Quality of Life Measure for Older Adolescents With Asthma
A small methodological study was conducted to modify the AAP HRQL measure, CHSA–Child version, to an older teen version, CHSA-T. Feasibility, reliability, and validity properties were obtained over three time points (baseline, Day 14, and Day 16) for 24 older teens to follow the AAP's original design for the child version (Radecki, Olson, Frintner, & Weiss, 2008). The agreement with the AAP stated that psychometric testing methods of the child version would be replicated when possible. A follow-up study used 46 older teens to further quantify substance use behaviors.
The primary clinical setting was an urban pediatric primary care practice on the East Coast drawing from the metropolitan area of seven cities. As a result of a geographical relocation by the nurse practitioner (NP) who was enrolling adolescents into the study, enrollment was stopped earlier than anticipated at the primary site. To obtain a more representative picture of drug use in the 17- to 19-year age group, a second cohort of teens who had been given the CHSA-T in the follow-up Teen Asthma Study (TAS) was added to the sampling plan. This addition was made because there were skewed race/ethnicity characteristics (largely African American), with only a small number of teens reporting current and lifetime substance use in the primary site sample. The TAS was a multisite randomized clinical trial with two sites at university pediatric clinics, as well as a community consortium of pediatric practices, funded by a state foundation grant.
Using two cohorts, the primary private practice cohort (n = 24) and another from the TAS follow-up study (n = 46), a total of 70 adolescents ages 17 to 19 years were assessed. Three eligibility criteria were used: (a) a diagnosis of asthma by health care providers, (b) 17 to 19 years of age, and (c) the ability to read and speak English. The only exclusion criterion was having other major physical or psychiatric comorbidities, as documented on the medical record.
Clinical status for asthma has traditionally been measured by symptom scores, spirometry, medication use, utilization of health care resources, and/or HRQL outcome measures. A similar battery of measures was used in this study. Instructions and completion of the battery of measures for the teen required less than 1 hour on the initial visit and less than that for subsequent visits.
A Medical Record Review Form was adapted from a record review form used in a previous study (O'Laughlen et al., 2008). It captured information on the history of the disease (including level of severity, whether or not the teen's asthma was persistent or intermittent in the past year, and general health status). Severity was determined on the basis of NAEPP diagnostic criteria (National Heart, Lung, and Blood Institute, 2007) rather than parental report because it was expected that some teens could be emancipated or away at college.
A Teen Medication Report Form, a brief two-item questionnaire, was used to examine the relationship of CHSA-T scores to oral corticosteroid use in the past year and rescue or quick-relief β-agonist use over the 2-week study period (Radecki, Olson, Frintner, & Weiss, 2008).
The CHSA-T is a 31-item measure that was adapted from the CHSA-C by adding a full substance use subscale; the other three domains (physical health, activities, and emotional health) were left unchanged (Radecki, Olson, Frintner, & Weiss, 2008). The original CHSA-C had 25 core scale items, computed into three domain scores to assess asthma-related child physical health, child activity, and child emotional well-being. These items are scored by a Likert-type scale (0 = "none of the time," 1 = "little of the time," 2 = "some of the time," 3 = "most of the time," and 4 = "all of the time"). These 25 questions became part of the CHSA-T, with only slight alterations to the wording of some questions to make them more appropriate for teens. The scoring of the 25 items was exactly the same for the CHSA-T as for the CHSA-C. Scoring syntax in both SPSS and SAS is provided in the User's Guide for the CHSA and CHSA-C, which is available online at http://www.aap.org/en-us/professional-resources/Research/pediatrician-surveys/Documents/CHSA_CHSA-C_UserGuide_2008.pdf. The resulting scores for each scale have a possible range from 0 to 100. Those scores were transformed so that higher scores would indicate better HRQL.
Six additional items were added to the original 25 items, creating a fourth domain of substance use behaviors as a 6-item summary index for this developmental age. It captures overall cognitive-behavioral HRQL outcomes based on self-perception of disease severity.
