Background

Asthma is a chronic disease affecting over 16 million adults and 5 million children in the United States alone, with nearly 20% requiring hospitalization annually. Worldwide estimates place the number of asthmatics at over 300 million.¹ In a recent report from the Centers for Disease Control and Prevention, it was estimated that from 2011 to 2016, among 161 million working adults in the United States, 7% had asthma. Among asthmatic, 45% experienced an asthma attack, and 10% had an asthma‐related emergency department (ED) visit in the prior year.² Asthma prevalence is highest among workers employed in the health care and social assistance industry (8.8%) and in health care support occupations (8.8%). Efforts to improve outpatient care for asthma have been promoted through the development of treatment guidelines, but many patients remain either undertreated or undiagnosed. Data from 2016 to 2018 demonstrated there were more than 1.6 million ED visits in the United States per year. The rate of ED visits per 10,000 population is considerably higher among children (88) than adults (42).³ It is also higher among women (50) than men (31). Rates of ED visits in asthma decrease with increasing age. In addition, blacks have significantly higher rates of visits for asthma (164), as compared to Hispanics (60) and whites (31).

While standard ED asthma management is straightforward (including bronchodilators, steroids, and occasionally adjunctive treatments including magnesium), a certain percentage of refractory asthmatics will require hospitalization and monitoring.⁴ Since this decision is often the crux of the ED evaluation of a patient with asthma, studies have sought to classify asthma severity and predict which patients will require hospitalization and which, if discharged, will relapse.

Clinical question

Are there scoring systems for ED patients that accurately and reliably predict asthma severity, the need for hospitalization, or relapses after treatment?

Distinct pediatric and adult acute asthma short‐term outcome predictors and severity assessment scores exist. Therefore, this section presents pediatric and adult instruments separately.

Reproducibility of the clinical exam

In 2003, Stevens et al.⁵ examined the interrater reliability of physical exam findings in children with asthma to address the concern that findings from clinical trials may not represent actual clinical practice because specially trained staff are usually involved in the research setting. Reproducibility of key physical findings and an overall global gestalt about the severity of a patient’s condition by different people across different levels of training is the first step in establishing that any scoring system can be successfully developed and used clinically. The observers in this study included pediatric emergency physicians (n = 20), pediatric ED nurses (n = 50), and hospital respiratory therapists (n = 50). The observers received no prior specialized training on physical exam assessment for the study. Patients aged 1–16 years from a large urban children’s hospital presenting with acute asthma were the examination subjects. Observers independently and simultaneously rated the following aspects on a standardized form on a scale of one to three or four: work of breathing, wheezing, decreased air entry, increased expiratory time, breathlessness, mental status, and respiratory rate. A global assessment of “overall” severity (options included asymptomatic, mild, moderate, and severe) was also presented, as was a composite total score.

Weighted kappa statistics for each component of the score (used to determine interrater reliability) for 230 pairs of exams ranged from 0.61 to 0.74, while the overall severity (weighted kappa 0.80) and total scores (weighted kappa 0.82) had excellent agreement. Paired observers who were practitioners of the same profession (physicians, nurses, or respiratory therapists) had slightly better interrater reliability for the elements assessed. The authors felt this supported the use of structured and standardized formats to assess pediatric asthmatic patients. More reassuring was that, among a diverse group of care providers, a high level of agreement was found in the clinical assessment of acute asthmatics.

Severity scores in asthmatic children

Numerous pediatric asthma scores have been developed to assess asthma, including discriminative scores (designed to gauge severity at a single point in time), predictive scores (aimed at predicting particular outcomes), and evaluative scores (allowing documentation of changes over time). Many of the early scoring systems were developed using small numbers of select subjects, impacting the generalizability of the results. Of the 16 pediatric asthma scores as of 1994, 11 had sample sizes of <100 study subjects, and only one had >300 study subjects.⁶ More concerning and critical is that most were not (and have not been) thoroughly evaluated; none have been validated or assessed in terms of their impact on clinical behavior.⁷

In 2004, Gorelick et al.⁸ published a pediatric asthma score, the Pediatric Assessment Severity Score (PASS), that was developed and vigorously validated on a large diverse pediatric asthma population, with no exclusions based on severity or disposition (home or admission). The PASS was validated and tested for reliability and responsiveness on a group of 1221 pediatric asthmatics in two EDs with an enrollment rate of 89% (out of 1379 eligible patients). Forty‐one percent (n = 503) were admitted to an inpatient service. Clinical items examined during the study for inclusion in the final score had been included in prior clinical asthma scores and were acceptable to and pertinent by the clinicians at the study sites. The final three‐item score included an assessment of wheezing (none or mild, moderate, or severe or absent due to poor air exchange), work of breathing (none or mild, moderate, or severe), and prolongation of expiration (normal or mildly prolonged, moderately prolonged, or severely prolonged). Items assessed but not included in the final score included air entry, tachypnea, and mental status.

