Colorectal cancer is the third most prevalent cancer in the US and leads to the second-highest number of cancer deaths. Early detection and prevention are ranked as the most effective methods of reducing mortality rates. Early detection encompasses a wide range of screening methods such as colonoscopy, guaiac-based fecal occult blood testing (FOBT), sigmoidoscopy, and fecal immunochemical tests (FIT) ( Williams et al., 2016, p.2). New York City records an estimated 2,000 annual colorectal deaths and 3,500 new cases. The city rolled out the plan to increase colorectal cancer awareness and colonoscopy screening among persons aged above 50 years to respond to the threat. The process entailed public education, targeted screening among minority ethnic communities such as blacks, and free screening services at public and volunteer hospitals. Since the implementation of the strategy, the city has recorded increases in colorectal cancer screening. Williams et al. (2016) sought to determine the effects of colorectal cancer incidence and death rate on the number of persons screened and their racial/ethnic diversity between 2003 and 2016. The expectation was that an increase in screening across the diverse groups would cause a decline in colorectal cancer incidence and mortality rate.
The null hypothesis assumed a reduction in screening disparity across the different ethnic groups was related to colorectal incidence and deaths across the ethnic groups. The study utilized multiple data sources to test the hypothesis, including New York State Cancer Registry data, New York City Vital Statistics data, and New York City Community Health Survey (CHS) data. While secondary data from both the New York State Cancer Registry and NYC Vital Statistics data were publicly available via the organization’s websites, the use of New York City Community Health Survey data required approval from the NYC Health Department’s IRB.
Delegate your assignment to our experts and they will do the rest.
The study collected colorectal cancer incidence data covering the stage of the disease at diagnosis from a secondary source, the New York State Cancer Registry, in the period between 1975 and 2016. The sample size of 65,550 and consisting of the race/ethnic and borough of residence was collected from the data of 173,388 persons. A sample size of 19,241 on the colorectal deaths was obtained from the NYC Office of Vital Statistics for the period between 2003 and 2016, including the race/ethnicity and borough of residence. New York City Community Health Survey involved a sample of 9,000 respondents aged above 50 years old. The survey was conducted annually to track the colonoscopy status of the New York population. The New York City Community Health Survey is a random sample conducted through random-digit-dial to persons living in the city. The calls are made both through landline and cellular telephone. Also, the survey uses a disproportionate stratified random sample design to capture the diverse population. The data obtained from the survey include sociodemographic and health behavior information such as colonoscopy, health risk behavior, healthy diet, and access to healthcare.
The data was analyzed using a regression model. The analysis generated curves for mortality and incidence based on the racial/ethnic groups of the participants. The analysis also considered the statistical significance of annual percent change and average percent change in incidence and mortality between 2003 and 2016 using a two-sided t-test. The annual percentage change across racial/ethnic groups was also tested using regression to analyze the difference in trends and identify disparities. The overall association between screening rates and the overall mortality and incidence was also analyzed using regression. All analyses were conducted using the SAS software Version 9.4.
Williams et al. (2016) established a significant decline in the age-adjusted colorectal cases at all stages in the period between 2000 and 2016. The decline in annual percentage within the period was from 57.5% to 37.3% in a population of 100,000 (t = − 2.79, p < 0.0001 sample size = 65,550). The declines in annual percentage incidence change were also significant in all the involved boroughs with the t statistic for Bronx= − 3.3, Brooklyn = − 3.1, Manhattan= − 3.6, Queens= − 2.5, and Staten Island = − 2.7 and the p-values < 0.0001. When tested for parallelism, the results showed insignificant differences in the colorectal incidence and mortality rate. The incidence rate was, however, higher among the blacks. The rate average incidence rate of colorectal cancer in 2016 was 42.5 per 100,000 with a 95% confidence interval of between 39.7 and 45.4 (95% CI: 39.7–45.4) as compared to the incidence rate among whites which was 38.0 with a 95% confidence interval of between 35.9 and 40.1 (95% CI: 35.9–40.1, p = 0.01), Latinos with an average of 31.7 and a 95% confidence interval of between 29.4 and 34.1 (95% CI: 29.4–34.1, p < 0.0001) and Asians with an average of 30 and a 95% confidence of between 27.2 and 33.2 0 (95% CI: 27.2–33.2, p < 0.0001).
