Guidelines for Lipid Screening in Children and Adolescents: Bringing Evidence to the Debate Available to Purchase

The Expert Panel guidelines for cardiovascular health and risk reduction in childhood (1), commissioned by the NHLBI, are a valuable resource for pediatric care providers, addressing the major risk factors associated with development of atherosclerosis. Endorsed by the AAP, the recommendations correspond with the age and developmental stages in Bright Futures (2) so they can be incorporated into routine health maintenance. The recommendations are presented with a summary of the evidence, allowing clinicians to apply their own knowledge and experience in deciding what is necessary for each child and family.

The Expert Panel was selected to include representatives from pediatrics, family medicine, cardiology, nutrition and nursing. Each brought their expertise to evaluation of the evidence and recommendations that can be readily integrated into practice. For example, there are specific recommendations for safely transitioning to lower intake of energy-dense foods, such as fat free milk. Recommended daily calories by age, sex and activity level are provided as well as age-specific diet recommendations practitioners can use to promote cardiovascular health. Practical and specific recommendations like this are provided for each of the major risk factors.

Universal lipid screening is the most discussed issue in the guidelines. It was also widely debated within the subgroup reviewing the evidence, risks and benefits, and the entire panel. The consensus recommendation is to assess all children between 9 and 11 years of age with a non-fasting non-HDL cholesterol (non HDL-C = TC - HDL-C). The primary purpose of screening is to identify the approximately 1:500 children who are heterozygous for familial hypercholesterolemia (FH), realizing that other forms of important dyslipidemia would be identified as well. Children with FH have elevated total and LDL cholesterol levels from birth and are at risk for early cardiovascular disease; 5% of individuals with this condition will have a coronary artery event before 30 years of age. Previous guidelines have relied on family history to initiate screening, but the evidence shows this approach to be insufficient. (3,4) The Panel concluded that universal screening was necessary to detect this important, common family condition. Non-fasting non-HDL-cholesterol level is an accurate screen for dyslipidemia, and elimination of the need to be fasting should make testing easier. Age 9 to 11 years was selected because most children entering 5th or 6th grade are required to have a health maintenance exam and because LDL-cholesterol levels fall with puberty before rising to pre-puberty levels. An additional benefit to screening of children is the identification of parents who are unaware that they have FH. Knowing a child's cholesterol level can initiate family screening and a targeted lifestyle intervention.

Providers of children's healthcare are familiar with screening to identify disease states that do not present on physical examination. As the medical home for children and families, we employ behavioral screening to identify conditions such as maternal depression, developmental delay and autism. We assess for disease states and behaviors throughout childhood, although their sequelae may not be manifest until adulthood. An example is screening for tobacco use. Pediatric care providers are, therefore, well positioned to identify children with dyslipidemia who need early intervention to prevent development of premature cardiovascular disease.

The guidelines have only been available for a year and it will take more time to become familiar and comfortable with the recommendations. Following the risk factor algorithms makes it clear that the guidelines only rarely recommend specialist referral. Rather, the approach is risk identification and management by the primary care practitioner. Just as those of us who care for children have learned to use various treatment modalities for diagnoses like ADHD, implementation of these new guidelines will help us to optimize cardiovascular health for children and their families.

(1) Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: full report. 2011. Available at: http://www.nhlbi.nih.gov/guidelines/cvd_ped/peds_guidelines_full.pdf. (2) Hagan JF, Shaw JS, Duncan P, eds. 2008. Bright Futures: Guidelines for Health Supervision of Infants, Children, and Adolescents, Third Edition. Elk Grove Village, IL: American Academy of Pediatrics. (3) Griffin TC, Christofel KK, Binns HJ, McGuire PA; Pediatric Practice Research Group. Family history evaluation as predictive screen for childhood hypercholesterolemia. Pediatrics 1989;84(2):365-373. (4) Ritchie SK, Murphy EC, Ice C, Cottrell LA, Minor V, Elliott E, Neal W. Universal versus targeted blood cholesterol screening among youth: The CARDIAC project. Pediatrics 2010;126(2):260-265.

