The UK CKD Guidelines, published on the Renal Association website
Established renal failure (ERF) is relatively rare, but treatment with dialysis or transplantation is very expensive. The number of patients receiving renal replacement therapy (RRT) in the UK is rising rapidly and is unlikely to reach steady state for another 25 years [1], costing over 2% of the total NHS budget. These figures make any improvement in the cost-effective treatment of early kidney disease highly desirable.
Late referral of patients with ERF requiring RRT to specialist renal services is associated with significant cost and poor clinical outcomes. The great majority of patients starting RRT have progressed from earlier stages of chronic kidney disease (CKD), and most could therefore have been identified and referred earlier. Early CKD is common, however, and referral of all patients with early CKD would completely overwhelm existing specialist services. The great majority of patients with early CKD do not progress to established renal failure, but do have increased risks of cardiovascular disease. Optimal management of the risk factors for cardiovascular disease also reduces the risk of progression from early CKD to ERF. These guidelines were therefore developed to promote the optimal management of patients with CKD within the NHS, including the identification of those who would benefit from referral to specialist services.
These guidelines were instigated at the suggestion of the Joint Specialty Committee on Renal Disease of the Royal College of Physicians of London (RCPL) and the Renal Association (RA), and were developed jointly with the Royal College of General Practitioners (RCGP), the Association of Clinical Biochemists (ACB), the Society for District General Hospital Nephrologists (SDGHN), the British Geriatrics Society (BGS), the Professional Advisory Council of Diabetes UK, and the National Kidney Federation (NKF). The guidelines were developed in parallel with, but independently from, the consultation process that accompanied the development of advice to Ministers for part 2 of the National Service Framework (NSF) for Renal Services for England. Expenses for attending meetings, and accommodation for meetings, were met by the Department of Health for England. However, the guidelines are intended for use throughout the United Kingdom, as applies to the Renal Association Standards Document [2].
Recommendations were based wherever possible on existing systematic reviews of the relevant literature. We searched the Cochrane Renal Group and the NHS Centre for Reviews and Dissemination DARE (Database of Abstracts of Reviews of Effects) databases for reviews. Where relevant, individual members performed purposive literature searches in their areas of expertise, using Medline, and then narrative synthesis. We have not had the resources to perform systematic quality assessment and grading on the literature identified. As part of the NSF process, the NHS Centre for Reviews and Dissemination at the University of York undertook searching and narrative synthesis on several topics, and we were given access to the reports; these included screening, kidney stones; weight loss, smoking, and exercise and CKD; influence of lower blood pressure on rate of decline of kidney function; early referral; use of angiotensin converting enzyme inhibitors (ACEI) in those with normal kidney function; risk of progression amongst people with a single kidney; renal function following relief of obstructive nephropathy; nutrition, anaemia, and CKD; and renal bone disease. We have also drawn on existing evidence based guidelines both UK and international, notably those by the National Institute for Clinical Excellence (NICE) on management of Type 1 and Type 2 diabetes mellitus and of hypertension, NSFs for coronary heart disease, diabetes mellitus, older people, cancer and renal services, as well as the Kidney Disease Outcomes Quality Initiative (K/DOQI) series on CKD and the Caring for Australians with Renal Impairment (CARI) guidelines. The guidelines were designed to meet, as far as possible, the criteria suggested by the Appraisal of Guidelines Research and Evaluation Collaboration [3-4].
We recognise that whilst there is significant randomised evidence for many of the interventions used to manage CKD (such as the use of ACEIs in diabetic nephropathy), many of the questions posed by this guidance are about surveillance for, or referral of, kidney disease. These questions have not been addressed by randomised controlled trials. For the specific details of surveillance and referral we have had to rely largely on expert opinion/consensus.
The Renal NSF used the following framework to grade evidence in line with other NSFs:
Level 1: Meta-analyses, systematic reviews of randomised controlled trials, or randomised controlled trials.
Level 2: Systematic reviews of case-control or cohort studies, or case-control or cohort studies.
Level 3: Non-analytic studies, e.g. case reports, case series.
Level 4: Expert opinion (in the absence of any of the above). This includes the views and experiences of people with renal failure and their carers.
NICE subdivides level 1 and 2 into three sub-categories depending on the quality of the studies, a system which is most applicable to treatment interventions; we have not done this. Our recommendations about the diagnosis of kidney disease have largely been based on observational diagnostic accuracy (DA) studies in which the test under consideration is compared with a reference standard. NICE has suggested grading these from 1-4; 1-3 being primary studies or systematic reviews at different levels of quality of such diagnostic accuracy studies, with level 4 being similar to level 4 above. For simplification we have put diagnostic studies into the single overall NSF evidence grading framework above but as level 3 DA to distinguish them from non-analytic intervention studies, and without subdivision by quality.
Many of the recommendations in this document relate to aspects of the organisation, or system, of care, rather than to therapeutic decisions. There are major methodological problems associated with grading levels of evidence for this type of recommendation, and for this reason many of our recommendations are level 4. All recommendations made in this document are graded as level 4 evidence unless otherwise stated.
