OBJECTIVES:

Drug dosing recommendations for children with obesity remain limited. This may lead to variability in medication dosing among children with obesity. Therefore, our objective was to determine differences in the prevalence of guideline-nonadherent systemic corticosteroid orders by weight category in children hospitalized for asthma.

METHODS:

We performed a retrospective cross-sectional study of children aged 2 to 17 years hospitalized with asthma and prescribed systemic corticosteroids between January 1, 2010, and December 31, 2017, using the Cerner Health Facts deidentified database. Weight categories ranging from underweight to class III obesity were defined on the basis of BMI percentiles by using CDC guidelines. Corticosteroid orders were categorized as guideline adherent or nonadherent on the basis of total body weight–based dosing guidelines from the National Heart, Lung, and Blood Institute. χ2 test and multivariable logistic regression models were used to determine differences in guideline adherence between weight categories.

RESULTS:

We identified 21 488 children prescribed systemic corticosteroids during asthma hospitalizations. Most (54.2%) had a healthy weight, and 23.8% had obesity. Almost one-quarter received guideline-nonadherent orders (22.2%), with increasing prevalence among higher weight categories (19.4% of healthy weight children versus 36.0% of those with class III obesity; P < .001). After controlling for demographic and clinical covariates, weight category remained significantly associated with receiving a guideline-nonadherent order (P < .001).

CONCLUSIONS:

The prevalence of guideline-nonadherent corticosteroid orders for children hospitalized with asthma increases linearly with weight category, disproportionately affecting children with severe obesity. Standardization of drug dosing guidelines for children with obesity may help reduce variability in drug doses prescribed that may increase risk of harm.

Despite continually rising obesity rates among hospitalized children,1  drug dosing recommendations for children with obesity remain limited, particularly for the most frequently prescribed drugs and commonly encountered diagnoses in the inpatient setting.2,3  Obesity impacts physiologic processes important to drug disposition, including changes in circulating blood volume and expression and function of drug-metabolizing enzymes.46  Yet, there is a paucity of pharmacokinetic data for the vast majority of drugs prescribed to children with obesity, such that few guidelines exist regarding dosing nuances applicable to this growing patient population.3,4,6  This knowledge gap may lead to substantial variability in prescribing practices for children with obesity, posing a potential safety risk for underexposure or overexposure to drugs.6,7 

Evidence-based guidelines help to decrease variability in drug dosing practices that negatively impact care quality and safety.8,9  On the basis of currently available evidence, the Pediatric Pharmacy Advocacy Group (a subgroup of the Pediatric Pharmacy Association focused on advocacy for special pediatric populations) has recommended that prescribers follow total body weight (TBW)-based dosing guidelines in all children unless the recommended adult dose for the specific indication is exceeded.10  However, recent drug-specific evidence also suggests that TBW-based dosing is not always appropriate for children with obesity,11  leaving prescribers with a dosing dilemma for drugs lacking obesity-specific guidance that are prescribed for many common conditions.

To improve understanding of how prescribers address this dosing dilemma for hospitalized children with obesity and how it may create variability in dosing, we examined corticosteroid medication orders during asthma hospitalizations. Children are frequently hospitalized with asthma exacerbations12  and are prescribed systemic corticosteroids to treat their illness, with dosing directed by well-established TBW-based drug dosing guidelines from the National Heart, Lung, and Blood Institute (NHLBI).13  However, the guidelines provide only maximum recommended TBW-based doses, with no recommendations specifically for patients with obesity. Therefore, our objective was to examine weight-based corticosteroid dosing variability on the basis of prevalence of NHLBI guideline nonadherence by weight status in children hospitalized with asthma treated with systemic corticosteroids.

This retrospective cross-sectional study included children hospitalized with an asthma exacerbation or status asthmaticus within the Cerner Health Facts (HF) database. The HF database is a large administrative database including deidentified Health Insurance Portability and Accountability Act–compliant data from 664 health care facilities across the United States representing ∼69 million unique patients (adults and children) and >500 million encounters over the past 2 decades. HF data are maintained by the Cerner Corporation (Kansas City, MO). The HF database contains administrative data and detailed patient-level clinical data, including demographic information, anthropometric measures, and drugs prescribed. Research with HF has been deemed non–human subjects research by our hospital’s institutional review board.

All children aged 2 to 17 years hospitalized with a principal discharge diagnosis of asthma to any HF health care facility during the years 2010–2017 were included (Supplemental Fig 4). Asthma exacerbation hospitalizations were identified by using principal diagnosis codes for asthma exacerbation or status asthmaticus by using International Classification of Diseases, Ninth Revision (ICD-9) and International Classification of Diseases, 10th Revision (ICD-10) (Supplemental Table 3).

We selected the systemic corticosteroids included in our analysis on the basis of the NHLBI asthma treatment guidelines and examination of the corticosteroids most frequently prescribed to children in the HF database: prednisone, prednisolone, and methylprednisolone. All included encounters had at least 1 corticosteroid ordered during their hospitalization. We used the prescription details for the first inpatient order of each corticosteroid. All encounters had at least 1 corticosteroid ordered during their hospitalization, but if an encounter had an inpatient order for >1 included corticosteroid, the first medication order for each drug was included in the analysis. Dexamethasone, although also commonly prescribed for treatment of asthma exacerbations,14  was excluded because it is not included in the NHLBI guideline for steroid dosing.

