Problem Set 2 Answers

docx

School

Saint Louis University *

*We aren’t endorsed by this school

Course

5000

Subject

Statistics

Date

Apr 3, 2024

Type

docx

Pages

Uploaded by SuperHarePerson908

Problem Set 2 I) Descriptive Epidemiology and Standardization Use CDC Wonder Compressed Mortality file, 1999-2016 to answer the following questions https://wonder.cdc.gov/mortSQL.html 1) What is the 2014 crude (unadjusted) all-cause mortality rate for a) Florida and b) California? Succinctly describe the comparison between the two crude rates. a) 934.8 deaths per 100,000 population in 2014 b) 633.8 deaths per 100,000 population in 2014 c) The crude, all-cause mortality rate in Florida was substantially higher than in California in the year 2014. 2) Examine the 2014 age-specific, all-cause mortality rates for both Florida and California? Use the chart function in CDC Wonder to generate a bar graph to help you examine the pattern of rates. Succinctly describe the pattern of mortality rates across age groups in Florida and California. When examining the pattern of mortality rates across age groups, rates start out relatively high in the <1 year age group, then are very low until they start to gradually increase after age 20, increasing at a higher in the 40’s and then at a much higher rate after age 65. The pattern is similar in Florida and California. Could also say the pattern is bimodal, with peaks at very young and very old ages.

3) Now calculate the 2014 all-cause mortality age-adjusted rates, using the 2000 US Std. Population as your standard population distribution, for a) Florida and b) California. a) 662.0 deaths per 100,000 population in 2014, adjusted for 2000 US Std. Pop. b) 605.7 deaths per 100,000 population in 2014, adjusted for 2000 US Std. Pop. 4) Summarize your findings with a succinct description of the crude, age-specific and age-adjusted results. What explains the difference in the comparison between Florida and California with respect to crude vs. age-adjusted rates (in other words, why are the state-specific crude and age- adjusted rates different)? Crude, all-cause mortality rates were substantially higher in Florida than in California in 2014 (935 vs 634 per 100,000 population). After adjusting for age using the 2000 US Std Population, rates were slightly higher in Florida than California (662 vs 606 per 100,000 population), and the pattern of age-specific rates was similar. The difference in age distributions of the two states, with a relatively larger proportion of elderly adults in Florida vs. California, would explain why the age-adjusted rates in Florida were lower than the crude rates, and thus more similar to California. 5) Are the age-adjusted mortality rates in this exercise an example of direct or indirect standardization? Why? Direct standardization; age-adjusted rates for both states were standardized to the same reference population (CDC WONDER default is to standardize using the 2000 US standard population) 6) Generate a table listing the top 15 causes of death for the most recent year of data available in the CDC WONDER data file "Underlying Cause of Death, 2018-2021", located at https://wonder.cdc.gov/Deaths-by-Underlying-Cause.htmlLinks to an external site. (note: you will group results by "15 Leading Causes of Death" and select "2021" for the year. 6a) List the top 3 causes of death in the US for 2021 with their corresponding mortality rates (note: be sure to include units with your rates).  Diseases of heart: 209.6 deaths per 100,000  Malignant neoplasms: 182.4 deaths per 100,000  COVID-19: 125.6 deaths per 100,000 6b) How do these causes of death compare to the top three causes of death prior to what has been described as the epidemiological transition during the 20th century? With the exception of COVID-19, we can see a shift to chronic diseases as the driving causes of death in the US population compared to infectious causes prior to the epidemiological transition during the 20 th century

6c) What type of mortality rate is the measure you described in question 1? Crude mortality rate 6d) Now generate a table listing the top 15 causes of death in Missouri in 2021. How do the mortality rates for the top 3 causes of death compare to those for the US overall? Missouri top 3 causes are the same as the US overall, but the rates differ.  Diseases of heart: 254.7 deaths per 100,000 (> US overall)  Malignant neoplasms: 213.2 deaths per 100,000 (> US overall)  COVID-19: 125.8 deaths per 100,000 (slightly > US overall) 6e) If you wanted to more formally compare the top causes of death in Missouri with another state, what additional measure would you want to include in your descriptive table? We would want to compare the age adjusted rates to account for differences in the age distribution of the US overall versus Missouri. II) Epidemiological Measures 1) Name three measures of disease frequency and write their formulas. Prevalence: Point Prevalence: # of individuals with the condition/# of individuals considered Period Prevalence: # that experienced the condition during a time interval/ Total # of individual considered during the time interval Incidence Proportion: # of new cases during a time period/ Total # of population at risk during that time period Incidence Density (Rate): # of new cases during a time period/ Total person-time of observation Mortality rate: The # of deaths/ Total population during certain period(person-year) 2) Describe two different ways to estimate person-time 1. Counting the person-time contributed by each person in the population during the study period. 2. multiplying the size of the population at the mid-point of the study period times the duration of the study period.

