Opinion Editorial
Ondansetron for Acute Gastroenteritis A Failure of Knowledge Translation Ron Keren, MD, MPH
Acute gastroenteritis (AGE) is one of the most common reasons for children seeking care in the emergency department (ED), accounting for nearly 2 million visits in the United States each year. For children with AGE and mild to Related article page 321 moderate dehydration, oral rehydration therapy (ORT) is effective and recommended by the Centers for Disease Control and Prevention1 and American Academy of Pediatrics as first-line therapy. However, vomiting is common in children with AGE and may prevent the success of ORT. Ondansetron, which became available in generic formulation in 2006, is efficacious in reducing vomiting and the need for intravenous (IV) rehydration and hospital admission,2 but little is known about its effectiveness in real-world practice. In this issue of JAMA Pediatrics, Freedman and colleagues3 report the results of a multicenter retrospective cohort study that used time series analysis of Pediatric Health Information System data to determine whether increasing use of ondansetron over the last 10 years in children treated for AGE in US EDs was associated with reduced use of IV hydration and admissions to hospital. Focusing their analysis on 18 hospitals with high-quality data and an adequate range of ondansetron use (n = 804 000), they found that ondansetron use increased dramatically from a median institutional rate of 0.11% in 2002 to 42.2% in 2011. However, as hospitals transitioned from low (25%) use of ondansetron—a therapy designed to facilitate oral rehydration and avoid IV rehydration—use of IV rehydration decreased only minimally from 18.7% to 17.8% (−0.9%; 95% CI, −0.7% to −1.1%), and admission to hospital actually increased from 6.0% to 6.7% (+0.7%; 95% CI, 0.6-0.8). In addition, median adjusted total hospital costs increased from $252 to $307. The only clear benefit was reduction in ED revisit within 3 days (−0.7%) and revisit associated with hospitalization (−0.4%). These findings were confirmed in adjusted time series analyses. The Pediatric Health Information System database has become a rich source of information for demonstrating variation in care across children’s hospitals and comparing the effectiveness of different approaches for managing conditions in the emergency and inpatient settings. But the Freedman et al3 study is decidedly different in nature. The authors showed that actually there was not much difference in the ondansetron use across children’s hospitals—over a 10-year period all 18 of the hospitals had incorporated it into their management for a large percentage of children presenting with acute gastroenteritis. And the study was predicated on prior randomized clinical trial evidence that ondansetron is in fact efficacious for reducing vomiting, the need for intravenous hydration, and admission to the hospital. The Freedman et al3 study is unique in that it identifies an effica308
cious therapy that has been widely adopted but not proving effective in the real world probably because of a failure of implementation. The reasons for failed implementation cannot be discerned from this study or data source, but we can speculate about potential culprits. It is possible that the medicine is not being given to the right patients—those who are most likely to benefit from it. That is, if a child with gastroenteritis is having minimal to no vomiting, then administering ondansetron is not likely to change outcomes and will only lead to inefficient use of resources. Another potential explanation is that the drug is not being properly incorporated into a bundle of care delivered in the ED. More specifically, it may be that the sequencing and/or timing of its delivery with respect to the evaluation and decision making for these children is not appropriate. For example, the ondansetron may not be given early enough (eg, at triage) to have the desired effect, and so ORT is abandoned for IV rehydration. Or the ondansetron is given but its effect is never used to make a decision about whether a trial of ORT should be attempted or continued (eg, the patient arrives in the ED, is given a dose of ondansetron, and immediately is given IV fluids). While the overall study results are discouraging, the subgroup analyses provide some hope for improvement (although, like any subgroup analysis, they should be interpreted with caution). A full one-third of the hospitals (n = 6) had clinically and statistically significant reductions in IV rehydration use between low and high ondansetron use periods, and 3 of those had reduced admission rates, confirming the efficacy observed in clinical trials. It appears that some hospital EDs have figured out which patients are most likely to benefit from ondansetron and how to incorporate its administration into the sequence of care for these patients. If we are to achieve maximum return on our investment in this therapy, research is needed to understand how those hospitals got the desired results. Failure to do so is a lost opportunity to leverage an inexpensive drug to prevent outcomes (eg, need for IV rehydration and admission to hospital) that are both costly (to the health system and families) and disruptive (to children and their families). This will require qualitative research to tease out the best practices used by successful hospitals in deciding how to use a medicine known to be efficacious and implementation science research to figure out how to disseminate that knowledge and implement it in a variety of unique clinical settings. Freedman and colleagues have, in effect, provided us with a perfect use case for addressing one of the fundamental challenges facing US health care—figuring out how to translate knowledge of research findings into practice to derive maximum value from evidence-based therapies.4
JAMA Pediatrics April 2014 Volume 168, Number 4
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a J H Quillen College User on 06/11/2015
jamapediatrics.com
Editorial Opinion
ARTICLE INFORMATION Author Affiliations: Division of General Pediatrics, Center for Pediatric Clinical Effectiveness, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia; Associate Editor, JAMA Pediatrics. Corresponding Author: Ron Keren, MD, MPH, Children’s Hospital of Philadelphia, 3535 Market St, Rm 1524, Philadelphia, PA 19104 (keren@email .chop.edu). Published Online: February 24, 2014. doi:10.1001/jamapediatrics.2013.5378.
