Medication for analgesia, anesthesia, and sedation is an issue in all pediatric age groups, and in a broad range of medical conditions and procedures. It is essential in emergency rooms, surgery, neonatal and pediatric intensive care, required in everyday occurrences such as fractures and appendicitis, and central to management of diseases such as sickle cell anemia and cancer. Drugs used in each setting vary depending on the intensity of pain, the age of the child, the need for sedation versus pain control, and other factors.1–3 Medications used for analgesia, anesthesia, and sedation are among the most frequently used medications in pediatric inpatients, with acetaminophen, lidocaine, fentanyl, ibuprofen, and morphine all among the top 10 medications received by children in the hospital.4 Many of these medications are without any pediatric drug labeling at all, and others have limited labeling for children.5 Acetaminophen, morphine, fentanyl, and propofol were among the 5 drugs most frequently used off-label in a pediatric intensive critical care unit.6 In the same setting, analgesics and anesthetics were the classes most likely to be used off-label.6 In an analysis of pediatric tertiary care hospitals in 2004, Shah et al.7 reported widespread off-label use of acetaminophen, morphine, fentanyl, midazolam, and propofol, with 27.8% of pediatric patients exposed to morphine, 23.2% exposed to fentanyl, and 19.6% exposed to midazolam. In a study of 248 medications dispensed in a pediatric intensive care unit, 24.2% were not Food and Drug Administration (FDA) labeled for any pediatric age group, and 42.7% were approved for use in limited age groups.8 A study in a tertiary care neonatal intensive care unit over a 3-year period found that 27 of 61 (45%) parenteral medications administered were used off-label, and that the most frequently used parenteral medications were analgesics.9
The wide need for analgesia, anesthesia, and sedation in children and the lack of medications labeled for pediatric use leads to the widespread off-label use of medications for analgesia, anesthesia, and sedation. Any attempt to address the lack of labeling will require national estimates of the numbers of children using each medication, their ages, and other factors, to understand the overall use of these medications, to identify the drugs used most frequently, and to describe the patterns of use of the most frequently used drugs. As with all medication use in children, data about inpatient and outpatient use generally reside in separate databases.10 In the analyses presented below, we describe use of analgesics, anesthetics, and sedatives in pediatric inpatients by result of conducting a statistical analysis of medication data from >800,000 pediatric hospitalizations in the United States. Our purpose is to provide a national estimate for the percentage of hospitalized children receiving specific analgesics, anesthetics, and sedatives and to describe their use by age group.
The study was approved by the University of Rhode Island IRB, and the requirement for written consent was waived by the board. We used data from the Premier Database, the largest hospital-based, service-level comparative database in the country, with >45 million records. Detailed service-level information is available for each hospital day and includes medication information and central supplies. Patient information collected includes, but is not limited to, patient demographics (age, gender, race/ethnicity), principal and secondary diagnoses, principal and secondary procedures, payer, length of stay, cost of care, drug use, department cost and charge detail, day-of-stay data, and physician specialty. In addition to the service-level data recorded in most standard hospital discharge files, the database provides a daily log of all billed items, including procedures, medications, laboratory tests, and diagnostic and therapeutic services, at the individual patient level. Hospitals are self-selected, choosing to provide their own data to Premier as part of agreements by which they receive access to analytic tools developed and offered by Premier. In the study year, 2008, 423 hospitals from 42 states and the District of Columbia participated (99.1% with pediatric hospitalizations). Data are entered into a hospital’s core information system, fed into their decision support system, and then sent to Premier on a monthly or quarterly basis. Upon receiving data from participating hospitals, Premier undertakes an extensive 7-phase data validation and correction process that includes >95 quality assurance checks. Deidentified data were extracted and used for statistical analysis.
We assessed external validity of the Premier sample by comparing characteristics of the sample with the Healthcare Cost and Utilization Project (HCUP) Kids’ Inpatient Database (KID) sample of pediatric hospitalizations for 2006 on the following patient characteristics: gender, major diagnostic category, admission source and admission type, length of stay, disposition, 3M™ All Patient Refined™ Diagnosis Related Group (APR-DRG) disease severity score, and APR-DRG mortality risk score, and on the following hospital characteristics: teaching versus nonteaching, urban versus rural, bed size, and region of the country. The samples in Premier and HCUP were similar except for differences in hospital size, teaching status, regionality, and admission source, but the interpretation of these differences and implications for analyses were unclear.4 The differences were statistically significant, but when analyzing large databases such as these, almost all differences are statistically significant. It is not possible to understand the clinical or public health implications until further research is done regarding the direction and magnitude of effect of these potential confounders and medication use.
