Abstract
Rationale: Although <5% of children hospitalized with an asthma exacerbation have pneumonia that can be radiographically confirmed, at some hospitals, an asthma–pneumonia co-diagnosis is so common that the term “asthmonia” is used to describe the phenomenon. High rates of asthmonia diagnosis may incur unwarranted health care costs and contribute to unnecessary antibiotic prescribing.
Objectives: To characterize hospital variation in rates of pediatric asthmonia diagnosis and analyze associations between hospitals’ asthmonia diagnosis rates and clinical outcomes.
Methods: We conducted a cross-sectional analysis of 274 hospitals contributing to the Premier Healthcare Database. Children and adolescents 2–17 years of age were included if they were hospitalized with an asthma exacerbation from 10/1/2015 to 6/30/2018. Asthmonia was defined as a discharge diagnosis of pneumonia in a patient with an asthma exacerbation. To compute hospital-level risk-standardized asthmonia rates, hierarchical generalized linear models with hospital random effects were estimated, adjusting for patient characteristics. The median odds ratio was calculated to quantify the effect of hospital-level clustering on asthmonia diagnosis. Hospitals were stratified into quartiles based on risk-standardized asthmonia diagnosis rates to identify associated hospital characteristics. Generalized linear models, adjusting for hospital characteristics, were developed to compute associations between hospital risk-standardized rates and clinical outcomes.
Results: Of 24,606 asthma exacerbations, 19,402 (78.9%) were diagnosed with asthma alone and 5,204 (21.1%) received asthma–pneumonia co-diagnoses. The hospital median risk-adjusted asthmonia diagnosis rate was 20.9% (interquartile range, 16.2–27.2%; range, 8.4–55.9%). The median odds ratio was 1.75 (95% confidence interval, 1.63–1.86). Compared with hospitals in the lowest quartile of asthma–pneumonia co-diagnosis, those in the highest quartile were more likely to be smaller, nonteaching, rural hospitals with minimal subspecialty support (all P < 0.001). Hospitals with high rates of risk-standardized asthmonia diagnosis had greater antibiotic use, more prolonged lengths of stay, and higher costs, with no significant differences in risk of transfer or readmission.
Conclusions: Marked variation exists in rates of asthmonia diagnosis, and the hospital of admission is one of the strongest predictors of diagnosis. Efforts to reduce rates of unwarranted asthmonia diagnosis are needed, particularly at small, rural, nonteaching hospitals with minimal pediatric specialty support.
Keywords: asthma, pneumonia, variation, quality of care, pediatrics
Respiratory illnesses are the most common reasons for pediatric hospitalization in the United States (1). Together, asthma and pneumonia account for over 1 million emergency department visits and 200,000 pediatric admissions annually, with direct costs of disease management exceeding six billion dollars each year (1–4). Since the publication of national guidelines for pediatric community-acquired pneumonia management in 2011, several studies have examined rates of guideline-concordant pneumonia management, with a particular focus on reducing unnecessary treatment with broad spectrum antibiotics (5–7). Similarly, numerous efforts have aimed to improve adherence to national and international guidelines for the management of asthma exacerbations; guidelines discourage routine performance of chest X-rays and laboratory tests and recommend against antibiotic treatment (8–11).
Accurate disease diagnosis is a prerequisite for adherence to treatment guidelines. Although several studies have shown that only 1–3% of pediatric asthma exacerbations cared for in emergency departments are complicated by radiographically confirmed pneumonia, in practice, misinterpretation of atelectasis as pulmonary consolidation on chest radiography may lead to asthma–pneumonia co-diagnosis rates considerably higher than 1–3% (12–14). In fact, at some hospitals, it is so common to see children with concomitant diagnoses of asthma and pneumonia that the term “asthmonia” has been coined to colloquially describe its presentation. Inappropriate co-diagnosis of asthma and pneumonia may result in overuse of antibiotics, adverse drug reactions, and increased treatment costs. In contrast, underdiagnosis may result in clinical deterioration or prolonged lengths of hospital stay.
The prevalence and consequences of asthmonia diagnosis are poorly understood, as the majority of research examining patterns of care has focused on asthma or on pneumonia, frequently excluding concurrent diagnoses from analysis. To address this knowledge gap, this study aimed to characterize variation in rates of pediatric asthmonia diagnosis across hospitals and analyze associations between the hospital rate of asthmonia diagnosis and clinical outcomes. We hypothesized that hospital rates of asthmonia diagnosis would vary substantially independent of patient characteristics and that increased rates of asthmonia diagnosis at the hospital level would be associated with higher health care costs and antibiotic use with no significant differences in rates of adverse outcomes.
Methods
Data Source
We conducted a cross-sectional study of pediatric asthma hospitalizations contained within the Premier Healthcare Database. Participating hospitals represent all regions of the United States and include a spectrum of hospital types, ranging from small community hospitals to freestanding children’s hospitals. Previous studies have validated the Premier sample of hospitalizations, demonstrating that the sex distribution, routine discharge status, and disease severity are similar between Premier-participating hospitals and a national sample (15, 16). The Premier Healthcare Database contains admitting and discharge diagnoses; hospital, physician, and patient characteristics; and a date-stamped log of items billed to the patient or insurer, including medications and laboratory and radiologic tests.