Encounters were excluded if weight and/or height were unavailable in a given encounter because of the inability to calculate BMI. In addition, children with an ICD-9 or ICD-10 code for pregnancy or childbirth, children who did not receive at least 1 of the included systemic corticosteroids (ie, prednisone, prednisolone, or methylprednisolone), children with incomplete medication order information, and children whose hospitalizations ended in mortality were excluded. Facilities with fewer than 20 encounters during the study period were also excluded to eliminate facilities seeing pediatric patients infrequently (facilities excluded n = 97 [63.8%], encounters excluded n = 509 [2.2%]).

Obesity was defined according to the current CDC classification system for children, which uses BMI percentile for age and sex to categorize patients into the following weight categories: underweight (BMI percentile <5% for age and sex), healthy weight (BMI percentile 5%–<85% for age and sex), overweight (BMI percentile 85%–<95% for age and sex), class I obesity (BMI 95%–<120% of the 95th BMI percentile for age and sex), class II obesity (BMI 120%–<140% of the 95th BMI percentile for age and sex), and class III obesity (BMI ≥140% of the 95th BMI percentile for age and sex).15,16 

The primary outcome measure of interest was the prevalence of guideline-nonadherent systemic corticosteroid orders by weight category. Appropriate weight-based dosing for prednisone, prednisolone, and methylprednisolone was selected by using NHLBI guidelines (Supplemental Table 4).13  For children aged ≤12 years, recommended daily corticosteroid dosing is 1 to 2 mg/kg TBW per day, with a maximum daily dose of 60 mg. For children aged >12 years, the minimum recommended daily dose is 40 mg, and the maximum is 80 mg.

Doses categorized as nonadherent to NHLBI guidelines were defined as the following: (1) total mg/kg per day or mg/kg per dose ≥110% of the maximum recommended TBW-based dose for children, (2) total mg/kg per day or mg/kg per dose ≤90% of the minimum recommended TBW-based dose for children, (3) a dose higher than the absolute recommended maximum daily dose, or (4) a dose lower than the absolute recommended minimum daily dose.17,18 

Data were analyzed at the encounter level; although some patients with asthma had repeat admissions during their childhood, each encounter represented a new and different opportunity to assess patient- and hospital-level variation in prescribing practices. Demographic data were analyzed as categorical variables, including age (ages 2–5, 6–10, 11–14, and 15–17 years). The HF database lacks reliable ethnicity information, notably for Hispanic patients. Because of this data quality issue, Hispanic patients were analyzed within the “other” race or ethnicity category and we have chosen not to present their data separately within the results. Insurance payers were categorized as private, government, other, and unknown. Some examples of insurance plans in the “other” insurance category included health maintenance organization and preferred provider organization plans that were not categorized as either government or commercial and self-pay. “Unknown” insurance types were unknown, unmapped in the HF database, or invalid. Hospital-level characteristics analyzed included region, urban or rural location, whether the hospital was a teaching institution, and number of encounters per facility. Regions were determined by using the definitions from the US Census Bureau.

For our primary outcome analysis, we stratified children by weight category and used the χ2 test to assess differences in the proportion of children who received guideline-adherent corticosteroid orders to those that received guideline-nonadherent orders. Similarly, we compared proportions of guideline-adherent versus guideline-nonadherent orders among demographic groups and by hospital-level characteristics. We also examined differences in overdosing and underdosing of corticosteroid orders by weight category and patient- and hospital-level characteristics. We performed multivariable logistic regression modeling adjusted for significant patient- and hospital-level variables and hospital clustering to determine the odds of receiving a guideline-nonadherent corticosteroid order by weight. All statistical analyses were performed by using SAS version 9.4 (SAS Institute, Inc, Cary, NC).

We identified 21 488 asthma exacerbation hospitalizations for patients aged 2 to 17 years during the years 2010–2017. The majority were aged 6 to 10 years (44.4%), were boys (60.2%), were non-Hispanic Black (51.1%), and had government insurance (55.6%) (Table 1).

