The off-label use of medications, defined as “the unauthorized use of a drug for a purpose other than that approved by the United States (US) Food and Drug Administration (FDA),”1 is common practice in pediatrics. The FDA establishes the intended indications, dosing, and age limitations for medications based on evidence from clinical trials provided by pharmaceutical companies. However, approximately 80% of medications licensed for use in the US have not been approved for use in the pediatric population.2 As a result, health care providers are left with the challenge of prescribing medications guided by clinical experience rather than data from controlled clinical trials.
The FDA approval process requires substantial resources on the part of drug manufacturers (Fig. 1). In the past, few FDA labeling studies were performed in the pediatric population because the market was considered too small, and the cost of conducting clinical trials in children was too large to make it justifiable. Furthermore, it was believed that dosing and efficacy from adult trials could be safely extrapolated to children. As a result, many medications that are approved by the FDA lack pediatric indications.
In 1979, the FDA began requiring that specific precautions with regard to usage in pediatric patients be included on product package inserts. That same year, regulations were enacted that required any statement on pediatric use of drugs to be based on substantial evidence unless a waiver was granted by the FDA.3 Subsequently, the percentage of medications listed in the Physician’s Desk Reference that had either no indication for pediatric use or age-specific limitations actually increased from 78% in 1971 to 81% in 1991.2 To remedy this situation, the FDA has instituted multiple initiatives and regulations since 1994, including the Pediatric Rule for Labeling, the FDA Modernization Act of 1997, the Best Pharmaceuticals for Children Act, the Pediatric Research in Equity Act, and the 2007 FDA Amendments Act.4–6 As a result of these initiatives, 118 drug labels have been revised to include new pediatric information.7
Despite these efforts, a number of medications continue to be routinely used off-label in children. Yang et al.8 found that 24.2% of medications dispensed in a pediatric intensive care unit were not FDA-approved for any pediatric age group. Similarly, Yoon et al. found that 27% of all medications listed in the Harriet Lane Handbook were not approved by the FDA for use in children.9 Off-label drug use in Canada, as defined by the Canadian Compendium of Pharmaceuticals and Specialties, was reviewed for pediatric anesthesia and intensive care by Doherty et al.,10 who found that 59.7% of prescriptions were off-label.
We sought to determine the incidence of off-label medication use in pediatric anesthesia and analgesia during the perioperative period in a US hospital. Specifically, we undertook an audit of drugs commonly used for anesthesia and pain management to determine which drugs had FDA labeling for pediatric use, which drugs were age-restricted or had no labeling for pediatric use, and which drugs were being used for nonapproved indications. Our hypothesis was that the majority of drugs used for pediatric patients (younger than age 18 years) by anesthesiologists are not FDA-labeled for the indicated use.
Drugs administered by anesthesiologists on a regular basis to pediatric patients during anesthesia or for pain management were identified by auditing the Johns Hopkins Hospital operating room pharmacy drug database and operating room Pyxis MedStations (CareFusion Corporation, San Diego, CA). One hundred six drugs were identified as being regularly administered by anesthesiologists in the operating room or postanesthesia care unit, or for pain management on the inpatient units. The case mix included a full range of pediatric surgical services (cardiac, neurosurgery, orthopedics, head and neck surgery, general surgery, plastic surgery, surgical oncology, and transplant).
FDA labeling and indications for pediatric patients younger than 18 years were determined by using the Thomson Micromedex online database (http://www.thomsonhc.com/micromedex2/librarian). The drugs were divided into 3 categories: not FDA-labeled for use in any pediatric age group, FDA-labeled for use in all pediatric age groups, and FDA-labeled for use in only limited age groups. The medications that did not have FDA-labeling for pediatric use were then further examined to identify the strength of the evidence and the strength of recommendation behind each indication according to the Micromedex website. Micromedex classifies strength of evidence into 3 categories: A, B, and C. Evidence classified as category A is based on meta-analyses of randomized control trials that have homogeneity among the individual studies with regard to the directions and degrees of results; this category is considered the strongest. Category B strength of evidence is based on data derived from meta-analysis of randomized controlled trials that conflict in regard to the directions and degrees of results and randomized controlled trials that involved small numbers of patients or had significant methodological flaws. Evidence classified as category C is based on expert opinion or consensus, case reports, or case series and is considered the weakest evidence. Micromedex classifies the strength of recommendations for the use of medications for specific indications as class I (proven to be useful and that should be used), class IIa (recommended in most cases), class IIb (recommended in some cases), and class III (not recommended). Both strength of recommendations and strength of evidence classifications were used to describe indications for drugs that were not FDA-labeled for pediatric use.