Participants
Children and adolescents 2–17 years of age were included if they were hospitalized between October 1, 2015, and June 30, 2018 with an asthma exacerbation. Asthma exacerbations were identified based on International Classification of Diseases, Tenth Edition, discharge diagnoses, including those with a primary diagnosis of asthma with exacerbation, those with a primary diagnosis of respiratory failure with a secondary diagnosis of asthma exacerbation, and those with a primary diagnosis of pneumonia with a secondary diagnosis of asthma with exacerbation. To maximize diagnostic specificity for exacerbation, eligible cases were required to have received systemic corticosteroids in the emergency department or on the first day of hospital admission. To focus on those who were not critically ill at the time of hospital admission, we excluded children who were cared for in an intensive care unit (ICU) or who received invasive or noninvasive ventilation on the first day of hospitalization. We also excluded cases that arrived at the hospital after interhospital transfer given our inability to evaluate the full course of treatment for these patients. Finally, to calculate stable hospital-level estimates, we excluded those who were admitted to hospitals with <10 pediatric asthma admissions during the study period.
Independent and Dependent Variables
Asthma–pneumonia co-diagnosis, hereafter referred to as asthmonia, was defined as an International Classification of Diseases, 10th Edition, diagnosis of pneumonia in a patient also diagnosed with an asthma exacerbation. This variable served as the primary independent variable in our analysis of clinical outcomes. In addition, we examined patient age, sex, race, ethnicity, primary form of insurance, the number of emergency department visits and hospitalizations for asthma at the index hospital in the year before admission, and the presence of co-occurring complex chronic conditions identified using a previously published algorithm (17). Admissions were categorized as taking place on a Level I medical–surgical bed or in a stepdown/intermediate care bed (Level II). The specialty of the admitting provider was categorized as either general pediatrics, specialty pediatrics, or other. Hospital characteristics included the number of beds, teaching status, census region, and urban versus rural location. Hospitals were categorized as freestanding children’s hospitals, defined as hospitals with a median patient age of <18 years, or general hospitals. Because the number of pediatric-specific services provided at general hospitals varies considerably, we calculated the number of pediatric subspecialty services available at each general hospital and stratified these hospitals into tertiles based on the number of these services. Hospitals with 0–3 pediatric specialty services were defined as having minimal pediatric subspecialty support, those with 4–10 were defined as having moderate subspecialty support, and those with more than 10 were defined as having extensive pediatric subspecialty support (18).
To examine differences in resource use between youth diagnosed with asthma alone relative to those with asthmonia, we determined frequencies and rates of diagnostic testing in the emergency department or on the day of admission. Diagnostic tests included receipt of one or more chest X-rays, complete blood counts, blood cultures, inflammatory markers (C-reactive protein and/or erythrocyte sedimentation rate), and electrolytes. Receipt of antibiotics was assessed through pharmacy charges.
To determine the relationships between asthmonia diagnosis and clinical outcomes, we examined 1) days of antibiotic therapy, a measure endorsed by the Centers for Disease Control and Prevention that sums the number of unique antibiotic agents administered to each patient, each day (such that a patient who received two different antibiotics each day for a total of 3 days would have a days of therapy measure of 6, thus accounting for antibiotic load in patients with a similar number of days of antibiotic treatment) (19); 2) total costs of the hospitalization, derived from hospitals’ cost accounting systems and reflecting actual hospital costs rather than charges; 3) transfer to a higher level of care during the hospitalization, defined as a transfer from a Level I to a Level II bed or from a Level I or II bed to the ICU or another hospital on or after the second day of hospitalization; 4) prolonged length of hospital stay, defined as a length of stay ≥5 days (>the 95th percentile for length of stay); and 5) all-cause readmission to the index hospital within 30 days of discharge.
Analysis
Our initial descriptive analyses were conducted at the patient level. To examine differences in baseline sociodemographic, clinical, and hospital characteristics between youth with asthma alone compared with those with asthmonia, we first calculated chi-square tests for categorical variables and t tests for continuous variables. We then examined unadjusted differences in rates of diagnostic testing, antibiotic administration, and healthcare costs.
At the hospital level, we computed the rate of asthmonia diagnosis as the proportion of each hospital’s asthma admissions that received a co-occurring pneumonia diagnosis. To calculate hospital-level risk-standardized asthmonia rates, we developed a hierarchical generalized logistic regression model with hospital random effects, adjusting for patient age, sex, race/ethnicity, payer, previous history of asthma hospitalizations or emergency department visits, and complex chronic conditions. For each hospital, the risk-standardized rate was computed as the ratio of the hospital predicted rate (using the hospital patient mix and hospital random effect) to the hospital expected rate (using the hospital patient mix and average effect), multiplied by the overall asthmonia rate. We determined the median, interquartile range (IQR), and range of the hospital-level rate of asthmonia diagnosis for both the unadjusted and risk-standardized rates. We also calculated the median odds ratio (MOR) to quantify the magnitude of the effect of hospital-level clustering on asthmonia diagnosis (20). For all pairs of patients with the same characteristics from different hospitals, the MOR is defined as the median value of the odds ratio between the individual with the greatest likelihood of receiving an asthmonia diagnosis and the individual least likely to be diagnosed with asthmonia, with differences in asthmonia risk based on hospital-level random effects.