TABLE 1

Cohort Characteristics by Weight Category

OverallUnderweightHealthyOverweightClass I ObesityClass II ObesityClass III ObesityP
Overall, n (%) 21 488 1608 (7.5) 11 643 (54.2) 3136 (14.6) 3242 (15.1) 1182 (5.5) 677 (3.2) — 
Age, n (%), y         
 2–5 7891 (36.7) 917 (57) 4607 (39.6) 939 (29.9) 1116 (34.4) 247 (20.9) 65 (9.6) <.001 
 6–10 9543 (44.4) 554 (34.5) 5221 (44.8) 1483 (47.3) 1395 (43) 560 (47.4) 330 (48.7) — 
 11–14 2813 (13.1) 94 (5.8) 1267 (10.9) 527 (16.8) 500 (15.4) 251 (21.2) 174 (25.7) — 
 15–17 1241 (5.8) 43 (2.7) 548 (4.7) 187 (6) 231 (7.1) 124 (10.5) 108 (16) — 
Sex, n (%)         
 Male 12 934 (60.2) 1012 (62.9) 7070 (60.7) 1807 (57.6) 1953 (60.2) 691 (58.5) 401 (59.2) .005 
 Female 8554 (39.8) 596 (37.1) 4573 (39.3) 1329 (42.4) 1289 (39.8) 491 (41.5) 276 (40.8) — 
Race, n (%) — — — — — — — — 
 Non-Hispanic white 6349 (29.5) 475 (29.5) 3405 (29.2) 946 (30.2) 1008 (31.1) 318 (26.9) 197 (29.1) <.001 
 Non-Hispanic Black 10 980 (51.1) 798 (49.6) 6140 (52.7) 1572 (50.1) 1532 (47.3) 581 (49.2) 357 (52.7) — 
 Other 1469 (6.8) 134 (8.3) 752 (6.5) 208 (6.6) 235 (7.2) 103 (8.7) 37 (5.5) — 
 Unknown 2690 (12.5) 201 (12.5) 1346 (11.6) 410 (13.1) 467 (14.4) 180 (15.2) 86 (12.7) — 
Payer, n (%)         
 Government 11 939 (55.6) 864 (53.7) 6497 (55.8) 1715 (54.7) 1811 (55.9) 672 (56.9) 380 (56.1) <.001 
 Private 4175 (19.4) 333 (20.7) 2299 (19.7) 609 (19.4) 626 (19.3) 205 (17.3) 103 (15.2) — 
 Other 2699 (12.6) 234 (14.6) 1421 (12.2) 402 (12.8) 387 (11.9) 162 (13.7) 93 (13.7) — 
 Unknown 2675 (12.4) 177 (11) 1426 (12.2) 410 (13.1) 418 (12.9) 143 (12.1) 101 (14.9) — 
Region, n (%)         
 South 10 857 (50.5) 863 (53.7) 5869 (50.4) 1555 (49.6) 1621 (50) 614 (51.9) 335 (49.5) <.001 
 Northeast 4243 (19.7) 379 (23.6) 2086 (17.9) 649 (20.7) 712 (22) 248 (21) 169 (25) — 
 Midwest 4162 (19.4) 221 (13.7) 2533 (21.8) 583 (18.6) 530 (16.3) 189 (16) 106 (15.7) — 
 West 2226 (10.4) 145 (9) 1155 (9.9) 349 (11.1) 379 (11.7) 131 (11.1) 67 (9.9) — 
Teaching institution, n (%)a         
 Yes 15 976 (87.8) 1300 (89.7) 8392 (87.9) 2317 (86.7) 2488 (87.2) 913 (86.4) 566 (91.3) .003 
Rural versus urban, n (%)         
 Urban 17 979 (83.7) 1340 (83.3) 9886 (84.9) 2591 (82.6) 2629 (81.1) 965 (81.6) 568 (83.9) <.001 
OverallUnderweightHealthyOverweightClass I ObesityClass II ObesityClass III ObesityP
Overall, n (%) 21 488 1608 (7.5) 11 643 (54.2) 3136 (14.6) 3242 (15.1) 1182 (5.5) 677 (3.2) — 
Age, n (%), y         
 2–5 7891 (36.7) 917 (57) 4607 (39.6) 939 (29.9) 1116 (34.4) 247 (20.9) 65 (9.6) <.001 
 6–10 9543 (44.4) 554 (34.5) 5221 (44.8) 1483 (47.3) 1395 (43) 560 (47.4) 330 (48.7) — 
 11–14 2813 (13.1) 94 (5.8) 1267 (10.9) 527 (16.8) 500 (15.4) 251 (21.2) 174 (25.7) — 
 15–17 1241 (5.8) 43 (2.7) 548 (4.7) 187 (6) 231 (7.1) 124 (10.5) 108 (16) — 
Sex, n (%)         
 Male 12 934 (60.2) 1012 (62.9) 7070 (60.7) 1807 (57.6) 1953 (60.2) 691 (58.5) 401 (59.2) .005 
 Female 8554 (39.8) 596 (37.1) 4573 (39.3) 1329 (42.4) 1289 (39.8) 491 (41.5) 276 (40.8) — 
Race, n (%) — — — — — — — — 
 Non-Hispanic white 6349 (29.5) 475 (29.5) 3405 (29.2) 946 (30.2) 1008 (31.1) 318 (26.9) 197 (29.1) <.001 
 Non-Hispanic Black 10 980 (51.1) 798 (49.6) 6140 (52.7) 1572 (50.1) 1532 (47.3) 581 (49.2) 357 (52.7) — 
 Other 1469 (6.8) 134 (8.3) 752 (6.5) 208 (6.6) 235 (7.2) 103 (8.7) 37 (5.5) — 
 Unknown 2690 (12.5) 201 (12.5) 1346 (11.6) 410 (13.1) 467 (14.4) 180 (15.2) 86 (12.7) — 
Payer, n (%)         
 Government 11 939 (55.6) 864 (53.7) 6497 (55.8) 1715 (54.7) 1811 (55.9) 672 (56.9) 380 (56.1) <.001 
 Private 4175 (19.4) 333 (20.7) 2299 (19.7) 609 (19.4) 626 (19.3) 205 (17.3) 103 (15.2) — 
 Other 2699 (12.6) 234 (14.6) 1421 (12.2) 402 (12.8) 387 (11.9) 162 (13.7) 93 (13.7) — 
 Unknown 2675 (12.4) 177 (11) 1426 (12.2) 410 (13.1) 418 (12.9) 143 (12.1) 101 (14.9) — 
Region, n (%)         
 South 10 857 (50.5) 863 (53.7) 5869 (50.4) 1555 (49.6) 1621 (50) 614 (51.9) 335 (49.5) <.001 
 Northeast 4243 (19.7) 379 (23.6) 2086 (17.9) 649 (20.7) 712 (22) 248 (21) 169 (25) — 
 Midwest 4162 (19.4) 221 (13.7) 2533 (21.8) 583 (18.6) 530 (16.3) 189 (16) 106 (15.7) — 
 West 2226 (10.4) 145 (9) 1155 (9.9) 349 (11.1) 379 (11.7) 131 (11.1) 67 (9.9) — 
Teaching institution, n (%)a         
 Yes 15 976 (87.8) 1300 (89.7) 8392 (87.9) 2317 (86.7) 2488 (87.2) 913 (86.4) 566 (91.3) .003 
Rural versus urban, n (%)         
 Urban 17 979 (83.7) 1340 (83.3) 9886 (84.9) 2591 (82.6) 2629 (81.1) 965 (81.6) 568 (83.9) <.001 

—, not applicable.

a

Teaching institution is missing data for 3587 encounters.

