EvidenceBased Nutrition Practice Applying the Concepts of EBP

Evidence-Based Nutrition Practice Applying the Concepts of EBP to Pediatric Nutrition Practice (with thanks to Donna Johnson)

What evidence-based medicine is: Evidence-based medicine is the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients. Sacket et al. BMJ 1996

What evidence-based medicine is: The practice of EBM requires the integration of l individual clinical expertise with the l best available external clinical evidence from systematic research.

Two Principles for Evidence Based Practice Evidence alone is never enough to make a clinical decision (or a decision about guidelines or evidence) l Decisions & Recommendations should be guided by a hierarchy of strength of evidence l

“If no randomized trials have been carried out for our patient’s predicament, we follow the trail to the next best external evidence and go from there. ” Sacket et al. BMJ 1996

Why Evidence-Base Practice? Clinicians need information l MDs said l » they need it twice a week, » they get it from our text books & journals. l RD’s said » They need it 5 times a week » They search the literature 5 times a month

Clinicians really need information! If shadowed: l they need it up to 60 times per week but only actually find 30% of what they need l and that comes from passers-by » “my textbooks are out of date” » “my journals too disorganized”

Medical textbooks are out-of-date l Fail to recommend Rx up to ten years after it’s been shown to be efficacious. l Continue to recommend therapy up to ten years after it’s been shown to be useless.

Many Traditional Interventions are Out of Date: Ex. Infant GERD Effective? Placing in elevated prone position No Placing in infant seat No Thickening formula with infant cereal Changing protein content of formula Use of pacifier No 5% vs. 10% dextrose water Slightly No Unclear Carroll et al. Arch Ped Adoles Med. 2002

Why don’t RDs seek out the best evidence? Lack of time l Lack of skills for critically appraising the literature l Discomfort with going against traditional institutional practice l Thomas et al. J Hum Nutr Diet. 2003

Three solutions 1 2 3 Carefully research each clinical question that comes our way Seek and apply evidence-based medical summaries generated by others Accept evidence-based practice protocols developed by our colleagues.

Process of EB Practice 1. Define the question 2. Plan and carry out search of the literature 3. Critically appraise the literature 4. Apply the results to your practice 5. Evaluate your performance

Step 1: Define Question -- PICO P - Patient and disease I - Intervention C - Comparative intervention (optional) O - Outcome

Step 2: Search for Evidence Translate PICO Question into a searchable question l Establish a search strategy l » key concepts » boolean operators » synonyms » prioritize » limit

Step 2: Search for Evidence

Step 3: Critically Appraise Basic Introduction and Tutorial at: http: //healthlinks. washington. edu/hsl/classes/evidence/

ADA Quality Indicators Design appropriate to hypothesis l Inclusion & exclusion criteria l Sample size l Key quality indicators for each study design type (ex randomized trials or meta-analysis) l

Step 4: Apply Results l Within context of individual patient preferences, values and rights (or population resources, values and culture)

Evidence, Values, and Resources Values Evidence Resources

The Strength of the Evidence Depends on the Rigor of the Studies Randomized Controlled Clinical Trial l Cohort Study l Case-control Study l Cross Sectional Study l Meta-analysis l

Randomized Controlled Clinical Trial Involves one or more test treatments and a control treatment l Specified outcome measures for evaluating the intervention l Bias free method for assigning treatment l

Randomized Controlled Clinical Trial

Confounding Variable “An extrinsic factor that is associated with the predictor variable and a cause of the outcome variable. ” Hulley and Cummings, Designing Clinical Research

Cohort Study l Identification of two groups » one received exposure of interest » one did not receive exposure l Follow cohort through time to observe the outcome of interest

Cohort Study

Case-control Study Identify patients who have the outcome of interest (cases) l Identify controls without the same outcome l Look back to see if they had the exposure of interest l

Case-control Study

Cross Sectional Study Observation of a defined population at a single point in time or time interval l Exposures and outcomes determined at same time l

Cross Sectional Study

Meta-analysis Quantitative method of combining the results of independent studies l Larger sample size and stronger summaries and conclusions l

The Five Strengths of Evidence 1. Strong Evidence from at least one systematic review of multiple well-designed RCT 2. Strong evidence of at least one well designed RCT of appropriate size 3. Evidence from well designed trials without randomization, single group pre-post, cohort, time series or matched case control 4. Evidence from well designed non-experimental studies from more than one research group 5. Opinions of respected authorities based on clinical evidence, descriptive studies or reports of expert committees

Systematic Reviews l l l l Identify problem area or area of uncertainty Formulate question Search for evidence Select relevant evidence Abstract findings and evaluate reports Form recommendations Summarize the strength of the evidence supporting the recommendation Disseminate the findings

EB Review: Example Huang R-C, Forbes DA, Davies MW Feed thickener for newborn infants with gastro-oesophageal reflux (Cochrane Review). In: The Cochrane Library, Issue 2, 2004. Chichester, UK: John Wiley & Sons, Ltd.

