Validation of the Hospital Asthma Severity Score (HASS)

Children aged 7-18 years will have a HASS score obtained by two providers. After scoring is complete, a pulmonary technician will perform the spirometry test.

Children aged 2 to 6 years old will concurrently (but separately) have their asthma severity graded by 2 individual measurers using the HASS tool and the PRAM tool. If consented to perform spirometry, this test will be performed after the tool(s) scores have been obtained.

The spirometry tests will be interpreted by a board-certified pediatric pulmonologist.

Descriptive statistics will be used for preliminary analysis to describe the sample characteristics, outliers, and representativeness of the data.

Primary Aim 1. To describe the inter-rater reliability for the HASS as scored by 2 individual health care providers Data Analysis 1. Use of an overall percentage of agreement, weighted and unweighted kappa and the intraclass correlation coefficient will show the inter-rater reliability of the HASS tool. For each measure, percent agreement between raters and Cohen's kappa statistic (rater agreement adjusted for chance agreement) will be reported.

Primary Aim 2. To evaluate construct validity of the HASS tool in a cohort of patients' ages 7 to18 years old as compared to forced expiratory volume through spirometry

Data Analysis 2. Treating the outcomes as continuous, a correlation coefficient will be computed for HASS and %Forced Expiratory Volume 1. Sub-analyses will also be performed by age group and by %Forced Expiratory Volume1 severity level to see whether the correlation of HASS and %Forced Expiratory Volume1 is consistent across age and asthma severity strata. Treating the outcomes as categorical, if the HASS and %Forced Expiratory Volume1 are each categorized into three levels (mild, moderate, severe), then the investigator could classify each of the nine cells of a 3 by 3 table (HASS X %Forced Expiratory Volume1) as either:

1. concordant (three cells; complete agreement: both tests say 'Severe', for instance),
2. discordant (four cells; e.g. one test says Moderate while the other say Mild) and
3. grossly discordant (two cells; one test says Severe when the other says Mild). For the Receiver Operating Curve analysis, the Forced Expiratory Volume will be dichotomized as Severe (%Forced Expiratory Volume1<60% predicted) versus Less Severe (%Forced Expiratory Volume1≥60% predicted). In analyzing these data, the correlation (continuous), 3 by 3 table (categorical), and Receiver Operating Curve (dichotomous %Forced Expiratory Volume1) treatments will be presented, with the continuous result considered primary. A high level of agreement is expected between the HASS and the spirometry severity score to make the case that the HASS can be used in place of the %Forced Expiratory Volume1 test.

Secondary Aim 3. To evaluate construct validity of the HASS tool in a cohort of patients' ages 2 to 6 years old as compared to forced expiratory volume through spirometry.

Data Analysis 3. Treating the outcomes as continuous, a correlation coefficient will be computed for HASS and %Forced Expiratory Volume1. Treating the outcomes as categorical, if the HASS and %Forced Expiratory Volume1 are each categorized into three levels (mild, moderate, severe), then the investigator could classify each of the nine cells of a 3 by 3 table (HASS X %Forced Expiratory Volume1) as concordant, discordant or grossly discordant (as described in Data Analysis 2). An adequate level of observed agreement would be 80% of observations in agreement cells, 18% discordant, and 2% in grossly discordant cells. In analyzing these data, both the correlation (continuous) and 3 by 3 table (categorical) treatments will be presented, with the continuous result considered primary. A high level of agreement is expected between the HASS and the spirometry severity score to make the case that the HASS can be used in place of the spirometry.

Secondary Aim 4. To evaluate construct validity of the HASS tool in a cohort of patients' ages 2 to 6 years old as compared to the Preschool Respiratory Assessment Measure (PRAM) Data Analysis 4. Treating the outcomes as continuous, a correlation coefficient will be computed for HASS and PRAM. If the HASS and PRAM are each categorized into three levels (mild, moderate, severe), then the investigator could classify each of the nine cells of a 3 by 3 table (HASS X PRAM) as either concordant, discordant, or grossly discordant (as described in Data Analysis 2). An adequate level of observed agreement would be 80% of observations in agreement cells, 18% discordant, and 2% in grossly discordant cells.