12 weeks, proteinuria was reduced by 76.2 +/- 17.0% for the tacrolimus group and 47.1 +/- 51.1% for the control group (P = 0.028). Serial change in renal function and SLEDAI score did not differ between the groups. During the study period, four patients of the tacrolimus group, and 11 of the control group, developed lupus flare (P = 0.027). There was no serious adverse effect in the tacrolimus group. CONCLUSIONS: A 6-month course of tacrolimus is a safe and effective treatment of pure class V (membranous) lupus nephritis. As compared with conventional cytotoxic treatment, tacrolimus possibly results in a faster resolution of proteinuria, and a lower risk of lupus flare within 1 yr. The long-term effect and optimal regimen of tacrolimus require further study.

Sevelamer treatment strategy in peritoneal dialysis patients: conventional dose does not make best use of resources.

BACKGROUND: The significant incremental expense in the use of conventional sevelamer dose prompted us to evaluate the role of prescribing a lower dose of sevelamer. METHODS: To determine the optimum strategy for prescribing sevelamer in peritoneal dialysis patients, we conducted an open-label randomized study comparing the treat-to-goal strategy (4.0-g daily sevelamer dose) with lower sevelamer dose (1.2-g daily dose). RESULTS: Twenty-seven peritoneal dialysis patients with serum calcium x phosphorus product above 55 mg2/dL2 were recruited. Eighteen were randomized to the low-dose treatment group (1.2 g daily), and 9 to the treat-to-goal (4.0 g daily) group. Overall, significantly lower calcium x phosphorus product and serum phosphorus levels at 6 months were achieved by the treat-to-goal treatment. The proportions of patients who attained the Kidney Disease Outcomes Quality Initiative (K/DOQI) treatment target, however, did not differ significantly between the treat-to-goal and low-dose treatment groups (66.7% +/- 30.8% vs. 33.3% +/- 21.8%, p=0.10). The numbers needed to treat to benefit 1 patient who attains the K/DOQI recommendation are 1.5 patients (95% confidence interval [95% CI], 1.0-2.8) in the 4.0-g daily dose and 3 patients (95% CI, 1.8-8.7) in the 1.2-g daily dose group. Therefore, an extra 66.7% of subjects would be able to attain the treatment recommendation within the same budget if the daily dose of sevelamer used was 1.2 g instead of the usual 4.0 g. Compared with a 1.2-g daily dose of sevelamer, the 4.0-g daily dose had an incremental cost-effectiveness ratio (ICER) of US $2,353 per additional patient achieving the K/DOQI target. Multivariate analysis showed that only the calcium x phosphorus product after 1 month of sevelamer treatment was predictive of treatment response. CONCLUSIONS: Low-dose sevelamer treatment might be a cost-effective approach, which is "good for many rather than best for a few."

Peritoneal dialysis-related peritonitis: can we predict it?

Peritonitis complicating peritoneal dialysis represents a major cause of technique failure, hospitalization, and increased mortality. Peritonitis tends to be recurrent and clustered within particular patients at risk. The aim of this review is to evaluate the potential predictive factors for development of peritoneal dialysis-associated peritonitis based on currently available evidence. Risk factors were divided into medical and non-medical ones, and characterized by a schema of fixed versus modifiable factors. A new direction in the landscape change of the risk factors of peritonitis appears to focus on psychosocial aspects and patient training. Identification of these factors have important clinical implications because of the hitherto lack of well-established strategies to prevent peritonitis complicating peritoneal dialysis. It is hoped that better understanding of the risk factors will allow us to take tangible steps toward minimizing the infectious burden from the Achilles' heel of peritoneal dialysis.

Expression of T-bet, a type 1 T-helper cell transcription factor, in the urinary sediment of lupus patients predicts disease flare.

BACKGROUND: Systemic lupus erythematosus (SLE) is characterized by the aberrant activation of T-lymphocytes. Since T-bet is the principal transcription factor for the differentiation of type-1 helper T-lymphocyte, we study the impact of urinary T-bet mRNA expression in clinically quiescent SLE patients on the risk of subsequent disease flare. METHODS: We studied 60 quiescent SLE patients. Urinary mRNA expression of T-bet was studied by the real-time quantitative polymerase chain reaction. Patients were followed for 4 yrs for disease flare. RESULTS: We studied 60 patients; 57 were female. The mean age was 38.8 +/- 11.2 yrs. Their baseline SLE disease activity index score was 1.63 +/- 1.64. During the follow-up, 28 patients (46.6%) developed lupus flare, of which 17 (28.3%) had severe flare. Receiver operating characteristic curves showed that urinary T-bet expression three times above the average level of healthy control had 64.3% sensitivity and 84.4% specificity of predicting all lupus flare. Using this cut-off, patients with a high urinary T-bet expression had a significantly higher risk of all lupus flare and severe flare than the patients with a low T-bet expression (log-rank test, P < 0.001 for both). With multivariate Cox proportional hazard model to adjust for potential confounding variables, urinary T-bet expression and patient's sex were the only independent predictors of all lupus flare and severe flare. It could be estimated that 1-fold increase in urinary T-bet expression would result in 8.4% excess risk of all lupus flare [95% confidence interval (CI), 4.1-13.0%, P < 0.001] and 12.9% excess risk of severe flare (95% CI 7.4-18.7%, P < 0.001). CONCLUSIONS: A high urinary T-bet expression was an independent predictor of lupus flare. Measurement of urinary T-bet may provide valuable information for the risk stratification of SLE patients.

