Site blog

Anyone in the world


Changes in Albuminuria Predict Mortality and Morbidity in Patients with Vascular Disease. Roland E. Schmieder et al. on behalf of the ONTARGET Investigators  (JASN July 2011) report a further analysis of the ONTARGET study related to changes in albuminuria and its impact on cardiovascular outcomes and death. You may recall the ONTARGET study showed little benefit of combining an ACE inhibitor (Ramipril) and ARB (Telmisartan) on CVD in older and high risk patients. Moreover, ONTARGET showed that those treated with a combination of Ramipril and Telmisartan had a faster rate of CKD decline in spite of more significant reduction in albuminuria. 
The current analysis by the ONTARGET study group shows that changes over 2 years in albuminuria impacts on CVD outcomes; decreased albuminuria being associated with improved outcomes and increased albumin excretion rate having negative impact. These changes are not entirely independent from changes in blood pressure levels. The changes in albuminuria had less of an impact on renal outcomes. This is an interesting observation albeit at some variance with the primary report of ONTARGET implying that the study negative outcomes were associated with decreased proteinuria. 
The most likely explanation remains that:
1. This is a study of older atherosclerotic high CVD risk patients.
2. These patients are at highest risk of renal harm from combining ACEi + ARB; faster decline of kidney function with the inherent increased CVD risk! 
3. ACEi + ARB may be beneficial on CVD outcomes in a subgroup of patients whose albuminuria also happens to decrease; concern about subgroup analyses and power as well as validity of such analysis.
4.  The price of CVD and albuminuria improved outcomes may still be worsening CKD! is it a price worth paying knowing that worsening CKD increases by itself CVD risk!?
5. Changes in albuminuria may be independent from treatment and merely reflecting the progression or otherwise of underlying atherosclerotic vascular disease. Albuminuria is a reliable marker of CVD risk and mortality. The question that is always worth asking is: Do changes in Albuminuria ADD to other more established CVD outcomes predictors such as the Framingham Risk Score. My guess is probably NOT!
[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Anyone in the world

Rosuvastatin in Diabetic Hemodialysis Patients. Hallvard Holdaas et al. , on behalf of the AURORA study group report, in JASN July 2011 issue, a posthoc analysis of the AURORA study on Rosuvastatin in ESRD patients treated by HD. The AURORA study was a negative one as Rosuvastatin failed to improve survival and related endpoints. In this posthoc analysis the AURORA study group focuses on the results of a subgroup analysis of the diabetic patients who received Rosuvastatin in this study. They claim that it "might" improves cardiac events outcomes.

Reading this report a number of comments come to mind:
1. Posthoc analysis is the way investigators or their sponsors...desperate to show that a negative result is not a negative result keep searching subgroupos for a glimmer of hope; a positive result. So if the study is negative, look again...and again...until you find a positive result.
2. The Nephrology community seems reluctant to accept that a negative trial outcome is a result as valid as a postive outcome?! Prehaps, the pharmaceutical sponsors are not as willing to accept a negative result that would seriously jeopardise their sales of the compound/drug under investigation...
3. Subgroup analysis is not statistically valid as it is by definition underpower with an inadequate sample size. AURORA required 2776 patients to show a result (Negative); how can a study with 731 be valid in analysing the same endpoints?! I guess hence the: conclusion that Rosuvastatin "might" reduce the risk of cardiac events.
4. The posthoc analysis for what it is worth also shwos that Rosuvastatin can cause considerable HARM in diabetic patients on HD: A 5 fold increase risk of hemorrhagic stroke!!!! hence my Blog title that Rosuvastatin "might" be harmful in diabetic haemodialysis patients. No matter how it is looked at there is little doubt that in most if not all statin studeis they increase risk of hemorrhagic stroke!
My conclusion from reading this analysis: 
"Better to accept that a study is negative than go on and on and on....trying to prove that it is positive!"
[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Anyone in the world

In the Lancet, this week publication of the results of two major and interesting studies on the control of DM and the impact on CV complications.