Of the included 21 488 encounters, 90.0% were patients with only a single hospitalization during the study period. Overall, encounters occurred at 55 facilities, of which 11 facilities (20.0%) contributed >500 encounters during the study period. This sample of asthma hospitalizations was geographically diverse; 50.5% were located in the southern US census region, 19.7% in the Northeast, 19.4% in the Midwest, and 10.4% in the West. A total of 15 976 encounters (87.8%) occurred at hospitals identified as teaching facilities, and 17 979 (83.7%) were located in urban areas (Table 1).

The majority of patients had a healthy weight status (n = 11 643; 54.2%) (Table 1). Approximately 38.4% had overweight or obesity, with 15.1% of patients having class I obesity, 5.5% having class II obesity, and 3.2% having class III obesity. Only 7.5% of patients were underweight.

A total of 26 254 systemic corticosteroid orders were included in the analysis: 48.9% for prednisolone, 24.9% for prednisone, and 26.2% for methylprednisolone. Overall, prednisolone was more likely to be prescribed to younger children and methylprednisolone and prednisone were more likely to be prescribed to older children (P < .001). Every patient encounter included at least 1 of these drugs, with some encounters having multiple orders (n = 5958; 29.9%).

For our primary outcome analysis, we found that 22.2% of encounters resulted in a guideline-nonadherent corticosteroid order. We found a significant increase in the proportion of children prescribed guideline-nonadherent doses as weight class increased (19.4% of children with a healthy weight to 36.0% of children with class III obesity [P < .001]) (Fig 1). In nearly one-half of all encounters (48.7%), methylprednisolone orders were guideline nonadherent. In contrast, prednisolone had the lowest rates of guideline-nonadherent orders (9.4%).

FIGURE 1

Unadjusted proportion of guideline-nonadherent corticosteroid orders by weight category.

FIGURE 1

Unadjusted proportion of guideline-nonadherent corticosteroid orders by weight category.

Close modal

Older patients (aged 15–17 years, 43.8% guideline-nonadherent orders; P < .001), girls (23.2%; P = .003), patients with non-Hispanic white race (24.0%; P < .001), and those with other insurance (24.1%; P < .001) had higher rates of guideline-nonadherent corticosteroid orders compared with children with other demographic characteristics (Table 2).

TABLE 2

Demographic and Hospital-Level Factors Associated With Guideline-Nonadherent Corticosteroid Orders

OverallUnderdoseNormal DoseOverdoseP
n (%) 21 488 1189 (5.5) 16 719 (77.8) 3580 (16.7) — 
Age, n (%), y      
 2–5 7891 (36.7) 226 (19) 6784 (40.6) 881 (24.6) <.001 
 6–10 9543 (44.4) 467 (39.3) 7547 (45.1) 1529 (42.7) — 
 11–14 2813 (13.1) 331 (27.8) 1690 (10.1) 792 (22.1) — 
 15–17 1241 (5.8) 165 (13.9) 698 (4.2) 378 (10.6) — 
Sex, n (%)      
 Male 8554 (39.8) 521 (43.8) 6566 (39.3) 1467 (41) .002 
 Female 12 934 (60.2) 668 (56.2) 10 153 (60.7) 2113 (59) — 
Race, n (%)      
 Non-Hispanic white 6349 (29.5) 511 (43) 8572 (51.3) 1897 (53) <.001 
 Non-Hispanic Black 10 980 (51.1) 463 (38.9) 4828 (28.9) 1058 (29.6) — 
 Other 1469 (6.8) 78 (6.6) 1197 (7.2) 194 (5.4) — 
 Unknown 2690 (12.5) 137 (11.5) 2122 (12.7) 431 (12) — 
Payer, n (%)      
 Government 11 939 (55.6) 602 (50.6) 9303 (55.6) 2034 (56.8) <.001 
 Private 4175 (19.4) 292 (24.6) 3203 (19.2) 680 (19) — 
 Other 2699 (12.6) 160 (13.5) 2050 (12.3) 489 (13.7) — 
 Unknown 2675 (12.4) 135 (11.4) 2163 (12.9) 377 (10.5) — 
Region, n (%)      
 South 10 857 (50.5) 601 (50.5) 8120 (48.6) 2136 (59.7) <.001 
 Northeast 4243 (19.7) 329 (27.7) 3275 (19.6) 639 (17.8) — 
 Midwest 4162 (19.4) 101 (8.5) 3577 (21.4) 484 (13.5) — 
 West 2226 (10.4) 158 (13.3) 1747 (10.4) 321 (9) — 
Teaching institution, n (%)a      
 No 2223 (12.2) 185 (15.8) 1632 (11.9) 406 (12.4) <.001 
 Yes 15 976 (87.8) 985 (84.2) 12 132 (88.1) 2859 (87.6) — 
Rural versus urban, n (%)      
 Rural 3509 (16.3) 208 (17.5) 2717 (16.3) 584 (16.3) .534 
 Urban 17 979 (83.7) 981 (82.5) 14 002 (83.7) 2996 (83.7) — 
OverallUnderdoseNormal DoseOverdoseP
n (%) 21 488 1189 (5.5) 16 719 (77.8) 3580 (16.7) — 
Age, n (%), y      
 2–5 7891 (36.7) 226 (19) 6784 (40.6) 881 (24.6) <.001 
 6–10 9543 (44.4) 467 (39.3) 7547 (45.1) 1529 (42.7) — 
 11–14 2813 (13.1) 331 (27.8) 1690 (10.1) 792 (22.1) — 
 15–17 1241 (5.8) 165 (13.9) 698 (4.2) 378 (10.6) — 
Sex, n (%)      
 Male 8554 (39.8) 521 (43.8) 6566 (39.3) 1467 (41) .002 
 Female 12 934 (60.2) 668 (56.2) 10 153 (60.7) 2113 (59) — 
Race, n (%)      
 Non-Hispanic white 6349 (29.5) 511 (43) 8572 (51.3) 1897 (53) <.001 
 Non-Hispanic Black 10 980 (51.1) 463 (38.9) 4828 (28.9) 1058 (29.6) — 
 Other 1469 (6.8) 78 (6.6) 1197 (7.2) 194 (5.4) — 
 Unknown 2690 (12.5) 137 (11.5) 2122 (12.7) 431 (12) — 
Payer, n (%)      
 Government 11 939 (55.6) 602 (50.6) 9303 (55.6) 2034 (56.8) <.001 
 Private 4175 (19.4) 292 (24.6) 3203 (19.2) 680 (19) — 
 Other 2699 (12.6) 160 (13.5) 2050 (12.3) 489 (13.7) — 
 Unknown 2675 (12.4) 135 (11.4) 2163 (12.9) 377 (10.5) — 
Region, n (%)      
 South 10 857 (50.5) 601 (50.5) 8120 (48.6) 2136 (59.7) <.001 
 Northeast 4243 (19.7) 329 (27.7) 3275 (19.6) 639 (17.8) — 
 Midwest 4162 (19.4) 101 (8.5) 3577 (21.4) 484 (13.5) — 
 West 2226 (10.4) 158 (13.3) 1747 (10.4) 321 (9) — 
Teaching institution, n (%)a      
 No 2223 (12.2) 185 (15.8) 1632 (11.9) 406 (12.4) <.001 
 Yes 15 976 (87.8) 985 (84.2) 12 132 (88.1) 2859 (87.6) — 
Rural versus urban, n (%)      
 Rural 3509 (16.3) 208 (17.5) 2717 (16.3) 584 (16.3) .534 
 Urban 17 979 (83.7) 981 (82.5) 14 002 (83.7) 2996 (83.7) — 