Influence of climate on the incidence of thiazide-induced hyponatraemia.

The role of hot temperature has been implicated in thiazide-induced hyponatraemia; however, it has never been studied in a systematic manner. The aim of this retrospective study is to correlate the incidence of thiazide-induced hyponatraemia and climate factors in a university teaching hospital from June 1996 to February 2002. We evaluated a representative sample of 201 subjects with thiazide-induced hyponatraemia. Overall, 2.9 +/- 2.2 (range 0-10, median 3) cases of thiazide-induced hyponatraemia were encountered each month during the study period. There was no seasonal variation in the rate of thiazide-induced hyponatraemia (overall chi(2) test, p = 7.0). Thiazide-induced hyponatraemia was not more frequently reported in summer. There was no discernible correlation between the monthly number of cases and average air temperature (r = -0.056, p = 0.65) and relative humidity (r = 0.103, p = 0.40). On the other hand, patients who presented with thiazide-induced hyponatraemia in July and August had significantly higher serum sodium concentration, 118 +/- 7 mmol/l vs. 114 +/- 8 mmol/l in other calendar months (p = 0.016). Temperature showed a statistically significant positive correlation with the level of serum sodium (r = 0.20, p = 0.004). These data demonstrate that there are no seasonal variations in thiazide-induced hyponatraemia disorders, at least in countries with subtropical climate. The question arises whether hypotonic sweat loss mitigates the risk of excessive water drinking in hot summer.

Enterobacteriaceae peritonitis complicating peritoneal dialysis: a review of 210 consecutive cases.

Enterobacteriaceae peritonitis is a serious complication in peritoneal dialysis (PD), but the clinical course of PD-related Enterobacteriaceae peritonitis remains unclear. We reviewed all Enterobacteriaceae peritonitis in our dialysis unit from 1995 to 2004. During this period, there were 1748 episodes of peritonitis recorded; 210 episodes (12.0%) in 123 patients were caused by Enterobacteriaceae. The most common species was Escherichia coli, accounting for 111 episodes. The primary response rate was 84.8% and complete cure rate was 58.1%. The presence of exit site infection was associated with a lower complete cure rate (43.2 versus 61.3%, P = 0.034). A total of 82 episodes (39.0%) did not respond to single antibiotic treatment despite sensitivity in vitro, and a second antibiotic was added. Patients treated with two antibiotics had a marginally lower risk of relapse and recurrence than those with one antibiotic (21.4 versus 36.1%, P = 0.051). The episodes that had recent antibiotic therapy had a marginally lower complete cure rate (49.3 versus 62.8%, P = 0.06). There was a gradual increase in the prevalence of resistance to several commonly used antibiotics over the years. Recent antibiotic therapy was associated with resistance to cefotaxime, ceftazidime, cefoperazone/sulbactam, and piperacillin/tazobactam. We conclude that Enterobacteriaceae peritonitis is a serious complication of PD. Recent antibiotic therapy is the major risk factor of antibiotic resistance. Exit site infection, and probably recent antibiotic therapy, is associated with poor therapeutic response. Contrary to the current recommendation, treatment with two antibiotics may reduce the risk of relapse and recurrence.

Imbalance of Th1/Th2 transcription factors in patients with lupus nephritis.

OBJECTIVE: Systemic lupus erythematosus (SLE) is characterized by the aberrant activation of T lymphocytes. Since T-bet and GATA-3 are the principal transcription factors for the differentiation of type-1 and type-2 helper T lymphocytes, respectively, we studied their mRNA expression in the urinary sediment of SLE patients and compared this with their urinary and intra-renal protein expression. METHODS: We studied 100 SLE patients and 10 healthy subjects. Urinary mRNA expression of T-bet and GATA-3 were studied by the real-time quantitative polymerase chain reaction. Intra-renal and urinary expressions of T-bet and GATA-3 were studied by immunohistochemistry and western blotting, respectively. RESULTS: The urinary mRNA and protein expressions of T-bet were significantly higher in SLE patients with active nephritis than those with inactive disease (mRNA: P < 0.001; protein: P = 0.004). The urinary mRNA expression of T-bet correlated with the SLE disease activity index (SLEDAI) score (r = 0.55, P < 0.001) and the histological activity index (r = 0.48, P = 0.03). On the other hand, the urinary mRNA and protein expressions of GATA-3 were significantly lower in SLE patients with active nephritis (mRNA: P < 0.001; protein: P = 0.006), and GATA-3 mRNA expression inversely correlated with the SLEDAI score (r = 0.38, P < 0.001). For the 22 SLE patients with kidney biopsy, tubular expressions of T-bet and GATA-3 significantly correlated with the histological activity index (T-bet: r = 0.57, P = 0.006; GATA-3: r = -0.79, P < 0.001). CONCLUSIONS: Patients with active lupus nephritis have increased T-bet and depressed GATA-3 expression in the urinary sediment and kidney tissue, indicating a predominant Th1 type of T-lymphocyte activation.

