Children with chronic illnesses have a doubled risk of developing psychosocial maladjustment--emotional problems, behavior disorder or difficulties in social relationships. Social work support and counselling aims to reduce this secondary morbidity, and is a common form of hospital-based psychosocial service. The first randomized controlled trial of this type of intervention was carried out to evaluate its effectiveness in treating and preventing maladjustment. This thesis describes how child behavior outcomes were assessed before and 4 months after a 6 month period of social worker assistance in 173 children randomized to intervention, and in 169 controls, all with chronic illnesses. / No significant difference between intervention and control groups in the overall prevalence of maladjustment was found. There was no evidence to support a therapeutic or preventive effect of social work counselling on child behavior outcomes, nor was there improvement in child perceived competence. A search for treatment interactions failed to reveal any sub-group that benefitted from the intervention, and restriction of the analysis to individuals who actually received the intervention does not alter any of these conclusions. / Measurement problems, co-intervention, or other forms of bias cannot account for the negative results. It is speculated that if social work support is to be effective, it should be targetted, potent, of adequate duration, and possibly integrated within specialist clinic services.

In order to understand and document the patterns of smoking among youth, the use of reliable and valid measures is essential. The current analysis uses data on 63 ever smokers (mean age 14.1 years) from the first three years of follow-up of the McGill University Study on the Natural History of Nicotine Dependence In Teens. To examine reliability, we compared responses to one-month recall to three-month recall of the subjects' smoking history. The questionnaire demonstrated very good reliability for smoking intensity (kappa = 0.75) and smoking frequency (kappa = 0.78). Number of cigarettes smoked per month showed good reliability (ICC = 0.64). Reliability remained acceptable when subjects who reported no cigarette consumption at both questionnaire administrations were excluded (kappa = 0.66; kappa = 0.70; ICC = 0.61, respectively). Older adolescents (14--16) were better at recalling their smoking history than younger adolescents (12--13). Female sex and smoking status (sporadic versus regular) were associated with adequate recall for one and two of the indicators investigated, respectively.

It has recently been demonstrated that a continuum of self-management (SM) for COPD patients provided by a trained health professional significantly reduce the utilisation of health care services (Bourbeau et al Arch Intern Med March 2003). The objective of the present study was to assess in COPD patients who received a self-management program of the beneficial effect of reduce hospitalisation could be maintained at long term. A multicentre, randomised clinical trial was carried out in 191 COPD patients from 7 hospitals. Patients who all had one hospitalisation in the year preceding study enrolment were assigned to a self-management program or to usual care. The self-management program "Living Well with COPD(c)" consisted of a multi-component patient education program administered through weekly visits by trained health professionals over a 2 months period with monthly telephone follow-up for 12 months. All cause hospitalisations, the main outcome, were marked using the provincial hospital database (MED-ECHO). / Most patients were elderly, not highly educated and had advanced COPD reflected by a mean FEV1 of 1L and 46% reported a dyspnea score of 5/5 (modified MRC). / The long-term effectiveness of the self-management program was assessed with an intention-to-threat analysis. The group-level chi-square analysis shows a statistically significant difference of 29% in the rate of all-cause hospitalization between the intervention group and the usual-care group. The individual-level proportional odds model analysis shows a statistically significant risk estimate of 0.566. / The study shows that patients with COPD, who received an education intervention with supervision and support based on disease specific self-management, still had a significant reduction in hospitalisations at 2 years. Within normal practice such program should be an integral part of the continuum of care of patients with COPD.

The consequences of stroke extend beyond the individual who has sustained the stroke to include family members, the main providers of care to community dwelling stroke survivors. Caregiver research, for the most part, has focused on burden and distress. Although there is a growing consensus among health care researchers concerning the relevance of quality of life (QoL) as an outcome, there is little research examining QoL among family caregivers and even less focused on conceptualization of the construct. This thesis was designed to address these gaps in theory and measurement through a series of related studies. / The development of a conceptual model of QoL for family caregivers, that includes the caregiving situation, caregiver characteristics, and environmental factors, and their proposed relationships with QoL, provided a framework for the measurement and study of QoL. Relationships between patient characteristics, caregiver burden, and health-related quality of life (HRQL) were examined over the first two years of caregiving and the results confirmed the impact of the caregiving role on the caregiver's emotional health, and additionally, supported the need for a specific measure of overall QoL to more fully capture the caregiving experience. / A subsequent study was conducted to develop this measure of QoL for family caregivers of stroke survivors. Data from several different samples of family caregivers contributed to the specific phases required in measure development and psychometric testing. Four sub-scales, 'health', 'resources', 'role mastery', and 'rewards', were constructed. Although the overall psychometric properties of this measure are promising, they also indicate the need for further development of selected items and the 'rewards' subscale measuring the positive dimension. / The specific hypotheses concerning the relationships among the variables in the conceptual framework of QoL of family caregivers were tested using structural equation modeling and they were partially supported. The results suggest several approaches for intervening with family caregivers. As restructuring of health care delivery is placing more responsibility for the care of disabled members on the family, we require effective interventions to ensure that we preserve this valuable commodity.

