Child/parent-assessed population health outcome measures: a structured review

Child Health Assessment at School Entry (CHASE) project: evaluation in 10 London primary schools

AIMS

To assess the feasibility of implementing the Child Health Assessment at School Entry (CHASE) questionnaire, developed to capture the multiple dimensions of the health of children in their first year at school, and to evaluate data quality, reliability and validity.

METHODS

Parents of 278 year-1 children, from 10 primary schools in two London boroughs, received a parent questionnaire and school nurses completed a separate questionnaire from health and education records for children whose parents consented. Additional data on free school meal eligibility and ethnicity were obtained from the two Local Education Authorities. The parent questionnaire included the Strengths and Difficulties Questionnaire (SDQ) and four dimensions of the Child Health Questionnaire Parent Form-28 (CHQ-PF28).

RESULTS

Response rate was 61%. The association between school free school meals eligibility and response rate in each school approached significance (r = -0.62, P = 0.05). Data completeness of the parent questionnaire was high (mean 98%). Data completeness of the school nurse questionnaire was more variable (mean 82%). Cronbach's Alpha was greater than 0.6 for four of the five SDQ scales and greater than 0.7 for the two CHQ-PF28 multi-item scales. Relative to parents with qualifications, parents with no qualifications rated their children as having significantly more conduct problems, peer problems, and overall mental health problems (P < 0.01) as assessed by the SDQ, and significantly lower global health (P < 0.01) as assessed by the CHQ-PF28. Children with special educational needs and children with long-standing illness or disability were rated as having significantly lower global health (P < 0.05) than children without these. Sample tables of inter-school and inter-borough comparison of key findings demonstrate considerable differences in physical and mental health status.

DISCUSSION

The questionnaire was acceptable to parents and school nurses, and feasible to implement within existing school resources. Initial tests of internal reliability and validity are satisfactory. These data have the potential to inform interventions and service provision at school and borough level, and public health trends over time.

Health-related quality of life in children and adolescents who have a diagnosis of attention-deficit/hyperactivity disorder

OBJECTIVE

The aim of treatment for attention-deficit/hyperactivity disorder (ADHD) is to decrease symptoms, enhance functionality, and improve well-being for the child and his or her close contacts. However, the measurement of treatment response is often limited to measuring symptoms using behavior rating scales and checklists completed by teachers and parents. Because so much of the focus has been on symptom reduction, less is known about other possible health problems, which can be measured easily using health-related quality-of-life (HRQL) questionnaires, which are designed to gather information across a range of health domains. The aim of our study was to measure HRQL in a clinic-based sample of children who had a diagnosis of ADHD and consider the impact of 2 clinical factors, symptom severity and comorbidity, on HRQL. Our specific hypotheses were that parent-reported HRQL would be poorer in children with ADHD than in normative US and Australian pediatric samples, in children with increasing severity of ADHD symptoms, and in children who had diagnoses of comorbid psychiatric disorders.

METHODS

Cross-sectional survey was conducted in British Columbia, Canada. The sample included 165 respondents of 259 eligible children (63.7% response rate) who were referred to the ADHD Clinic in British Columbia between November 2001 and October 2002. Children who are seen in this clinic come from all parts of the province and are diverse in terms of socioeconomic status and case mix. ADHD was diagnosed in 131 children, 68.7% of whom had a comorbid psychiatric disorder. Some children had >1 comorbidity: 23 had 2, 5 had 3, and 1 had 4. Fifty-one children had a comorbid learning disorder (LD), 45 had oppositional defiant disorder or conduct disorder (ODD/CD), and 27 had some other comorbid diagnosis. The mean age of children was 10 years (standard deviation: 2.8). Boys composed 80.9% (N = 106) of the sample. We used the 50-item parent version of the Child Health Questionnaire to measure physical and psychosocial health. Physical domains include the following: physical functioning (PF), role/social limitations as a result of physical health (RP), bodily pain/discomfort (BP), and general health perception (GH). Psychosocial domains include the following: role/social limitations as a result of emotional-behavioral problems (REB), self-esteem (SE), mental health (MH), general behavior (BE), emotional impact on parent (PTE), and time impact on parents (PTT). A separate domain measures limitations in family activities (FA). There is also a single-item measure of family cohesion (FC). Individual scale scores and summary scores for physical (PhS) and psychosocial health (PsS) can be computed. Symptom severity data (parent and teacher) came from the Child/Adolescent Symptom Inventory 4. These checklists provide information on symptoms for the 3 ADHD subtypes (inattentive, hyperactive, and combined). Each child underwent a comprehensive psychiatric assessment by 1 of 4 child psychiatrists. Documentation included a full 5-axis Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition diagnosis on the basis of a comprehensive assessment. Clinical information for each child was extracted from hospital notes.