To identify hospital characteristics associated with the rate of asthmonia diagnosis, we stratified hospitals into quartiles based on their risk-adjusted diagnostic rate and examined differences in hospital characteristics across the quartiles using chi-square and Wilcoxon rank sum tests. In addition, we examined the relationships between hospitals’ risk-standardized asthmonia diagnosis rates and our five hospital-level outcomes of interest using generalized linear models, adjusting for hospital characteristics including census region, hospital type, bed size, urban–rural region, and hospital teaching status.
This study was reviewed and approved by the Institutional Review Board at Baystate Medical Center. Analyses were conducted using SAS (version 9.4; SAS Institute) and Stata (version 16; StataCorp) software. All statistical testing was two sided, with P < 0.05 considered statistically significant.
Results
Participants
After applying our eligibility criteria, 24,606 hospitalizations for asthma exacerbation at 274 hospitals were included in our analysis (see Figure E1 in the online supplement). Of these, 19,402 (78.9%) were diagnosed with asthma alone, whereas 5,204 (21.1%) received an asthmonia diagnosis. Children diagnosed with asthmonia were younger, more likely to be non-Hispanic white, and more likely to have one or more complex chronic conditions than children without a concurrent pneumonia diagnosis (Table 1). Children with asthmonia were also significantly less likely to have been seen in the emergency department or admitted to the index hospital with asthma in the preceding year. Compared with children with asthma alone, those with an asthmonia diagnosis were significantly less likely to be admitted to teaching hospitals and more likely to be admitted to general hospitals with minimal or moderate levels of pediatric subspecialty support.
Table 1.
Characteristics of children hospitalized with asthma exacerbation, with and without a concurrent diagnosis of pneumonia
| Patient Characteristics | All Encounters (n = 24,606) | Asthmonia Diagnosis (n = 5,204) | Asthma Exacerbation Alone (n = 19,402) | P Value |
|---|---|---|---|---|
| Age in yr, median (IQR) | 6 (3–9) | 5 (3–8) | 6 (3–9) | <0.001 |
| Sex, female | 9,786 (39.8) | 2,364 (45.4) | 7,422 (38.3) | <0.001 |
| Race/ethnicity | <0.001 | |||
| Non-Hispanic White | 8,557 (34.8) | 2,006 (38.6) | 6,551 (33.8) | — |
| Non-Hispanic Black | 8,145 (33.1) | 1,370 (26.3) | 6,775 (34.9) | — |
| Hispanic | 3,754 (15.3) | 895 (17.2) | 2,859 (14.7) | — |
| Other/unknown | 4,150 (16.9) | 933 (17.9) | 3,217 (16.6) | — |
| Primary payer | <0.01 | |||
| Private | 7,281 (29.6) | 1,642 (31.6) | 5,639 (29.1) | — |
| Medicaid | 15,704 (63.8) | 3,217 (61.8) | 12,487 (64.4) | — |
| Other/unknown/uninsured | 1,621 (6.6) | 345 (6.6) | 1,276 (6.6) | — |
| Complex chronic conditions | <0.001 | |||
| Single organ system | 1,058 (4.3) | 353 (6.8) | 705 (3.6) | — |
| Multiple organ systems | 376 (1.5) | 184 (3.5) | 192 (1.0) | — |
| ≥1 ED/hospital encounters for asthma in year before index admission | 6,191 (25.2) | 935 (18.0) | 5,256 (27.1) | <0.001 |
| Hospital characteristics | ||||
| Attending provider specialty | <0.001 | |||
| General pediatrician | 17,479 (71.0) | 3,575 (68.7) | 13,904 (71.7) | — |
| Subspecialty pediatrician | 3,153 (12.8) | 853 (16.4) | 2,300 (11.9) | — |
| Other | 3,974 (16.2) | 776 (14.9) | 3,198 (16.5) | — |
| Admitting location | 0.59 | |||
| Level I medical–surgical bed | 23,307 (94.7) | 4,937 (94.9) | 18,370 (94.7) | — |
| Level II intermediate care | 1,299 (5.3) | 267 (5.1) | 1,032 (5.3) | — |
| Geographic region | <0.001 | |||
| South | 13,851 (56.3) | 3,075 (59.1) | 10,776 (55.5) | — |
| Northeast | 3,484 (14.2) | 620 (11.9) | 2,864 (14.8) | — |
| Midwest | 3,936 (16.0) | 795 (15.3) | 3,141 (16.2) | — |
| West | 3,335 (13.6) | 714 (13.7) | 2,621 (13.5) | — |
| Number of beds | <0.001 | |||
| <200 | 3,283 (13.3) | 803 (15.4) | 2,480 (12.8) | — |
| 201–400 | 7,847 (31.9) | 1,834 (35.2) | 6,013 (31.0) | — |
| >400 | 13,476 (54.8) | 2,567 (49.3) | 10,909 (56.2) | — |
| Urban region | 21,992 (89.4) | 4,476 (86.0) | 17,516 (90.3) | <0.001 |
| Teaching hospital | 12,724 (51.7) | 2,195 (42.2) | 10,529 (54.3) | <0.001 |
| Hospital type | <0.001 | |||
| General hospitals with minimal pediatric specialty support* | 3,468 (14.1) | 878 (16.9) | 2,590 (13.4) | — |
| General hospitals with moderate pediatric specialty support* | 4,283 (17.4) | 989 (19.0) | 3,294 (17.0) | — |
| General hospitals with extensive pediatric specialty support* | 13,566 (55.1) | 2,731 (52.5) | 10,835 (55.8) | — |
| Freestanding children’s hospital† | 3,289 (13.4) | 606 (11.6) | 2,683 (13.8) | — |
Definition of abbreviations: ED = emergency department; IQR = interquartile range.