The membership of the committee was as follows:
Dr C Tomson (Chair) (RCPL/RA)
Professor R Bilous (Professional Advisory Council, Diabetes UK)
Dr S Blades (RCGP)
Dr R Burden (co-opted, Renal Association)
Dr J Cunningham (co-opted, Renal Association)
Dr J Dennis (RCGP)
Mr D Gilbert (Observer, Department of Health for England)
Dr E Lamb (ACB) (from May 2003)
Dr D Newman (ACB) (until March 2003)
Mr G Nicholas (NKF)
Dr S O’Riordan (BGS)
Dr P Roderick (Public Health observer from External Reference Group for NSF for Renal Services)
Dr P Stevens (SDGHN)
Dr J Vora (Professional Advisory Council, Diabetes UK)
Dr David Newman died in March 2003. He had contributed enormously to British nephrology, with many original research contributions as well as active input into the UK Renal Registry and to this Committee.
Meetings were held on 17th Oct 2002, 17th Dec 2002, 5th Feb 2003, 12th May 2003, 2nd September 2003, 24th March 2004, 21st June 2004, 21st September 2004, 15th November 2004, and 18th March 2005.
Each committee member was asked to record any potential conflicts of interest according to guidance published by the British Medical Journal [5].
Dr Tomson: member of advisory board of Genzyme UK between April 2002 and May 2004. Honoraria from Janssen Cilag, MSD, Bayer, Novartis, Baxter, Amgen, Genzyme, AstraZeneca, Pfizer, Roche. Current member of international steering committee of the Study of Heart and Renal Protection.
Dr Bilous: honoraria and some time member of advisory boards for Takeda, Glaxo Smith Kline, Astra Zeneca, Pfizer, Lilly, Roche, Novo Nordisk and Aventis.
Dr Blades: no conflicts of interest
Dr Burden: honoraria from Astra Zeneca, Fujisawa, Novartis, Novo Nordisk, and Roche
Dr Cunningham: Lecture fees, consultancy fees and research grant support from Amgen and Genzyme; Consultancy fees from Abbott; Lecture fees from Shire.
Dr Dennis: no conflicts of interest
Mr Gilbert (observer): no conflicts of interest
Dr Lamb: no conflicts of interest
Mr Nicholas: no conflicts of interest
Dr O’Riordan: no conflicts of interest
Dr Roderick: honoraria from Glaxo Smith Kline and from meetings sponsored by Orthobiotech and Amgen. Small scale research sponsored by Roche and Pfizer in the past.
Dr Stevens: member of advisory board of Hoffman La Roche from 2002 to date. Honoraria in the past from Ortho Biotech, Bayer, Amgen, Pfizer and Hoffman La Roche. Current research in the field of chronic kidney disease supported by funding from Roche UK.
Dr Vora: honoraria and some time member of advisory boards for Takeda, Glaxo Smith Kline, Astra Zeneca, Lilly, Novonordisk and Aventis
The final draft of the guidelines was circulated electronically to each of the bodies represented (RA, RCPL, RCGP, ACB, BGS, Diabetes UK, NKF) and to the Royal College of Pathologists; the British Hypertension Society; the Renal Information Group; the National Screening Committee; the Departments of Health of England, Northern Ireland, Scotland and Wales; the Scottish Intercollegiate Guidelines Network (SIGN); and the National Coordinating Centre for Chronic Conditions; with the request that they be given as wide a circulation as possible, and inviting comments to be sent direct to the Chairman. Notices drawing attention to the draft guidelines were placed on the websites of the participating organisations. Individuals who commented on earlier drafts and those known to have an interest in referral or management of CKD were also invited to comment directly. A final meeting/teleconference was then held to discuss this feedback and consider revision by consensus. The agreed version was submitted for endorsement by the Joint Specialty Committee on Renal Disease of the RCPL and the RA, and the RCGP.
Evidence-based guidelines for management of CKD have been developed in other countries, notably the USA [6], Australia [7], and Canada [8]; and European guidelines for haemodialysis include guidance on referral for patients with CKD [9]. Many of these guidelines are directly applicable to the UK. We have, for instance, adapted the North American K/DOQI terminology and classification for CKD, which has many advantages [10]. However, these countries have very different health-care systems. For instance, many Canadian nephrologists provide primary care for chronic dialysis patients [11]. International guidelines are being developed [12], but are not yet available. Because of the unique NHS health care system, with separate primary and secondary care services and a strong gate-keeper role for primary care, we believe that UK guidelines are necessary. It is also highly desirable that guidelines for management of patients with CKD do not conflict with existing UK guidelines (for instance, those issued by NICE, SIGN, and other specialist bodies such as the British Hypertension Society). These guidelines must address two questions:
In addressing the second question, we acknowledge that extra demands on time and resources will be welcomed neither by GPs nor by nephrologists. We have addressed this as much as possible from the patient’s perspective, and assumed that patients will not wish to travel further than necessary. What can safely and reliably be done in primary care therefore ought to be done in that setting; any intervention that is better done in a hospital setting should be done there. This has relied on an analysis, based on UK practice, of what interventions and treatments require specialist training, and on when these interventions and treatments are likely to be necessary. We anticipate that the NHS will develop new ways of working to deliver the best care to patients with CKD, building on existing initiatives for diabetes mellitus and Coronary Heart Disease (CHD).