—, not applicable.

a

Teaching institution is missing data for 3587 encounters.

Hospital-level characteristics associated with guideline-nonadherent orders included the following: region, urban or rural location, and nonteaching institution location (all P < .001; Table 2). The South had the highest proportion of encounters receiving a guideline-nonadherent order (25.2%), and those in the Midwest had the lowest (14.1%). Encounters occurring at teaching institutions had lower rates of prescribing a guideline-nonadherent corticosteroid (24.1% vs 26.6% in nonteaching institutions). Facilities with larger numbers of encounters over the study period were less likely to have guideline-nonadherent orders; for each increase of 1000 patients seen over the study period, the percent of encounters receiving nonadherent orders decreased by 9.3% (P = .003).

The proportion of corticosteroid orders that were underdosed and the proportion that were overdosed both increased with increasing weight category (Fig 1). Overdoses were more common than underdoses across weight categories and increased from 14.7% of all orders for underweight children to 26.3% of all orders for children with class III obesity. The proportion of underdoses also increased across weight categories from 2.9% of all orders for underweight children to 9.7% for children with class III obesity. When examining the proportion of overdoses and underdoses by drug, we observed guideline-nonadherent prednisolone orders infrequently, with a comparable distribution of underdosed (3.1%) and overdosed (6.6%) orders (Fig 2). In contrast, orders for methylprednisolone were more likely to be guideline nonadherent than other included drugs (48.7%), with the majority of these being overdosed (7.0% underdosed versus 43.6% overdosed).

FIGURE 2

Proportion of corticosteroid orders overdosed and underdosed.

FIGURE 2

Proportion of corticosteroid orders overdosed and underdosed.

Close modal

After controlling for significant demographic (age group, sex, race, insurance), clinical (corticosteroid prescribed), and hospital-level (census region, teaching institution, and urban or rural status) factors, we found that weight status remained an independent risk factor for receiving a guideline-nonadherent corticosteroid order during hospitalization for asthma exacerbation (P < .001). Compared with patients with a healthy weight, those with obesity had increased odds of receiving guideline-nonadherent corticosteroid orders (adjusted odds ratio [OR] 1.35 [95% confidence interval (CI): 1.19–1.51] for class I obesity, OR 1.59 [95% CI: 1.38–1.82] for class II obesity, and OR 1.80 [95% CI: 1.48–2.18] for class III obesity) (Fig 3).

FIGURE 3

Adjusted odds of receiving a guideline-nonadherent steroid order by weight category. The model is adjusted for the following: age, sex, race, payer, drug received, census region, teaching facility designation, and urban or rural location.

FIGURE 3

Adjusted odds of receiving a guideline-nonadherent steroid order by weight category. The model is adjusted for the following: age, sex, race, payer, drug received, census region, teaching facility designation, and urban or rural location.

Close modal

Using a large, nationwide, deidentified clinical database, we found a substantial proportion of children received NHLBI guideline-nonadherent systemic corticosteroid orders, indicating variability in corticosteroid dosing for children hospitalized with asthma. Weight category was significantly associated with likelihood of receiving guideline-nonadherent corticosteroid orders (both underdoses and overdoses), with an increasing likelihood as weight category increased.

