Proton pump inhibitor use and progression to major adverse renal events: a competing risk analysis.

BACKGROUND: Proton pump inhibitors (PPIs) are associated with acute tubulointerstitial nephritis and there are reports associating their use with the development of chronic kidney disease (CKD). AIM: To determine if PPI use is associated with major adverse renal events (MARE) in patients with CKD. DESIGN: Observational cohort study comprising patients with CKD attending secondary care renal clinics from 1 January 2006 until 31 December 2016. METHODS: We collated baseline clinical, socio-demographic and biochemical data at start of PPI (PPI group) or study inception (control group). MARE was considered a composite of doubling of creatinine or end-stage renal disease. Association between PPI exposure and progression to MARE was assessed by cause-specific hazards competing risk survival analysis. RESULTS: There were 3824 patients with CKD included in the analyses of whom 1195 were prescribed a PPI. The PPI group was younger (64.8 vs. 67.0 years, P < 0.001), with lower estimated glomerular filtration rate (eGFR) (30 vs. 35 ml/min, P < 0.001) and more proteinuria (64 vs. 48 mg/mmol, P < 0.001). PPI use was associated with progression to MARE on multivariable adjustment (hazard ratio 1.13 [95% confidence interval 1.02-1.25], P = 0.021). Other factors significantly associated with progression to MARE were higher systolic blood pressure, lower eGFR, greater proteinuria, congestive cardiac failure and diabetes. Hypomagnesaemia was more common in the PPI group (39.5 vs. 18.9%, P < 0.001). CONCLUSION: PPI use was associated with progression to MARE, but not death in patients with CKD after adjusting for factors known to predict declining renal function, including lower eGFR, proteinuria and comorbidities. A prospective cohort study is required to validate these findings.

Collection and determinants of patient reported outcome measures in haemodialysis patients in Scotland.

BACKGROUND/INTRODUCTION: Patient reported outcome measures (PROMs) can evaluate the quality of health in patients with established renal failure. There is limited experience of their use within national renal registries. AIM: To describe the Scottish Renal Registry's (SRR) experience of collecting PROMS in the haemodialysis population and correlate PROMS to demographic and clinical parameters. DESIGN: Retrospective observational cross-sectional study. METHODS: Haemodialysis patients in Scotland were invited to complete the KDQOL™-36 questionnaire on the day of the annual SRR census in 2015 and 2016. Questionnaires were linked to census demographic and clinical variables. RESULTS: In 2016, 738 questionnaires were linked to census data (39% of prevalent haemodialysis population). Response rates differed with age (≥ 65 years 42%, < 65 years 36%) [χ2P = 0.006]; duration of renal replacement therapy (<1 year 46%, ≥1 < 5 years 38%, ≥ 5 years 33%) [χ2P = 0.002] and social class (Scottish Index of Multiple Deprivation (SIMD) Class 1 32%, Class 2 41%, Class 3 40%, Class 4 48%, Class 5 40%) [χ2P < 0.001]. There were significant differences in PROMs with age, SIMD quintile and primary renal diagnosis. Achieving a urea reduction ratio of >65% and dialysing through arteriovenous access were associated with significantly higher PROMs. PROMs were not affected by haemoglobin or phosphate concentration. DISCUSSION/CONCLUSIONS: Routine collection of PROMs is feasible and can identify potentially under-recognized and treatable determinants to quality of life. The association between attaining recommended standards of care and improved PROMs is striking. Individual and population-wide strategies are required to improve PROMs.

End-stage renal disease and survival in people with diabetes: a national database linkage study.