Clostridium difficile is a bacterium which causes gastro-intestinal infection. The association between non-steroidal anti-inflammatory drugs (NSAIDs) and Clostridium difficile-associated disease (CDAD) has not been studied. Several case reports link diclofenac, an NSAID, with CDAD. In this thesis, we conducted a case-control study, using data from the United Kingdom's General Practice Research Database (GPRD), to examine the risk of CDAD associated with NSAID use. We identified 1,360 cases and 13,072 matched controls from 1994 through 2005. Using conditional logistic regression, we found an increased risk of CDAD associated with diclofenac [adjusted rate ratio 1.35; 95% confidence interval: 1.10-1.67]. We did not observe an increased risk of CDAD with use of any other NSAID. In addition, no dose response for diclofenac was found. In conclusion, diclofenac was associated with an increased risk of CDAD. Several NSAIDs could be prescribed in place of diclofenac, reducing the risk of CDAD without additional inconveniences. / Le Clostridium difficile est une bactérie qui cause des infections gastro-intestinales. L'association entre les anti-inflammatoires non-stéroïdiens (AINS) et le C. difficile n'a jamais été étudiée. Plusieurs rapports de cas lient le diclofenac (AINS) au C. difficile. Dans cette étude cas-témoin, basée sur des données du United Kingdom's General Practice Research Database, le risque de C. difficile suite à l'utilisation d'AINS a été évalué. 1360 cas et 13072 contrôles ont été identifiés entre 1994 et 2005. À l'aide de régression logistique, nous trouvons une augmentation du risque de C. difficile chez les utilisateurs de diclofenac [Risque relatif ajusté 1.35; intervalle de confiance 95%: 1.10-1.67]. Aucun autre AINS n'est associé à une augmentation d'infection par le C. difficile. De plus, le diclofenac n'a pas d'effet de dose. En conclusion, le diclofenac est associé à un risque accru de C. difficile. Ce risque peut être éliminé en remplaçant le diclofenac par d'autres AINS.

Inuit infants throughout the Arctic experience higher mortality and poorer health than their non-Inuit counterparts, and suffer disproportionately from bacterial and viral infections. This research initially reviews the health status of these infants, with a focus on Canadian Inuit communities and reference to other circumpolar regions, as appropriate. It also discusses the wide range of inter-related factors that affect their health and their susceptibility to infection: their demographic, social, economic and physical environment, as well as personal health practices and the availability of high quality, culturally-appropriate health services within their communities. / Data were then analyzed from a cohort study of 46 healthy Inuit infants that had been previously conducted in Iqaluit, Nunavut from December 1995 to November 1997. Hospitalization and morbidity patterns were examined over their first year of life. Infants experienced an average of four respiratory tract infections (RTIs) annually, which accounted for half of the hospitalizations in the cohort. Some interesting trends were evident from assessment of risk factors for hospitalization and infections using multiple linear regression. Infants of mothers with higher educational attainment spent six fewer days in hospital per year (95% CI: -14.6, 2.9), after adjustment for confounding variables. Adoption appeared to have adverse health effects in addition to those that would be expected due to lack of breastfeeding alone; among infants who were not breast-fed, adopted infants had three more RTIs per year than non-adopted infants (95% CI: 0.5, 5.1). These results provide support for undertaking larger epidemiological studies in order to clarify the role of these risk factors, so that future preventive efforts can be informed and effective.