Data are presented as n (%) unless otherwise stated.
Pediatric specialty support defined by number of pediatric subspecialty services at general hospitals (minimal: 0–3; moderate: 4–10; extensive: >10).
Freestanding hospitals defined as hospitals with a median patient age <18 years.
Resource Use and Asthmonia Diagnosis
Children diagnosed with asthmonia were significantly more likely to receive all diagnostic tests, either on the day of admission or multiple times during the hospital admission (Table 2). In addition, they were significantly more likely to be treated with at least one antibiotic (n = 4,740, 91.1%) compared with 18.9% (n = 3,667) of those with asthma alone (P < 0.001). Children with asthmonia were more likely to be transferred to a higher level of care, to experience prolonged lengths of stay, to be readmitted to the hospital within 30 days of discharge, and to incur significantly higher hospital costs.
Table 2.
Resource use among children hospitalized with asthma exacerbation, with and without a concurrent diagnosis of pneumonia
| All Encounters (n = 24,606) | Asthmonia Diagnosis (n = 5,204) | Asthma Exacerbation Alone (n = 19,402) | P Value | |
|---|---|---|---|---|
| Diagnostic testing at hospital admission | ||||
| Chest X-ray | 14,845 (60.3) | 4,011 (77.1) | 10,834 (55.8) | <0.001 |
| Complete blood count | 10,706 (43.5) | 3,632 (69.8) | 7,074 (36.5) | <0.001 |
| Inflammatory markers | 2,097 (8.5) | 1,044 (20.1) | 1,053 (5.4) | <0.001 |
| Electrolytes | 10,601 (43.1) | 3,172 (61.0) | 7,429 (38.3) | <0.001 |
| Blood culture | 4,567 (18.6) | 2,347 (45.1) | 2,220 (11.4) | <0.001 |
| Recurrent diagnostic testing during hospitalization | ||||
| Chest X-ray | 1,116 (4.5) | 680 (13.1) | 436 (2.2) | <0.001 |
| Complete blood count | 1,172 (4.8) | 633 (12.2) | 539 (2.8) | <0.001 |
| Inflammatory markers | 321 (1.3) | 238 (4.6) | 83 (0.4) | <0.001 |
| Electrolytes | 1,553 (6.3) | 596 (11.5) | 957 (4.9) | <0.001 |
| Blood culture | 69 (0.3) | 42 (0.8) | 27 (0.1) | <0.001 |
| Antibiotic administration (any) | 8,407 (34.2) | 4,740 (91.1) | 3,667 (18.9) | <0.001 |
| Antibiotic class when antibiotics given | <0.001 | |||
| Aminopenicillin alone | 1,699 (20.2) | 632 (13.3) | 1,067 (29.1) | — |
| Cephalosporins alone | 1,470 (17.5) | 762 (16.1) | 708 (19.3) | — |
| Macrolide alone | 2,394 (28.5) | 1,251 (26.4) | 1,143 (31.2) | — |
| Other individual antibiotic | 243 (2.9) | 145 (3.1) | 98 (2.7) | — |
| >1 antibiotic | 2,601 (31.0) | 1,950 (41.1) | 651 (17.8) | — |
| Days of antibiotic therapy, median (IQR) | 2 (1–4) | 3 (2–4) | 2 (1–3) | <0.001 |
| Total costs of hospitalization, USD, median (IQR) | 3,217 (2,209–4,973) | 4,050 (2,628–6,383) | 3,029 (2,127–4,637) | <0.001 |
| Transfer to higher level of care | 441 (1.8) | 161 (3.1) | 280 (1.4) | <0.001 |
| Prolonged length of stay* | 767 (3.1) | 417 (8.0) | 350 (1.8) | <0.001 |
| 30-d readmission | 950 (3.9) | 229 (4.4) | 721 (3.7) | 0.02 |
Definition of abbreviations: IQR = interquartile range; USD = United States dollar.
Data are presented as n (%) unless otherwise stated.
Defined as length of stay ≥5 days (>the 95th percentile for length of stay).