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

3. Actuarial adjustment, multiply number of cases by half of the study follow up period, added to the number of non-cases multiplied by full study follow-up period 3) A cohort study starts with 10,000 women. Of these, 500 had experienced breast cancer sometime in their past. The remaining 9,500 are followed for five-years. Two-hundred and fifty (250) breast cancer cases occurred during the follow-up period. Assume there is no loss to follow-up and no competing risk in this cohort. a. What is the five-year risk (a.k.a., cumulative incidence, or, incidence proportion) of breast cancer? Report the risk per 1,000 people. Incidence Risk: (250*1000)/9500 = 26.32 per 1000 people for 5 years. b. What is the rate (a.k.a., incidence density) of breast cancer in this cohort? Report the rate per 1,000 person-years. Incidence Rate: (250*1000)/ (9250*5+250*2.5) = 5.33 person-year. NOTE: this calculation used actuarial adjustment to estimate person-time in the denominator. c. Show how risk~=rate*time using the results from a. and b. of this problem. Risk (26.32) ~= Rate (5.33) * Year (5) III) Case study and Multiple choices 1) Which measure is most appropriate for estimating the frequency of disease or health-related outcome in each scenario: prevalence, cumulative incidence or incidence density (consider the numerator and denominator!)? a. The frequency of 16-year-old driver’s license applicants in 2007 who are found to have visual acuity impairment severe enough to require corrective lenses while driving. Prevalence at age 16 b. The frequency of developing carpal tunnel syndrome among all grocery clerks who are employed by a large chain of supermarkets during 2007-2009. Incidence density (or disease rate) NOTE: this is incidence density because if the time of developing carpal tunnel syndrome is known the at-risk time, or person-time, for this cohort of employees could also be calculated. c. The frequency of newborns dying within seven days after birth among those born at a tertiary hospital during January 2008. Cumulative incidence (or disease risk) (if you assumed you would have person- time at risk for this group you may have answered Incidence density)

2) Here are some fictitious data describing the frequency of a specific disease among high school students in a particular school district: Annual incidence Prevalence Year per 100,000 per 100,000 1995 24.5 41.8 1996 24.9 41.2 1997 23.8 40.9 1998 24.6 40.1 1999 24.1 38.4 2000 24.7 37.9 2001 24.2 35.3 2002 23.9 33.2 2003 25.1 29.8 2004 24.5 27.2 Which of the following statements (can choose more than one) correctly explain these data? a. Aggressive treatment over time has resulted in earlier and more frequent cures. b. Adoption of a new treatment, that diminishes the severity of the symptoms, suppresses the immune responses and thereby prolongs the clinical course of the disease. c. Efforts to prevent new cases of the disease have been successful. d. A shift toward the occurrence of more aggressive disease has resulted in earlier and more frequent deaths among afflicted students. Inspection of the data shows that incidence density (ID) was stable, while prevalence (P) was declining. Hence, the average duration of disease (D) must also have been declining according to the relationship P = ID x D. Explanations (a) and (d) are compatible with this assertion. Explanation (b) is not, since it implies an increase in duration of disease, not a decrease. Explanation (c) suggests that incidence should have been dropping, which it was not. 3) The figure below depicts a fictitious sample of a cohort being used to investigate the incidence of neurological disorders at a local retirement home. The data below is for Alzheimer’s disease diagnoses in 2008.

a. What is the total person time contributed in the cohort (denominator)? (Hint: Include in your count the month where event or loss to follow up occurs) (5+9+7+12+2+5+12+11+4+12) 79 person-months, 6.5833 p-y (if draw a line at beginning of July and count 7 person-years, that is fine) b. How many incident cases of Alzheimer’s were identified? 4 c. What is the incidence density of Alzheimer’s disease for this sample cohort in 2008? 4/6.5833 = 0.61 cases / person-year x 100 (or, 4/79=.051 cases/person-months) = 61 cases / 100 person-years 4) The major disadvantage of crude rates is that: A. they may not allow for comparison of populations that differ in size. B. they do not permit comparison of populations that vary in composition. C. they are difficult to calculate from available data sources.

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

D. all of the above 5) To express the burden or extent of some condition or attribute in a population. A. This is a use primarily for incidence data. B. This is a use primarily for prevalence data. C. This application could apply equally for both incidence and prevalence data. D. This is a use for neither incidence data nor prevalence data. 6) To provide a direct estimate of the risk of developing a disease. A. This is a use primarily for incidence data. B. This is a use primarily for prevalence data. C. This application could apply equally for both incidence and prevalence data. D. This is a use for neither incidence data nor prevalence data.

Problem Set 2 Answers

Related Documents