The primary objective of this study was to assess the effectiveness of a hygiene programme in reducing the incidence of respiratory and diarrheal diseases in toddlers attending day care centres (DCCs). A randomized field trial was conducted in 52 DCCs in Quebec. One or two toddler groups in each DCC were followed between September 1st, 1996 and November 30, 1997. Absences for any reasons and the daily occurrence of colds and/or diarrhea in toddlers were recorded on calendars by the DCC educators. Questionnaires to the DCC director and on-site visits were used to measure potential confounding variables. The number of fecal coliforms (FC) on children's hands, on educators' hands, in the sandbox and in the playarea of each DCC was measured during three unannounced visits. Participating parents were asked to record on a 15-month calendar the occurrence of colds and diarrhea in the child and the actions taken when s/he was ill. Overall, 1,729 children were followed in 47 DCCs for a total of 153,643 child-days. The incidence rate of diarrhea was significantly reduced by the effect of monitoring alone (IRR = 0.73, 95% BCI = 0.54, 0.97) but not significantly reduced by the intervention whereas the intervention had an effect in reducing the IR of upper respiratory tract infections (URTI) (IRR = 0.80, 95% BCI = 0.68, 0.93) but the monitoring did not. Monitoring also had a significant effect in reducing the level of bacterial contamination on children's and educators' hands. Overall estimates of the IRs of respiratory and diarrheal infections based on parents' data were higher than those based on educators' data (difference of 0.81 episode of diarrhea and 2.2 episodes of URTI per child-DCC year-at-risk). During the 6-month pre-intervention period, the average direct, indirect and total costs per child due to the study illnesses were $132.06, $235.58 and $367.64, respectively. These results indicate that both a hygiene intervention program and monitoring alone play a signifi

To determine the effects of exposure to nonsteroidal antiinflammatory drugs (NSAIDs) on the risk of invasive female breast cancer we carried out a nested case-control study, using the beneficiaries of the Saskatchewan Prescription Drug Plan (SPDP) from 1981 to 1995 with no history of cancer since 1970 as the source population. Four age-matched controls were randomly selected for each case using incidence density sampling. Dispensing rates, calculated over successive time periods, characterized NSAID exposure according to dosage and the timing of exposure. / We accrued 5,882 cases and 23,517 controls. Increasing NSAID exposure during the period 7--12 months preceding diagnosis was associated with a trend towards an increasing rate ratio (RR) for the diagnosis of breast cancer (p trend = 0.003), which was attributed to detection bias because it was markedly reduced by eliminating from the analysis the cases with tumours most likely to have been detected by screening (≤ 2 cm diameter). Increasing NSAID exposure 2--5 years preceding diagnosis was associated with a trend towards a decreasing RR (p-trend = 0.003); for the highest exposure level RR = 0.76, 95% CI: 0.63--0.92. This protective effect could not be attributed to confounding by other risk factors. Confounding by risk factors not in the SPDP database was studied in analyses adjusted with data obtained by interviewing samples of subjects accrued from mid1991 to mid-1995, selected according to both disease and exposure status. / Increasing NSAID exposure 2--5 years and 6--10 years before diagnosis was associated with trends towards decreasing RRs for having a primary tumour >5 cm diameter at diagnosis (p-trend = 0.06 and 0.02, respectively). NSAID exposure did not alter the risk of regional lymph node metastasis. However, increasing NSAID exposure 2--5 years and 6--10 years before diagnosis was associated with trends towards decreasing RRs for having distant metastases at diagnosis (p-trend = 0.0003 and 0.03, respectively). / The use of NSAIDs may retard the growth of established, undiagnosed breast cancers and may decrease the risk of distant metastasis.

The province of Quebec screens for autosomal recessive phenylketonuria (PKU) and other forms of hyperphenylalaninemia due to phenylalanine hydroxylase deficiency and tetrahydrobiopterin variants in newborns. A review of the annual reports of the Quebec Newborn Screening Program and of the clinical files of individuals with hyperphenylalaninemia born in Quebec since 1970 was undertaken. The Newborn Screening Program was evaluated for its ability to detect and identify individuals with hyperphenylalaninemia, to characterize their phenotype, and to continue surveillance. Less than universal participation in the screening (98.6%) and loss to follow-up of individuals not on treatment are causes for concern in the context of maternal hyperphenylalaninemia. Characteristics of individuals with PKU or non-PKU HPA including ethnicity, age at screening test, administrative region of birth, and month of birth were analyzed.

This study evaluates the prevalence and incidence rates of infection with Human immunodeficiency Virus (HIV), hepatitis B (HBV) and hepatitis C (HCV), in a cohort of drug users (DU) on methadone maintenance treatment in Geneva, Switzerland. Over 700 DU participated between 1988 and 1995; the follow-up rate was high. The prevalence rate at entry into treatment declined dramatically over time for all 3 viruses. Comparing DU born before 1961 to those born after 1970 the prevalence rate of HIV was 29. 1% versus 2.0%, of HBV 71.3% versus 2.2%, and of HCV 83.6% versus 17.9%. The incidence rates for HIV and HBV were low (0.6 and 2.1 per 100 person years of follow up). For HCV the rate was high (4.2) with a slightly higher rate among women. These data suggest that DU have changed HIV risk taking behaviour in response to HIV prevention campaigns. Current prevention efforts should focus on improvement of HCV prevention and maintaining safe behaviour.