Social Determinants in Chronic Disease in British Columbia

Overview

Anyone who practices clinical medicine will understand that socially disadvantaged children will have worse health outcomes, no matter what the underlying condition might be. There is limited prospective data on the effects of social deprivation on children in BC and there is none concerning the effects of social deprivation on children with chronic diseases. In order to generate relevant data for those who manage children with chronic diseases in BC, the investigators wish to perform an observational study of the relationship between questionnaire-derived social variables and measured outcomes in children with cystic fibrosis, type 1 diabetes, and chronic kidney disease. Our working hypothesis is that there is an association between social determinants of health (income, education, race) and health outcomes in children with cystic fibrosis, type 1 diabetes and chronic renal failure, that is independent of access to health care (assessed by distance to nearest specialty clinic and number of clinic visits in the last year).

1. Cystic fibrosis (CF). Cystic fibrosis is a genetic condition that causes slowly progressive lung disease. It occurs in people of all social class so the disease itself is independent of social factors. There is considerable variation in the outcome of CF in different children. There are many potential causes for these differences but the available literature would suggest that a significant part of this variability is due to social factors. While there are no prospective clinical studies, the limited information from database reviews suggest that socially deprived children have significantly worse outcomes. In the US, children living in poverty are twice as likely to be underweight and over three times more likely to die (25). UK data is similar - disadvantaged children have worse growth parameters and lower lung function compared to richer children (26). 2. Type 1 diabetes (T1D). Type 1 diabetes is a multi-factorial disease. It is not an autosomal recessive disease like CF but genetic predisposition can be one of its predisposing features. As with the data on CF, socio-economic deprivation has been shown to have a significant adverse effect on outcome in type 1 diabetes. In the US, glycemic control amongst young adults was significantly worse amongst disadvantaged young adults (27). A retrospective review of New Zealand children with type 1 diabetes also showed that poor socio-economic status and Polynesian ethnicity were significantly associated with poor glycemic control and long term complications (28). 3. Chronic kidney disease (CKD). Chronic kidney disease may be due to congenital, acquired, or hereditary causes. Within all these groups, socioeconomic status, gender, and race may influence disease progression and outcomes. It has been shown in children with chronic kidney disease, primarily from the United States, that it was more common in low- and middle- income families to have an abnormal birth history such as prematurity, low birth weight or small for age which predisposes to the development of renal disease (29). In addition, blood pressure control and height deficits improve faster in children of families with higher income. In Canada, Aboriginals have a higher prevalence of severe chronic kidney disease and a 77% increased mortality risk (30,31) and a lower likelihood of nephrology clinic visits (32). Aboriginal children and young adults with chronic kidney disease are more likely to have glomerulonephritis as a cause of kidney disease compared to Caucasians and are more likely to reach end stage renal disease (33). In contrast, lower mortality risks are seen in East Asian and Indo Asian Canadian adults starting dialysis (34). There is still much to be studied on the influence of these various factors in renal health and disease. STUDY DESIGN. 1. Study Objective. We wish to determine the influence of education, income, race and access to health care upon health outcomes in children with chronic diseases in British Columbia. 2. Justification. While the effects of social deprivation on health outcomes are commonly discussed, it is surprising how little prospective research there is to quantify the adverse effects of social inequity. There is limited prospective data on the effects of social deprivation on children in BC and there is none concerning the effects of social deprivation on children with chronic diseases. In order to generate firm data relevant for those who manage children with chronic diseases in BC, we wish to perform a cross sectional observational study of the relationship between social variables and outcome in children with cystic fibrosis, type 1 diabetes and chronic kidney disease. 3. General design. All families in the three clinics will initially be informed of the study by letter. When the child is due for regular follow-up assessment, the parents will be approached during the clinic visit and invited to join the study. Their participation will only involve the completion of a questionnaire. The research assistant will collect the data at interview and will subsequently (with parental permission) retrieve laboratory results from records of previous clinic visits to assess measured health trends over time. The study protocol does not add any further investigations other than the questionnaire. Apart from the variable time needed for parent information and consent, the questionnaire itself will require no more than about 30 minutes to complete. 4. Questionnaire. We will use a socio-economic assessment questions that we are already using in our study of quality of life in parents caring for children needing home ventilation. It is a conventional form assessing education level, employment, income, partnership status and ethnicity. This last is very important because of the poorer health outcomes already known to be associated with Aboriginal background. The proxy for access to health care will be distance to the main specialty clinic and number of visits in the last year. 5. Analysis. The relationship between socio-economic variables plus ethnicity and outcome will be examined using multiple linear regression analysis. In the absence of relevant literature, there are too many unknowns to allow any reliable form of power analysis or patient number calculation to be performed. One accepted 'rule of thumb' is that there should be at least 10 patients for each predictive variable in multiple linear regression analysis. We will concentrate on 6 to 8 of the main social factors and will try to enroll over 100 patients within each sub-specialty - enough to ensure 15 to 20 patient outcomes for each variable.