Hospital Variation in Asthmonia Diagnosis
We observed significant variation in rates of asthmonia diagnosis across the 274 institutions included in the hospital-level analysis (Figure 1). The median unadjusted rate of asthma–pneumonia co-diagnosis was 22.3%, with an IQR of 15.4–33.3% and a range of 4.0–70.0%. When adjusted for patient-level risk factors, the median risk-standardized asthmonia diagnosis rate was 20.9% (IQR, 16.2–27.2%; range 8.4–55.9%). The MOR, calculated to quantify the magnitude of the effect of hospital of admission on asthmonia diagnosis, was 1.75 (95% confidence interval [CI], 1.63–1.86). This odds ratio was higher than all of the odds ratios for patient-level characteristics included in our risk adjustment model (Table 3), with the exception of complex chronic conditions, indicating that the hospital of admission was a stronger predictor of asthmonia diagnosis than almost all other patient characteristics.
Figure 1.
Hospitals’ unadjusted and risk-standardized rates of asthmonia diagnosis at 274 hospitals with 10 or more asthma exacerbations during study period with associated 95% confidence intervals, ordered from lowest to highest rate of asthma–pneumonia co-diagnosis. (A) Unadjusted rates and (B) Risk-standardized rates.
Table 3.
Results of hierarchical generalized logistic regression indicating factors associated with hospital rate of asthmonia diagnosis
| Odds Ratio (95% CI) | |
|---|---|
| Patient characteristics | |
| Age (yr) | 0.94 (0.94–0.95) |
| Female sex (male is referent) | 1.40 (1.31–1.50) |
| Race/ethnicity (non-Hispanic White is referent) | |
| Non-Hispanic Black | 0.74 (0.67–0.81) |
| Hispanic | 0.98 (0.88–1.10) |
| Other/unknown | 0.87 (0.78–0.97) |
| Primary payer (private is referent) | |
| Medicaid | 0.95 (0.88–1.03) |
| Other/unknown/uninsured | 0.99 (0.86–1.15) |
| Complex chronic conditions (none is referent) | |
| Single organ system | 2.23 (1.94–2.57) |
| Multiple organ systems | 4.25 (3.42–5.28) |
| ≥1 ED/hospital encounters for asthma in year before index admission | 0.66 (0.60–0.71) |
| Median odds ratio* | 1.75 (1.63–1.86) |
Definition of abbreviations: CI = confidence interval; ED = emergency department.
The median odds ratio is defined for all pairs of patients from different hospitals with the same characteristics as the median value of the odds ratio between the individual with the greatest likelihood of receiving an asthmonia diagnosis and the individual least likely to be diagnosed with asthmonia, with differences in asthmonia risk based on hospital-level random effects.
Hospital Characteristics and Asthmonia Diagnosis Rates
We observed several differences in hospital characteristics across the four quartiles of risk-standardized rates of asthmonia diagnosis (Table 4). Compared with hospitals in the lowest quartile of asthmonia diagnosis, those in the highest quartile were more likely to be smaller, nonteaching, rural hospitals. We also observed a relationship between hospital volume of pediatric asthma hospitalizations and asthmonia diagnosis rates, with lower volume hospitals having significantly higher diagnostic rates.
Table 4.
Hospital characteristics associated with risk-standardized asthmonia diagnosis rates
| Hospital Risk-Standardized Asthma–Pneumonia Co-diagnosis Rate |
P Value* | ||||
|---|---|---|---|---|---|
| Quartile 1 Asthmonia Diagnosis Rate 8.4–16.1% | Quartile 2 Asthmonia Diagnosis Rate 16.2–20.9% | Quartile 3 Asthmonia Diagnosis Rate 21.0–27.2% | Quartile 4 Asthmonia Diagnosis Rate 27.4–55.9% | ||
| Number of asthma hospitalizations during study period, median (IQR) | 107 (44–186) | 49 (22–122) | 31 (16–60) | 36 (17–55) | <0.001 |
| Geographical region | — | — | — | — | 0.77 |
| West | 12 (17.7) | 13 (18.8) | 16 (23.2) | 11 (16.2) | — |
| South | 31 (45.6) | 34 (49.3) | 34 (49.3) | 38 (55.9) | — |
| Northeast | 11 (16.2) | 8 (11.6) | 8 (11.6) | 4 (5.9) | — |
| Midwest | 14 (20.6) | 14 (20.3) | 11 (15.9) | 15 (22.1) | — |
| Number of beds | — | — | — | — | <0.001 |
| <200 | 9 (13.2) | 17 (24.6) | 21 (30.4) | 29 (42.7) | — |
| 201–400 | 17 (25.0) | 28 (40.6) | 27 (39.1) | 28 (41.2) | — |
| >400 | 42 (61.8) | 24 (34.8) | 21 (30.4) | 11 (16.2) | — |
| Urban vs. rural | — | — | — | — | 0.002 |
| Rural | 8 (11.8) | 14 (20.3) | 21 (30.4) | 26 (38.2) | — |
| Urban | 60 (88.2) | 55 (79.7) | 48 (69.6) | 42 (61.8) | — |
| Teaching status | — | — | — | — | <0.001 |
| Nonteaching | 29 (42.7) | 41 (59.4) | 44 (63.8) | 57 (83.8) | — |
| Teaching | 39 (57.4) | 28 (40.6) | 25 (36.2) | 11 (16.2) | — |
| Hospital type | — | — | — | — | 0.001 |
| General hospitals with minimal pediatric specialty support† | 8 (11.8) | 13 (18.8) | 23 (33.3) | 28 (41.2) | — |
| General hospitals with moderate pediatric specialty support† | 17 (25.0) | 20 (29.0) | 23 (33.3) | 21 (30.9) | — |
| General hospitals with extensive pediatric specialty support† | 39 (57.4) | 34 (49.3) | 21 (30.4) | 18 (26.5) | — |
| Freestanding children’s hospitals‡ | 4 (5.9) | 2 (2.9) | 2 (2.9) | 1 (1.5) | — |
Definition of abbreviation: IQR = interquartile range.