We identified all uses of a given medication, selected the first use for each child, and calculated the prevalence of use of specific medications among hospitalized children in 2008. We decided, a priori, to consider the hospitalization as the unit of analysis rather than the patient. This arose out of our original work validating the Premier data against the HCUP KID, a database that does not permit individuals to be linked across multiple hospitalizations. Hospitalizations were categorized by use of a specific drug and prevalence (percentage) was calculated as the number of hospitalizations in which the drug was used per 100 hospitalizations. Dose and number of doses were not considered in these analyses. We did not conduct our own validation of the billing data, but accept the validity based on 2 assumptions: (1) hospitals have incentives to bill for medications and will not underbill, and (2) payers have incentives to audit and prevent overbilling, and will institute measures to minimize overbilling.
We reviewed guidelines on pediatric analgesia, anesthesia, and sedation to develop a list of medications that may be used for analgesia or procedural sedation in pediatric inpatients, and grouped the medications into the following categories: nonsteroidal antiinflammatory drugs (NSAIDs) and aspirin, local and regional anesthetics, opioids, benzodiazepines, sedative-hypnotics, combinations, barbiturates, and others.1,3 Acetaminophen was considered in these analyses as an NSAID for the sake of categorical simplicity and in keeping with FDA categories; however, pharmacologically, the antiinflammatory activity of acetaminophen is minimal, such that some do not consider it a true NSAID. Combinations were medications that were administered, billed, and recorded in the database as a combination. Simultaneous administration of 2 drugs was not considered to be a combination if the drugs were billed and recorded separately. Statistical analysis was performed using SAS 9.2.11
The dataset contained records for 877,201 hospitalizations of children younger than 18 years of age at the time of admission with 50,879 (5.8%) being repeat admissions. A total of 12,040,196 medication charges for these stays was identified. The study population was 50.9% male and 49.1% female. Of these, 48.5% were Caucasian, 16.1% African American, 12.2% Hispanic, and the remainder of Asian/Pacific Islander, American Indian, or other racial backgrounds. The average length of stay was 3.7 days (median 2 days, interquartile range 2–3 days). Private insurance paid for 46.2% of the hospital stays, government paid for 45.9%, and self-pay, no charge, or other sources accounted for 7.9% of the hospital stays. Most of the hospitalizations took place in urban areas (89.2%) compared with rural areas (10.8%).
Table 1 summarizes the use of individual medications as a percentage of pediatric hospitalizations (prevalence). Thirty-three medications and an additional 11 combinations were administered in this population. The individual medications included NSAIDs, local and regional anesthetics, opioids, benzodiazepines, sedative-hypnotics, barbiturates, and others. The combinations included the individual NSAIDs, local and regional anesthetics, opioids, and combinations with epinephrine, caffeine, and chlorpheniramine, which were considered in our study for their analgesic agent components despite their use for amelioration of other symptoms as well. Twenty-one medications and 7 combinations were administered in fewer than 1% of hospitalizations. Within each category of medications, it was possible to identify 1 or 2 medications that were more widely used than the others. Acetaminophen and ibuprofen were the most widely used NSAIDs in pediatric hospitalizations; lidocaine the leading local/regional anesthetic; fentanyl and morphine the most frequently used opiates; midazolam the most frequently used benzodiazepine; propofol the leading sedative-hypnotic; and phenobarbital the most frequently used barbiturate. The 10 most frequently administered analgesic, anesthetic, and sedative medications in this population were acetaminophen (14.7%), lidocaine (11.0%), fentanyl (6.6%), ibuprofen (6.3%), morphine (6.2%), midazolam (4.5%), propofol (4.1%), lidocaine/prilocaine (2.5%), hydrocodone/acetaminophen (2.1%), and acetaminophen/codeine (2.0%) (Table 2).