The first by Lind and colleagues in 20,985 patients with T1DM showing that tighter DM control (lowe HbA1c) led to a significant decreased incidence of new heart failure; patients with HbA1c of less than 6.5% had four times less heart failure over 9 years compared to those with HbA1c of at least 10.5%. The authors conclude that poor glycemic control promotes heart failure in T1DM. Whilst this study gives support to tight glycemia ocntrol in T1DM, it does not argue for a similar reduction in glycemia in T2DM. In fact, the relationship between glycemia control and moratlity assumes a U-shaped curve with high mortality at Low and High HbA1c levels; (Opie et al. Heart 2011;97:6-14).

The second study published in the Lancet this month by Griffin and colleagues reports data from ADDITION-Europe study of early and multifactorial management of T2DM patients versus standard care in a primary care setting. The intensive treatment group had better reduction of BP and LDL cholesterol. Overall, over ~5 years follow-up there was a non-significant reduction in the risk of composite cardiovascular endpoints in the intensive multifactorial group compared to routine care. Small differences in glycemia, BP and LDL cholesterol levels in favour of the multifactorial intensive treatment group may have contributed to the trend towards improved outcome in this group although it did not reach statistical significance. Of note statin intake was comparable in the two groups.

 Unfortunately, this study has a number of limitations including:

1. Insufficient number of patients and most likely an inadequate power due to the lower than expected rate of complications in the control group.

2. Event rates in the routine group were much lower than expected due to overall improved care of people with DM. This may be explained by improved BP and dyslipidemia in the routine care group based on the recommendations of national and international guidelines .Data from a number of studies imply that BP and lipids control are the most important to reduce CVD risk in DM2.

3. The small difference between the groups in Glycemia, BP and lipids control.

Overall, these two studies argue for more efforts in the management of glycemia (T1DM) but also agressive reduction in BP and control of dyslipidemia (T2DM). ADDITION-Europe shows that such an approach apears to reduce overall events rate making further distinction between routine an dintensive care difficult.

Perhaps, good metabolic and BP control in T2DM is good enough...!!!!  


[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Picture of Meguid El Nahas
by Meguid El Nahas - Monday, 4 July 2011, 4:35 PM
Anyone in the world

Gale and colleagues from London and Cyprus reported in 2010 the association between Complement factor H-Related protein 5 (CFHR-5) nephropathy and endemic CKD in Cyprus. They identified a novel mutation of the CFHR5 gene in cypriot families causing an autosomal form of isolated mesangial C3 deposition and a clinical picture very similar to mesangial IgA nephropathy (Lancet 2010). In the June issue of cJASN Athanasiou and colleagues describe the clinical entity associated with the CFHR5 mutation. The majority of patients prsent with isolated microscopic hematuria with around 38% also shoiwng proteinuria. Impaired kidney function an dprogressive CKD is more likely to develop in older patients (>50 years) where 80% of men  and 21% of women developed CKD. 20% developed ESRD; mostly men. Of interest, almost all male patients  with CFHR5 nephropathy who progressed to ESRD had recurrent episodes of macroscopic hematuria during childhood associated with URTI.

This disease shows that dysregulation of the alternative complement pathway following a mutation of CFHR5 leads to isolated C3 mesangial deposition and a clinical phenotype similar to mesangial IgA nephropathy. 

[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Anyone in the world

A small RCT is reported in the June issue of cJASN relating to technique survival in peritoneal dialysis patients  with diabetic nephropathy comparing standard glucose solutions (4x2L/24h) and (3x2L +) Icodextrin dwell (Takatori et al. Okayama, Japan).

The authors conclude that over 24months technique survival on CAPD was better in the group receiving an additional Icodextrin dwell. This was related primarily to an improved fluid balance control and a lower switch to HD due to fluid overload. There was no difference between the two groups in dialysis adequacy; KT/V and D/P Creatinine. There was no impact on residual renal function.

This is a small RCT with 21 and 20 patient in the two groups. Such a small number always raises concern over the possibility of the small sample size and inadequate power may lead to statistical bias and type1 error (False positive result). Also, it appears as if the analysis was per protocol (PP) confined to a small number of patients and not as intention to treat (ITT) including all those who were randomised.

Finally, the study confirms the obvious that icodextrin has a more sustained osmotic effect and improves fluid balance by mobilising more fluid. As patients with diabetic nephropathy have a tendency to fluid retention, the result is not suprising. Also, no data is given on the treatment modality of patients based on their transporter status; high transporters would have been better off on APD.  