RESULTS

Compared with both population samples, children with ADHD had comparable physical health but clinically important deficits in HRQL in all psychosocial domains, FA, FC, and PsS, with effect sizes as follows: FC = -0.66, SE = -0.90, MH = -0.97, PTT = -1.07, REB = -1.60, BE = -1.73, PTE = -1.87, FA = -1.95, and PsS = -1.98. Poorer HRQL for all domains of psychosocial health, FA, and PsS correlated significantly with more parent-reported inattentive, hyperactive, and combined symptoms of ADHD. Children with > or =2 comorbid disorders differed significantly from those with no comorbidity in most areas, including RP, GH, REB, BE, MH, SE, PTT, FA, and PsS, and from those with 1 comorbid disorder in 3 domains, including BE, MH, and FA and the PsS. The mean PsS score for children in the ODD/CD group (mean difference: -12.9; effect size = -1.11) and children in the other comorbidity group (-9.0; effect size = -.77) but not children in the LD group were significantly lower than children with no comorbid disorder. Predictors of physical health in a multiple regression model included child's gender (beta = .177) and number of comorbid conditions (beta = -.197). These 2 variables explained very little variation in the PhS. Predictors of psychosocial health included the number of comorbid conditions (beta = -.374) and parent-rated combined ADHD symptoms (beta = -.362). These 2 variables explained 31% of the variation in the PsS.

CONCLUSIONS

Our study shows that ADHD has a significant impact on multiple domains of HRQL in children and adolescents. In support of our hypotheses, compared with normative data, children with ADHD had more parent-reported problems in terms of emotional-behavioral role function, behavior, mental health, and self-esteem. In addition, the problems of children with ADHD had a significant impact on the parents' emotional health and parents' time to meet their own needs, and they interfered with family activities and family cohesion. The differences that we found represent clinically important differences in HRQL. Our study adds new information about the HRQL of children with ADHD in relation to symptom severity and comorbidity. Children with more symptoms of ADHD had worse psychosocial HRQL. Children with multiple comorbid disorders had poorer psychosocial HRQL across a range of domains compared with children with none and 1 comorbid disorder. In addition, compared with children with no comorbidity, psychosocial HRQL was significantly lower in children with ODD/CD and children in the other comorbidity group but not in children with an LD. The demonstration of a differential impact of ADHD on health and well-being in relation to symptom severity and comorbidity has important implications for policies around eligibility for special educational and other supportive services. Because the impact of ADHD is not uniform, decisions about needed supports should incorporate a broader range of relevant indicators of outcome, including HRQL. Although many studies focus on measuring symptoms using rating scales and checklists, in our study, using a multidimensional questionnaire, we were able to show that many areas of health are affected in children with ADHD. We therefore argue that research studies of children with ADHD should include measurement of these broader domains of family impact and child health.

Reviewing measures of outcome: reliability of data extraction

RATIONALE, AIMS AND OBJECTIVES

Although there is wide acceptance of the necessary domains for the evaluation of measures of health outcome, there has been little development or evaluation of methods of data extraction, such as checklists. This study aimed to assess the inter-rater reliability of data extraction from published articles using a new electronic database developed for structured reviews of health outcome measures.

METHODS

Domains considered important in the evaluation of measures of health outcome provided the foundation for a checklist for data extraction using a new electronic database. The database utilized Microsoft Access 2000 software (Microsoft Corporation, 2000). Following training, three independent reviewers extracted data from seven selected articles using the electronic database, and assessed article and measure quality against pre-defined criteria. Data extraction was compared for consistency and item content. Inter-reviewer agreement for categorical data was assessed by multiple kappa correlation.

RESULTS

Analysis demonstrated strong agreement between reviewers for all aspects of data extraction. However, lack of clarity in published articles affected the ability to clearly identify measures of outcome.

CONCLUSIONS

The high level of inter-reviewer agreement supports the use of multiple trained reviewers in data extraction for reviews of measures of outcome using the checklist and Access software described. The electronic database supports standardized data extraction from published articles, benefiting from the combination of data extraction and data entry in a single step. Adaptation of the database to support structured reviews of measures of outcome adopted in the evaluation of other health states is proposed.

Health status and health-related quality of life in a population-based sample of neonatal intensive care unit graduates

OBJECTIVE

To measure the health status (HS) and health-related quality of life (HRQL) of preschoolers who were admitted to a neonatal intensive care unit (NICU) at birth and their family caregivers and to investigate differences in HS and HRQL in relation to gestational age and major morbidity experienced during the NICU stay.