Conflict of Interest:

Irwin Benuck: no disclosures; Patrick E. McBride: no disclosures

The 3 eLetters to the Editor highlight concerns regarding the Expert Panel recommendation for universal screening of lipids in children. (1) These concerns continue to be debated as noted in the commentary by Daniels and Gillman, both Panel members with opposing viewpoints on this issue. (2) The evidence for an important role of elevated LDL cholesterol in accelerated atherosclerosis beginning in childhood is extensive, and for adults the benefits of lipid-lowering therapy are definitive. However, for children, the evidence for benefit is inferential, the long- term risks are unknown, and costs over a lifetime have not been projected. The number of assumptions that would have to be made in order to provide an estimate quantifying net benefit versus risks/costs would render conclusions suspect. Nonetheless, as stated in the guidelines, the Panel agrees wholeheartedly that this should be pursued and informed by more definitive evidence.

Our management of potential conflicts of interest did not conform to the recommendations of the 2009 Institute of Medicine report which post- dated formation of our Panel by 3 years. The Panel was formed by inclusion of members with the greatest expertise, and did include family medicine practitioners, nurse-practitioners and pediatricians as well as specialists in preventive cardiology. In pediatrics, expertise in lipid management in childhood is limited and elimination of all those with potential conflicts of interest would have precluded formation of an "expert" panel. No Panel member had a fiduciary or promotional relationship with industry. Some Panel members had potential financial conflicts of interest that were declared but did not preclude participation in deliberations or drafting recommendations. Uy and Agawu bring up the potential of intellectual conflicts of interest, which are much harder to recognize and to avoid. Should guidelines be developed by content experts or strictly by methodologists? We recognize that some of our recommendations would have been given lower evidence grades, and some areas would have no recommendation in the absence of sufficient evidence had the Panel been composed exclusively of methodologists. Indeed, the U.S. Preventive Services Task Force in 2007 made no recommendation for lipid screening during childhood, concluding that there was insufficient evidence. (3) The differences in approach to that of our Panel are discussed in the commentary by Daniels and Gilman.(2) By the next time guidelines are developed in this area, we anticipate that the process will be more informed and may utilize the IOM COI recommendations.

We agree that the production of a guideline alone is insufficient. It is best if accompanied by strategies for knowledge translation and implementation, driven by a research program aimed at resolving evidence gaps and assessing impact, and entail a feedback loop for revision; all of these have been explicitly advocated by the Panel. (4) The National Heart, Lung and Blood Institute is conducting a randomized trial of guideline implementation in pediatric practice, to be completed in January 2013, and will utilize the findings in developing forthcoming implementation materials. We feel that our guideline was not paternalistic, but concede that it would have been preferable for knowledge translation and implementation materials to be released coincident with the guidelines.

Finally, the Expert Panel guidelines provide an integrated approach to detection, assessment and management of all cardiovascular risk factors, aimed at both primordial and primary prevention of atherosclerotic cardiovascular disease. We remain seriously concerned that an epidemic of unhealthy behaviors and risk factors exists now in youth, the future in terms of health and costs can be predicted, and must be addressed by evolving effective population-based and targeted solutions aimed at prevention in youth. The controversy over universal lipid screening does not invalidate the guideline, and should not preclude practitioners from scrutiny and implementation of the larger strategy encompassed in the Expert Panel guidelines. Accumulating evidence supports the yield of universal screening. (5) It would be imprudent to "throw the baby out with the bathwater."

1. Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: full report. 2011 [updated Jan 5, 2011] [January 5, 2012]; p. S3. Available from: http://www.nhlbi.nih.gov/guidelines/cvd_ped/index.htm. 2. Gillman MW, Daniels SR. Is universal lipid screening justified? JAMA 2012; 307:259-260 3. Screening for lipid disorders in children: US Preventive Services Task Force recommendation statement. Pediatrics. 2007; 120(1):e215-e219 4. Gidding SS, Daniels SR, Kavey RE. Developing the 2011 Integrated Pediatric Guidelines for Cardiovascular Risk Reduction. Expert Panel on Cardiovascular Health and Risk Reduction in Youth. Pediatrics. 2012; 129:e1311-e1319 5. Ritchie SK, Murphy EC, Ice C, Cottrell LA, Minor V, Elliott E, Neal W. Universal versus targeted blood cholesterol screening among youth: The CARDIAC project. Pediatrics. 2010; 126:260-265