A more logical approach to these patients would have been:

1. Define patients' transporter status.

2. Treat High and high average groups by APD

3. Treat the low transporters by CAPD

4. Use standard Glucose versus glucose + icodextrin solutions and evaluate outcomes

A much larger number of patients would be warranted for this type of RCT. 

In the meanwhile, nephrology journals continue to publish small and meaningless clinical trials?!

[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Anyone in the world

Progressive kidney disease remains a key clinical problem facing nephrologists all over the world. Despite the plethora of in vivo studies suggesting new therapies there has been an abject failure to translate these findings into meaningful clinical studies. This is partly because pharma have been reluctant to take on CKD as a therapeutic area given the inherent difficulties and expense of conducting a trial in this area.  So Warnock and colleagues should be congratulated on completing a phase 2, randomized control trial assessing the effects of Bardoxolone Methyl (an oral anti-inflammatory and anti-oxidant) on progression of kidney disease in Type 2 Diabetes, which were presented this week at the EDTA and have now been published online in the New England Journal of Medicine (see here).

Bardoxolone methyl is an antioxidant inflammation modulator that activates the Keap-Nrf2 pathway, which is a key regulator of the body’s natural antioxidant, anti-inflammatory response to injury. Oxidative stress and inflammation are thought to be key drivers of the fibrotic process in CKD inducing structural changes within the glomerulus.  In vivo data indicates that the Keap-Nrf2 pathway is suppressed in models of kidney fibrosis whilst there is a simultaneous increase in oxidative stress and inflammation. Thus the hypothesis underpinning the trial was that Bardoloxone Methyl could be protective in progressive kidney disease by inducing an antioxidant, anti-inflammatory response that would inhibit proinflammatory transcription factors such as nuclear factor κB

The BEAM study involved assigning 227 adults with type 2 diabetes and an eGFR of 20 to 45 mL/min per 1.73 m2 of body surface area to 1 of 4 groups: bardoxolone 25 mg, 75 mg, or 150 mg once daily, or placebo. Virtually all patients were on ACE-I or ARBs. At 24 weeks, there was significant improvement in eGFR in all bardoxolone methyl groups, as compared with the placebo group, with mean differences per minute per 1.73 m2 of 8.2±1.5 ml in the 25-mg group, 11.4±1.5 ml in the 75-mg group, and 10.4±1.5 ml in the 150-mg group (P<0.001 for all comparisons). The improvement in eGFR occurred within 4 weeks, peaked at 12 weeks and was maintained at 24 and 52 weeks. Bardoxolone Methyl therapy was associated with lower blood urea, uric acid, serum phosphorous and magnesium concentrations. It was generally well tolerated though muscle spasms and elevations in alanine aminotransferase were an issue. There was a slight, but significant increase in albuminuria with Bardoxolone Methyl.

Whilst the results of this trial are promising a number of issues arise from the study:

  1. The improvement in serum creatinine may simply reflect increased tubular excretion of creatinine rather than improvement in kidney function per se. The authors acknowledge that future trials need to formally measure GFR rather than just use the serum creatinine – though previous studies suggest that Bardoxolone Methyl does not impact on creatinine production or excretion. The changes in magnesium concentration suggest tubular toxicity may be an issue however the improvement in serum urea would suggest that the improvement in serum creatinine truly reflects improvements in glomerular filtration. This is a key issue for all trials in progressive kidney disease – how to measure kidney function and clearly the serum creatinine has significant limitations. 
  2. The mechanism of action of improvement in eGFR is not clear – the fact that eGFR improved within 4 weeks suggest that the observed effects cannot have been due changes in glomerular structure but rather actions on glomerular inflammation and haemodynamics. No renal biopsies were performed to address this question however the data suggests eGFR is not determined solely by glomerular structure.
  3. There was no decline in kidney function in the placebo group (perhaps reflecting low levels of albuminuria) and its difficult to interpret the clinical significance of these results. A double-blind, placebo-controlled Phase 3 outcome study (BEACON) with cardiovascular disease and dialysis as key endpoints will help address this.
[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Anyone in the world

Recent data from the "ESPN/ERA-EDTA Regisrty" demonstrate improved survival of the renal function as well as better patient survival in a large European cohort of nephropathic cystinosis (NC) patients (145 patients from 18 European countries) over the last two decades. NC is the most common inherited cause of renal Fanconi syndrome.