We also examined the rate of off-label drug use in patients younger than 18 years old in operating rooms between July 1, 2010, and August 31, 2011. Operating room sites included 8 general pediatric operating rooms, 1 cardiac operating room, and 6 outpatient operating rooms. Cases in which anesthesia was provided outside the operating room were excluded because those records were not available electronically. We determined the number of cases in which medications were administered off-label as well as the number of times an off-label medication was administered to a patient. Using these data, we calculated the rate of off-label medication administration by age group. The results were then graphed and analyzed by linear regression to assess the relationship between off-label drug administration and patient age. The residuals, calculated as the difference between the actual value of the rate and the value predicted by the best-fit linear regression model, were then graphed to assess whether the rate of off-label drug use for a given age varied significantly from that predicted by the least squares regression analysis. Excel 2007 (Microsoft Corp, Redmond, WA) was used for statistical anaylsis.
The 106 drugs identified as being available to anesthesiologists were reviewed for pediatric FDA-labeling status (Fig. 2). Of those, 40 drugs (38%) are FDA-labeled for patients of all ages (Table 1, column 1) and 36 (34%) have no FDA-labeled indications for patients younger than age 18 years (Table 1, column 2). Thirty (28%) of the 106 drugs are FDA-labeled with age-specific limitations (Table 2).
Of the 36 drugs without FDA labeling for pediatric use (Table 3), 15 (41.7%) have indications for which the strength of evidence is category C, meaning the evidence for use in pediatric patients is based on expert opinion, consensus, case reports, or case series, but not clinical trials. Two (5.6%) of the 36 drugs without FDA approval for pediatric use have a class III recommendation, i.e., not recommended for the specific indication for which they are often used. The first, aminocaproic acid, is commonly used for perioperative blood conservation11–13 even though it is not recommended for postoperative hemorrhage prophylaxis. The other is esmolol, which is not recommended for postoperative hypertension or tachyarrythmias but is often used for those purposes.
Medications were used off-label in 7911 of the 10,782 cases (73.4%) performed during the study’s time period. Of the 36 medications without labeling for pediatric use, 21 were administered in 7911 cases for a total of 14,306 doses. The remaining 15 drugs without pediatric FDA labeling were not used in the operating rooms. Thirteen of those 15 drugs were used primarily for pain management in children outside of the operating room; those data were not captured. The remaining 2 (potassium chloride and esmolol) were used primarily in the intensive care unit. The 3 drugs most commonly administered off-label were ondansetron, hydromorphone, and neostigmine, which when combined accounted for 86% of off-label usage. Aminocaproic acid, 1 of 3 drugs not approved for pediatric use with a class III recommendation, accounted for 2% of off-label use.
Off-label medication use was present in every age group. The rate of off-label medication use and patient age were found to have a strong positive association (R2 = 0.8548; Fig. 3). In addition, compared to all other age groups, the rate of off-label drug administration in children younger than 1 years old was significantly greater (i.e., it was more than 2 SD from the mean) than the value predicted by our regression analysis (Fig. 4).
Our findings support our hypothesis that the majority of drugs used for pediatric patients are not FDA-labeled for the indicated use. This practice continues despite multiple initiatives and regulations proposed and instituted by the FDA to improve drug labeling for infants, children, and adolescents. Our findings are consistent with those of Yang et al.,8 who found that 24% of drugs dispensed in a pediatric intensive care unit were not FDA-approved for use in any pediatric age group and that 43% were approved for use in limited age groups. Additionally, our analysis revealed that the off-label use of drugs in pediatric anesthesia is not limited to obscure or rarely administered drugs. In fact, many of the drugs used off-label are among the most commonly used in routine practice, such as neostigmine for antagonism of neuromuscular blockade and hydromorphone for acute perioperative pain management. These results underscore the need for continuing oversight and research to improve drug labeling for the pediatric population.