Change in bacterial aetiology of peritoneal dialysis-related peritonitis over 10 years: experience from a centre in South-East Asia.

This study reviewed 1787 episodes of peritoneal dialysis (PD)-related peritonitis in 544 patients between 1994 and 2003. The overall rate of peritonitis was 0.68 episodes/year of PD, but decreased from 1.10 to 0.46 episodes/year between 1994 and 2003. The incidence of peritonitis caused by coagulase-negative staphylococci declined between 1994 and 1998 from 0.21 to 0.06 episodes/year of PD, coinciding with a reduction in the use of spike PD sets. There was a 60.1% response rate to antibiotics throughout the period, but the percentage of cases that required modification of the initial empirical antibiotic regimen rose from 13.6% to 58.7%, indicating that treatment should be individualised.

Risk factors for thiazide-induced hyponatraemia.

BACKGROUND: Thiazide-induced hyponatraemia is common and potentially life threatening. In the absence of well-defined risk factors for this complication, guidelines for prescribing cannot be established. AIM: To examine whether a subgroup of patients is particularly susceptible to this complication. DESIGN: Retrospective case-control study. METHODS: We defined and recruited cases of symptomatic hyponatraemia that necessitated hospitalization from January 1996 to April 2002. Controls were selected from 8420 patients being prescribed thiazides and seen at the same institution during that period of time. RESULTS: There were 223 cases and 216 controls, with a median 115 days thiazide use. Cases were older than controls (76 +/- 9 vs. 66 +/- 13 years, p < 0.001) and lighter (52.3 +/- 10.3 vs. 63.4 +/- 3 kg, p < 0.001). By univariate analysis, serum potassium level, use of indapamide, elderly home institutionalization and physical immobility were risk factors for thiazide-induced hyponatraemia, but gender, duration of thiazide use, concomitant therapy with loop diuretics, angiotensin-converting enzyme inhibitors or non-steroidal anti-inflammatory drugs, and renal function were not. By stepwise logistic regression analysis, patient age, body weight and serum potassium were the only independent predictive factors. Each 10-year increment of age was associated with a two-fold increase in risk (hazards ratio 2.14, 95%CI 1.59-2.88). For a 5 kg increment in mass, there was a 27% decrease in odds ratio (odds ratio 0.77, 95%CI 0.68-0.87). One SD increase in serum potassium (0.84 mmol/l) was associated with a 63% decrease in risk (odds ratio 0.37, 95%CI 0.27-0.50; p < 0.0001). DISCUSSION: Hyponatraemia is a common problem after thiazide therapy. Extra caution and close monitoring are warranted when prescribing thiazides for elderly patients with low body mass.

Extrapolation of reciprocal creatinine plot is not reliable in predicting the onset of dialysis in patients with progressive renal insufficiency.

BACKGROUND: Reciprocal creatinine plot is often used to monitor patients with progressive renal insufficiency and to predict the onset of dialysis, although the latter practice has not been validated. OBJECTIVE: We examined whether extrapolating the reciprocal creatinine plot can predict the onset of dialysis. SETTING: Single centre study in the dialysis unit of a University teaching hospital. DESIGN: We studied 170 consecutive patients with progressive renal insufficiency referred to a single nephrology unit and subsequently dialysed. Reciprocal creatinine plot was constructed by all available serum creatinine values before dialysis (the 'definitive plot'). Four 'interim plots' were constructed for each patient by using serum creatinine below 400, 500, 600 and 700 micromol L(-1). Interim plots with at least five points and Pearson's r > 0.9 were analysed. The date of dialysis was predicted from the least squares linear regression formula and a target serum creatinine level cor- responding to estimated creatinine clearance of 7 mL min-1, at which dialysis was recommended. RESULTS: The median duration of observation was 25 months. After serum creatinine 500 micromol L(-1), the slope of the interim plot remained stable and extrapolation was possible in 117 patients (68.8%). However, the limits of agreement for predicting the onset of dialysis were wide (from -11.7 to +9.5 months). At this creatinine level, the onset of dialysis fell within 1 month of the predicted onset in only 41 patients (24.1%). The limits of agreement for prediction narrowed when time points of higher serum creatinine were included into the plot. However, nine patients (5.3%) required dialysis within 1 month at creatinine 600 micromol L(-1) and the dialysis was not predicted by the reciprocal creatinine plot. Target serum creatinine did not correlate with acute serum creatinine at which dialysis was started (r = 0.051, P = 0.51). A slower decline in renal function was associated with a higher prediction error (r = 0.212, P = 0.014). CONCLUSIONS: The onset of dialysis cannot be predicted by extrapolation of the reciprocal creatinine plot because of individual variation in the renal function that require dialysis. Dialysis would be almost imminent in some patients by the time serum creatinine reaches a level that allows accurate construction and extrapolation of a plot.