Trends in Socioeconomic Inequalities in Breast Cancer Incidence Among Women in Canada.

INTRODUCTION: Breast cancer is the most common cancer among females in Canada. This study examines trends in socioeconomic inequalities in the incidence of breast cancer in Canada over time from 1992 to 2010. METHODS: A census division level dataset was constructed using the Canadian Cancer Registry, Canadian Census of the Population and National Household Survey. A summary measure of the Concentration index (C), which captures inequality across socioeconomic groups, was used to measure income and education inequalities in breast cancer incidence over the 19-year period. RESULTS: The crude breast cancer incidence increased in Canada between 1992 and 2010. Age-standardized C values indicated no income or education inequalities in breast cancer incidence in the years from 1992 to 2004. However, the incidence was significantly concentrated among females in high income and highly educated neighbourhoods almost half the time in the 6 most recent years (2005-2010). The trend analysis indicated an increase in breast cancer incidence among females living in high income and highly educated neighbourhoods. CONCLUSION: Breast cancer incidence in Canada was associated with increased socioeconomic status in some more recent years. Our study findings provide previously unavailable empirical evidence to inform discussions on socioeconomic inequalities in breast incidence.

Socioeconomic inequalities in colorectal cancer incidence in Canada: trends over two decades.

PURPOSE: Colorectal cancer is the third most commonly diagnosed cancer in Canada. This study aimed to measure and examine trends in socioeconomic inequalities in the incidence of colorectal cancer in Canada. METHODS: This study is a time trend ecological study based on Canadian Census Division level data constructed from the Canadian Cancer Registry, Canadian Census of Population, and National Household Survey. We assessed trends in income and education inequalities in colorectal cancer incidence in Canada from 1992 to 2010. The age-standardized Concentration index ([Formula: see text]), which measures inequality across all socioeconomic groups, was used to quantify socioeconomic inequalities in colorectal cancer incidence in Canada. RESULTS: The average crude colorectal cancer incidence was found to be 61.52 per 100,000 population over the study period, with males having a higher incidence rate than females (males: 66.98; females: 56.25 per 100,000 population). The crude incidence increased over time and varied by province. The age-standardized C indicated a higher concentration of colorectal cancer incidence among lower income and less-educated neighborhoods in Canada. Income and education inequalities increased over time among males. CONCLUSION: The concentration of colorectal cancer incidence in low socioeconomic neighborhoods in Canada has implications for primary prevention and screening.

Socioeconomic Gradients in Prostate Cancer Incidence Among Canadian Males: A Trend Analysis From 1992 to 2010.

INTRODUCTION: Understanding the effects of socioeconomic status on cancer incidence and their trends over time will help inform public health interventions for cancer control. This study sought to investigate trends in socioeconomic inequalities in prostate cancer incidence among Canadian males. METHODS: Using a census division level dataset (n = 280) constructed from the Canadian Cancer Registry, Canadian Census of Population (1992, 1996, 2001, 2006) and 2011 National Household Survey, we examined the effect of socioeconomic status on prostate cancer incidence among Canadian males between 1992 and 2010. The age-adjusted concentration index was used to quantify education/income-related inequalities in prostate cancer incidence. RESULTS: The crude prostate cancer incidence increased from 115 to 137 per 100 000 males in Canada from 1992 to 2010 with a peak in 2007. The rate increased significantly in all but three of four western provinces. The age-adjusted concentration index showed a higher concentration of prostate cancer diagnoses among males living in high-income neighbourhoods in Canada in particular from 1996 to 2005. In contrast, the index was higher among males living in less-educated neighbourhoods in the most recent study years (2006-2010). CONCLUSIONS: The concentration of new prostate cancer cases among high-income populations in Canada may be explained by the rise of opportunistic screening of asymptomatic males; however, this should be studied in further detail. Since we found a higher incidence rate of prostate cancer among less-educated males in Canada in recent years, risk-benefit investigation of primary prevention and opportunistic screening for less-educated males is advised.

Population surveillance of navigation frequency and palliative care contact before death among cancer patients.

Cancer patient navigation in Canada began in 2002 in Nova Scotia with oncology nurses providing support to patients from diagnosis up to and including end of life. This novel study was carried out to determine navigation frequency and palliative care contact rates, and variations in these rates among adults who were diagnosed with cancer, navigated, and then died between 2011 and 2014. Among the 2,532 study subjects, 56.7% were navigated for more than one month and 30.6% had palliative care contact reported. Variations were observed by geographic area, cancer stage, time from diagnosis to death, and whether the person died of cancer. Further study of the role of navigation is advised for persons at end of life.