Our finding that 73.4% of pediatric patients were administered drugs off-label is consistent with that of a Canadian study that examined off-label drug use in pediatric anesthesia and intensive care. Doherty et al.10 found that over a 1-month period, 77% to 99% of patients received at least 1 off-label drug in the operating room or postanesthesia care unit. In that study, the percentage was a range rather than a single number because the definition of off-label was determined by 3 different sources: 2 from national formularies and the third from contemporary pediatric references. Overall, they found that drug use was less likely to be considered off-label if contemporary pediatric references were used rather than national formularies. Our finding may be an improvement from historical data in which >80% of drugs used in pediatrics were administered off-label.
Infants and neonates, the most vulnerable patients, have the least number of drugs labeled for use.14 Thus, off-label use of anesthetics and analgesics is common in this age group. Our findings showed that the rate of off-label drug use increased with age beginning at age 1 year. Children younger than 1 year had a higher rate of off-label drug use than did children of all ages up to 8 years. Additionally, when a regression analysis was performed to assess the relationship between age and rate of off-label drug use, it was determined that the rate of off-label drug administration was significantly greater in children younger than age 1 year than the value predicted by the model. Although the Pediatric Research in Equity Act and Best Pharmaceuticals for Children Act legislation has led to increased pediatric testing for newly developed and marketed drugs, there are large gaps in our understanding of how off-label use affects the infant and neonate. Recent clinical studies suggest a potential association between exposure to surgery/anesthesia early in life and learning disabilities, but whether this is a true risk remains to be seen.15 The FDA has recognized that the potential link between anesthesia use and learning disabilities is an area of great concern, and its National Center for Toxicological Research and the Center for Drug Evaluation and Research have implemented in vitro and in vivo models to study ketamine and eventually other anesthetic drugs. Designing epidemiologic studies in neonatal and pediatric surgery patients is complicated by the need to adjust for level of prenatal care, surgical and medical comorbidities, anesthetic and sedative exposure, and level of health care in pre- and postnatal periods, including early childhood. Until sufficient data are available to make reasonable recommendations, we are left to use our best judgment when providing anesthesia to our young patients.
Challenges associated with conducting drug research in the pediatric population have resulted in a paucity of data. FDA labeling requires data from “adequate and well controlled studies,” which typically means 2 randomized, double-blind, placebo-controlled trials.16 However, children cannot be denied “standard treatment,” thus calling into question the ethics of trials with a placebo arm. However, not conducting such research in pediatric patients may be an inappropriate and unethical response to the challenges. In discussing the ethical principal of justice, Tobin et al. stated that “The principal of justice is violated when they children are not afforded access to effective treatment because the necessary research has not been conducted to support pediatric labeling.”17The fact that off-label drug use is disclosed for presentations of data in scientific meetings and research manuscripts does not mean that the use of the drug is inappropriate. Off-label administration of drugs in children is common practice and often clinically appropriate because the benefits outweigh potential risks, and no suitable alternative is available. The risk of adverse drug reactions with off-label use of drugs has been studied both prospectively and retrospectively, but the magnitude of the risk remains unclear. A recent review of the literature showed some association between adverse drug reactions and off-label use in children, but studies varied in their definitions of adverse drug reaction and off-label use.18 Conversely, using drugs on-label may also have risks, as exemplified by the case of rapacuronium. After receiving approval from the FDA for use in adults and children ages 1 month to 12 years, rapacuronium was found to be associated with life-threatening bronchospasm in the pediatric population and was removed from the market.19 The true risk of adverse events for both on- and off-label drug use during anesthesia in children would be of great interest and warrants further investigation.