Conflict of Interest:

Brian W. McCrindle, MD MPH: Consultant or Advisory Board Member: Genzyme (Sanofi), Merck, Bristol Myers Squibb, Eli Lilly; Data and Safety Monitoring Board Member: Medpace; Research Funding: Schering Plough, Astra Zeneca, National Institutes of Health; Peter O. Kwiterovich, MD: Consultant or Advisory Board Member: Merck; Patrick E. McBride, MD MPH: No disclosures: Stephen R. Daniels, MD PhD: Consultant or Advisory Board Member: Merck, QLT; Rae-Ellen W. Kavey, MD MPH: No disclosures

McCrindle et al. titled their response to our commentary, "bringing evidence to the debate." However, they primarily reiterated the rationale already in the guidelines, rather than bringing new evidence to address our concerns.

One concern was that the guideline did not address the cost-efficacy of its recommendations. McCrindle et al cited studies of the cost-efficacy of screening for the rare (1/500) genetic condition familial hypercholesterolemia (FH). However, such a narrowly focused screening program was not recommended in the guideline. The $8700 per year gained that they quote is irrelevant because it refers to a program to screen family members of known FH cases,(1) not to the population-wide screening program they recommend, which would be far less cost effective.

McCrindle et al. are right that our commentary contained opinions. So did theirs. It may help to highlight areas where we agree and disagree.

We agree that:

1. Childhood lipid levels can identify children at increased risk of arteriosclerosis decades later.

2. Clinical trials have shown that treating the one child in 500 who has FH can lead to improvements in intermediate outcomes such as coronary atherosclerosis.

3. Trials of whether treating the much larger number of children with high lipid levels as recommended by the proposed guidelines reduces future coronary events have not been done and are unlikely ever to be feasible.

Our areas of disagreement relate both to the aggressive nature of the NHLBI guidelines, and to the process by which they were produced.

1. We disagree that it is acceptable to make screening recommendations without estimating the health benefits, harms, and costs that might result. Because such estimates are essential for informed decision making, we disagree that the "guidelines provide clinicians with the necessary evidence ... to make their own informed judgment as to the utility and role for these recommendations."

2. In the absence of randomized trial evidence of clinical event benefits, we disagree with making a "strong recommendation," requiring a "compelling rationale for an alternative approach" (quoted from Table 1-3, Evidence Grading System" Strength of Recommendations).(2)

3. Most important, we disagree that it is appropriate for panel members with extensive conflicts of interest to have leading roles in creating practice guidelines.

Conflicts of interest among authors of guidelines were discussed in a recent report(3) from the Institute of Medicine (IOM). With the exception of disclosing conflicts, none of the panel's recommendations were followed (Table).

The Panel members, however well-meaning, are only human and it is unreasonable to believe that the large body of research on conflicts of interest that led to the IOM recommendations does not apply to them. A flawed process led to overly aggressive guidelines in which the strength of the evidence was misrepresented and key evidence needed to evaluate the guidelines was lacking. We can and should do better. Let's start by following this key IOM recommendation: scientists with extensive conflicts of interest should not be permitted to have leadership or voting roles on guideline panels.

++++++++++++++++

Table: Institute of Medicine(3) Recommendations for Managing Potential Conflicts of Interest Among Panels writing Clinical Practice Guidelines.

RECOMMENDATION 7.1 Groups that develop clinical practice guidelines should generally exclude as panel members individuals with conflicts of interest...In the exceptional situation in which avoidance of panel members with conflicts of interest is impossible... groups should:

* publicly document that they made a good-faith effort to find experts without conflicts of interest by issuing a public call for members and other recruitment measures;

* appoint a chair without a conflict of interest;

* limit members with conflicting interests to a distinct minority of the panel;

* exclude individuals who have a fiduciary or promotional relationship with a company that makes a product that may be affected by the guidelines;

* exclude panel members with conflicts from deliberating, drafting, or voting on specific recommendations; and

* publicly disclose the relevant conflicts of interest of panel members.