P value calculated using chi-square and Wilcoxon rank sum tests examining differences across all four quartiles.
Pediatric specialty support defined by number of pediatric subspecialty services at general hospitals (minimal: 0–3; moderate: 4–10; extensive: >10).
Freestanding hospitals defined as hospitals with a median patient age <18 years.
Relationship between Hospital Asthmonia Diagnostic Rates and Clinical Outcomes
In our adjusted models of hospital-level risk-standardized asthmonia diagnosis rates and clinical outcomes, higher rates of asthmonia diagnosis were associated with significantly higher median days of antibiotic therapy (β = 0.02; 95% CI, 0.01 to 0.03; P < 0.001), proportion of patients with prolonged lengths of stay (β = 0.12; 95% CI, 0.03 to 0.21; P < 0.01), and median hospital costs (β = 0.26; 95% CI, 0.00 to 0.51; P < 0.05), with no significant differences in risk of transfer to higher level of care (β = 0.06; 95%CI, −0.01 to –0.13; P = 0.09) or 30-day readmission (β = 0.02; 95% CI,−0.03 to 0.07; P = 0.32). Expressed in terms of a 10% increase in the hospital asthmonia rate, this is equivalent to a $260 higher median hospitalization cost. These findings are summarized in Table E1.
Discussion
Although research indicates that <5% of pediatric asthma exacerbations are complicated by radiographically confirmed pneumonia, this study found that more than one in five children with an asthma exacerbation were also diagnosed with pneumonia. Rates of asthmonia diagnosis varied substantially across hospitals, even after accounting for patient characteristics. Although the average risk-standardized hospital rate of asthmonia diagnosis was 20%, at one in four hospitals, rates of asthma–pneumonia co-diagnosis exceeded 27%, and at some sites, it reached 56%. Hospitals with high risk-standardized asthmonia diagnosis rates used significantly more antibiotics and incurred significantly higher costs, and their patients were more likely to experience prolonged lengths of stay. In contrast, rates of transfer to a higher level of care and hospital readmission, outcomes that would be expected to vary with disease severity, were not associated with the hospital asthmonia rate after risk adjustment for observable patient sociodemographic and clinical characteristics. Taken together, our findings suggest that improving diagnostic practices regarding the presence or absence of pneumonia at the time of an asthma exacerbation represents an opportunity to enhance the value of hospital-based care.
A recent meta-analysis demonstrated that children with asthma are at significantly increased risk of pneumonia and invasive pneumococcal disease, perhaps explained by their higher prevalence of oropharyngeal colonization by Streptococcus pneumoniae when receiving inhaled corticosteroid therapy and/or impaired airway clearance (21). Correspondingly, consideration of pneumonia is prudent in children and adolescents admitted with an asthma exacerbation. As expected, children and adolescents diagnosed with co-occurring asthma and pneumonia were, on average, sicker than those not diagnosed with pneumonia, evidenced by prolonged lengths of stay and higher rates of transfer to a higher level of care and 30-day readmission. Paradoxically, these associations did not hold true at the hospital level, where we saw substantial variation in rates of asthmonia diagnosis, even after adjustment for observable patient characteristics. With the exception of co-occurring complex chronic conditions, the MOR indicates that the hospital of admission was the strongest predictor of asthma–pneumonia co-diagnosis. This finding, together with the weak associations between the hospital asthmonia rate and clinical outcomes, suggests that the variation across hospitals is at least partially attributable to local diagnostic practices rather than underlying differences in the patient populations that hospitals serve.
As expected, asthmonia diagnosis was associated with increased antibiotic use. Only one in eight children diagnosed with asthmonia received aminopenicillin monotherapy (the first line antibiotic recommended for community-acquired pneumonia management), and more than one-third received multiple antibiotics, increasing risks of adverse drug reactions and antibiotic resistance (5). In addition, we observed very high rates of diagnostic testing that well exceeded benchmarks of care for hospitalized children with asthma (22, 23). These patterns of diagnostic testing illustrate the need to implement guidelines by the Global Initiative for Asthma and the National Heart, Lung, and Blood Institute to minimize routine chest radiography and laboratory screening for asthma exacerbations (8, 9).
