Guidelines Have Done More Harm than Good - …

1 Blood Purif 2008;26:73 76 Published online: January 10, 2008. DOI: Guidelines have done more harm than Good Richard Amerling a James F. Winchester a Claudio Ronco b a Division of Nephrology and Hypertension, Beth Israel Medical Center, New York, , USA;. b Department of Nephrology and Intensive Care, St. Bortolo Hospital, Vicenza, Italy Key Words Introduction Practice Guidelines Compliance Renal bone disease, treatment Anemia, treatment Arteriovenous fistula Practice Guidelines are a recent fad in medicine [1]. They came from non-existence 20 years ago, to over 2,000. strong today, covering all aspects of medical care. Guide- Abstract lines can be educational and resemble medical textbooks Practice Guidelines have proliferated in medicine but their in certain ways.

2 There are key differences, however. Text- impact on actual practice and outcomes is difficult, if not book chapters are usually single or double authored and impossible, to quantify. Though Guidelines are based largely are as authoritative as the individual authors. Different on observational data and expert opinion, it is widely be- textbooks exist covering more or less the same subjects, lieved that adherence to them leads to improved outcomes. giving a spectrum of opinion. While many authors share Data to support this belief simply does not exist. If Guidelines their personal clinical experience and make suggestions are universally ignored, their impact on treatment and out- regarding treatment, they shy away from being overly comes is minimal.

3 The incorporation of Guidelines into treat- specific, leaving treatment decisions to the readers. Text- ment protocols and performance measures, as is now com- book entries can be updated until literally just pre-publi- mon practice in nephrology, increases greatly the likelihood cation. Authors are free, indeed encouraged, to scan that Guidelines will influence practice and hence, outcomes. ahead and make predictions about where the field is go- Practice patterns set up this way may be resistant to change, ing, based on as yet unpublished data. Textbooks are fi- should new evidence emerge that contradicts certain rec- nanced by publishers who then market the product in the ommendations.

4 Even if Guidelines are entirely appropriate, hope of making a profit, or at least, recoup expenses. a one-size-fits-all' approach is likely to benefit some, but not In contrast, Guidelines result from the deliberations all. Certain patients may be harmed by adherence to spe- and contributions of a panel of experts' formed into a cific Guidelines . Guidelines certainly do not encourage clini- work group'. Panelists frequently have financial relation- cians to consider and treat each patient as an individual. ships with industry with actual or apparent conflicts of They are unlikely to stimulate original research. They are cre- interest with the outcome of deliberations.

5 Specific treat- ated by a process that is artificial, laborious and cumber- ment recommendations are made, and with great author- some. This all but guarantees many Guidelines are obsolete ity. Literature to be reviewed by panelists is pre-selected by the time they are published. Guidelines are produced by staffers. The process requires a strict cutoff date ar- with industry support and recommendations often have a ticles published after this date cannot be considered major impact on sales of industry products. (though they may be relevant). The upshot is that guide- Copyright 2008 S. Karger AG, Basel lines are hopelessly vulnerable to being out of date when 2008 S.

6 Karger AG, Basel Richard Amerling, MD. 0253 5068/08/0261 0073$ Division of Nephrology and Hypertension, Beth Israel Medical Center Fax +41 61 306 12 34 Baird Hall, 18th floor, 350 East 17th Street E-Mail Accessible online at: New York, NY 10003 (USA). Tel. +1 212 420 4070, E-Mail ultimately released. Guidelines are generally sponsored Kt/V or URR. implicitly or explicitly by industry, via funding of spe- cialty societies. These companies are not in the altruism 80% of patients in 1993 had URR !70; it was down to business; they expect to recoup their investment through 35% in 2004. We can remember the disappointment when increased sales of their products, based on guideline rec- this significant improvement in average URR was associ- ommendations (which are distributed for free).

7 Ated with such an unimpressive change in mortality [2]. Most of the comments in this paper will refer to guide- There are several possible explanations: to the extent that lines produced by the National Kidney Foundation's Kid- the boost in average URR was due to increases above al- ney Disease Outcome and Quality Initiative (KDOQI), ready acceptable numbers, this would not be expected to since this is our main area of expertise. We believe our lead to improved survival. It may also reflect the inherent critique is applicable to Guidelines in general. It is no se- limitations of URR as a predictor of survival, or be due to cret that clinical research is increasingly dominated by unintended consequences of limiting dialysis time in pa- industry-supported studies, mostly of pharmaceutical tients with adequate' URR.

8 Some of these will be small products. It should not come as a surprise that most guide- or malnourished, others will have unmet needs for phos- lines' bottom line is a recommendation for pharmaceuti- phorous clearance or volume removal. Both groups have cal treatment of one type or another. It should be self- poorer survival. A last possibility is that a significant ef- evident that changes in target levels for blood pressure, fect of improved clearance was masked by a deleterious cholesterol, and glycosylated hemoglobin in diabetics, effect elsewhere. will have a profound effect on sales of medications used to achieve these targets. We note in passing the inexorable decline in guideline definitions of normal blood pressure Fistula Use and cholesterol, and optimal HbA1C in diabetics.

9 have Guidelines done more harm than good? It is im- A-V fistula use increased from under 30% to over 40%. possible to answer this question with hard evidence. That from 1998 to 2004. This was accompanied by a marked Guidelines have been accepted on faith should be trou- decline in the use of A-V grafts [2, 3]. The superiority of bling to scientists. A-V fistulae is assumed based purely on observational data (of which there is a prodigious amount) [4]. How- ever, A-V shunts increase cardiac output, work, and pul- Impact of KDOQI Guidelines on the Practice of monary arterial pressures [5]. Small, radiocephalic fistu- Nephrology lae in younger patients with healthy hearts are generally well tolerated.

10 Larger brachiocephalic shunts in older pa- The KDOQI Guidelines have almost certainly influ- tients, or those with significant cardiac disease, are not. enced the practice of nephrology. Increases in Kt/V (or Indiscriminate or premature placement of A-V shunts URR), hemoglobin (Hb), fistula use, as well as reductions could have a negative impact on survival. in Ca ! Phos and PTH, have been seen and may be par- tially attributed to Guidelines [2]. more patients are being diagnosed with chronic kidney disease (CKD) as a result Treatment of Anemia of the classification system based on MDRD GFR. Since this equation has a built-in age bias, a significant number None of the KDOQI Guidelines or recommendations of otherwise healthy elderly folk now have a diagnosis of has caused more controversy than the recent anemia CKD.