The legality of off-label drug use is another topic of concern. The FDA, under the authority of the Federal Food, Drug and Cosmetic Act, U.S.C. Title 21 (The Act), regulates the pharmaceutical industry and its manufacturing and marketing of drugs, but it does not regulate the practice of medicine.20,21 The Act requires accurate labeling of drugs and permits advertising based on data approved by the FDA for inclusion in a drug’s labeling.22 Physicians are free to use prescription drugs in any manner that they believe to be beneficial to the patient’s welfare. The FDA allows physicians to prescribe approved drugs for unapproved uses under the “practice of medicine exemption.” Federal law also supports the ability of the physician to prescribe a different dosage or otherwise vary the conditions of use from the approved labeling without informing or obtaining approval of the FDA (United States versus Evers).23 Any such off-label use of a drug should be based on the practitioner’s own expert medical opinion and be supported by peer-reviewed literature and the opinions of the practitioner’s local colleagues.24–26
This study has several limitations. We reviewed only medications available to and used by our anesthesiologists. Other institutions may have alternative or additional drugs available for the anesthesiologists’ use that we did not review. We did not address drug labeling by drug regulatory agencies in other countries, only the FDA. The drugs used during anesthesia care are relatively limited and do not reflect all of the medications used in pediatrics. We also did not include all of the medications used in the intensive care unit or on the wards during the immediate postoperative period. The degree of off-label use may have been greater had we reviewed all of the potential drugs given to patients during the perioperative period.
In conclusion, despite the efforts of the FDA, much work remains to be done in drug labeling for children. It is in the best interest of our young patients that we continue to push for resources to study the safety and efficacy of drugs in this population. Input from the anesthesiology community can potentially guide the direction and future of this research area and ultimately impact the safety of children who receive anesthesia for surgery and diagnostic procedures.
Name: Michael C. Smith, MPH.
Contribution: This author helped conduct the study, collect the data, analyze the data, and prepare the manuscript.
Conflicts of Interest: Michael C. Smith has no conflicts of interest to declare.
Name: Julie Williamson, DO.
Contribution: This author helped design the study, collect the data, and analyze the data.
Conflicts of Interest: Julie Williamson has no conflicts of interest to declare.
Name: Myron Yaster, MD.
Contribution: This author helped design the study and prepare the manuscript.
Conflicts of Interest: Myron Yaster currently serves on the Data and Safety Monitoring Board (DSMB) for Purdue and Endo Pharmaceuticals. Over the past 3 years, he has been both a consultant and/or a primary investigator on several clinical trials sponsored by Purdue Pharma, Endo Pharmaceuticals, Hospira, Cadence Pharmaceuticals, and AstraZeneca Pharmaceuticals. Additionally, he or a family member is a stockholder (5K) for Sucampo Pharmaceuticals, Johnson and Johnson, Cadence, and Pfizer.
Name: Geoffrey J. C. Boyd, BS Pharm, JD.
Contribution: This author helped prepare the manuscript.
Conflicts of Interest: Geoffrey J. C. Boyd has no conflicts of interest to declare.
Name: Eugenie S. Heitmiller, MD.
Contribution: This author helped conduct the study, collect the data, analyze the data, and prepare the manuscript.
Conflicts of Interest: Eugenie S. Heitmiller has no conflicts of interest to declare.
This manuscript was handled by: Peter J. Davis, MD.
The authors want to thank Sharon Paul for her effort in extracting the medication data, Gregory Latham for his contribution, and Claire Levine for her expert medical editing in preparing this manuscript.