In recent years, a number of quality improvement initiatives at freestanding children’s hospitals have successfully reduced chest radiography and antibiotic administration for pediatric asthma exacerbations (10, 11, 24, 25). We found that the highest rates of asthmonia diagnosis occurred in small, rural, nonteaching hospitals with lower patient volumes, indicating a need for focused quality improvement in these settings. Although past research has identified relationships between clinical volumes and patient outcomes, particularly for surgical conditions, this analysis represents one of the first to demonstrate a similar association for the care of common pediatric medical conditions (26–28). Quality improvement initiatives to reduce chest radiography and antibiotic administration in rural, nonteaching hospitals will need to be tailored to the unique challenges experienced in these settings, including a relative lack of pediatric specialists and resources, substantial physical distances from pediatric ICUs, and limited infrastructure for quality improvement. Dissemination of clinical decision support tools from children’s hospitals to smaller hospitals in the region, such as clinical practice guidelines and reminders embedded within the electronic health record, may be a potential mechanism to improve the quality of care in lower volume nonpediatric hospitals.
The results of this study should be interpreted in light of its limitations. First, by including only diagnostic codes for asthma exacerbation in combination with oral or intravenous corticosteroid administration, we may have omitted exacerbations that did not receive guideline-endorsed corticosteroids. Second, our analyses of diagnostic testing and antibiotic administration are limited to the index hospitalization, as the dataset does not contain information about resources used in other settings. Correspondingly, our results may underestimate resource use. Third, our risk adjustment models cannot account for unmeasured confounding factors that may be associated with asthmonia diagnosis. Related to this, we have limited ability to account for disease severity in our models as diagnostic test results are not available within the dataset. In particular, we acknowledge the potential for confounding by indication in our analysis of associations between asthmonia diagnosis and clinical outcomes. Finally, studies have shown poor interrater reliability in identifying definitive pneumonia on chest X-ray; although three past studies suggest that the “correct rate” of asthma–pneumonia co-diagnosis is 1.6–3.2%, this is challenging to validate given the limitations of chest radiography and lack of other validated diagnostic tools (29).
In conclusion, our finding that more than 1 in 5 children and adolescents admitted for asthma exacerbations were diagnosed with co-occurring pneumonia far exceeds previous studies and indicates an important area for quality improvement. At the patient level, asthmonia diagnosis was associated with worse clinical outcomes relative to an asthma diagnosis alone. However, our hospital-level analysis indicates that hospitals vary widely in their patterns of diagnostic testing and coding practices; the hospital of admission was one of the strongest predictors of asthmonia diagnosis. Efforts to reduce rates of diagnostic testing and rates of asthmonia diagnosis are needed, particularly at small, rural, nonteaching hospitals with minimal pediatric specialty support.
Footnotes
Supported by National Health, Lung, and Blood Institute (NHLBI) grant K24 HL132008 (P.K.L.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Author Contributions: J.K.L., M.-S.S., P.S.P., and P.K.L. made substantial contributions to the conception or design of the work or the analysis and interpretation of the data. J.K.L. drafted the manuscript, and all authors reviewed the manuscript in detail and made valuable revisions to its content. M.-S.S. conducted all analyses under the supervision of P.S.P. and P.K.L. All authors provide final approval of this paper for publication and agree to be accountable for all aspects of the work.
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.
Author disclosures are available with the text of this article at www.atsjournals.org.
References
- 1. Leyenaar JK, Ralston SL, Shieh MS, 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;11:743–749. doi: 10.1002/jhm.2624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Perry R, Braileanu G, Palmer T, Stevens P. The economic burden of pediatric asthma in the United States: literature review of current evidence. Pharmacoeconomics. 2019;37:155–167. doi: 10.1007/s40273-018-0726-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Chamberlain JK, Cull WL, Melgar T, Kaelber DC, Kan BD. The effect of dual training in internal medicine and pediatrics on the career path and job search experience of pediatric graduates. J Pediatr. 2007;151:419–424. doi: 10.1016/j.jpeds.2007.04.064. [DOI] [PubMed] [Google Scholar]
- 4.McDermott KW, Stocks C, Freeman WJ. Healthcare Cost and Utilization Project (HCUP) statistical briefs. Rockville, MD: Agency for Healthcare Research and Quality; 2018. Overview of pediatric emergency department visits, 2015: Statistical Brief #242. [PubMed] [Google Scholar]
- 5. Bradley JS, Byington CL, Shah SS, Alverson B, Carter ER, Harrison C, et al. Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. The management of community-acquired pneumonia in infants and children older than 3 months of age: clinical practice guidelines by the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. Clin Infect Dis. 2011;53:e25–e76. doi: 10.1093/cid/cir531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Ambroggio L, Thomson J, Murtagh Kurowski E, Courter J, Statile A, Graham C, et al. Quality improvement methods increase appropriate antibiotic prescribing for childhood pneumonia. Pediatrics. 2013;131:e1623–e1631. doi: 10.1542/peds.2012-2635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Parikh K, Biondi E, Nazif J, Wasif F, Williams DJ, Nichols E, et al. Value in Inpatient Pediatrics Network Quality Collaborative For Improving Care In Community Acquired Pneumonia. A multicenter collaborative to improve care of community acquired pneumonia in hospitalized children. Pediatrics. 2017;139:e20161411. doi: 10.1542/peds.2016-1411. [DOI] [PubMed] [Google Scholar]
- 8.Global Initiative for Asthma. 2020.
