Comparability of Acute Asthma Severity Assessments by Parents and Respiratory Therapists FREE

THERE IS a growing recognition of the importance of determining clinical outcomes after acute exacerbations of asthma.¹ Although most children discharged to home after emergency department (ED) treatment do well, a considerable minority experiences prolonged symptoms.¹ Many studies of ED treatment have focused on relapse, or return for unscheduled care. However, differences in care-seeking thresholds may influence the decision to return. Contact with patients or families for follow-up after ED discharge may therefore be useful to determine which patients have symptoms that need further evaluation. For those patients discharged to home, information on severity must usually be obtained by self-report of the patient or parent. Studies of home measurement of peak expiratory flow rate (PEFR) have demonstrated variations in accuracy and compliance.^2- 3 Moreover, many children have difficulty in performing PEFR.⁴ Thus, disease severity must often be based on parental evaluation of clinical findings such as the severity of wheezing or work of breathing. Although there is extensive literature on various clinical asthma scores applied by health care professionals,⁵ the validity of clinical assessments by parents has not previously been reported to our knowledge.

This study was undertaken as part of a cohort study of predictors of short-term outcomes of acute asthma in which children were assessed within 24 hours of an ED visit either by the respiratory therapist (RT) in the hospital for admitted patients or by telephone for those discharged to home. To ensure comparability of these assessments, the goal of this validation study was to compare parent evaluation of their child's wheezing with an assessment performed at home by a RT. We specifically wished to evaluate whether children who were reported by their parents to be well enough not to require further acute care would also be so judged by the RT.

Patients in this study were enrolled as part of a larger prospective cohort study of predictors of short-term outcomes of acute asthma. Children aged 2 to 17 years treated in an urban pediatric ED (Children's Hospital of Philadelphia) for acute asthma and discharged to home were eligible. Acute asthma was defined as wheezing or respiratory distress in a patient with a previous physician diagnosis of asthma or with at least 1 prior episode of wheezing treated with inhaled bronchodilators. Throughout the course of 12 months, eligible patients were identified on randomly selected study days, and a 15% random sample of all eligible patients was invited to have a home care visit, at no charge to the family, within 24 hours of discharge. Standard discharge therapy and discharge instructions were given to all patients. Informed consent was obtained from parents, and assent was obtained from children 8 years or older. The study was approved by the institutional review board.

A registered RT went to the child's home and performed a clinical assessment, including 4 clinical variables: wheezing, work of breathing, quality of air entry, and prolongation of expiration. Each variable was scored as 0 (none or mild), 1 (moderate), or 2 (severe). A clinical score was calculated by summing the 4 variables (total possible range, 0-8).⁶ This clinical assessment was performed at least 3 hours after the last inhaled bronchodilator treatment. The child's parent was asked his or her overall assessment of the severity of the child's asthma compared with when the child was discharged from the ED and the severity of wheezing and work of breathing (also on a 0-2 scale). Because the parent was only asked to rate 2 clinical variables, combined scores were calculated from the sum of the wheezing and work of breathing assessments by the parent and the RT. A difference of 2 or more points out of a possible 4 was considered to represent a clinically important discrepancy.

Categorical variables were described as proportions, and continuous variables as mean ± SD. When appropriate, 95% confidence intervals (CIs) were calculated. The χ² test was used to test significance of differences in categorical variables, and the unpaired t test (2-tailed) or Wilcoxon rank sum test was used to compare continuous and ordinal variables. Interrater agreement between RTs and parents was measured with the weighted κ statistic. The main outcome measure was the percentage of patients with an important discrepancy in scores. A target sample size of 54 was chosen based on the ability to estimate the proportion to within ±10%, assuming a rate of 10%.

A total of 578 patients were eligible during the study period, and 90 patients were invited to participate in the home care visit; 51 visits (57%) were successfully completed. The patients with a home care visit were similar to those who did not have a home care visit regarding age, chronic asthma severity, and clinical findings at the time of ED discharge (Table 1).

For 43 of the children (84%), the parent rated their asthma as improved compared with the previous day, and 8 (16%) of the patients were considered unchanged. None was felt to be worse. The median 4-item clinical score measured by the RT at the time of the home care visit was 0 compared with a median score of 1 at the time of their discharge from the ED. Six patients (12%) had a higher clinical score at follow-up. The difference was 1 point in 3 cases and 2 points in 3. No patient was considered by the RT or the parent to require additional acute care.

Respiratory therapist–assigned scores for wheezing and work of breathing, by patient age, are given in Table 2. Three patients had incomplete documentation of 1 or both assessments and are excluded. No patients were rated as severe for either finding by the RT. Older children had milder findings, but the distribution of scores in the 2 younger groups was similar.

Figure 1 shows the difference in the combined assessment of wheezing and work of breathing between parents and RTs. Overall chance-adjusted agreement was poor, with a κ statistic of 0.27 (95% CI, 0.05-0.47). However, both raters assigned the same score in 31 cases (65%). Parents underestimated the clinical severity in only 3 cases (6%) (95% CI, 1.3%-17.4%), and in only 1 patient was the parent score more than 1 point less than the RT's score. We examined the details of these cases of parent underestimation. The 1 patient with a 2-point difference was a 2-year-old scored as 0 on both wheezing and work of breathing by the parent but as 1 on both items by the RT. In both cases of a single-point discrepancy, the parent scored a 0 for work of breathing, whereas the RT scored a 1. Both of these children were scored as 0 for wheezing by both observers. Parents underestimated severity for 2 patients (11%) in the youngest age group (2-5 years), 1 (5%) in the middle age group (6-11 years), and none in the oldest age group (12-17 years). These differences are not statistically significant (χ²₄, 3.04; P = .55).