METHODS

Retrospective cross-sectional survey was conducted in the province of British Columbia, Canada. A total of 1140 of 2221 children who were admitted at birth to the 3 tertiary care NICUs in the province and 393 of 718 healthy full-term children recruited from 2 of these hospitals were studied. The main outcome measures were Infant and Toddler Quality of Life Questionnaire (ITQOL), Health Status Classification System Preschool Version (HSCS-PS), and Child Behavior Checklist/1.5-5 (CBCL) RESULTS: The overall response rate was 55%; the response rate for families that we located was 67.1%. NICU children differed from healthy children on the ITQOL in physical abilities, growth and development, temperament/moods, behavior, and general health perceptions, and caregivers differed on both parent-impact scales. On the HSCS-PS, proportionally more NICU children had a health problem in the following areas: sight, speech, getting around, using hands and fingers, taking care of self, learning and remembering, thinking and solving problems, pain and discomfort, general health, and behavior. The NICU sample reported more behavioral problems on the CBCL/1.5-5. Poorer HS and HRQL were reported for infants who were born at <27 weeks' gestation and for children who experienced > or =1 major morbidities during their NICU stay.

CONCLUSIONS

Preschool-aged children with conditions that require NICU care and their family caregivers had poorer HS and HRQL in a range of domains compared with healthy children. There were also differences within the sample by gestational age and major morbidity. The differences in health were small using the ITQOL and CBCL/1.5-5 but larger using the HSCS-PS.

Health related quality of life in 3 and 4 year old children and their parents: preliminary findings about a new questionnaire

Background

Few measures of health related quality of life exist for use with preschool aged children. The objective of this study was to assess reliability and validity of a new multidimensional generic measure of health-related quality of life developed for use with preschool children.

Methods

Cross-sectional survey sent to parents as their child turned 3 1/2 years of age. The setting was the province of British Columbia, Canada. Patients included all babies admitted to tertiary level neonatal intensive care units (NICU) at birth over a 16-month period, and a consecutive sample of healthy babies. The main outcome measure was a new full-length questionnaire consisting of 3 global items and 10 multi-item scales constructed to measure the physical and emotional well-being of toddlers and their families.

Results

The response rate was 67.9%. 91% (NICU) and 84% (healthy baby) of items correlated with their own domain above the recommended standard (0.40). 97% (NICU) and 87% (healthy baby) of items correlated more highly (≥ 2 S.E.) with their hypothesized scale than with other scales. Cronbach's alpha coefficients varied between .80 and .96. Intra-class correlation coefficients were above .70. Correlations between scales in the new measure and other instruments were moderate to large, and were stronger than between non-related domains. Statistically significant differences in scale scores were observed between the NICU and healthy baby samples, as well as between those diagnosed with a health problem requiring medical attention in the past year versus those with no health problems.

Conclusions

Preliminary results indicate the new measure demonstrates acceptable reliability and construct validity in a sample of children requiring NICU care and a sample of healthy children. However, further development work is warranted.

Health-related quality of life in children with epilepsy: development and validation of self-report and parent proxy measures

PURPOSE

To answer a need to include and measure accurately the impact and burden of epilepsy as outcomes of interventions with affected children, we developed and validated self-report and parent-proxy respondent health-related quality of life (HRQL) instruments for preadolescent children with epilepsy.

METHODS

We combined qualitative and quantitative research methods. Items were extracted from focus group discussions involving children with epilepsy and their parents. We created scales formatted with alternative paired options of forced responses and used factor analysis to generate relevant subscales and reduce the number of items. We checked internal consistency, assessed test-retest reliability 10-14 days apart, and documented construct validity.

RESULTS

A sample of 381 children with epilepsy, age 6-15 years, and their parents independently completed a 67-item questionnaire, from which we chose five items for each subscale. The measures share four subscales, but each measure has an additional distinct subscale. The children and parents could discern differences and report differentially between the various aspects of the HRQL. Internal consistency measured with Cronbach's alpha was acceptable for all subscales; construct validity has been demonstrated from the testing of several hypotheses. Test-retest reliability examined with the intraclass correlation coefficient was satisfactory for the parents and for children age 8 years and older. The correlations between the mothers' and children's responses was poor to moderate.

CONCLUSIONS

The data demonstrate sound psychometric properties for both related measures, which are easy to administer for children with epilepsy who are 8 years and older and their parents. The subscales encompass HRQL dimensions judged most important by children with epilepsy for the self-report measure and by parents for the proxy response measure. The parent-proxy measure should be useful as a complement to the child self-report measure in evaluating the validity of parental assessment of the child's health status; in longitudinal outcome research; and in HRQL assessment of children who are unable to respond independently.