1. United States Public Law 105-115e Nov 21, 1997. . Title 1 Improving Regulations of Drugs Sec. 101-131. Food and Drug Modernization Act of 1997 21 USC 301 note
2. Gorman RL. The march toward rational therapeutics in children. Pediatr Infect Dis J. 2003;22:1119–23
3. Department of Health, Education, and Welfare, Food and Drug Administration. . Prescription drug advertising; content and format for labeling of human prescription drugs Fed Reg 44:21 CFR 201.57(f) (9)(iv). June 26, 1979:37433
4. Department of Health and Human Services, Food and Drug Administration. . Specific requirements on content and format of labeling for human prescription drugs; revision of “Pediatric Use” subsection in the labeling Fed Reg 59(No. 238). 21 CFR section 201.57(f)(9) December 13. 1994:64240–50
5. Department of Health and Human Services, Food and Drug Administration. . Best Pharmaceuticals for Children Act Public Law 107-109. January 4, 2002
6. Department of Health and Human Services, Food and Drug Administration. . Pediatric Research Equity Act Public Law 108-155. December 16, 2003
7. Ward RM, Kauffman R. Future of pediatric therapeutics: reauthorization of BPCA and PREA. Clin Pharmacol Ther. 2007;81:477–9
8. Yang CP, Veltri MA, Anton B, Yaster M, Berkowitz ID. Food and Drug Administration approval for medications used in the pediatric intensive care unit: a continuing conundrum. Pediatr Crit Care Med. 2011;12:e195–9
9. Yoon EY, Davis MM, El-Essawi H, Cabana MD. FDA labeling status of pediatric medications. Clin Pediatr (Phila). 2006;45:75–7
10. Doherty DR, Pascuet E, Ni A, Stewart P, Splinter W, Vaillancourt R. Off-label drug use in pediatric anesthesia and intensive care according to official and pediatric reference formularies. Can J Anaesth. 2010;57:1078–8
11. Tzortzopoulou A, Cepeda MS, Schumann R, Carr DB. Antifibirinolytic agents for reducing blood loss in scoliosis surgery in children. Cochrane Database Syst Rev. 2008;3:CD006883
12. Chauhan S, Das SN, Bisoi A, Kale S, Kiran U. Comparison of epsilon aminocaproic acid and tranexamic acid in pediatric cardiac surgery. J Cardiothorac Vasc Anesth. 2004;18:141–3
13. Kenyon CA, Flick R, Moir C, Ackerman MJ, Pabelick CM. Anesthesia for videoscopic left cardiac sympathetic denervation in children with congenital long QT syndrome and catecholaminergic polymorphic ventricular tachycardia—a case series. Pediatr Anesth. 2010;20:465–70
14. Tobin JR. Use of pharmaceuticals ‘off-label’ in the neonate. Best Pract Res Clin Anaesthesiol. 2010;24:451–60
15. Davidson AJ. Anesthesia and neurotoxicity to the developing brain: the clinical relevance. Paediatr Anaesth. 2011;21:716–21
16. Department of Health and Human Services, Food and Drug Administration. . Title 21 Code of Federal Regulation, April 1, 2011 Part 201, §201.57
17. Tobin JR, Shafer SL, Davis P. Pediatric research and scholarship: another Gordian Knot. Anesth Analg. 2006;103(1):43–8
18. Mason J, Pirmohamed M, Nunn T. Off-label and unlicensed medicine use and adverse drug reactions in children: a narrative review of the literature. Eur J Clin Pharmacol 2011 Jul 22. Epub ahead of print
19. Rajchert DM, Pasquariello CA, Watcha MF, Schreiner MS. Rapacuronium and the risk of bronchospams in pediatric patients. Anesth Analg. 2002;94(3):488–93
20. Pub. L. No. 75-717, § 201(g)(1)(B), ch. 675, 52 Stat. 1040 (June 25, 1938) (as amended 1962) (codified as amended at 21 U.S.C. § 321(g)(1)(B)).
21. 52 Fed. Reg. 19,466 (1987) (codified at 21 C.F.R. § 312 (1993).
22. . Requirements on Content and Format of Labeling for Human Prescription Drug and Biological Products Fed Reg. 2006;71(15):3922–3997
23. . FDA, Legal Status of Approved Labeling of Prescription Drugs; Prescribing for Uses Unapproved by the FDA, 37 Fed Reg. 16503, 16503; Aug. 15, 1972
24. Henry V. Off-label prescribing. Legal implications. J Leg Med. 1999;20:365–83
25. Cohen PJ. ‘Off-label’ use of prescription drugs: legal, clinical and policy considerations. Eur J Anesthesiol. 1997;14:231–5
© 2012 International Anesthesia Research Society
26. Hill P. Off license and off label prescribing in children: litigation fears for physicians. Arch Dis Child. 2005;90(Suppl I):i17–8