Conflict of Interest Disclosures: Dr Keren has received research grant funding in the past from Children’s Hospital Association. REFERENCES 1. King CK, Glass R, Bresee JS, Duggan C; Centers for Disease Control and Prevention. Managing acute gastroenteritis among children: oral rehydration, maintenance, and nutritional therapy. MMWR Recomm Rep.2003;52(RR-16):1-16.
gastroenteritis in children and adolescents. Cochrane Database Syst Rev. 2011;(9):CD005506. 3. Freedman SB, Hall M, Shah SS, et al. Impact of increasing ondansetron use on clinical outcomes in children with gastroenteritis [published online February 24, 2014]. JAMA Pediatr. doi:10.1001/jamapediatrics.2013.4906. 4. Grimshaw JM, Eccles MP, Lavis JN, Hill SJ, Squires JE. Knowledge translation of research findings. Implement Sci. 2012;7:50.
2. Fedorowicz Z, Jagannath VA, Carter B. Antiemetics for reducing vomiting related to acute
Appetite for Prevention Genetics and Developmental Epidemiology Join Forces in Obesity Research Daniel W. Belsky, PhD
The obesogenic environment does not affect all children equally. Diminished opportunities for physical activity in daily life and increasing availability and declining cost of caloriedense foods are primary culprits in the obesity epidemic. Related articles pages 338 But there is more to obesity and 345 than the environment. Even children raised together in the same family may experience diverging trajectories of body mass.1 The fact that children confronted with similar environmental circumstances experience disparate outcomes has been attributed to genetic factors.2 And family-based and molecular genetic methods indicate substantial genetic contributions to obesity etiology.3 But just what these genetic factors are and just how they contribute to individual differences in response to the obesogenic environment remains, if not entirely a mystery, an enduring puzzle.4 Solving this puzzle is a public health priority. There is no going back to a world in which calories are scarce and obtaining them is physically demanding. And governments and their publics have shown little enthusiasm for regulations restricting access to palatable, calorie-dense foods. Policies and public health messaging that reframe health-behavior decision making provide a population-level approach to promoting healthier diets and more active lifestyles.5 Complementary practices are needed that can more directly address the most vulnerable individuals, preferably early in life, before obesity develops. Rapid growth in the first months and years of life is associated with increased obesity risk across the life course.6 Genomewide association studies (GWASs) have discovered genetic variants that predispose to higher adult body mass index.7 These variants contribute to the etiology of obesity partly by increasing the rate of growth in childhood.8 The next step in
translating these discoveries to develop interventions that can prevent obesity is to identify mediating mechanisms that connect the DNA sequence a child inherits with the rapid growth trajectory that places them at risk for obesity and its adverse health consequences. In this issue of JAMA Pediatrics, articles by van Jaarsveld and colleagues9 and Llewellyn and colleagues10 suggest one mechanism: individual differences in satiety responsiveness during childhood. Specifically, these studies suggest that diminished satiety responsiveness is a conduit through which genetic risk accelerates weight gain in early life, contributing to obesity pathogenesis. The study by van Jaarsveld and colleagues9 examined longitudinal associations among infant satiety, food responsiveness, and weight gain in a sample of 228 same-sex dizygotic twin pairs followed up from birth through age 15 months. Satiety and food responsiveness were measured for the first 3 months of life, when all children were exclusively breastfed. The twins were selected from a larger cohort on the basis of high within-pair discordance in appetite measures. The researchers then conducted a co-twin control analysis to test whether infant appetite phenotypes predicted the rate of weight gain. The co-twin control analysis asked whether the twin with the higher appetite gained more weight than their co-twin with the lower appetite. The co-twin design ruled out confounding by any factors that influence both appetite and weight gain and are shared by twins in a family. For example, the constitution of the breast milk, the feeding environment, and the maternal biases in reporting on appetite are all controlled. Within this design, the researchers found that the twin with the greater appetite—lower satiety responsiveness and higher food responsiveness—gained weight more rapidly. Appetite was not related to weight at birth. From birth, differences emerged quickly; by age 3 months, the twin with the higher appetite was already heavier. This finding parallels what
jamapediatrics.com
JAMA Pediatrics April 2014 Volume 168, Number 4
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a J H Quillen College User on 06/11/2015
309
Ondansetron for Acute Gastroenteritis A Failure of Knowledge Translation Ron Keren, MD, MPH
