Describe the demographic characteristics of the two study populations. Are there any meaningful differences?

Part I: Measures of Disease Frequency

In 2009, President Obama launched a nationwide initiative to end homelessness in the U.S. by 2020. The homeless are a vulnerable population with limited access to health care and poor health outcomes. In order to allocate sufficient federal and local resources to eliminate homelessness, U.S. cities conduct an annual survey to estimate the number of homeless persons living within major cities. The City of Boston’s Emergency Shelter Commission (ESC) conducted a survey of homelessness on the night of January 25, 2017. Volunteers counted the number of homeless persons living on the streets, in emergency shelters for individuals or families, in domestic violence programs, in residential mental health or substance abuse programs, transitional housing, and in specialized programs.

 

1. Which of the following best describes the homeless population in the City of Boston?
2. Dynamic population
3. Fixed population

 

2. Which of the following describes the homeless population that took part in the ESC survey on January 25^th?
3. Dynamic population
4. Fixed population

 

3. The 2015 Homeless Census counted 3,456 homeless persons in Boston. The 2016 homeless census counted 3,384 homeless persons in Boston. The size of the population in Boston was 665,984 in 2015 and 673,184 in 2016. From 2015-2016, did the burden of homelessness:
4. Increase
5. Decrease
6. Stay the same (2015: .52%, 2016 0.50%)
7. Cannot determine from this information

 

4. What do you consider to be the biggest limitation in the homeless survey and why?

5. Boston Health Care for the Homeless Program (BHCHP) analyzed deaths among homeless adults who were seen at their program from 1/1/2003 through 12/31/2008 (Baggett et al., 2014). A total of 1,302 deaths occurred during the observation period among the 28,033 study population. The table below gives the race and gender of the decedents and total population.

Characteristic	Decedents (N=1,302)	Total Study Population (N=28,033)
Gender Male Female	1,055 247	18,612 9,421
Race White, non-Hispanic Black, non-Hispanic Hispanic Other/Unknown	784 301 131 86	11,912 8,066 5,301 2,754

 

1. Calculate the prevalence of Hispanics in the total study population.

1. Calculate the ratio of males to females in the total study population.

1. Calculate the cumulative incidence of mortality in the total study population over the 6-year period.

1. Calculate the cumulative incidence of mortality according to gender and race over the 6-year period.

1. The authors stated that the total study population accrued 90,054 person-years of follow-up. What was the overall incidence rate of mortality per 100,000 person-years?

1. What was the average length of follow-up in the study population?

1. Fill in the number of person-years accrued by the five individuals in the table below.

 

Study Subject	First Seen at Health Care for the Homeless	Vital Status at End of Follow-up  12/31/08	Other Relevant Information	Person-Years of Follow-Up
1	1/1/2003	Alive	Never lost to follow-up
2	1/1/2005	Dead	Died on 1/1/2006
3	1/1/2006	Unknown	Lost to follow-up on 1/1/2007
4	1/1/2006	Alive	Lost to follow-up from 1/1/2007 to 12/31/2007. Showed up alive at BHCHP on 1/1/2008 and known to be alive at end of follow-up.
5	1/1/2003	Unknown	Lost to follow-up on 12/31/2006

Part II: Comparing Measures of Disease Frequency

The homeless are a vulnerable population with limited access to health care and poor health outcomes. Boston Health Care for the Homeless Program (BHCHP) analyzed deaths among homeless adults who were seen at their program from 1/1/2003 through 12/31/2008 (Baggett et al., 2014).  A total of 1,302 deaths occurred during the observation period among the study population of 28,033 individuals. Mortality rates during 2003-2008 were compared to an earlier mortality study of homeless adults seen by BHCHP during 1988-1993 (Hwang et al., 1997). In the earlier cohort, there were 606 deaths among 17,292 individuals. Below are the demographic characteristics of participants in the two studies.

 

	1988-1993 Study N= 17,292		2003-2008 Study N=28,033
	Percent
Sex Male Female	68 32	66 34
Age 18-24 25-44 45-64 >=65	9 62 25 4	13 49 35 3
Race White Black Hispanic Other	45 36 17 2	43 29 19 9

1. Describe the demographic characteristics of the two study populations. Are there any meaningful differences?

2. The two studies collected only demographic data on the study participants. Suppose that you could have interviewed participants at enrollment to gather information on their illicit drug use. Exactly what information would you obtain on this behavior, and how would you summarize the data into categories for analysis? Be sure to specify your exposure and referent groups.

 

 

 

3. Below are some crude cause-specific mortality rates from the two studies.

	Crude mortality rate per 100,000 person-years
	1988-1993	2003-2008
Drug Overdose	80.7	242.1
Substance Use Disorder	54.8	109.5
HIV Disease	280.0	84.0

 

Using both relative and absolute measures of association, compare the three cause-specific mortality rates for the 2003-2008 and 1988-1993 study populations. Use the 1988-1993 study population as the referent group. State in words your interpretation of each of these measures.

4. Do the demographic characteristics of the two study populations influence your interpretation of these crude measures of association? Why or why not?

Below are some crude mortality rates from the 2003-2008 study among 25-44 year old men and women.

	Crude mortality rate per 100,000 person-years
	25-44 Year Old Men	25-44 Year Old Women
Drug Overdose	346.9	172.6
Substance Use Disorder	90.5	43.1
HIV Disease	79.2	43.1

5. Using the relative measure of association, compare the three cause-specific mortality rates among men to that among women (i.e., use women as the referent group). State in words your interpretation of each of these measures.

 

6. What is the main advantage of limiting the male-female comparisons to 25-44 year olds?