BACKGROUND: Increasing prevalence of diabetes worldwide is projected to lead to an increase in patients with end-stage renal disease (ESRD) requiring renal replacement therapy (RRT). AIM: To provide contemporary estimates of the prevalence of ESRD and requirement for RRT among people with diabetes in a nationwide study and to report associated survival. METHODS: Data were extracted and linked from three national databases: Scottish Renal Registry, Scottish Care Initiative-Diabetes Collaboration and National Records of Scotland death data. Survival analyses were modelled with Cox regression. RESULTS: Point prevalence of chronic kidney disease (CKD)5 in 2008 was 1.63% of 19 414 people with type 1 diabetes (T1DM) compared with 0.58% of 167 871 people with type 2 diabetes (T2DM) (odds ratio for DM type 0.97, P = 0.77, on adjustment for duration. Although 83% of those with T1DM and CKD5 and 61% of those with T2DM and CKD5 were receiving RRT, there was no difference when adjusted for age, sex and DM duration (odds ratio for DM type 0.83, P = 0.432). Diabetic nephropathy was the primary renal diagnosis in 91% of people with T1DM and 58% of people with T2DM on RRT. Median survival time from initiation of RRT was 3.84 years (95% CI 2.77, 4.62) in T1DM and 2.16 years (95% CI: 1.92, 2.38) in T2DM. CONCLUSION: Considerable numbers of patients with diabetes continue to progress to CKD5 and RRT. Almost half of all RRT cases in T2DM are considered to be due to conditions other than diabetic nephropathy. Median survival time for people with diabetes from initiation of RRT remains poor. These prevalence data are important for future resource planning.

The delivery of renal replacement therapy in Scotland--why the geographic variation?

BACKGROUND: The incidence of patients starting renal replacement therapy (RRT) for established renal failure (ERF) in Scotland has fallen from 2005 to 2011 due to a reduction in older patients starting RRT; there are significant differences between NHS Health board areas. AIM: To understand the apparent inequality in provision of RRT between NHS board areas in Scotland. DESIGN: Retrospective population analysis of Scottish renal registry (SRR) data, population statistics and quality outcomes framework summary statistics. RESULTS: The incidence of patients starting RRT for ERF in Scotland fell from 123 per million population (pmp) in 2005 to 96 pmp in 2011. The incidence of ≥75 year olds fell from 406 to 274 pmp. There are significant differences between NHS board areas when standardized for age and social deprivation. There is no relationship between the population prevalence of CKD as reported by QOF and the incidence of RRT for ERF. Those areas with high incidence rates of ≥75 year olds have higher 90-day [Spearman's rank correlation: coefficient = 0.662; P = 0.03] and 1-year [Spearman's rank correlation: coefficient = 0.776; P = 0.003] mortality rates. CONCLUSION: The significant variation in provision of RRT for ERF between Scottish NHS Board areas is not explained by age or social deprivation. There is evidence of change in practice towards RRT for patients aged ≥75 years but variation between NHS Board areas. This disparity must be further investigated to ensure equity of access to RRT for those who will benefit from it, and to non-dialytic care for those who would not.

Stratifying risk in chronic kidney disease: an observational study of UK guidelines for measuring total proteinuria and albuminuria.

BACKGROUND: Proteinuria predicts poor renal and cardiovascular outcomes. Some guidelines recommend measuring proteinuria using albumin:creatinine ratio (ACR), while others recommend total protein:creatinine ratio (TPCR). AIM: To compare renal outcomes and mortality in the populations identified by these different recommendations. DESIGN: Retrospective longitudinal cohort study. METHODS: Baseline ACR and TPCR measurements were obtained from 5586 patients with chronic kidney disease (CKD) attending a Scottish hospital nephrology clinic. The cohort was divided into three groups with concordant results by ACR and TPCR (no proteinuria; low proteinuria; significant proteinuria) and one group with discordant results (significant proteinuria with TPCR, but not ACR). Outcomes were assessed using Kaplan-Meier plots and Cox proportional hazards models. RESULTS: Median follow-up was 3.5 years [interquartile range (IQR) 2.1-6.0]; 844 (15%) died at 3.0 years (IQR 1.8-4.7) and 468 (8%) started renal replacement therapy (RRT) at 1.7 years (IQR 0.6-3.4). Proteinuria was associated with a substantially increased risk of RRT and death. Patients with significant proteinuria by TPCR, but not ACR (n = 231) had high renal risk, and the highest all-cause mortality (log-rank P < 0.001). With multivariate analysis the risk fell below those with significant proteinuria with concordant results by ACR and TPCR but remained considerably higher than those without significant proteinuria. CONCLUSION: Proteinuria screening with TPCR identifies an additional 16% of patients with significant proteinuria, not identified using ACR. This subgroup has high renal risk, and high risk of all-cause mortality and therefore warrant identification. Guideline recommendations on proteinuria screening in CKD should be reconsidered.