A single question on smoking ("Do you smoke?") appeared in the CHSA-C survey (Part IV: About You), but it was not scored or included in any of the CHSA-C domains. That question was included in the CHSA-T, along with five additional questions on substance use. Those five questions were: (1) "If you do smoke, how often?" (2) "Do you drink alcohol?" (3) "If you do drink alcohol, how often?" (4) "Do you use recreational drugs (marijuana, cocaine, inhalants)?"and (5) "If you do use recreational drugs, how often?"
Lifetime use and current use, both scored by yes/no responses, were computed for smoking, alcohol, and recreational drugs in the following way. Lifetime use of one of the substance types (smoking, alcohol, and recreational drugs of marijuana, cocaine, or inhalants) was considered positive if the teen gave a "yes" response to the substance type (e.g., to "Do you smoke?") and/or the teen responded with a positive frequency to the corresponding question on "how often" the substance was used. Lifetime substance use, a composite score, was scored positive if there was lifetime use for one or more of the three substance types. Current use was considered positive for one of the three substance types if the teen gave a "yes" response to the "Do you…" question and/or responded to the "how often" question with one of these responses: (a) "occasionally," (b) "regularly in the past 12 months," or (c) "regularly now." Current substance use, a composite score, was considered positive if there was current use for one or more of the three substance types.
The Asthma Symptom Day questionnaire is a four-item symptom activity questionnaire that categorizes symptom burden based on the number of days the teen was affected out of the past 14 days: low (0–2 days), medium (3–10 days), and high (11+ days). AAP had to combine medium and high; this step was repeated for this study (Olsen, Radecki, Frintner, Weiss, & Korfmacher, 2007).
The Symptom Recall for 48 Hours questionnaire is a one-item questionnaire scored dichotomously that asks if the teen had any asthma-related problems between visit 2 and visit 3 (Olsen, Radecki, Frintner, Weiss, & Korfmacher, 2007).
Lung function was captured by the Vitalograph Asma-1 electronic peak flow meter (Vitalograph, 2013). In measuring lung function, the percent predicted values for forced expiratory volume in 1 second (FEV1) was categorized into two groups (≤ 79% and ≥ 80%). Regarding precision and accuracy, the performance standards of the Vitalograph Asma-1 meter meets or exceeds both European Norm International Organization for Standardization 23747:2007 and ATS/European Respiratory Society 2005 Guidelines (Vitalograph, 2013).
Periodic Assessment of Drug Use Among Youth (PADU), a self-report instrument for adolescents developed by Barnes and colleagues, was used to assess frequency and amount of substance use risk behaviors (Barnes and Welte, 1986, Barnes et al., 1993). Five substances were reported for the present study: (1) smoking, (2) alcohol consumption, and (3) recreational drug use (marijuana, cocaine, and inhalants). The PADU was pretested for readability in slow-reading 7th-grade classes in New York City and was found to be acceptable. The PADU has been used in three representative school samples from New York State, which included more than 20,000 in each survey conducted by Barnes and colleagues, thereby providing substantial normative data. For this study, lifetime substance use was defined as having ever engaged in at least one type of substance (smoking cigarettes; alcohol in beer, wine, or liquor; or recreational drug use, such as marijuana). Current substance use was defined as having engaged in any use of the three types of substances in the past 30 days for cigarettes, and the past 12 months for any type of alcohol or recreational drug use.
This study was approved by the Institutional Review Board at two universities and by boards of the private practices. Eligible research participants provided written consent/assent according to institutional policy.