REFERENCES

1. Wonderling D, Umans-Eckenhausen MA, Marks D, Defesche JC, Kastelein JJ, Thorogood M. Cost-effectiveness analysis of the genetic screening program for familial hypercholesterolemia in The Netherlands. Semin Vasc Med. 2004 Feb;4(1):97-104.

2. Kavey R, Simons-Morton DG, de Jesus J. Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: full report. 2011 [updated Jan 5, 2011January 5, 2012]; p. S3. Available from: http://www.nhlbi.nih.gov/guidelines/cvd_ped/index.htm.

3. IOM (Institute of Medicine). Conflict of Interest in Medical Research, Education, and Practice. Washington, DC: The National Academies Press; 2009.

Conflict of Interest:

Dr. Pletcher has NIH funding to support research on targeting of cholesterol-lowering medications for prevention of cardiovascular disease. Drs. Hulley and Newman have no financial relationships relevant to this article to disclose.

To the Editor

I was very interested to read two commentaries (1, 2) on the recent guidelines on lipid screening in children(3) because the differences highlighted several important paradigm differences between the authors.

The first is that the guideline and the defense of the guideline reflect an attitude of paternalism towards both primary care physicians and the patient. It is unfortunate that the guideline recommends an intervention without quantifying the benefit or the harm. I recognize that the weight of evidence suggests lipid screening is a good idea but when a mother asks, "How effective is this? What difference is this going to make?" how is the primary care doctor supposed to answer? Instead of giving the information needed to have a collaborative discussion that would quantify the risks and benefits (such as number needed to screen- NNS, number needed to harm-NNH), the primary care doctor is left to simply answer that it seems like a good idea but no one knows how big a difference this is going to make. In the past, a paternalistic guideline may have been readily accepted, but we are moving towards an era where patients and families, with increasing health literacy, are expecting to make decisions with the physician in a collaborative matter. A guideline that does not give the physician an ability to explain why an intervention is worth it will undermine the authority and credibility of physician to provide tangible information. I understand that without a specific screening trial that tests these guidelines in a real population, the current evidence does not lend itself to calculating a NNS/NNH but that should be recognized as a weakness of the guidelines. For harms data, telling a parent that a two year trial showed safety when the child may be on a medication for decades is unlikely to be reassuring. It is better to have data. I think guidelines need to reflect that both the primary care physician and the patient would like to know not just if an intervention is effective, but how effective it is.

Second, the effectiveness and harms of a guideline can not be assumed without testing. Guidelines are applied both too aggressively (such as colonoscopy guidelines for cancer screening (4)) and ignored (such as HIV screening guidelines (5)) leading to a different outcome than expected by the writers of guidelines. If a guideline is extrapolated from circumstantial information, the true efficacy and harm of a guideline will differ from the originating evidence. This will require further studies to see if the guideline actually achieves what was intended with the expected amount of harm. It can not be assumed from disease oriented data.

Third, it is very complicated to mandate a new universal screening to an already packed well child check when the intervention has unquantified risks and benefits and will likely trigger an incomplete discussion. There is already not enough time to fully practice preventive care (6) or chronic disease management (7) in primary care. Looking at other topics in the same issue of Pediatrics, there are articles about tonsillectomy, genetics, child abuse, lyme disease, maternal infant feeding issues etc. Without time or clear information to have a collaborative discussion, there is the potential that this guideline could be implemented in a coercive manner disrupting the patient doctor relationship or ignored altogether because it is too time consuming to explain the guideline and the logistics of a fasting test. This recommendation for screening simply may be placed on the large pile of things primary care physicians should be doing but do not have time to do.

Fourth, the conflict of interest that I am concerned about has little to do with ties to industry. My concern is that for some of the authors, their careers are focused on lipids. They have a vested interest in highlighting the importance of lipid problems. In the same way that to a hammer everything looks like a nail, to a lipid specialist, managing lipids may seem to take priority over the myriad of issues facing the primary care pediatrician. I would also suspect there is a lack of general primary care clinician input. A lipid specialist may not recognize the need to justify a new intervention to a primary care clinician who is juggling multiple priorities.