Acute gastroenteritis (AGE) is one of the most common reasons for children seeking care in the emergency department (ED), accounting for nearly 2 million visits in the United States each year. For children with AGE and mild to Related article page 321 moderate dehydration, oral rehydration therapy (ORT) is effective and recommended by the Centers for Disease Control and Prevention1 and American Academy of Pediatrics as first-line therapy. However, vomiting is common in children with AGE and may prevent the success of ORT. Ondansetron, which became available in generic formulation in 2006, is efficacious in reducing vomiting and the need for intravenous (IV) rehydration and hospital admission,2 but little is known about its effectiveness in real-world practice. In this issue of JAMA Pediatrics, Freedman and colleagues3 report the results of a multicenter retrospective cohort study that used time series analysis of Pediatric Health Information System data to determine whether increasing use of ondansetron over the last 10 years in children treated for AGE in US EDs was associated with reduced use of IV hydration and admissions to hospital. Focusing their analysis on 18 hospitals with high-quality data and an adequate range of ondansetron use (n = 804 000), they found that ondansetron use increased dramatically from a median institutional rate of 0.11% in 2002 to 42.2% in 2011. However, as hospitals transitioned from low (25%) use of ondansetron—a therapy designed to facilitate oral rehydration and avoid IV rehydration—use of IV rehydration decreased only minimally from 18.7% to 17.8% (−0.9%; 95% CI, −0.7% to −1.1%), and admission to hospital actually increased from 6.0% to 6.7% (+0.7%; 95% CI, 0.6-0.8). In addition, median adjusted total hospital costs increased from $252 to $307. The only clear benefit was reduction in ED revisit within 3 days (−0.7%) and revisit associated with hospitalization (−0.4%). These findings were confirmed in adjusted time series analyses. The Pediatric Health Information System database has become a rich source of information for demonstrating variation in care across children’s hospitals and comparing the effectiveness of different approaches for managing conditions in the emergency and inpatient settings. But the Freedman et al3 study is decidedly different in nature. The authors showed that actually there was not much difference in the ondansetron use across children’s hospitals—over a 10-year period all 18 of the hospitals had incorporated it into their management for a large percentage of children presenting with acute gastroenteritis. And the study was predicated on prior randomized clinical trial evidence that ondansetron is in fact efficacious for reducing vomiting, the need for intravenous hydration, and admission to the hospital. The Freedman et al3 study is unique in that it identifies an effica308
cious therapy that has been widely adopted but not proving effective in the real world probably because of a failure of implementation. The reasons for failed implementation cannot be discerned from this study or data source, but we can speculate about potential culprits. It is possible that the medicine is not being given to the right patients—those who are most likely to benefit from it. That is, if a child with gastroenteritis is having minimal to no vomiting, then administering ondansetron is not likely to change outcomes and will only lead to inefficient use of resources. Another potential explanation is that the drug is not being properly incorporated into a bundle of care delivered in the ED. More specifically, it may be that the sequencing and/or timing of its delivery with respect to the evaluation and decision making for these children is not appropriate. For example, the ondansetron may not be given early enough (eg, at triage) to have the desired effect, and so ORT is abandoned for IV rehydration. Or the ondansetron is given but its effect is never used to make a decision about whether a trial of ORT should be attempted or continued (eg, the patient arrives in the ED, is given a dose of ondansetron, and immediately is given IV fluids). While the overall study results are discouraging, the subgroup analyses provide some hope for improvement (although, like any subgroup analysis, they should be interpreted with caution). A full one-third of the hospitals (n = 6) had clinically and statistically significant reductions in IV rehydration use between low and high ondansetron use periods, and 3 of those had reduced admission rates, confirming the efficacy observed in clinical trials. It appears that some hospital EDs have figured out which patients are most likely to benefit from ondansetron and how to incorporate its administration into the sequence of care for these patients. If we are to achieve maximum return on our investment in this therapy, research is needed to understand how those hospitals got the desired results. Failure to do so is a lost opportunity to leverage an inexpensive drug to prevent outcomes (eg, need for IV rehydration and admission to hospital) that are both costly (to the health system and families) and disruptive (to children and their families). This will require qualitative research to tease out the best practices used by successful hospitals in deciding how to use a medicine known to be efficacious and implementation science research to figure out how to disseminate that knowledge and implement it in a variety of unique clinical settings. Freedman and colleagues have, in effect, provided us with a perfect use case for addressing one of the fundamental challenges facing US health care—figuring out how to translate knowledge of research findings into practice to derive maximum value from evidence-based therapies.4
JAMA Pediatrics April 2014 Volume 168, Number 4
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a J H Quillen College User on 06/11/2015
jamapediatrics.com
Editorial Opinion
ARTICLE INFORMATION Author Affiliations: Division of General Pediatrics, Center for Pediatric Clinical Effectiveness, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia; Associate Editor, JAMA Pediatrics. Corresponding Author: Ron Keren, MD, MPH, Children’s Hospital of Philadelphia, 3535 Market St, Rm 1524, Philadelphia, PA 19104 (keren@email .chop.edu). Published Online: February 24, 2014. doi:10.1001/jamapediatrics.2013.5378.