Table 3 summarizes the use of analgesic, anesthetic, and sedative medications in children by age subgroup. The percentage of hospitalizations with use of these medications generally increased with age; however, because of the larger number of hospitalizations in the younger than 2 years age group (which included healthy newborns), low percentages translate into large numbers of children younger than 2 years who are exposed to these medications. Figure 1 shows the percentage and numbers of pediatric hospitalizations with use of different NSAIDs. Acetaminophen, considered here as part of the NSAID group for simplicity, is the NSAID used most frequently in all age groups, administered in 10.3% of hospitalizations of children younger than 2 years, 39.7% age 2 to 4 years, 34.9% age 5 to 11 years, and 28.7% age 12 to 17 years. Whereas the highest prevalence occurred in hospitalizations of children age 2 to 4 years, and 5 to 11 years, the largest number of hospitalizations with acetaminophen use occurs in children younger than 2 years (72,334), and in children age 12 to 17 years (27,010). The use of ketorolac shows a linear increase in both percentage and absolute numbers of hospitalizations occurring across age groups, increasing in prevalence from 0.1% in children younger than 2 years up to 14.7% in children age 12 to 17 years, and increasing in absolute numbers from 1025 to 13,801 in those age groups.
Figure 2 shows the percentage and numbers of pediatric hospitalizations involving use of opiates. Fentanyl and morphine were the most frequently used opiates in all age groups, administered in 2.1% and 1.6% of hospitalizations of children younger than 2 years old, 13.8% and 15.2% age 2 to 4 years, 21.3% and 25.5% age 5 to 11 years, and 29.6% and 27.6% age 12 to 17 years. Both opiates showed a linear increase in prevalence of use with increasing age. Children younger than 2 years had the lowest prevalence of fentanyl or morphine use, but the absolute numbers of children younger than 2 years with fentanyl or morphine use (14,951 and 11,397) were second only to the numbers of 12 to 17 year olds receiving fentanyl (27,821) or morphine (26,011). The distribution of fentanyl and morphine use in absolute numbers was u-shaped, in contrast to the distribution in percentages, which was linear. Other opiates were used to a lesser degree than either fentanyl or morphine. The next most frequently used opiates were hydromorphone, meperidine, and oxycodone. Again, prevalence of use increased with age, but absolute numbers were high in children younger than 2 years, lowest in 2 to 4 year olds, and highest in children 12 to 17 years. Hydromorphone and meperidine were administered in fewer than 0.1% of hospitalizations of children younger than 2 years, lower than the 0.7% and 0.9% observed in hospitalizations of children 2 to 4 years, the 2.4% and 2.5% in children 5 to 11 years, and the 8.8% and 6.6% in children 12 to 17 years. The numbers of hospitalizations involving hydromorphone and meperidine in children younger than 2 years were second only to the numbers in children 12 to 17 years.
Our purpose was to describe the use of analgesics, anesthetics, and sedatives in pediatric inpatients by conducting a statistical analysis of medication data from >800,000 pediatric hospitalizations in the United States, to provide national estimates for the percentage of hospitalized children receiving these medications, and to describe their use by age group. The analyses presented here document extensive use of medications for analgesia, anesthesia, and sedation in children, much of which is off-label. Although many of the findings were unsurprising, some questions are raised. For example, aspirin was administered in 0.1% (680) of hospitalized children younger than 2 years, although it is not labeled for that age group. Although some of these children may have had Kawasaki disease, this needs to be confirmed in future analyses, which could also describe dose and duration, and other medications given to children with Kawasaki disease. Analyses by medication, as well as by condition, will be needed.
As with many other medications, use changed greatly with age groups, and the direction of change (increases and decreases) and the type of change (linear, u-shaped, or other) appeared to be specific to each drug. For example, use of fentanyl appeared to increase linearly with age group, whereas acetaminophen increased and then decreased sharply. The variation in patterns of use reflects the heterogeneity of the dataset, comprising a wide range of ages and conditions in which analgesia, anesthesia, and sedation might be required. Given this heterogeneity, it is not possible to assess whether use of a specific medication was appropriate or not; however, the widespread exposure of many children to these medications occurred despite lack of pediatric labeling for some of the age subgroups.