Rules to Identify Persons with Frailty in Administrative Health Databases.

This study sought to develop frailty "identification rules" using population-based health administrative data that can be readily applied across jurisdictions for living and deceased persons. Three frailty identification rules were developed based on accepted definitions of frailty, markers of service utilization, and expert consultation, and were limited to variables within two common population-based administrative health databases: hospital discharge abstracts and physician claims data. These rules were used to identify persons with frailty from both decedent and living populations across five Canadian provinces. Participants included persons who had died and were aged 66 years or older at the time of death (British Columbia, Alberta, Ontario, Quebec, and Nova Scotia) and living persons 65 years or older (British Columbia, Alberta, Ontario, and Quebec). Descriptive statistics were computed for persons identified using each rule. The proportion of persons identified as frail ranged from 58.2-78.1 per cent (decedents) and 5.1-14.7 per cent (living persons).

Identifying persons with diabetes who could benefit from a palliative approach to care.

OBJECTIVE: To determine the need for diabetes mellitus palliative care, we identified persons with a diagnosis of diabetes who accessed palliative care programs and those who may have benefited from a palliative approach to care. METHODS: This retrospective, descriptive research used 6 linked databases comprising 66 634 Nova Scotians from 3 health districts who died between 1995 and 2009, each with access to a palliative care program and diabetes centres. RESULTS: The percentage of persons with diabetes enrolled in palliative care increased from 3.2% in 1995 to 34.3% in 2009; 31.5% were enrolled within their last 2 weeks of life. Most did not have their diabetes recorded in palliative data. Among the 5353 persons with a diagnosis of diabetes who died between 2005 and 2009, 61.0% were in the Diabetes Care Program of Nova Scotia registry. An additional 19.6% were identified in the Cardiovascular Health Nova Scotia registry, and a further 3.7% in palliative data. Applying the criteria of Rosenwax et al to the 5353, 65.8% to 97.9% may have benefitted from a palliative approach. CONCLUSIONS: Rates of palliative enrollment for persons with diabetes are increasing. Diabetes care providers need to prepare patients and their families for changes in diabetes management that will be beneficial as end of life approaches. Collaboration among chronic disease programs, palliative care and primary care is advised to identify persons at end of life who have diabetes and to develop and implement care guidelines for this population.

Co-morbidities of persons dying of Parkinson's disease.

INTRODUCTION: Disease interactions can alter functional decline near the end of life (EOL). Parkinson's disease (PD) is characterized by frequent occurrences of co-morbidities but data challenges have limited studies investigating co-morbidities across a broad range of diseases. The goal of this study was to describe disease associations with PD. METHODS: We conducted an analysis of death certificate data from 1998 to 2005 in Nova Scotia. All death causes were utilized to select individuals dying of PD and compare with the general population and an age-sex-matched sample without PD. We calculated the mean number of death causes and frequency of disease co-occurrence. To account for the chance occurrence of co-morbidities and measure the strength of association, observed to expected ratios were calculated. RESULTS: PD decedents had a higher mean number of death causes (3.37) than the general population (2.77) and age-sex-matched sample (2.88). Cancer was the most common cause in the population and matched sample but fifth for those with PD. Cancer was one of nine diseases that occurred less often than what would be expected by chance while four were not correlated with PD. Dementia and pneumonia occurred with PD 2.53 ([CI] 2.21-2.85) and 1.83 (CI 1.58-2.08) times more often than expected. The strength of association for both is reduced but remains statistically significant when controlling for age and sex. DISCUSSION: Those with PD have a higher number of co-morbidities even after controlling for age and sex. Individuals dying with PD are more likely to have dementia and pneumonia, which has implications for the provision of care at EOL.

Identifying population groups with low palliative care program enrolment using classification and regression tree analysis.

Classification and regression tree (CART) analysis was used to identify subpopulations with lower palliative care program (PCP) enrolment rates. CART analysis uses recursive partitioning to group predictors. The PCP enrolment rate was 72 percent for the 6,892 adults who died of cancer from 2000 and 2005 in two counties in Nova Scotia, Canada. The lowest PCP enrolment rates were for nursing home residents over 82 years (27 percent), a group residing more than 43 kilometres from the PCP (31 percent), and another group living less than two weeks after their cancer diagnosis (37 percent). The highest rate (86 percent) was for the 2,118 persons who received palliative radiation. Findings from multiple logistic regression (MLR) were provided for comparison. CART findings identified low PCP enrolment subpopulations that were defined by interactions among demographic, social, medical, and health system predictors.