The rate of mortality declined significantly in the period between 2003 and 2016 from 21.0 to 13.9 per 100,000 of the New York City population. The average percentage change declined significantly (t=-2.92, p < 0.0001 and sample size = 19,241). The mortality rate due to colorectal cancer in 2016 was 17.9 per 100,000 with a 95% confidence interval of between 16.1 and 19.7 deaths (95% CI: 16.1–19.7) and significantly above the mortality rate among whites which was 15.2 deaths with a 95% confidence interval of between 13.9 and 16.4 (95% CI: 13.9–16.4, p = 0.01), Latinos with an average of 10.4 and a 95% confidence interval of between 9.0 and 11.8 (95% CI: 9.0–11.8, p < 0.0001) and Asians with an average of 8.8 and a 95% confidence of between 7.1 and 10.4 (95% CI: 7.1–10.4, p < 0.0001).The declines in annual percentage mortality rate were not significant in all the involved boroughs with the t statistic for Bronx= − 3.1, Brooklyn = − 3.3, Manhattan= − 3.4, Queens= − 2.5 and Staten Island = − 2.5. The 2016 colorectal cancer deaths in Staten Island were 14.5 per 100,000 with a 95% confidence interval of between 11.6 and 18.0 (95% CI: 11.6–18.0), Bronx had a mortality rate of 14.0 per 100,000 and 95% confidence interval of between 12.0 and 15.9 (95% CI: 12.0–15.9), and Brooklyn had a mortality rate of 13.3 per 100,000 and a 95% confidence interval of between 11.9 and 14.6 (95% CI: 11.9–14.6). The colorectal mortality rate was significantly higher in Staten Island and the Bronx relative to those in Queens and Manhattan. The mortality rate in Queens was 11.5 per 100,000 with a 95% confidence interval of between 10.2 and 12.7 (95% CI: 10.2–12.7), and Manhattan had a mortality rate of 11.5 per 100,000 with a 95% confidence interval of between 10.0 and 13.0 (95% CI: 10.0–13.0). Colorectal cancer mortality also varied across the neighborhoods in New York City. The
Williams et al. (2016) mapped the New York City regions with higher rates of colorectal incidence and mortality. The colorectal cancer incidence rates were higher in regions with a high proportion of the black population, although the mortality rate was not related to the proportions of other racial groups, including Whites, Latinos, and Asians. The research established that higher colorectal death rates were associated with the African American race with a p-value < =.0001). Also, there were higher instances of lack of fruits and vegetables in the diet among blacks (p < =.0001), a higher proportion of smokers (p = 0.0026), lack of health insurance (p < =.0001), and lack of primary care provider (p = .0254).
The level of colonoscopy screening increased from 41.7% in 2003 to 68.5% in 2016. According to the 2016 data, an estimated 1.6 million New York residents aged 50 or above had a colonoscopy. The annual percentage change was higher in the period between 2003 and 2008 at 7.55 as compared to 0.56 in the period between 2008 and 2016. The research evaluated the screening rates across the various ethnic/racial groups in 2016. African Americans recorded the highest screening rate of 72.2%. The screening rate among blacks was significantly higher than that of Asians 60.9% at p-value 0.0045 and Latinos, 71.1%, and p-value 0.0092. The screening rate among blacks was, however, insignificantly different from the screening rate among whites, 67.2% at p-value 0.0885. The average percentage increase in screening rates was highest among blacks in the period between 2003 and 2007. The average percentage increase from 2003 to 2007 among blacks was 28.43 relative to Latinos’ 11.55, Whites’ 6.36, and Asian’s 2.92.
Williams et al. (2016) also established insignificant variations in the rate of colonoscopy among persons aged 50 and above across the boroughs in the city. The rate of colonoscopy among eligible persons in 2016 was Staten Island 62.4%, Manhattan 71.9%, Queens 67.4%, Bronx 70.8%, and Brooklyn 67.2%. Stool-based colorectal screening is also a popular method of testing, although not popular in New York City. In 2003 increased stool-based screening accounted for a significant proportion of the total number of tests. Including the number of stool-based screenings increased the total percentage of annual tests from 11.9% to 53.6%. The use of stool-based screening, however, declined significantly between 2003 and 2012 and only accounted for 1.6% of the total screenings out of the total 70.1% in 2012. Therefore, the stool-based testing decline in New York City declines as the level of colonoscopy testing increased.