The prevalence of chronic kidney disease in rheumatology outpatients.

BACKGROUND: The introduction of routine reporting of estimated glomerular filtration rate coupled with a new definition of chronic kidney disease (CKD) has led to an unprecedented focus on kidney disease in many patient groups. In light of this, we performed an audit of patients attending the rheumatology clinics to assess the prevalence of CKD in this population. METHODS: Over a four week period, we reviewed the renal function of all patients attending the rheumatology clinics and day ward at our hospital (n=351). Renal function was assessed using the 4-variable MDRD formula. We then interviewed those patients with estimated glomerular filtration rate (eGFR) of 59 ml/min or lower. RESULTS: We found a prevalence rate of 18% for stage 3 CKD or lower in our audit population. Surprisingly, 60.3% of patients in this category were not aware of any problems with their kidneys (n=38). CONCLUSIONS: The prevalence rate of 18% for stage 3 CKD or lower is significantly higher than the five per cent reported within the general population. As a result of this audit, we now plan to ensure that these patients undergo measurement of blood pressure, eGFR, and urinalysis on a six to twelve monthly basis.

Gastric carcinoid: germline and somatic mutation of the neurofibromatosis type 1 gene.

Neurofibromatosis type 1 (NF1) is one of the most common autosomal dominantly inherited conditions. A range of complications has been described, including gastrointestinal manifestations. Gastric carcinoid tumours are associated with multiple endocrine neoplasia, atrophic gastritis and pernicious anaemia but have not been reported in NF1 in the absence of other predisposing factors. We report the occurrence and investigation of a gastric carcinoid tumour in a 23-year-old woman with previously uncomplicated NF1. Analysis of the tumour tissue revealed loss of heterozygosity at the NF1 gene locus but a normal karyotype and an absence of microsatellite instability. A germline NF1 gene nonsense mutation in exon 37 was detected by denaturing high-performance liquid chromatography and DNA sequence analysis. This is the first reported occurrence of a gastric carcinoid tumour in a patient with NF1 in the absence of other predisposing factors such as pernicious anaemia. The analyses indicate that the carcinoid arose through NF1 gene inactivation but in the absence of an inherited NF1 gene microdeletion. This case adds to the range of gastrointestinal tumours that may be encountered in patients with NF1, particularly in those who present with upper gastrointestinal haemorrhage.

Assessing the utility of the stop dialysate flow method in patients receiving haemodiafiltration.

BACKGROUND: The stop dialysate flow (SDF) method of post-dialysis urea sampling is the most commonly used method in the UK. It can also be used with a published formula to predict 30 min equilibrated urea accurately. The method has not been validated in patients undergoing haemodiafiltration (HDF). Given the increased use of HDF across Europe, we felt it prudent to assess the utility of the SDF method and prediction equation in this modality. METHODS: Fourteen patients from two renal units were studied. Blood samples were taken at 1 min intervals from the arterial side of the dialysis circuit in the first 5 min after HDF had ceased whilst blood circulation continued. A peripheral sample was taken from the contralateral arm immediately after HDF had ceased and a 30 min sample was taken from the arterial needle. These samples were used to assess the utility of 5 min arterial blood urea and the 30 min prediction formula, respectively. RESULTS: Blood urea measured from the arterial circuit at 5 min correlated closely with the contralateral sample taken immediately post-HDF, with no significant difference (6.45+/-2.11 vs 6.52+/-2.19 mmol/l, P = 0.39). The use of 5 min arterial blood urea and prediction formula allowed an accurate prediction of 30 min urea (R2 = 0.96). CONCLUSIONS: The use of the SDF method with a 5 min post-HDF arterial sample is valid in patients receiving HDF. The previously published prediction formula for estimating 30 min urea is also valid using the 5 min post-HDF sample.