- 9.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. [Google Scholar]
- 10. Watnick CS, Arnold DH, Latuska R, O’Connor M, Johnson DP. Successful chest radiograph reduction by using quality improvement methodology for children with asthma. Pediatrics. 2018;142:e20174003. doi: 10.1542/peds.2017-4003. [DOI] [PubMed] [Google Scholar]
- 11. Dorzin SE, Halaby C, Quintos ML, Noor A, El-Chaar G. Antimicrobial stewardship program using plan-do-study-act cycles to reduce unjustified antibiotic prescribing in children admitted with an asthma exacerbation. J Pediatr Pharmacol Ther. 2017;22:436–443. doi: 10.5863/1551-6776-22.6.436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Narayanan S, Magruder T, Walley SC, Powers T, Wall TC. Relevance of chest radiography in pediatric inpatients with asthma. J Asthma. 2014;51:751–755. doi: 10.3109/02770903.2014.909459. [DOI] [PubMed] [Google Scholar]
- 13. Gershel JC, Goldman HS, Stein RE, Shelov SP, Ziprkowski M. The usefulness of chest radiographs in first asthma attacks. N Engl J Med. 1983;309:336–339. doi: 10.1056/NEJM198308113090603. [DOI] [PubMed] [Google Scholar]
- 14. Florin TA, Carron H, Huang G, Shah SS, Ruddy R, Ambroggio L. Pneumonia in children presenting to the emergency department with an asthma exacerbation. JAMA Pediatr. 2016;170:803–805. doi: 10.1001/jamapediatrics.2016.0310. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. 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. doi: 10.1002/pds.2063. [DOI] [PubMed] [Google Scholar]
- 16. Lasky T, Greenspan J, Ernst FR, Gonzalez L. Pediatric vancomycin use in 421 hospitals in the United States, 2008. PLoS One. 2012;7:e43258. doi: 10.1371/journal.pone.0043258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Feudtner C, Feinstein JA, Zhong W, Hall M, Dai D. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatr. 2014;14:199. doi: 10.1186/1471-2431-14-199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Jewell MJ, Leyenaar J, Shieh M-S, Pekow PS, Stefan M, Lindenauer PK. Unnecessary antibiotic prescribing in children hospitalised for asthma exacerbation: a retrospective national cohort study. BMJ Qual Saf. doi: 10.1136/bmjqs-2019-010792. [DOI] [PubMed] [Google Scholar]
- 19.Centers for Disease Control. 2021.
- 20. Austin PC, Merlo J. Intermediate and advanced topics in multilevel logistic regression analysis. Stat Med. 2017;36:3257–3277. doi: 10.1002/sim.7336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Castro-Rodriguez JA, Abarca K, Forno E. Asthma and the risk of invasive pneumococcal disease: a meta-analysis. Pediatrics. 2020;145:e20191200. doi: 10.1542/peds.2019-1200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Parikh K, Hall M, Mittal V, Montalbano A, Mussman GM, Morse RB, et al. Establishing benchmarks for the hospitalized care of children with asthma, bronchiolitis, and pneumonia. Pediatrics. 2014;134:555–562. doi: 10.1542/peds.2014-1052. [DOI] [PubMed] [Google Scholar]
- 23. Knapp JF, Hall M, Sharma V. Benchmarks for the emergency department care of children with asthma, bronchiolitis, and croup. Pediatr Emerg Care. 2010;26:364–369. doi: 10.1097/PEC.0b013e3181db2262. [DOI] [PubMed] [Google Scholar]
- 24. Gildenhuys J, Lee M, Isbister GK. Does implementation of a paediatric asthma clinical practice guideline worksheet change clinical practice? Int J Emerg Med. 2009;2:33–39. doi: 10.1007/s12245-008-0063-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Bekmezian A, Fee C, Weber E. Clinical pathway improves pediatrics asthma management in the emergency department and reduces admissions. J Asthma. 2015;52:806–814. doi: 10.3109/02770903.2015.1019086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Sewalt CA, Wiegers EJA, Venema E, Lecky FE, Schuit SCE, Den Hartog D, et al. The volume-outcome relationship in severely injured patients: a systematic review and meta-analysis. J Trauma Acute Care Surg. 2018;85:810–819. doi: 10.1097/TA.0000000000002043. [DOI] [PubMed] [Google Scholar]
- 27. Phillips P, Poku E, Essat M, Woods HB, Goka EA, Kaltenthaler EC, et al. Systematic review of carotid artery procedures and the volume-outcome relationship in Europe. Br J Surg. 2017;104:1273–1283. doi: 10.1002/bjs.10593. [DOI] [PubMed] [Google Scholar]
- 28. Gooiker GA, van Gijn W, Wouters MW, Post PN, van de Velde CJ, Tollenaar RA. Signalling Committee Cancer of the Dutch Cancer Society. Systematic review and meta-analysis of the volume-outcome relationship in pancreatic surgery. Br J Surg. 2011;98:485–494. doi: 10.1002/bjs.7413. [DOI] [PubMed] [Google Scholar]
- 29. Florin TA, Gerber JS. Sticking by an imperfect standard: chest radiography for pediatric community-acquired pneumonia. Pediatrics. 2020;145:e20193900. doi: 10.1542/peds.2019-3900. [DOI] [PubMed] [Google Scholar]