Exploring generalizability in a study of costs for community-based palliative care.

CONTEXT: Palliative care researchers face challenges recruiting and retaining study subjects. OBJECTIVES: This article investigates selection, study site, and participation biases to assess generalizability of a cost analysis of palliative care program (PCP) clients receiving care at home. METHODS: Study subjects' sociodemographic, geographic, survival, disease, and treatment characteristics were compared for the same year and region with those of three populations. Comparison I was with nonstudy subjects enrolled in the PCP to assess selection bias. Comparison II was with adults who died of cancer to assess study site bias. Comparison III was with study-eligible persons who declined to participate in order to assess participation bias. RESULTS: Comparison I: When compared with the other 1010 PCP clients, the 50 study subjects were on average 3.6 years younger (P=0.03), enrolled 70 days longer in the PCP (P<0.001), lived 6.7 km closer to the PCP (P<0.0001), and were more likely to have cancer (96.0% vs. 86.4%, P=0.05). Comparison II: Compared with all cancer decedents, the 45 study subjects who died of cancer were on average 7.0 years younger (P<0.001), lived 2.7 km closer to the PCP (P<0.001), and were more likely to have had radiotherapy (62.2% vs. 33.8%, P<0.0001) and medical oncology (28.9% vs. 14.8%, P=0.01) consultations. Comparison III: The 50 study subjects lived on average 42 days longer after their diagnosis (P=0.03) and 2.6 km closer to the PCP (P=0.01) than the 110 eligible persons who declined to participate. CONCLUSION: If the study findings are applied to populations that differ from the study subjects, inaccurate conclusions are possible.

Variation in the use of palliative radiotherapy at end of life: examining demographic, clinical, health service, and geographic factors in a population-based study.

Palliative radiotherapy (PRT) can improve quality of life for people dying of cancer. Variation in the delivery of PRT by factors unrelated to need may indicate that not all patients who may benefit from PRT receive it. In this study, 13,494 adults who died of cancer between 2000 and 2005 in Nova Scotia, Canada, were linked to radiotherapy records. Multivariate logistic regression was used to examine the relationships among demographic, clinical, service, and geographic variables, and PRT consultation and treatment. Among the decedents, 4188 (31.0%) received PRT consultation and 3032 (22.3%) treatment. PRT declined with increased travel time and community deprivation. Females, older persons, and nursing home residents also had lower PRT rates. Variations were observed by cancer site and previous oncology care. Variations in PRT use should be discussed with referring physicians, and improved means of access to PRT considered. Benchmarks for optimal rates of PRT are needed.

A population-based study of age inequalities in access to palliative care among cancer patients.

BACKGROUND: Inequalities in access to palliative care programs (PCP) by age have been shown to exist in Canada and elsewhere. Few studies have been able to provide greater insight by simultaneously adjusting for multiple demographic, health service, and socio-cultural indicators. OBJECTIVE: To re-examine the relationship between age and registration to specialized community-based PCP programs among cancer patients and identify the multiple indicators contributing to these inequalities. METHODS: This retrospective, population-based study was a secondary data analysis of linked individual level information extracted from 6 administrative health databases and contextual (neighborhood level) data from provincial and census information. Subjects included all adults who died due to cancer between 1998 and 2003 living within 2 District Health Authorities in the province of Nova Scotia, Canada. The relationship between registration in a PCP and age was examined using hierarchical nonlinear regression modeling techniques. Identification of potential patient and ecologic contributing indicators was guided by Andersen's conceptual model of health service utilization. RESULTS: Overall, 66% of 7511 subjects were registered with a PCP. Older subjects were significantly less likely than those <65 years of age to be registered with a PCP, in particular those aged 85 years and older (adjusted odds ratio: 0.4; 95% confidence interval: 0.3-0.5). Distance to the closest cancer center had a major impact on registration. CONCLUSIONS: Age continues to be a significant predictor of PCP registration in Nova Scotia even after controlling for the confounding effects of many new demographic, health service, and ecologic indicators.

End-of-life care for nursing home residents dying from cancer in Nova Scotia, Canada, 2000-2003.