Williams et al. (2016) established that the incidence and mortality of colorectal cancer decline among persons living in New York City for the period between 2000 and 2016. The number of persons with local stage colorectal cancer increased between 2002 and 2007, owing to the increase in colonoscopy testing and the decline in stool-based testing. The increase in the local stage confirms the effectiveness of colonoscopy in identifying the early stages of colorectal cancer. Increased colonoscopy screening among persons aged 50 and above contributed to a higher decline in the incidence and mortality rates in New York City relative to the national average. In 2016, the incidence declined by 2.8%, while mortality dropped by 2.9% compared to 2.4% and 2.2% national reduction rates, respectively. Owing to the diverse population of New York City, the analysis indicated no significant differences in the change of annual percentage deaths due to colorectal cancer between white and black communities. New York City’s population consists of 32.1% White, 24.3% Black, 29.1% Latino, and 14% Asian (Williams et al., 2016). The results were contrary to national data findings between 1975 and 2012, which indicate a significant decline in colorectal cancer incidence among Whites than Blacks. According to the national data, colorectal cancer incidence declined by 1.4% among the white population relative to 0.5% among the black community. Studies attribute the disparities to availability screening. Reduced screening rates mean that Blacks are more likely to be tested with advanced stages of colorectal cancer, which limits the treatment options, reduces survival, and causes higher mortality. Other possible causes of higher colorectal cancer rates include genetic predisposition, health risk behavior such as smoking, access to preventive health care services, and timely treatment.
Although the decline in incidence did not vary significantly between white and black communities living the New York City, the study showed significant disparities in colorectal incidence and mortality. The incidence rate among blacks was 11% higher, while the mortality rate was 18% above the mortality rate among whites. In 2016, colorectal cancer rates in New York City were similar to the national rates. The national incidence rate was 37.7 per 100,000 as compared to the New York City rate of 37.3 per 100,000. However, the national data for the period between 2007 and 2014 showed significant disparities, with the blacks having a 32% risk above that of whites. Data from the US CONCORD-2 study also showed a significant colorectal burden among blacks with a survival rate lower than whites. The mortality rates also varied across neighborhoods. Higher death rates were recorded among neighborhoods with higher proportions of black populations. Black neighborhoods were also characterized by higher rates of high-risk factors such as smoking and lack of fruits and vegetables in the diet.
New York City screening campaign named Citywide Colon Cancer Control Coalition helped to significantly reduce the incidence and mortality rates. The campaign caused a 64% increase in the number of persons reporting a timely diagnosis of colorectal cancer in the 50 and above age bracket. In 2016, the New York department of health and mental hygiene reported a 68.5% timely diagnosis rate of colorectal cancer among persons aged 50 years and above. Since the department of health is primarily involved in tracking the timely diagnosis through colonoscopy, the 68.5% timely screening rate may be lower than the actual owing to the availability of other colorectal cancer screening methods such as stool-based testing. The campaign focused on promoting colonoscopy apart from instances when the involved were unwilling to undertake colonoscopy. The timely colorectal cancer screening was not significantly different across the different boroughs in New York City between 2012 and 2016.
There are multiple sources of data on colorectal cancer screening, prevalence, and mortality from the health departments at various levels of government. Health departments have the role of progressively tracing the prevalence, mortality rate, and possible disparities in respective populations. Although the primary data is available from healthcare facilities, data involving patient details is protected under the patient privacy data act (HHS, 2013). Since the research focused on the status of colorectal cancer in New York City, the article utilized the data from the New York health department. The sample data covers a large proportion of the population and is therefore a proper representation of colorectal status in New York City. Apart from New York State Cancer Registry and New York City Vital Statistics data, the article also utilized data from the New York City Community Health Survey (CHS), which consists of the social demographic information of the participants. Like the data from the health department, the community health service data covers a significant proportion of the population, meaning that the data is an appropriate representation of the New York City population. Williams et al. (2016) adhered to the ethical standards by seeking approval from the NYC Health Department’s IRB where necessary.
Quality of data involves numerous factors such as accuracy, validity, missing values, errors, and appropriateness of the data. Since the utilized data was obtained from health facilities or broad surveys, the data may be considered a true representation of the status of colorectal cancer in New York City. Any missing value and errors are eliminated in the data cleansing stage, for example, through deleting or estimation. Appropriate data should be representative of the population and therefore must be normally distributed. The normality of data is tested by calculating the descriptive statistics and graphical representation. The common descriptive statistics include measures of central tendency and measures of variation.