The primary cohort (n = 24) had three time points scheduled over 16 days ( Table 1 ). The first time point included consent/assent for the parent and child, as well as instructions for the Vitalograph Electronic peak expiratory flow (PEF)/FEV1 diary. The NP gave directions to the teens on use of the electronic peak flow meters (twice a day) with a return demonstration, as well as the daily diary for 14 days. A medication report form was also completed (Teen Medication Report). At the second time point (day 14), the teen turned in the electronic peak flow meter for retrieval of data. Using a secure laptop, the teen was helped by the NP with the initial battery of online measures. The teen completed both a paper and online CHSA-T to ensure avoidance of technical issues. At the third time point (day 16), the teen went online to complete the CHSA-T and a 24-hour symptom recall, avoiding the necessity of another clinic visit. As study incentives, the NP provided a short resume statement related to study participation, the teens were allowed to keep the electronic peak flow meters, and a total of $75 was given over the course of the study.
The teens from TAS (n = 46) followed the initial visit procedures with two exceptions: the teens completed the CHSA-T online at the baseline visit, and FEV1 values were not collected because spirometry was cost-prohibitive for this initial study. No other psychometric measures were obtained because the primary aim of the TAS was changing behavior through the use of a decision aid. They were given a short resume statement of participation and $25 as incentives.
All statistical analyses were conducted using SPSS version 21 (IBM Corp., 2013). Descriptive statistics were generated for individual items and for each subscale. There was very little missing data, so no imputation was used. Footnotes to the tables indicate where there were missing values. When possible, analysis methods of the CHSA-T followed the original methods of the CHSA-C.
The classic guidelines of Nunnally and Bernstein (1994)) were used to interpret the reliability coefficients; > 0.70 is acceptable for new measures (Ferketich, 1991, Nunnally and Bernstein, 1994). Using another classic guideline for validity (Anastasi, 1988), three criteria were used as a foundation for estimation of support for validity: (1) multiple procedures; (2) sequential employment (e.g., reliability is a must before construct validity); and (3) assessment of validity at various stages of instrument development (e.g., when newly developed; when moving toward established status). More current guidelines for HRQL measures were also used (Frost et al., 2007). Psychometricians believe that validity is a question of degree, with no absolute standard for the magnitude of a validity coefficient; support of validity through multiple testing accrues evidence.
Methods
Design/Setting
A small methodological study was conducted to modify the AAP HRQL measure, CHSA–Child version, to an older teen version, CHSA-T. Feasibility, reliability, and validity properties were obtained over three time points (baseline, Day 14, and Day 16) for 24 older teens to follow the AAP's original design for the child version (Radecki, Olson, Frintner, & Weiss, 2008). The agreement with the AAP stated that psychometric testing methods of the child version would be replicated when possible. A follow-up study used 46 older teens to further quantify substance use behaviors.
The primary clinical setting was an urban pediatric primary care practice on the East Coast drawing from the metropolitan area of seven cities. As a result of a geographical relocation by the nurse practitioner (NP) who was enrolling adolescents into the study, enrollment was stopped earlier than anticipated at the primary site. To obtain a more representative picture of drug use in the 17- to 19-year age group, a second cohort of teens who had been given the CHSA-T in the follow-up Teen Asthma Study (TAS) was added to the sampling plan. This addition was made because there were skewed race/ethnicity characteristics (largely African American), with only a small number of teens reporting current and lifetime substance use in the primary site sample. The TAS was a multisite randomized clinical trial with two sites at university pediatric clinics, as well as a community consortium of pediatric practices, funded by a state foundation grant.
Sample/Sampling Plan
Using two cohorts, the primary private practice cohort (n = 24) and another from the TAS follow-up study (n = 46), a total of 70 adolescents ages 17 to 19 years were assessed. Three eligibility criteria were used: (a) a diagnosis of asthma by health care providers, (b) 17 to 19 years of age, and (c) the ability to read and speak English. The only exclusion criterion was having other major physical or psychiatric comorbidities, as documented on the medical record.
Instruments
Clinical status for asthma has traditionally been measured by symptom scores, spirometry, medication use, utilization of health care resources, and/or HRQL outcome measures. A similar battery of measures was used in this study. Instructions and completion of the battery of measures for the teen required less than 1 hour on the initial visit and less than that for subsequent visits.