Our findings align with previous work describing increased risk of aberrant or variable prescribing practices for children with obesity.3,17,19,20  In one previous study, researchers described substantial dosing variability for children with obesity for commonly prescribed drugs in an outpatient setting (eg, analgesics, asthma medications).3  The association of guideline-nonadherent corticosteroid prescriptions with weight may be related to the fact that no specific dosing guidelines exist for patients with obesity for the vast majority of commonly prescribed drugs in pediatrics, leading to confusion and variability in current prescriber practices.21  Additionally, we found many demographic and hospital-level factors other than obesity were also associated with nonadherence to corticosteroid-prescribing guidelines, resulting in nearly one-fourth of all orders overall being guideline nonadherent.13  Importantly, non–evidence-based variability in practice is known to negatively impact the quality and safety of medical care.8,9  The implementation of guidelines for children hospitalized with asthma has been associated with improvements in hospitalization-related outcomes.8  Our findings indicate that more work may be needed to standardize corticosteroid-prescribing practices universally for children hospitalized with asthma, especially regarding methylprednisolone use.

We found that even when TBW-based guidelines that include daily maximum doses exist (eg, NHLBI asthma guidelines), the odds of receiving a guideline-nonadherent order is increased for children with obesity. We observed both overdosing and underdosing (as defined by currently accepted TBW-based dosing guidelines) increased with rising weight category, with some variability depending on the corticosteroid prescribed and overdosing being most common for children with obesity. On the basis of previous pharmacokinetic studies revealing evidence and risk of adverse drug events due to over- or underexposure to drugs in children with obesity,5  this variability in dosing could place patients with obesity at greater risk of experiencing adverse drug events during asthma hospitalization, including toxicities if corticosteroids are overdosed (eg, hyperglycemia, hypertension) or therapeutic failures if underdosed (eg, increased risk of readmission, requirement of more inhaled β agonist therapy).5,7  Obesity has been shown to be associated with adverse outcomes and events in general during hospitalization in children.22,23  The results of our study highlight the need to elucidate the rates of potential adverse events related to drug dosing variability and current prescribing recommendations in patients with obesity, including associated negative outcomes or increased cost.

Although our study and previous literature illuminate the need for obesity-specific drug dosing guidelines, creation of detailed drug dosing recommendations for children with obesity remains difficult because of a paucity of pharmacokinetic and pharmacodynamic evidence in this population to support evidenced-based, obesity-specific guidelines.6  In children with asthma, for example, there are known associations with obesity and poor health outcomes, including poor response to inhaled therapies and more frequent exacerbations.2426  These differences in outcomes may be a result of physiologic differences occurring in children with obesity, including higher degrees of systemic inflammation.27,28  However, other related factors that may be contributing are pharmacokinetic and/or pharmacodynamic differences in children with obesity that may necessitate alterations in dosing to achieve asthma control comparable to that seen in healthy weight children. Specific dosing guidelines for children with obesity and asthma would give providers tools to make more consistent, evidence-based decisions when prescribing drugs to this population of children.

Until obesity-specific pharmacokinetic and pharmacodynamic knowledge is generated through computer model simulation and clinical trials,29,30  providers must make dosing decisions on the basis of available expert opinion or consider dose adjustments to desired treatment response, carefully weighing potential treatment benefits against risks. For example, the Pediatric Pharmacy Advocacy Group published a position statement acknowledging the physiologic alterations and pharmacokinetic differences that may exist for children with obesity but recommended that prescribers follow typical TBW-based dosing in pediatric patients who weigh <40 kg and that TBW-based dosing be used for patients >40 kg unless the recommended adult dose for the specific indication is exceeded.10  At this time, because of the lack of necessary information to formulate drug dosing guidelines for children with obesity, it is advisable for providers to adhere to current NHLBI TBW-based guidelines for children of all weight categories.

Although timely and clinically important, this study has several limitations. Importantly, the HF database lacks illness severity level information typically used for analysis of common clinical outcomes, including patient location (eg, ICU or otherwise). This limited our ability to meaningfully assess use and other clinical outcomes of interest, a primary limitation of this study and an area in need of further investigation in the future. Additionally, illness severity may play a role in the degree of steroid dosing variability and guideline-nonadherent order rates because patients with more severe symptoms may be more likely to receive larger or intravenous doses (eg, intravenous methylprednisolone use is more likely in the ICU setting) despite NHLBI guidelines stating there is no known advantage for larger steroid doses in severe exacerbations.13  Although the HF database provides a wealth of patient-level information not available in many other large data sets, the data collected may be incomplete or inconsistent across contributing organizations.31  We believe the quality of the data included in our analysis is sound on the basis of demographic similarities to other published works using nationally representative samples of hospitalized children with asthma,32  with a few notable exceptions. For example, there were few individuals identified as Hispanic (only 2.8% of our cohort), which is substantially lower than expected on the basis of other publications studying children hospitalized with asthma.33  There was also a relatively large proportion of patients with “other” or “unknown” insurance types and fewer patients with private insurance, a known limitation of the HF database.28 

BMI calculated from height and weight also has some limitations worth noting, including variations in accuracy across different races and inaccuracies in height measurements due to certain conditions (eg, scoliosis, cerebral palsy) or measuring error.34  Anthropometric data are also often incompletely recorded during hospitalizations.35  Of the initial group of patients hospitalized with asthma exacerbation in the study years, ∼21% were excluded because of lack of complete anthropometric data. However, the use of measured anthropometric measures is an advantage of the HF database and strength of our study. Other commonly used data sets contain only diagnosis coding information for obesity, which is frequently poor.36  Lastly, our definition of guideline adherence may not allow for institutional- and provider-level practice variation that may be appropriate, but different from, the NHLBI guidelines.