The mean, mode, and median measure the central location of data. When data is normally distributed, the three measures of central tendency are similar (Zikmund et al., 2013). The mean is the total of all data values divided by sample size, while the median is the most central location when data is organized in the order of size. The mode estimates the center of data as the most frequent data entry and is often inapplicable based on the nature of a data set. When the mean is greater than the median, it means that the data is skewed to the right. On the contrary, a median > the mean suggests that the data is skewed to the left. When the data is skewed, the median is a more accurate measure of the center because it is less affected by the outliers (Camm et al., 2018). The most common measures of spread include variance, standard deviation, and the interquartile range. The interquartile range measures the spread of data from the median by calculating the bulk of data that lies between the 25 th and 75 th quartiles (Camm et al., 2018). A low interquartile range suggests that a significant proportion of data is arranged around the median. Like the interquartile range, the standard deviation measures the deviation of data from the mean. A low standard deviation means that the data is closely arranged in the center. The interquartile range is a better estimate of the center for skewed as it is not influenced by the outliers.
The spread of data between the minimum and maximum values in data is estimated through the range. The range is the difference between the least and highest values in data that describes the appropriateness of descriptive statistics in explaining the data. A very high range may indicate the presence of outliers. Visuals such as histograms and box plots are commonly used to describe the shape of the data. A histogram summarizes the arrangement of data using bars. The bar graphs of normally distributed data assume a bell-shaped curve, while skewed data shows the presence of a tail either to the left or right of the data (Zikmund et al., 2013). The box-plot shows a visual description of the five-number summary, consisting of the minimum, maximum, median, and the first and third quartiles (Camm et al., 2018). The arrangement of the five-number summary shows whether the data is evenly distributed or not. Outliers are also shown in the box plot.
Prior to statistical analysis, a researcher must eliminate any outliers. Outliers refer to the data points that lie unreasonably far from the other data points. A data point is considered as an outlier if it lies more than 1.5 interquartile range below the first quartile or above the third quartile. The first quartile less 1.5* interquartile range is referred to as the lower fence, while the third quartile plus 1.5 * interquartile range is the upper fence. Alternatively, outliers may be identified using the z-score method. Z-score estimates the number of standard deviations that a data entry is below or above the mean. The z-score method assumes that data assumes a bell-shaped curve around the mean. Z-scores are calculated using the formula:
Z-score = ((x-x ̅))/s (Zikmund et al., 2013).
Where x ̅ is the sample mean and s is the sample standard deviation
A z-score of zero implies that a data point is equal to the mean. Similarly, a z-score above 1 means that a data point is greater than the mean, while less than one means that a data point is less than the mean. Normal data points lie between a z-score of -3 and + 3 (Camm et al., 2018). All data points lying below -3 or above +3 are considered too extreme and hence outliers.
Although the results of the analysis indicate that regression analysis was appropriate, a prior analysis should be conducted to test the appropriateness of the model. Regression analysis assumes linearity and that the error terms are identically, independently, and normally distributed. The researcher ought to have tested the appropriateness of regression analysis by estimating the linearity of data to avoid breach of the assumptions, inaccurate results, and interpretations (Theobald et al., 2019). Linearity of data is estimated by calculating correlation coefficients or plotting a scatter graph. A correlation of between 0 and 1 means that variables are positively correlated, while that of between 0 and -1 means that data variables are negatively correlated. A problem arises when multicollinearity exists. This is a strong correlation between the independent variables. Multicollinearity occurs if the correlation between two independent variables is very strong, mainly above 0.8. Such a strong correlation suggests that the independent variables may be estimating the same phenomenon, which means that the researcher should consider using one of the independent variables in such cases. The effect of multicollinearity in a model is estimated using tolerance value and the VIF (Theobald et al., 2019). A tolerance value that is less than 10 and a VIF of near 1 means that multicollinearity has an insignificant effect on a regression model.