Nephropathic cystinosis (NC) (MIM 219800) is the most common inherited cause of renal Fanconi syndrome. It is a rare autosomal recessive disorder (incidence in 1:100,000– 200,000 live births), which presents with renal proximal tubular dysfunction (Fanconi syndrome) during infancy and progresses to end-stage renal failure during the first decade of life if not treated with cysteamine. Over time, patients with NC also present with dysfunction of several organs, including eyes, thyroid, liver, pancreas, male gonad, muscles, and brain. NC is caused by mutations in the CTNS gene (maps to chromosome 17p13) that encodes for a cystine transmembrane transporter, cystinosin. Lack of cystinosin activity causes cystine accumulation and results in formation of intralysosomal cystine crystals [Gahl et al, 2002].

The Study Group, ESPN/ERA-EDTA Registry analyzed data available on a total of 245 NC patients from 18 European countries collected in the ESPN/ERA-EDTA registry. NC patients on renal replacement therapy (RRT) were matched to comparable RRT children. Between 1979 and 2008, the mean age at start of RRT among NC children gradually increased from 8.8 years to 12.7 years, while this was not observed in other RRT children. This effect can be probably attributed to a wide use of cysteamine therapy in Europe starting from the early 1990s. Five-year survival after start of RRT improved in NC patients from 86.1 percent (started before 1990) to 100 percent (after 2000) as compared to control population (89.6 and 94.0 %, respectively). NC on dialysis were less hypertensive than children matched on age, country, and dialysis modality, and had lower, close to normal PTH levels. Whilst their height at start of RRT slightly improved during the past decade, they were still significantly shorter than other children at the start of RRT.

NC is commonly overlookedand should be suspected in all patients with failure to thrive and signs of renal Fanconi syndrome, as it is the most common cause of inherited renal Fanconi syndrome in children. Despite the awareness of the medical community, the diagnosis of cystinosis is frequently delayed, because of the initial incomplete presentation and the rarity of the disorder [Soliman et al. 2009; Wilmer et al, 2011]. Timely diagnosis and treatment is crucial since early treatment results in more favorable prognosis. It is estimated that for every month of treatment prior to 3 years of age, 14 months' worth of later renal functions were preserved [Gahl, 2003]. Because systemic cysteamine treatment has no effect on corneal cystine crystals, topical 0.5 % cysteamine eye drops are indicated [Gahl et al, 2000].

The true diagnosis of cystinosis rests on the measurement of elevated cystine content in blood cells, slit lamp examination for corneal cystine crystals (late, after 2 years of age), and molecular analysis of the CTNS gene. Cysteamine is currently the only available treatment interfering with the disease pathogenesis; however, novel potential therapeutic modalities, syngeneic bone marrow and hematopoietic stem cell transplantation [Syres et al, 2009], may become available in the future.

Do you have cystinosis patients under your care?
Do they have access to oral cysteamine/cysteamine eye drops?


  2. Gahl WA, Thoene JG, Schneider JA. Cystinosis. N Engl J Med 2002;347:111–121
  3. Soliman NA, El-Baroudy R, Rizk A, Bazaraa H, Younan A. Nephropathic cystinosis in children: An overlooked disease. Saudi J Kidney Dis Transpl 2009;20:436-42
  4. Wilmer MJ, Schoeber JP, van den Heuvel LP, Levetchenko EN. Cystinosis: practical tools for diagnosis and treatment .Pediatr Nephrol 2011;26:205–215
  5. Gahl WA. Early oral cysteamine therapy for nephropathic cystinosis. Eur J Pediatr 2003;162 Suppl 1:S38-41
  6. Gahl WA, Kuehl EM, Iwata F, Lindblad A, Kaiser-Kupfer MI. Corneal crystals in nephropathic cystinosis: natural history and treatment with cysteamine eyedrops. Mol Genet Metab 2000;71:100–120
  7. Syres K, Harrison F, Tadlock M, Jester JV, Simpson J, Roy S, Salomon DR, Cherqui S (2009) Successful treatment of the murine model of cystinosis using bone marrow cell transplantation. Blood 114:2542–2552


[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Anyone in the world

Santín et al (CJASN, May 2011) compiled the clinical and genetic data of eight podocyte genes [NPHS1,NPHS2, TRPC6, CD2AP, PLCE1, INF2, WT1 (exons 8 and 9), and ACTN4 (exons 1 to 10)] analyzed in their large Spanish cohort of patients with NS (125 patients belonging to 110 familiesranging from congenital to adult onset) and provided a seemingly practical guideline for genetic testing in SRNS.