Obesity is associated with receipt of NHLBI guideline-nonadherent systemic corticosteroid orders for patients hospitalized with asthma exacerbation, disproportionately affecting children with severe obesity. The development of obesity-focused drug dosing guidelines, or standardization of current TBW-based dosing guidelines to be inclusive of children with obesity, is imperative to reduce variability in drug doses prescribed to this growing, at-risk population. In future studies, researchers should attempt to address differences in clinical outcomes between patients with and without obesity on the basis of drug dosing differences as a first step toward generating obesity-specific prescription knowledge for pediatric asthma.

Deidentified individual participant data will not be made available.

Dr Kyler proposed the study idea, participated in the study design and analysis and interpretation of the data, and was the primary author of the manuscript; Drs Bettenhausen, Hall, Glynn, Hoffman, and Shakhnovich participated in the study design and analysis and interpretation of the data and were authors of the manuscript; Drs Smolderen and Davis participated in the study design and analysis and interpretation of the data, were authors of the manuscript, and served as mentor authors; and all authors have provided critical intellectual content in the revision of the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.

FUNDING: Partially supported by the Sarah Morrison Student Research Award, a student research grant from the University of Missouri-Kansas City.

1
Kyler
KE
,
Bettenhausen
JL
,
Hall
M
,
Hampl
S
.
Prevalence and trends in obesity among hospitalized children
.
Hosp Pediatr
.
2019
;
9
(
11
):
897
902
2
Kyler
KE
,
Wagner
J
,
Hosey-Cojocari
C
,
Watt
K
,
Shakhnovich
V
.
Drug dose selection in pediatric obesity: available information for the most commonly prescribed drugs to children
.
Paediatr Drugs
.
2019
;
21
(
5
):
357
369
3
Gade
C
,
Christensen
HR
,
Dalhoff
KP
,
Holm
JC
,
Holst
H
.
Inconsistencies in dosage practice in children with overweight or obesity: a retrospective cohort study
.
Pharmacol Res Perspect
.
2018
;
6
(
3
):
e00398
4
Brill
MJE
,
Diepstraten
J
,
van Rongen
A
,
van Kralingen
S
,
van den Anker
JN
,
Knibbe
CAJ
.
Impact of obesity on drug metabolism and elimination in adults and children
.
Clin Pharmacokinet
.
2012
;
51
(
5
):
277
304
5
Harskamp-van Ginkel
MW
,
Hill
KD
,
Becker
KC
, et al
;
Best Pharmaceuticals for Children Act–Pediatric Trials Network Administrative Core Committee
.
Drug dosing and pharmacokinetics in children with obesity: a systematic review
[
published correction appears in JAMA Pediatr. 2015;169(12):1179
].
JAMA Pediatr
.
2015
;
169
(
7
):
678
685
6
Kendrick
JG
,
Carr
RR
,
Ensom
MHH
.
Pediatric obesity: pharmacokinetics and implications for drug dosing
.
Clin Ther
.
2015
;
37
(
9
):
1897
1923
7
Callaghan
LC
.
Prescribing in paediatric obesity: methods to improve dosing safety in weight-based dose calculations
.
Arch Dis Child Educ Pract Ed
.
2018
;
103
(
5
):
274
277
8
Johnson
DP
,
Arnold
DH
,
Gay
JC
, et al
.
Implementation and improvement of pediatric asthma guideline improves hospital-based care
.
Pediatrics
.
2018
;
141
(
2
):
e20171630
9
Rutman
L
,
Atkins
RC
,
Migita
R
, et al
.
Modification of an established pediatric asthma pathway improves evidence-based, efficient care
.
Pediatrics
.
2016
;
138
(
6
):
e20161248
10
Matson
KL
,
Horton
ER
,
Capino
AC
;
Advocacy Committee for the Pediatric Pharmacy Advocacy Group
.
Medication dosage in overweight and obese children
.
J Pediatr Pharmacol Ther
.
2017
;
22
(
1
):
81
83
11
Shakhnovich
V
,
Abdel-Rahman
S
,
Friesen
CA
, et al
.
Lean body weight dosing avoids excessive systemic exposure to proton pump inhibitors for children with obesity
.
Pediatr Obes
.
2019
;
14
(
1
):
e12459
12
Leyenaar
JK
,
Ralston
SL
,
Shieh
M-S
,
Pekow
PS
,
Mangione-Smith
R
,
Lindenauer
PK
.
Epidemiology of pediatric hospitalizations at general hospitals and freestanding children’s hospitals in the United States
.
J Hosp Med
.
2016
;
11
(
11
):
743
749
13
National Asthma Education and Prevention Program, Third Expert Panel on the Diagnosis and Management of Asthma
.
Expert Panel Report 3: Guidelines for the Diagnosis and Management of Asthma
.
Bethesda, MD
:
National Heart, Lung, and Blood Institute
;
2007
14
Paniagua
N
,
Lopez
R
,
Muñoz
N
, et al
.
Randomized trial of dexamethasone versus prednisone for children with acute asthma exacerbations
.