A Medical Record Review Form was adapted from a record review form used in a previous study (O'Laughlen et al., 2008). It captured information on the history of the disease (including level of severity, whether or not the teen's asthma was persistent or intermittent in the past year, and general health status). Severity was determined on the basis of NAEPP diagnostic criteria (National Heart, Lung, and Blood Institute, 2007) rather than parental report because it was expected that some teens could be emancipated or away at college.
A Teen Medication Report Form, a brief two-item questionnaire, was used to examine the relationship of CHSA-T scores to oral corticosteroid use in the past year and rescue or quick-relief β-agonist use over the 2-week study period (Radecki, Olson, Frintner, & Weiss, 2008).
The CHSA-T is a 31-item measure that was adapted from the CHSA-C by adding a full substance use subscale; the other three domains (physical health, activities, and emotional health) were left unchanged (Radecki, Olson, Frintner, & Weiss, 2008). The original CHSA-C had 25 core scale items, computed into three domain scores to assess asthma-related child physical health, child activity, and child emotional well-being. These items are scored by a Likert-type scale (0 = "none of the time," 1 = "little of the time," 2 = "some of the time," 3 = "most of the time," and 4 = "all of the time"). These 25 questions became part of the CHSA-T, with only slight alterations to the wording of some questions to make them more appropriate for teens. The scoring of the 25 items was exactly the same for the CHSA-T as for the CHSA-C. Scoring syntax in both SPSS and SAS is provided in the User's Guide for the CHSA and CHSA-C, which is available online at http://www.aap.org/en-us/professional-resources/Research/pediatrician-surveys/Documents/CHSA_CHSA-C_UserGuide_2008.pdf. The resulting scores for each scale have a possible range from 0 to 100. Those scores were transformed so that higher scores would indicate better HRQL.
Six additional items were added to the original 25 items, creating a fourth domain of substance use behaviors as a 6-item summary index for this developmental age. It captures overall cognitive-behavioral HRQL outcomes based on self-perception of disease severity.
A single question on smoking ("Do you smoke?") appeared in the CHSA-C survey (Part IV: About You), but it was not scored or included in any of the CHSA-C domains. That question was included in the CHSA-T, along with five additional questions on substance use. Those five questions were: (1) "If you do smoke, how often?" (2) "Do you drink alcohol?" (3) "If you do drink alcohol, how often?" (4) "Do you use recreational drugs (marijuana, cocaine, inhalants)?"and (5) "If you do use recreational drugs, how often?"
Lifetime use and current use, both scored by yes/no responses, were computed for smoking, alcohol, and recreational drugs in the following way. Lifetime use of one of the substance types (smoking, alcohol, and recreational drugs of marijuana, cocaine, or inhalants) was considered positive if the teen gave a "yes" response to the substance type (e.g., to "Do you smoke?") and/or the teen responded with a positive frequency to the corresponding question on "how often" the substance was used. Lifetime substance use, a composite score, was scored positive if there was lifetime use for one or more of the three substance types. Current use was considered positive for one of the three substance types if the teen gave a "yes" response to the "Do you…" question and/or responded to the "how often" question with one of these responses: (a) "occasionally," (b) "regularly in the past 12 months," or (c) "regularly now." Current substance use, a composite score, was considered positive if there was current use for one or more of the three substance types.
The Asthma Symptom Day questionnaire is a four-item symptom activity questionnaire that categorizes symptom burden based on the number of days the teen was affected out of the past 14 days: low (0–2 days), medium (3–10 days), and high (11+ days). AAP had to combine medium and high; this step was repeated for this study (Olsen, Radecki, Frintner, Weiss, & Korfmacher, 2007).
The Symptom Recall for 48 Hours questionnaire is a one-item questionnaire scored dichotomously that asks if the teen had any asthma-related problems between visit 2 and visit 3 (Olsen, Radecki, Frintner, Weiss, & Korfmacher, 2007).