The committee is critically aware of the difficulty in translating guidelines into implementation, particularly when they refer to a condition that is seen by many as relatively rare and that is often thought of as requiring specialist care. We are also aware of the danger of disease-specific guidelines when applied to patients with multiple conditions [13]. Many patients with kidney disease have diabetes, hypertension, or cardiovascular disease. For this reason, these guidelines have been specifically developed to be consistent, wherever possible, with existing UK guidelines on the management of these conditions. Our recommendations are designed to be integrated into existing management systems, and in particular into the management of cardiovascular risk and diabetes in the NHS. Full implementation will require
We have deliberately not addressed the question of which individuals should be responsible for the “care plan” for CKD outlined here. The Department of Health’s publication ‘National Standards, Local Action’ sets out the 2005/2006 to 2007/2008 national priorities for the NHS, including a new national target for the development of personalised care plans for all people with long-term conditions. We anticipate that a variety of models will emerge, including conventional “shared care” between GPs and hospital-based nephrologists; geriatricians, diabetologists, and other secondary care physicians; specialist GPs working within primary care trusts [17-18]; specialist nurses working at General Practice or Primary Care Trust (PCT) level; and computer-based shared care, including prompting systems to trigger clinical actions [19]. It is clear that disease registers and an adequate IT infrastructure will be an essential pre-requisite for delivery of the care plan for CKD. In the longer term, the development of “community nephrologists” with roles similar to community diabetologists [20] may further help to break down unnecessary barriers to the delivery of comprehensive chronic disease management caused by the divide between primary and secondary care.
Implementation of these guidelines will carry cost implications, particularly for the community-based treatment of patients with anaemia, which is not covered by existing funding streams. It is important that the NHS develops a clear strategy for equitable funding of the management of CKD.
It is also clear that there is a pressing need for an educational package for GPs, hospital physicians and surgeons, and community-based nurses, together with clear and concise patient information, on the recognition and management of CKD.
Any system for implementation should be designed to reduce existing ethnic and socioeconomic differences in the burden and consequences of CKD [21-23].
These recommendations include numerical values for biological variables – such as haemoglobin concentration, blood pressure, serum calcium, phosphate and parathyroid hormone concentration – that are directly influenced by treatment. In most guidelines published to date, these numerical values represent the desired outcome of treatment. These values are commonly referred to as “targets” or “standards”. However, if clinicians use the same values as intervention thresholds (the point at which treatment should be changed), the inevitable outcome is that the results of treatment of a population of patients will be distributed around that numerical value. If the results are normally distributed, this means that half of all patients’ values will be above and half below the “target”. For instance, if the aim of treatment of anaemia complicating CKD is to ensure that a patient’s haemoglobin concentration is above 11 g/dL, increasing the epoetin dosage only when the Hb level falls below 11 g/dL will guarantee that half of all patients so treated will have a Hb below 11 g/dL at any given time. To ensure that nearly all patients being treated for anaemia have a Hb > 11 g/dL at any given time, the treatment strategy should ensure that the mean Hb level should exceed 11 g/dL by 1.34 x the standard deviation of the distribution of Hb values. In clinical practice, this means that clinicians need to recognise the difference between intervention thresholds and targets - a distinction that is not recognised by existing clinical practice guidelines [24-25]. It would be preferable to specify target ranges, with an “ideal” value being in the middle of that range, but no previous guidelines have done this, and only a few of the RCTs whose results inform current practice did so – possibly accounting for the failure of many trials to achieve adequate separation between two groups when these groups are allocated to different “targets”, the goals of clinical intervention.
In these recommendations, we have not been able to specify intervention thresholds for each variable, nor a strategy that will result in reliable achievement of the “target” value in the great majority of patients. “Target” values here mean the values that should be achieved in the great majority of patients.
These guidelines are intended to apply to adults (aged >18y) of all ages. They are designed primarily for use in General Practice, although they are also applicable to many patients managed jointly with other disciplines, including diabetic medicine, urology, and geriatrics. They do not cover all eventualities, and nothing in these guidelines should discourage clinicians in any specialty from seeking advice (e.g. by letter, email or telephone conversation) from a nephrologist about the care of a specific patient. They are designed primarily to improve the care of, and outcome amongst, patients with CKD, including those who eventually develop ERF and require RRT. We recognise that there might be other reasons for referral to a nephrologist, including acute renal failure (ARF) (see p 30), nephrotic syndrome, unexplained electrolyte or acid-base abnormalities, metabolic disorders, and refractory hypertension without other evidence of kidney disease, and have included limited guidance in some of these areas.
We hope NICE will develop guidelines for CKD management in future, replacing this document. Failing that, we plan revision in 2009.
...
Published on the Renal Association website. This page created June 10th 2005, last amended
.
.