Authors found out that 34% (37/110) of SRNS patients could be explained by mutations in one of these genes, of whom 67% (16/24) were familial cases and 25% (21/86) were sporadic. They found disease-causing mutations in 100% of congenital onset cases and in 57% of infantile-onset cases, which is higher than previously described. Interestingly, in childhood onset they found a lower rate of mutations than previously, but in adolescent and adult-onset patients, a higher than expected rate was observed. In conclusion, they indicate that two main criteria that determine the appropriate genes to test: (1) age at onset and (2) familial/sporadic status.

Although the increasing knowledge of the molecular basis of NS represents a milestone in nephrology, nevertheless it also adds greater complexity to clinical nephrologists’ decisions with as to when and which genetic tests should be indicated.

In medical practice, the genetic heterogeneity of  SRNS, the significant phenotypic variability, and the lack of knowledge of the relative frequency of mutations in these genes in pediatric and adult patients with NS hinder the routine genetic analysis for SRNS. There are only a few studies searching for mutations in several of these genes and only in pediatric patients.

Multicenter collaborative research studies including larger numbers of SRNS pediatric and adult patients from various ethnic groups might help to answer the question as to when and which genetic tests should be indicated particularly in adult onset SRNS.

It still remains unclear if unraveling the underlying genetic mutation in adult onset SRNS patient would have its clinical implication on the individually tailored management plan (cost effectiveness/clinical validity):

1.     Would you order genetic testing for your SRNS patient, assuming it is available, in the first place?

        2.     Would you avoid immunosuppressive treatment in  documented genetic SRNS?

[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Picture of Meguid El Nahas
by Meguid El Nahas - Friday, 24 June 2011, 5:46 PM
Anyone in the world


Nephron Digest June 2011

IgA-Dominant Postinfectious Glomerulonephritis: A New Twist on an Old Disease (Samih H. Nasr, Rochester, Minn. and Vivette D. D’Agati, New York, N.Y.;Nephron Clin Pract 2011;119:c18-c26). The review by Nasr and D’Agati draws the reader’s attention to the changing nature of postinfectious glomerulonephritis. IgA-dominant postinfectious glomerulonephritis is increasingly being recognized. The article highlights the fact that this form of acute glomerulonephritis is more likely to occur in the elderly, the immunocompromised and those with diabetes. It also stresses the association with staphylococcal infections and the range of histological changes including the mesangioproliferative form associated with IgA deposits. In developing countries, including the Far East, where IgA nephropathy is quite common, it is important to distinguish between postinfectious IgA nephropathy and an acceleration of idiopathic IgA nephropathy or a superimposed infection. In addition to hypocomplementemia and subepithelial immune deposits characteristic of postinfectious glomerulonephritis, patients with postinfectious IgA-dominant glomerulonephritis tend to present frequently with AKI when compared to those with idiopathic IgA nephropathy. The poor prognosis and lower recovery rate associated with IgA-dominant postinfectious glomerulonephritis is attributed to the associated comorbidities. Nephrologists should be aware of this entity and include it in the differential diagnosis of elderly patients presenting with AKI.


The ‘Centre Effect’ in Nephrology: What Do Differences between Nephrology Centres Tell Us about Clinical Performance in Patient Management?(Hodsman and colleagues, (Bristol and Southampton; Nephron Clin Pract 2011;119:c10-c17). This review by Hodsman and her colleagues from the UKRR Centre in Bristol examines an important and timely issue, that of the 'centre effect' in healthcare delivery. Increasingly, commissioners of healthcare assess the quality of care delivery by drawing comparisons to national and international 'standards' and link the care delivered to the resources allocated. National/international comparisons are not always applicable in view of a number of variables that impact on a given healthcare centre. The review discusses the different factors affecting healthcare delivery including the patients' case mix and organisational characteristics ('compositional effect'). Moreover, geographical and sociodemographic factors may affect a centre performance, including poverty with associated poor compliance and difficulties in accessing available services. The authors discuss how the gap and linkage between centre provisions and patients' outcome can be studied and closed. Focusing on a multi-variable quality performance, rather than on individual parameters/processes, may ultimately provide a more holistic and successful approach to quality and outcomes.