Conflict of Interest Disclosures: Dr Keren has received research grant funding in the past from Children’s Hospital Association. REFERENCES 1. King CK, Glass R, Bresee JS, Duggan C; Centers for Disease Control and Prevention. Managing acute gastroenteritis among children: oral rehydration, maintenance, and nutritional therapy. MMWR Recomm Rep.2003;52(RR-16):1-16.
gastroenteritis in children and adolescents. Cochrane Database Syst Rev. 2011;(9):CD005506. 3. Freedman SB, Hall M, Shah SS, et al. Impact of increasing ondansetron use on clinical outcomes in children with gastroenteritis [published online February 24, 2014]. JAMA Pediatr. doi:10.1001/jamapediatrics.2013.4906. 4. Grimshaw JM, Eccles MP, Lavis JN, Hill SJ, Squires JE. Knowledge translation of research findings. Implement Sci. 2012;7:50.
2. Fedorowicz Z, Jagannath VA, Carter B. Antiemetics for reducing vomiting related to acute
Appetite for Prevention Genetics and Developmental Epidemiology Join Forces in Obesity Research Daniel W. Belsky, PhD
The obesogenic environment does not affect all children equally. Diminished opportunities for physical activity in daily life and increasing availability and declining cost of caloriedense foods are primary culprits in the obesity epidemic. Related articles pages 338 But there is more to obesity and 345 than the environment. Even children raised together in the same family may experience diverging trajectories of body mass.1 The fact that children confronted with similar environmental circumstances experience disparate outcomes has been attributed to genetic factors.2 And family-based and molecular genetic methods indicate substantial genetic contributions to obesity etiology.3 But just what these genetic factors are and just how they contribute to individual differences in response to the obesogenic environment remains, if not entirely a mystery, an enduring puzzle.4 Solving this puzzle is a public health priority. There is no going back to a world in which calories are scarce and obtaining them is physically demanding. And governments and their publics have shown little enthusiasm for regulations restricting access to palatable, calorie-dense foods. Policies and public health messaging that reframe health-behavior decision making provide a population-level approach to promoting healthier diets and more active lifestyles.5 Complementary practices are needed that can more directly address the most vulnerable individuals, preferably early in life, before obesity develops. Rapid growth in the first months and years of life is associated with increased obesity risk across the life course.6 Genomewide association studies (GWASs) have discovered genetic variants that predispose to higher adult body mass index.7 These variants contribute to the etiology of obesity partly by increasing the rate of growth in childhood.8 The next step in
translating these discoveries to develop interventions that can prevent obesity is to identify mediating mechanisms that connect the DNA sequence a child inherits with the rapid growth trajectory that places them at risk for obesity and its adverse health consequences. In this issue of JAMA Pediatrics, articles by van Jaarsveld and colleagues9 and Llewellyn and colleagues10 suggest one mechanism: individual differences in satiety responsiveness during childhood. Specifically, these studies suggest that diminished satiety responsiveness is a conduit through which genetic risk accelerates weight gain in early life, contributing to obesity pathogenesis. The study by van Jaarsveld and colleagues9 examined longitudinal associations among infant satiety, food responsiveness, and weight gain in a sample of 228 same-sex dizygotic twin pairs followed up from birth through age 15 months. Satiety and food responsiveness were measured for the first 3 months of life, when all children were exclusively breastfed. The twins were selected from a larger cohort on the basis of high within-pair discordance in appetite measures. The researchers then conducted a co-twin control analysis to test whether infant appetite phenotypes predicted the rate of weight gain. The co-twin control analysis asked whether the twin with the higher appetite gained more weight than their co-twin with the lower appetite. The co-twin design ruled out confounding by any factors that influence both appetite and weight gain and are shared by twins in a family. For example, the constitution of the breast milk, the feeding environment, and the maternal biases in reporting on appetite are all controlled. Within this design, the researchers found that the twin with the greater appetite—lower satiety responsiveness and higher food responsiveness—gained weight more rapidly. Appetite was not related to weight at birth. From birth, differences emerged quickly; by age 3 months, the twin with the higher appetite was already heavier. This finding parallels what
jamapediatrics.com
JAMA Pediatrics April 2014 Volume 168, Number 4
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a J H Quillen College User on 06/11/2015
309