Background Gastro-oesophageal reflux (GOR) is common in newborn infants. A common first line management is the use of feed thickeners. The prevalence of excessive GOR in children is approximately 8%, as diagnosed on 24 hour ambulatory p. H manometry studies in an unselected healthy infant population. Symptomatic regurgitation alone is more common and has been found to occur in 18% of the general infant population

Step 1: Define Question P - Patient and disease I - Intervention C - Comparative intervention (optional) O - Outcome

P = Newborn infants with GOR & preterm infants up to 44 weeks I = Thickeners of all types including rice, gum, or flour based, added to all types of milk including formula and human milk C = Non-thickened feeds O = signs and symptoms of reflux, 24 hour ambulatory p. H monitoring and/or oesophagitis on biopsy

Step 2. Search for Evidence We searched MEDLINE from 1966 to December 2001, the Cochrane Controlled Trials Register, The Cochrane Library, Issue 1, 2002. CINAHL from 1982 to December 2001, and conference and symposia proceedings published in Pediatric Research 1990 to 1994. We also searched conference proceedings for the European Society for Paediatric Gastroenterology and Nutrition (ESPGAN) and the North American Society for Pediatric Gastroenterology and Nutrition (NASPGAN) from 1994 to December 2001. We did not restrict the searches to the English language.

Selection Criteria All randomised controlled trials that examine the effects of thickening formulas on treating gastro-oesophageal reflux in neonates. The eligible studies were to compare thickened feeds to no intervention (unthickened feeds).

Step 3. Critically appraise the literature Key Criteria: l blindness of randomisation l blindness of intervention l completeness of follow up l blinding of outcome measurement l For individual trials, mean differences (and 95% confidence intervals) were reported for continuous variables l For categorical outcomes the relative risk and risk difference (and 95% confidence intervals) were reported

Data collection and analysis Two independent reviewers identified potential studies from the literature search. Quality was independently assessed by two independent reviewers.

Study Reason for exclusion Bailey 1987 Patients' ages ranged from 4 days to 14 months. This was a cross over study in which each patient received both thickened and unthickened feeds, but it does not appear to be a randomised cross over trial. Carcassonne 1975 No control group was used. The age group ranged from 40 days to five years. The patients used had anatomical abnormalites, severe burns or brain tumours. Gouyon 1989 This study evaluated smectite in newborn infants with gastroesophageal reflux. It was rejected because of the lack of randomisation and use of further "thickeners" in some patients in both intervention and placebo group on the basis of undefined symptoms. Gouyon 1988 is an abbreviated report of the same study. Greally 1992 The study population was aged between 2 and 18 months. Infants were randomised to receive either cisapride or gaviscon/carobel. There was no placebo group. Khoshoo 2000 Age range was from 4 to 10 months. Not a RCT. Le Luyer 1992 The study group ranged in age from 2 weeks to 57 months. Not a RCT (no control group). Patients were divided into two groups using different doses of a thickener (sodium alginate) without randomisation. Miller 1999 This study was rejected on the basis of the age group encompassing 0 to 12 months. The outcome data for newborn infants could not be separated from the data of older infants. Orenstein 1987 Ages were between 4 to 34 weeks of age. This was a cross over study, not a RCT. Each patient received both thickened and unthickened feeds. Vandenplas 1994 Ages ranged from 1 week to 4 months old. The outcome data for newborn infants could not be separated from the data of older infants. Weldon 1972 Not a RCT. No control group.

Main Results No studies fulfilled the requirements for inclusion in the systematic review. Reviewers' conclusions There is no evidence from randomised controlled trials to support or refute the efficacy of feed thickeners in newborn infants with GOR. Given the absence of evidence, we cannot recommend using thickening agents for management of GOR in newborn infants

Step 4. Apply results Implications for practice At present, there is no evidence from randomised controlled trials to support or refute the efficacy of feed thickeners in newborn infants with GOR. Although thickening feeds is a simple and cheap manoeuvre, there are some theoretical side effects of this treatment such as delayed gastric emptying with increased caloric density of feed. Therefore, given the absence of evidence, we do not recommend using thickening agents for management of GOR in the neonatal population.

Medline http: //www. ncbi. nlm. nih. gov/entrez/query. fcgi? db=Pub. Med l Under “Limits” can select: l » Review » Meta-analysis » Practice Guidelines » Randomized Controlled trial

Cochran Database of Systematic Reviews http: //www. updatesoftware. com/clibng/cliblogon. htm Can search and review abstracts for free l Full text requires subscription l

Cochran Database of Systematic Reviews (examples – 51 nutrition) Dietary interventions for PKU l Carnitine supplementation of parenterally fed neonates l Feed thickener for newborn infants with GER l Vitamin A supplementation for preventing morbidity and mortality in very low birthweight infants l Formula milk vs. preterm human milk for feeding preterm or LBW infants l

Cochran Database of Systematic Reviews (examples – 51 nutrition) Growth monitoring in children l Fat supplementation of human milk for promoting growth in preterm infants l Gastrostomy feeding versus oral feeding alone for children with cerebral palsy l Multicomponent fortified human milk for promoting growth in preterm infants l Enteral nutritional therapy for induction of remission in Crohn's disease l

National Guideline Clearinghouse http: //www. guideline. gov/ 340 with “nutrition”; 60 with peds nutrition examples: l Early discharge of the term newborn (bottle and breast-feeding; National Assoc Neonatal Nurses) l l Nutrition practice guidelines for type 1 and type 2 diabetes mellitus (ADA) Guidelines for the evaluation of food allergies (American Gastroenterological Association) l Bariatric surgery for severely overweight adolescents: concerns and recommendations.

National Guideline Clearinghouse http: //www. guideline. gov/ ADA currently has 12 guidelines here (ex. ): l chronic kidney disease l gestational diabetes l hyperlipidemia MNT l type 1 and type 2 diabetes) l hypertension – older adults Most expected to be added in the future

Haynes & Haines, BMJ 1998