The three‐item PASS score discriminated admitted versus discharged patients with a high level of confidence: area under the curve (AUC) for the receiver operator characteristic (ROC) curves for each of the two EDs was 0.83 (confidence interval [CI] 0.80–0.86) and 0.85 (CI 0.81–0.89). The new score also was responsive to changes (e.g., improvements) on serial assessment; scores improved by 51–79% among those discharged home, whereas in those asthmatics admitted for inpatient care the scores improved only 25–32%. By comparison, the peak expiratory flow rate (PEFR) also improved by 25–32%, but the change was similar between admitted and discharged patients.

In 2008, Ducharme et al.^9,10 assessed whether their previously published Preschool Respiratory Assessment Measure (which had been developed and internally validated in children aged 3–6 years) could be expanded as the Pediatric Respiratory Assessment Measure (PRAM). They examined the validity, responsiveness, and reliability of PRAM in children from 2 to 17 years old, and found that it had predictive validity in that higher values of PRAM were associated with higher rates of admission, both at triage and after initial bronchodilation. PRAM was also responsive to change, as demonstrated by a Guyatt responsiveness coefficient of 0.7 (used to detect the ability of an evaluative instrument to detect change over time; 0.5 and 0.8 are considered moderate to large effect sizes, respectively), which was calculated by determining the ratio of the change in PRAM after initial bronchodilation among discharged patients to the standard deviation of the change in PRAM of the patients who were admitted. Last, the overall score was found to have high interrater reliability when assessed by a physician and a nurse (κ = 0.78) for 254 children aged 2–17 years. PRAM, which contains more components than PASS, measures scalene muscle contraction, suprasternal retractions, wheezing, air entry, and oxygen saturation (SpO2).

When Gouin et al.¹¹ performed a prospective external validation of both PRAM and PASS on 283 patients in 2010, they found that both scores had good initial discriminative ability. The scores were measured at triage and again after 90 minutes. Their initial performance as predictors of a length of stay of greater than 6 hours and/or admission was equal (PRAM AUC = 0.69 [CI 0.59–0.79] and PASS AUC = 0.70 [CI 0.60–0.80]) but, at 90 minutes, PRAM improved (AUC = 0.82 [CI 0.73–0.90]) while PASS remained steady (AUC = 0.72 [CI 0.62–0.82]). Nevertheless, given that this is the only external comparison of both scores, PASS remains a valid prognostic screening instrument.

Predicting hospitalization in asthmatic children

A large prospective multicenter study by the Multicenter Airway Research Collaboration (MARC) investigators examined risk factors and predictors of hospital admission among children aged 2–17 years seen in 44 EDs from 1997 to 1998.¹² Enrolling sites were from 37 general hospitals and 7 children’s hospitals across 18 US states and four Canadian provinces. Prospective enrollment occurred 24 hours per day over a median of 2 weeks. Repeat ED visits and patients discharged from the ED against medical advice were excluded.

A total of 1601 eligible children presenting to EDs with acute asthma were identified, and 1178 patients were included in the analysis (74%). The admission rate was 23% (CI 21–26%) with an interquartile range of 11–31% across the 44 EDs. Multivariate logistic regression modeling produced patient variables that were independently predictive of hospital admission (Table 54.1). PEFR was not included in the logistic regression model because it could be measured in only 23% of children. However, in those 23% with PEFR measurements, admitted children had lower initial PEFRs compared to nonadmitted children (percentage predicted 36% versus 50%, mean difference: 14%). Demographic factors were not predictive of admission.

Related posts:

Adult Septic Arthritis Testicular Torsion Palpitations Penetrating Trauma to the Extremities and Vascular Injuries First‐Episode Seizure Small Bowel Obstruction