The strength of this analysis is the large number of pediatric hospitalizations included, larger than any other analysis published to date, from 423 hospitals in 42 states and the District of Columbia. The analysis included more than one-eighth of all the hospitalizations in the United States (in 2009, the HCUP KID estimated a total of 6393,803 hospitalizations). The narrow confidence intervals produced reliable estimates of the numbers of children exposed to each of the medications, and these estimates can be used to plan further retrospective data analyses, prospective observational studies, or clinical trials. The analysis has several limitations that can be addressed in more extensive analyses in the future. Analysis was confined to 1 year of data, and does not reflect changes in practice between 2008 and the present. Heaton et al.12 reported a 3-fold increase in pediatric propofol use between 2001 and 2007, using the same database, and if rapid changes in practice take place with respect to the use of other medications, it will be necessary to repeat the analyses with subsequent years of data. This would be helpful in understanding the impact of medication shortages, and may also help prevent medication shortages by providing more accurate forecasts of demand. Future analyses are needed to describe the different medications received by 1 child; the current analyses only described medication combinations when these were administered and billed as a combination. It will also be important to quantify dose and duration. By definition, the analyses were limited to medication used in the inpatient setting. Separate analyses of other datasets are needed to estimate the outpatient use of the medications described here.
The large sample size presents a potential resource for comparative effectiveness and drug safety research. Although we did not describe diagnoses or procedures associated with each type of medication, future analyses may identify subgroups of patients with similar diagnoses and characteristics, and describe short-term outcomes after medication use. The role of retrospective or prospective observational studies is critical given the well-noted challenges involved in conducting randomized clinical trials in children.13-18 Even if randomized clinical trials have been or can be conducted, it is unlikely that they will be large enough to provide a good understanding of the safety profile. Smith et al.19 point out the need for continued monitoring during the postmarketing period for drugs given exclusivity under the Best Pharmaceuticals for Children Act.
The widespread off-label use of these medications in children raises a range of questions regarding the quality of care for children. The lack of labeling, resulting from a lack of evidence, poses barriers to development of practice guidelines and prevents implementation of computer algorithms in ordering and prescribing, contributing to pediatric medication errors.20 The lack of evidence may also lead to conflicting guidelines about dosage, as in the case of oxycodone, a drug that is not FDA approved for use in children. Despite the lack of evidence to inform dosing, and the lack of pediatric labeling, the medication is used in children. Four reputable sources provide dosing recommendations and the upper limit of appropriate for one source is nearly 2-fold the overdose by another source.21 Presumably, such discrepancies in recommendations would disappear if data were available.
Without evidence, labeling, and guidelines, it is difficult to define medication errors and set goals for patient safety. The widespread off-label use of these medications also raises questions for practitioners and information provided to parents of children receiving such medications. In its discussion about whether use of a drug off-label constitutes experimentation and research, the American Academy of Pediatrics Committee on Drugs noted that, “Discussion about the off-label status of a drug may, as a matter of professional judgment, be part of the information provided to the patient or parents.”22 Our documentation of the large numbers of children requiring analgesia, anesthesia, and sedation underscores the need for further study and labeling of these drugs, and echoes growing concern over the need for studying medications used for analgesia, anesthesia, and sedation in children.5,23–26
Name: Tamar Lasky, PhD.
Contribution: This author helped conduct and review the original study data and data analysis, and approved the final manuscript.
Name: Frank R. Ernst, PharmD, MS.
Contribution: This author helped review the original study data and data analysis, and approved the final manuscript.
Name: Jay Greenspan, MD, MBA.
Contribution: This author helped review the data analysis and approved the final manuscript.
This manuscript was handled by: Peter J. Davis, MD.