Lung function was captured by the Vitalograph Asma-1 electronic peak flow meter (Vitalograph, 2013). In measuring lung function, the percent predicted values for forced expiratory volume in 1 second (FEV1) was categorized into two groups (≤ 79% and ≥ 80%). Regarding precision and accuracy, the performance standards of the Vitalograph Asma-1 meter meets or exceeds both European Norm International Organization for Standardization 23747:2007 and ATS/European Respiratory Society 2005 Guidelines (Vitalograph, 2013).
Periodic Assessment of Drug Use Among Youth (PADU), a self-report instrument for adolescents developed by Barnes and colleagues, was used to assess frequency and amount of substance use risk behaviors (Barnes and Welte, 1986, Barnes et al., 1993). Five substances were reported for the present study: (1) smoking, (2) alcohol consumption, and (3) recreational drug use (marijuana, cocaine, and inhalants). The PADU was pretested for readability in slow-reading 7th-grade classes in New York City and was found to be acceptable. The PADU has been used in three representative school samples from New York State, which included more than 20,000 in each survey conducted by Barnes and colleagues, thereby providing substantial normative data. For this study, lifetime substance use was defined as having ever engaged in at least one type of substance (smoking cigarettes; alcohol in beer, wine, or liquor; or recreational drug use, such as marijuana). Current substance use was defined as having engaged in any use of the three types of substances in the past 30 days for cigarettes, and the past 12 months for any type of alcohol or recreational drug use.
Procedures in Clinic
This study was approved by the Institutional Review Board at two universities and by boards of the private practices. Eligible research participants provided written consent/assent according to institutional policy.
The primary cohort (n = 24) had three time points scheduled over 16 days ( Table 1 ). The first time point included consent/assent for the parent and child, as well as instructions for the Vitalograph Electronic peak expiratory flow (PEF)/FEV1 diary. The NP gave directions to the teens on use of the electronic peak flow meters (twice a day) with a return demonstration, as well as the daily diary for 14 days. A medication report form was also completed (Teen Medication Report). At the second time point (day 14), the teen turned in the electronic peak flow meter for retrieval of data. Using a secure laptop, the teen was helped by the NP with the initial battery of online measures. The teen completed both a paper and online CHSA-T to ensure avoidance of technical issues. At the third time point (day 16), the teen went online to complete the CHSA-T and a 24-hour symptom recall, avoiding the necessity of another clinic visit. As study incentives, the NP provided a short resume statement related to study participation, the teens were allowed to keep the electronic peak flow meters, and a total of $75 was given over the course of the study.
The teens from TAS (n = 46) followed the initial visit procedures with two exceptions: the teens completed the CHSA-T online at the baseline visit, and FEV1 values were not collected because spirometry was cost-prohibitive for this initial study. No other psychometric measures were obtained because the primary aim of the TAS was changing behavior through the use of a decision aid. They were given a short resume statement of participation and $25 as incentives.
Statistical Analyses
All statistical analyses were conducted using SPSS version 21 (IBM Corp., 2013). Descriptive statistics were generated for individual items and for each subscale. There was very little missing data, so no imputation was used. Footnotes to the tables indicate where there were missing values. When possible, analysis methods of the CHSA-T followed the original methods of the CHSA-C.
The classic guidelines of Nunnally and Bernstein (1994)) were used to interpret the reliability coefficients; > 0.70 is acceptable for new measures (Ferketich, 1991, Nunnally and Bernstein, 1994). Using another classic guideline for validity (Anastasi, 1988), three criteria were used as a foundation for estimation of support for validity: (1) multiple procedures; (2) sequential employment (e.g., reliability is a must before construct validity); and (3) assessment of validity at various stages of instrument development (e.g., when newly developed; when moving toward established status). More current guidelines for HRQL measures were also used (Frost et al., 2007). Psychometricians believe that validity is a question of degree, with no absolute standard for the magnitude of a validity coefficient; support of validity through multiple testing accrues evidence.