INTRODUCTION: With our population aging, an increasing proportion of cancer deaths will occur in nursing homes, yet little is known about their end-of-life care. This paper identifies associations between residing in a nursing home and end-of-life palliative cancer care, controlling for demographic factors. METHODS: For this population-based study, a data file was created by linking individual-level data from the Nova Scotia Cancer Centre Oncology Patient Information System, Vital Statistics, and the Halifax and Cape Breton Palliative Care Programs for all persons 65 years and over dying of cancer from 2000 to 2003. Multivariate logistic regression was used to compare nursing home residents to nonresidents. RESULTS: Among the 7,587 subjects, 1,008 (13.3%) were nursing home residents. Nursing home residents were more likely to be female [adjusted odds ratio (OR) 1.4, 95% confidence interval (CI) 1.2-1.7], older (for > or = 90 vs 65-69 years OR 5.4, CI 4.1-7.0), rural (OR 1.5, CI 1.2-1.8), have only a death certificate cancer diagnosis (OR 4.2, CI 2.8-6.3), and die out of hospital (OR 8.5, CI 7.2-10.0). Nursing home residents were less likely to receive palliative radiation (OR 0.6, CI 0.4-0.7), medical oncology consultation (OR 0.2, CI 0.1-0.4), and palliative care program enrollment (Halifax OR 0.2, CI 0.2-0.3; Cape Breton OR 0.4, CI 0.3-0.7). CONCLUSION: Demographic characteristics and end-of-life services differ between those residing and those not residing in nursing homes. These inequalities may or may not reflect inequities in access to quality end-of-life care.

Radiotherapy wait times for patients with a diagnosis of invasive cancer, 1992-2000.

PURPOSE: To study the wait times for cancer patients from the time of diagnosis to consultation with a radiation oncologist (T1), from consultation to radiotherapy (T2) and from diagnosis to radiotherapy (T3) in the context of treatment practices and measurement issues. METHODS: From 1992 to 2000, we studied 6585 Nova Scotian patients over the age of 24 years with a diagnosis of breast, lung, colorectal or prostate cancer who received radiotherapy within 1 year of diagnosis. Multivariate analyses examined associations between wait time and diagnosis year, age, sex, median household income (MHI), distance to the cancer centre and extent of disease. Univariate findings reported are median times and interquartile ranges. RESULTS: The T3 was 16 weeks for breast and colorectal cancer, 6 weeks for lung cancer and 18 weeks for prostate cancer. The greatest T1 decrease over time was for prostate cancer: 13-8 weeks (hazards ratio [HR] = 1.07, 95% confidence interval [CI] = 1.05-1.10). The T2 increased for all cancers, and the T3 increased from 5 to 7 weeks for lung cancer, from 17 to 22 weeks for prostate cancer and from 10 to 18 weeks for breast cancer, with no change for colorectal cancer. The T3 decreased by age for breast cancer (HR = 1.12, CI = 1.10-1.14) and prostate cancer (HR = 1.07, CI = 1.02-1.11), showed no consistent association with distance to a cancer centre and varied by extent of disease. Patients with localized lung disease had a longer T3 than those with distant disease, but the opposite results were noted for patients with breast cancer. The T3 was greater for regional than distant disease in lung and breast cancers. Sex and MHI had no effect. CONCLUSION: Wait times reflected clinical practice, and there were no adverse patterns related to age, sex, income or distance from a cancer centre.

Community-based cultural predictors of Pap smear screening in Nova Scotia.

BACKGROUND: Pap smear screening is effective in reducing the incidence of cervical cancer. However, some subgroups of women are less likely to be screened than others. Since Canadian provincial health databases do not contain data fields identifying ethnicity or language, analyses employing these variables are typically not available. This paper overcomes this problem by using community- rather than person-based measures. Associations with having had a recent Pap smear are reported by community income, language, ethnic group, and urban/rural status, as well as the woman's age. METHODS: The provincial Health Card Number and Cytology Registries were linked to ascertain the screening status of women in mainland Nova Scotia and Cape Breton. Postal codes were linked to census enumeration areas and then to Statistics Canada census data to create community-based cultural measures for each woman. RESULTS: Women in mainland Nova Scotia were more likely to have had a recent Pap smear (Odds Ratio (OR)=1.36; 95% Confidence Interval (CI):1.33-1.39). Women living in low income (OR=1.19; CI:1.15-1.22), Aboriginal (OR=1.60; CI:1.46-1.76), mixed Black (OR=1.25; CI:1.19-1.30) and rural (OR=1.09; CI:1.07-1.11) communities and who were older were less likely to have had a recent Pap smear. DISCUSSION: These findings were not unexpected. In the United States and elsewhere, associations between Pap screening status and women with low income, rural residence, Aboriginal and Black heritage have been reported using person-based methods. Our findings demonstrate a method of providing measures of ethnicity and language that should be considered for use in Canadian studies of service utilization, disease status, and well-being.