1. Anand KJ, Hall RW. Pharmacological therapy for analgesia and sedation in the newborn. Arch Dis Child Fetal Neonatal Ed. 2006;91:F448–53
2. Doyle L, Colletti J. Pediatric procedural sedation and analgesia. Pediatr Clin North Am. 2006;53:279–92
3. Krauss B, Green SM. Procedural sedation and analgesia in children. Lancet. 2006;367:766–80
4. Lasky T, Ernst FR, Greenspan J, Wang S, Gonzalez L. Estimating pediatric inpatient medication use in the United States. Pharmacoepidemiol Drug Saf. 2011;20:76–82
5. Gregoire MC, Finley GA. Why were we abandoned? Orphan drugs in paediatric pain. Paediatr Child Health. 2007;12:95–6
6. Hsu B, Brazelton T. Off-label medication use in an academic hospital pediatric critical care unit. WMJ. 2009;108:343–8
7. Shah SS, Hall M, Goodman DM, Feuer P, Sharma V, Fargason C Jr, Hyman D, Jenkins K, White ML, Levy FH, Levin JE, Bertoch D, Slonim AD. Off-label drug use in hospitalized children. Arch Pediatr Adolesc Med. 2007;161:282–90
8. Yang C, Veltri M, Anton B, Yaster M, Berkowitz I. Food and Drug Administration approval for medications used in the pediatric intensive care unit: a continuing conundrum. Pediatr Crit Care Med. 2011;12:195–9
9. Kumar P, Walker J, Hurt K, Bennett K, Grosshans N, Fotis M. Medication use in the neonatal intensive care unit: current patterns and off-label use of parenteral medications. J Pediatr. 2008;152:412–5
10. Lasky T. Estimates of pediatric medication use in the United States: current abilities and limitations. Clin Ther. 2009;31:436–45
11. SAS Institute Inc. . Base SAS 9.2 Procedures Guide: Statistical Procedures. 20092nd ed Cary, NC SAS Institute Inc
12. Heaton P, Schuchter J, Lannon C, Kemper A. Impact of drug label changes on propofol use in pediatrics for moderate conscious sedation. Clin Ther. 2011;33:886–95
13. Gorman RL. The march toward rational therapeutics in children. Pediatr Infect Dis J. 2003;22:1119–23
14. Kauffman RE. Status of drug approval processes and regulation of medications for children. Curr Opin Pediatr. 1995;7:195–8
15. Roberts R, Rodriguez W, Murphy D, Crescenzi T. Pediatric drug labeling: improving the safety and efficacy of pediatric therapies. JAMA. 2003;290:905–11
16. Steinbrook R. Testing medications in children. N Engl J Med. 2002;347:1462–70
17. Ward RM, Benitz WE, Benjamin DK Jr, Blackmon L, Giacoia GP, Hudak M, Lasky T, Rodriguez W, Selen A. Criteria supporting the study of drugs in the newborn. Clin Ther. 2006;28:1385–98
18. Thomson D, Hartling L, Cohen E, Vandermeer B, Tjosvold L, Klassen T. Controlled trials in children: quantity, methodological quality and descriptive characteristics of pediatric controlled trials published in 1948-2006. PLoS One. 2010;5:313106
19. Smith P, Benjamin D, Murphy M, Johann-Liang R, Iyasu S, Gould B, Califf R, Li J, Rodriguez W. Safety monitoring of drugs receiving pediatric marketing exclusivity. Pediatrics. 2008;122:628–33
20. Miller M, Robinson K, Lubomski L, Rinke M, Pronovost P. Medication errors in paediatric care: a systematic review of epidemiology and an evaluation of evidence supporting reduction strategy recommendations. Qual Saf Health Care. 2007;16:116–26
21. McPhillips H, Stille C, Smith D, Pearson J, Stull J, Hecht J, Andrade S, Miller M, Davis RHenriksen K, Battles J, Marks E. Methodological challenges in describing medication dosing errors in children and methodology. In: Advances in Patient Safety: From Research to Implementation. 2005 Rockville, MD Agency for Healthcare Research and Quality
22. American Academy of Pediatrics Committee on Drugs. . Uses of drugs not described in the package insert (off-label uses). Pediatrics. 2002;110:181–3
23. Berde C, Sethna N. Analgesics for the treatment of pain in children. N Engl J Med. 2002;347:1094–103
24. American Academy of Pediatrics Committee on Psychosocial Aspects of Child and Family Health and American Pain Society Task Force on Pain in Infants, Children, and Adolescents. . The assessment and management of acute pain in infants, children, and adolescents. Pediatrics. 2001;108:793–7
25. Howard RF. Current status of pain management in children. JAMA. 2003;290:2464–9
26. Waisel D. Moral responsibility to attain thorough pediatric drug labeling. Pediatr Anesth. 2009;19:989–93