Williams et al. (2016) conducted research into significantly important topics of colorectal cancer and health disparities across racial/ethnic groups. According to the (CDC, 2021), colorectal cancer is the third most prevalent type of cancer among males and females and is currently ranked third in terms of death-causing cancers. Despite numerous efforts, colorectal cancer incidence continues to increase among persons aged between 50 to 75 years. In 2018, the number of screenings increased by 1.4% to 68.8%, accounting for a 3.5 million increase in the number of new colorectal screenings. To contribute to new research information, Williams et al. (2016) expanded the research to establish the significant declines in the incidence and mortality rate of colorectal cancer in New York City, across boroughs in the city, and across racial/ethnic groups. The findings of the article provide a basis for further research into the topic and propose new ways of improving the quality of health outcomes.
The research concluded that survival chances increase with the increase in colonoscopy screening. The incidence and mortality of colorectal cancer declined among persons living in New York City for the period between 2000 and 2016 owing to the Citywide Colon Cancer Control Coalition campaign that increased awareness and provide free screening services. Colonoscopy proved effective in identifying the early stages of cancer. According to Williams et al. (2016), the number of persons with local stage colorectal cancer increased between 2002 and 2007, owing to the increase in colonoscopy testing and decline in stool-based testing. Increased colonoscopy screening among persons aged 50 and above may be used to decrease the incidence and mortality as established that the colonoscopy screening campaign helped to decrease New York City’s below the national average.
Williams et al. (2016) also mapped the New York City regions to indicate areas with higher rates of colorectal incidence and mortality. The findings showed that the mortality is higher in areas with higher proportions of African Americans. The incidence and mortality rates of colorectal cancer also proved to be more significant among blacks as compared to other racial or ethnic groups. The findings may be used as a basis for future study and to reduce health disparities across racial/ethnic groups. Increase colonoscopy screening among black populations may, for example, be used as a way of early identification, which in turn allows for better patient outcomes. The research also identified that instances of lack of fruits and vegetables in the diet among blacks, a higher proportion of smokers, lack of health insurance, and lack of primary care are significant among black communities. Public education on health risk behavior, providing health insurance among vulnerable groups, and increased access to healthcare may help reduce the health disparities.
The research is well structured, with a clear introduction that highlights the objectives and goals of the research. The literature review is, however, shallow and does not focus on comparing results in related topics. The article may be improved by discussing findings in similar research and their relation to the current study. The methodology is clearly defined, and the sources of data are well explained. Quality secondary data was used as it covers the wide population proportion in New York City. Other than the secondary data, the survey data also capture a wide population hence the validity and accuracy of generalization. Lack of incidence and mortality rate data that contains social demographic information such as racial/ethnic affiliation and age created deficiencies requiring the need to pair data from different sources. The study may be improved by conducting a random online sample that contains questions about incidence, mortality, and social demographic questions to reduce the chances of error during the pairing of data. The findings and discussion sections of the paper are sufficiently detailed and answer a wide range of fundamental questions. The discussion part also relates the findings of the article to national data results to show the similarity.
In conclusion, Williams et al. (2016) is a high-quality piece of work that requires minimal improvements. The literature review part may, however, be improved by including findings from a wide range of related topics. The researcher should have also tested the data for normality to ensure they eliminate outliers to avoid the risk of skewed results. Notably, the research’s results are consistent with national data results hence supporting the reliability of the research findings. The size of data used in the study was large and wide-ranging to support the generalizability of and applicability of the research findings.
References
Camm, J. D., Cochran, J. J., Fry, M. J., Ohlmann, J. W., and Anderson, D. R. (2018). Essentials of business analytics . Cengage Learning.
CDC. (2021, June 8). Colorectal cancer statistics . Centers for Disease Control and Prevention. https://www.cdc.gov/cancer/colorectal/statistics/
HHS. (2013, July 26). Summary of the HIPAA Privacy Rule . HHS.gov. https://www.hhs.gov/hipaa/for-professionals/privacy/laws-regulations/index.html
Theobald, E. J., Aikens, M., Eddy, S., & Jordt, H. (2019). Beyond linear regression: A reference for analyzing common data types in discipline-based education research. Physical Review Physics Education Research , 15(2), 020110.
Williams, R., White, P., Nieto, J., Vieira, D., Francois, F., & Hamilton, F. (2016). Colorectal Cancer in African Americans: An Update: Prepared by the Committee on Minority Affairs and Cultural Diversity, American College of Gastroenterology. Clinical and translational gastroenterology , 7(7), e185. https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-021-11330-6
Zikmund, W. G., Carr, J. C., & Griffin, M. (2013). Business Research Methods . Cengage Learning