Conversely, parents overestimated severity in 14 cases (29%) (95% CI, 17.0%-44.1%); however, in only 5 (10%) (95% CI, 3.5%-22.5%) was the discrepancy greater than 1 point. Overestimation by the parent occurred in 6 (33%) of the youngest patients, 4 (20%) of the patients in the middle age group, and 4 (40%) of the oldest patients.

Most patients were scored as 0 by either the parent, RT, or both. To examine whether agreement was similar between those with somewhat more significant findings, we looked at the 9 patients who were scored as at least 1 by both observers. Seven of these (78%) were scored the same by the parent and the RT; in 2 cases, the parent assigned a higher score than the RT.

Prior studies of self-reported asthma severity have compared subjective severity ratings with PEFR. Kotses et al⁷ found poor correlation between symptom ratings and PEFR in 29 children, as did Yoos and McMullen⁸ in their study of 28 children and parent pairs. A study of adult asthmatics by Kendrick et al⁹ similarly found that 60% of patients showed no significant correlation between subjective severity score and PEFR. In contrast, Silverman et al¹⁰ studied 18 patients with asthma aged 7 to 43 years and found a significant correlation (r = 0.97) between perceived and actual PEFR. In all of these studies, however, self-reported subjective estimates of disease severity were compared with the results of pulmonary function testing, not with another clinical assessment. Peak expiratory flow rate may be of limited usefulness in the evaluation of airflow obstruction, especially in young children or during acute exacerbations,^4,11 leaving clinical and symptom measures as important parts of the assessment of severity. Thus, while clinical score is not the gold standard for measuring acute asthma severity, for many children recovering from acute asthma, it may be the best available gauge of disease status.

In two thirds of cases, parents and RTs assigned the same severity rating. In the remaining one third, the discrepancies tended to be small and clinically unimportant. Moreover, from the standpoint of patient care, underestimation would have greater clinical consequences, as it could lead to undertreatment and increased morbidity. Parental underestimation of severity was quite uncommon, occurring in only 3 cases (6%) overall. Underestimation tended to occur more often in the youngest children, those younger than 6 years. Although the differences in age-specific rates were not statistically significant, this study did not have sufficient power to examine such subgroup differences. This suggests that parent evaluation may be more problematic in young children; further study is warranted to clarify this point.

We defined a clinically important difference as more than 1 point. We believed that a smaller difference would, in general, be too sensitive and would be unlikely to alter management. However, given the narrow range of possible values, from 0 to 4, it is possible that a 1-point difference might, in fact, be relevant. This would be true particularly if a parent rated the child as moderate and the RT rating was severe. It is therefore reassuring that in no case did this type of discrepancy occur. Moreover, regardless of the score assigned, in all cases parents and RTs agreed that the child did not require any change in treatment or a return to the ED.

There are several limitations to this study. First, we did not use a validated clinical asthma score for parent assessment because such scores are based on findings from auscultation as well as observation of the patient. Because the rationale for this study was based on the desire to ensure that outpatients assessed by their parents would be evaluated comparably to inpatients who were being assessed by RTs using clinical scores, we chose to focus on 2 items from the score that we believed parents would be able to evaluate. Since these are commonly elicited clinical findings that parents are frequently asked to evaluate, we believed these parental assessments would be applicable to usual clinical practice. It is, however, possible that parents may be more or less accurate in their assessment of other features of acute asthma. Second, although the assessments were performed independently, the RTs may not have been blinded to the parents' impression of the child's degree of illness, which may have influenced the RTs' evaluation. Finally, all of the patients had relatively mild acute asthma, with most patients having a score of 0. This study was specifically targeted at a group of patients—those being successfully cared for at home—expected to have minimal findings. While the agreement was similar in the most symptomatic children, this subgroup was small. Therefore, our findings may not be applicable to a populations with a wider range of acute severity.

We conclude that parents and RTs provide comparable assessments of children's clinical severity within 24 hours of discharge from the ED after treatment of acute asthma exacerbation. In particular, the parent's judgment that a child does not require return for acute care corresponds well to that of the RT. Parent-reported severity may be a useful short-term outcome measure, particularly in children in whom PEFR measurement is problematic.

Accepted for publication July 11, 2002.

This work was supported by grant R40-MC00097 from the Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Services Administration, US Department of Health and Human Services, Rockville, Md.

Corresponding author: Marc H. Gorelick, MD, MSCE, Children's Hospital of Wisconsin, Mail Stop 677, 9000 W Wisconsin Ave, Milwaukee, WI 53226 (e-mail: mgorelic@mail.mcw.edu).

Patients are frequently contacted by telephone for follow-up after emergency department treatment for asthma. This study demonstrates that assessments by parents and RTs are comparable and, specifically, that underestimation of severity by parents is uncommon. Parental report that a child with asthma does not require further acute care seems to be reliable.