J Pediatr
.
2017
;
191
:
190
196.e1
15
Skinner
AC
,
Ravanbakht
SN
,
Skelton
JA
,
Perrin
EM
,
Armstrong
SC
.
Prevalence of obesity and severe obesity in US children, 1999-2016
[
published correction appears in Pediatrics. 2018;142(3):e20173459
].
Pediatrics
.
2018
;
141
(
3
):
e20173459
16
Flegal
KM
,
Wei
R
,
Ogden
CL
,
Freedman
DS
,
Johnson
CL
,
Curtin
LR
.
Characterizing extreme values of body mass index-for-age by using the 2000 Centers for Disease Control and Prevention growth charts
.
Am J Clin Nutr
.
2009
;
90
(
5
):
1314
1320
17
Miller
JL
,
Johnson
PN
,
Harrison
DL
,
Hagemann
TM
.
Evaluation of inpatient admissions and potential antimicrobial and analgesic dosing errors in overweight children
.
Ann Pharmacother
.
2010
;
44
(
1
):
35
42
18
Burke
CN
,
Voepel-Lewis
T
,
Wagner
D
, et al
.
A retrospective description of anesthetic medication dosing in overweight and obese children
.
Paediatr Anaesth
.
2014
;
24
(
8
):
857
862
19
Christian-Kopp
S
,
Sinha
M
,
Rosenberg
DI
,
Eisenhart
AW
,
McDonald
FW
.
Antibiotic dosing for acute otitis media in children: a weighty issue
.
Pediatr Emerg Care
.
2010
;
26
(
1
):
19
25
20
Kuhle
S
,
Fung
C
,
Veugelers
PJ
.
Medication use in normal weight and overweight children in a nationally representative sample of Canadian children
.
Arch Dis Child
.
2012
;
97
(
9
):
842
847
21
Rowe
S
,
Siegel
D
,
Benjamin
DK
 Jr
;
Best Pharmaceuticals for Children Act – Pediatric Trials Network Administrative Core Committee
.
Gaps in drug dosing for obese children: a systematic review of commonly prescribed emergency care medications
.
Clin Ther
.
2015
;
37
(
9
):
1924
1932
22
Bechard
LJ
,
Rothpletz-Puglia
P
,
Touger-Decker
R
,
Duggan
C
,
Mehta
NM
.
Influence of obesity on clinical outcomes in hospitalized children: a systematic review
.
JAMA Pediatr
.
2013
;
167
(
5
):
476
482
23
Halvorson
EE
,
Irby
MB
,
Skelton
JA
.
Pediatric obesity and safety in inpatient settings: a systematic literature review
.
Clin Pediatr (Phila)
.
2014
;
53
(
10
):
975
987
24
McGarry
ME
,
Castellanos
E
,
Thakur
N
, et al
.
Obesity and bronchodilator response in black and Hispanic children and adolescents with asthma
.
Chest
.
2015
;
147
(
6
):
1591
1598
25
Quinto
KB
,
Zuraw
BL
,
Poon
K-YT
,
Chen
W
,
Schatz
M
,
Christiansen
SC
.
The association of obesity and asthma severity and control in children
.
J Allergy Clin Immunol
.
2011
;
128
(
5
):
964
969
26
Savas
M
,
Wester
VL
,
Staufenbiel
SM
, et al
.
Systematic evaluation of corticosteroid use in obese and non-obese individuals: a multi-cohort study
.
Int J Med Sci
.
2017
;
14
(
7
):
615
621
27
Vijayakanthi
N
,
Greally
JM
,
Rastogi
D
.
Pediatric obesity-related asthma: the role of metabolic dysregulation
.
Pediatrics
.
2016
;
137
(
5
):
e20150812
28
Norris
AL
,
Steinberger
J
,
Steffen
LM
,
Metzig
AM
,
Schwarzenberg
SJ
,
Kelly
AS
.
Circulating oxidized LDL and inflammation in extreme pediatric obesity
.
Obesity (Silver Spring)
.
2011
;
19
(
7
):
1415
1419
29
Marsousi
N
,
Desmeules
JA
,
Rudaz
S
,
Daali
Y
.
Usefulness of PBPK modeling in incorporation of clinical conditions in personalized medicine
.
J Pharm Sci
.
2017
;
106
(
9
):
2380
2391
30
Knibbe
CAJ
,
Brill
MJE
,
van Rongen
A
,
Diepstraten
J
,
van der Graaf
PH
,
Danhof
M
.
Drug disposition in obesity: toward evidence-based dosing
.
Annu Rev Pharmacol Toxicol
.
2015
;
55
:
149
167
31
Glynn
EF
,
Hoffman
MA
.
Heterogeneity introduced by EHR system implementation in a de-identified data resource from 100 non-affiliated organizations
.
JAMIA Open
.
2019
;
2
(
4
):
554
561
32
Barrett
ML
,
Wier
LM
,
Washington
R
.
Trends in Pediatric and Adult Hospital Stays for Asthma, 2000-2010: Statistical Brief #169
.
Rockville, MD
:
Agency for Healthcare Research and Quality (US)
;
2014
33
Hasegawa
K
,
Tsugawa
Y
,
Brown
DFM
,
Camargo
CA
 Jr
.
Childhood asthma hospitalizations in the United States, 2000-2009
.
J Pediatr
.
2013
;
163
(
4
):
1127
1133.e3
34
Dugas
LR
,
Cao
G
,
Luke
AH
,
Durazo-Arvizu
RA
.
Adiposity is not equal in a multi-race/ethnic adolescent population: NHANES 1999-2004
.
Obesity (Silver Spring)
.
2011
;
19
(
10
):
2099
2101
35
Katzow
M
,
Homel
P
,
Rhee
K
.
Factors associated with documentation of obesity in the inpatient setting
.
Hosp Pediatr
.
2017
;
7
(
12
):
731
738
36
Woo
JG
,
Zeller
MH
,
Wilson
K
,
Inge
T
.
Obesity identified by discharge ICD-9 codes underestimates the true prevalence of obesity in hospitalized children
.
J Pediatr
.
2009
;
154
(
3
):
327
331

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data