Guidelines rarely survive their first encounter with the real world, but I would hope that newer guidelines would take into account issues that are important to the practicing physician, not just a mandate for an intervention. Guidelines should be able to help the clinician and patient understand the tangible risks and benefits of a proposed intervention as well as being able to be realistically implemented.

1. Thomas B. Newman, Mark J. Pletcher, and Stephen B. Hulley. Overly Aggressive New Guidelines for Lipid Screening in Children: Evidence of a Broken Process Pediatrics 2012; 130:349-352

2. Brian W. McCrindle, Peter O. Kwiterovich, Patrick E. McBride, Stephen R. Daniels, and Rae-Ellen W. Kavey. Guidelines for Lipid Screening in Children and Adolescents: Bringing Evidence to the Debate. Pediatrics 2012; 130:353-356

3. Kavey R, Simons-Morton DG, de Jesus J. Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: full report. 2011 [updated Jan 5, 2011 January 5, 2012]; p. S3. Available from: http://www.nhlbi.nih.gov/guidelines/cvd_ped/index.htm.

4. James S. Goodwin, MD; Amanpal Singh, MD, MS; Nischita Reddy, MD; Taylor S. Riall, MD, PhD; Yong-Fang Kuo, PhD. Overuse of Screening Colonoscopy in the Medicare Population. Arch Intern Med. 2011;171(15):1335-1343. doi:10.1001/archinternmed.2011.212

5. Centers for Disease Control and Prevention. Revised Recommendations for HIV Testing of Adults, Adolescents, and Pregnant Women in Health-Care Settings. MMWR 2006;55(No. RR-14)

6. Yarnall KS, Pollak KI, ?stbye T, Krause KM, Michener JL. Primary care: is there enough time for prevention? Am J Public Health. 2003 Apr;93(4):635-41.

7. Truls ?stbye, MD, PhD, Kimberly S. H. Yarnall, MD, Katrina M. Krause, MA, Kathryn I. Pollak, PhD, Margaret Gradison, MD, and J. Lloyd Michener, MD. Is There Time for Management of Patients With Chronic Diseases in Primary Care? Ann Fam Med. 2005 May; 3(3): 209-214.

Conflict of Interest:

None declared

In the August 2012 issue of Pediatrics, Newman et al discuss the recent expert panel recommendations released by the National Heart Lung and Blood Institute and endorsed by the American Academy of Pediatrics; McCrindle et al issue a reply. I have three points I wish to bring up in regards to the latter.

1. Newman et al discuss the logistical implications of having up to 40% of children going for two fasting blood tests, and up to 10% of girls going for at least yearly fasting tests indefinitely thereafater; McCrindle et al do not address this concern in their reply.

2. McCrindle et al in their article suggest that 1.3% of fifth graders would meet criteria for long-term medication to lower lipids, and do not appear to realize that many people (myself included) may find this number inordinate for a not well-proven therapy. The concept that guidelines determined by expert medical opinion, when the evidence is not clear-cut, should be tempered by societal consensus (as well explored in the article "Choriophobia: A 1-Act Play" in the same issue by Taylor and Opel) takes on an added emphasis here.

3. Newman et al bring up the issue of financial conflict of issue. McCrindle et al take exception to this, beliveing it to 'contest the integrity of the panel members'. Nevertheless, this is a legitmiate concern; it is not meant to imply that the authors are unethical, but subconscious bias can creep in when there is such a conflict. McCrindle et al state that 'declared relationships reflect participation as consultants in the design of studies relevant to lipids and children, service on data and safety monitoring boards, and participation in industry-funded trials, ' and that 'the remuneration received covered costs and time, without significant financial incentive.' Their list of declared relationships omits being on a speaker's bureau, which my (admittedly anecdotal) experience suggests is not an insignificant source of income when viewed from an hourly basis.

Jon Matthew Farber

Conflict of Interest:

None declared