Parallel Deterioration of Albuminuria, Arterial Stiffness and Left Ventricular Mass in Essential Hypertension: Integrating Target Organ Damage (E. Andrikou and colleagues, (Athens; Nephron Clin Pract 2011;119:c27-c34). Andrikou and colleagues report on 428 non-diabetic untreated hypertensives' ACR (urine albumin:creatinine ratio) and c-f PWV (pulse wave velocity, a measure of arterial stiffness). Age, male sex, 24-hour systolic BP, ACR and c-f PWV were independent predictors of left ventricular hypertrophy (LVMI). Increased ACR in conjunction with pronounced arterial stiffness is accompanied by augmented LV mass and higher LVH rates. The close interrelationships between albuminuria, c-f PWV and LVMI suggest parallel target organ damage progression. This is not surprising since untreated hypertension is likely to have a detrimental effect on a number of end-organs including arteries, heart and kidneys. I have put forward the Cardio-Kidney-damage (C-K-D) concept a few years ago to highlight such a close link between hypertension, systemic vascular damage and albuminuria in the community. It is often said that patients with CKD have higher CVD risk, but this may simply reflect that patients with CKD in the community already suffer from underlying CVD and therefore would naturally have higher CVD mortality. Of interest, the prognosis of these patients has recently been shown to be predictable by conventional CVD risk scores (like the Framingham Risk Score), with little added value of adding albuminuria or GFR to the scoring system!



Prevalence of Chronic Kidney Disease and Associated Risk Factors, and Risk of End-Stage Renal Disease: Data from the PREVADIAB Study (J. Vinhas and colleagues, Coimbra; Nephron Clin Pract 2011;119:c35-c40). The authors analyzed data from a nationally representative sample of 5,167 subjects, and estimated the prevalence of CKD stages 3 to 5 to be around 6.1. The prevalence of risk factors such as diabetes (11.7%), obesity (33.7%), and metabolic syndrome (41.5%) was similar to that in the US, but greater than in most European countries. The risk of ESRD was greater than in other European countries, but lower than in the US. The authors conclude that the high incidence of ESRD among the Portuguese population is not due to a greater prevalence of CKD but may be associated with a higher rate of CKD progression. Caution in the interpretation of CKD prevalence in populations should be exerted when evaluating such reports as cross-sectional, single test analyses have been associated with over-reporting of CKD. Also, a high ESRD/CKD ratio may be due to faster progression but also to lower mortality as the majority of CKD patients do not reach ESRD, but instead die from CVD. It would be interesting to know if Portugal has a lower CKD3-4 mortality and also if current global efforts to reduce CVD mortality will lead to an increased burden of ESRD worldwide..?!



[ Modified: Thursday, 1 January 1970, 1:00 AM ]
Anyone in the world

A posthoc analysis by Mallamaci and colleagues (JASN June 2011 issue) of the REIN study suggests that CKD patients suffering from obesity are at higher risk of ESRD and that Ramipril can abolish this increased risk of renal progression and ESRD. This is of considerable interest, if confirmed, considering the increasingly prevalence of obesity and CKD. Also a growing number of publications has suggested a link between obesity and CKD and its progression. Also there is good evidence that the adipose tissue itself can generate excess of RAS, justifying the inhibition of this system in obese individuals with CKD.

However a few words of caution:

1. REIN was a partially blinded study with unblinding taking place during the course of the trial (week 27)!?

2. IN the overall population group REIN showed little difference in teh rate of CKD progression between placebo (0.26ml/min/month) and Ramipril (0.26ml/min/month)! It was only effective at reducing the number of ESRD hence teh benficial combined endpoint; difficult to understand why and how there was increased incidence of ESRD in control group although its rate of progression did not differ from the ramipril group? Starting RRT can be a subjective decision in an unblinded study.

3. Caution with the interpretation of posthoc and subgroup analyses as studies are not laways powered to address these points.




[ Modified: Thursday, 1 January 1970, 1:00 AM ]