Do not get sick and go to the hospital on the weekend!
Part I: Consider the following article.
Hospital death rates higher on weekends
BY JIM RITTER, HEALTH REPORTER You have a greater chance of dying in the hospital if you are admitted on the weekend, when fewer doctors and nurses are working researchers have found. In 23 of the 100 leading causes of death, mortality rates for patients admitted on weekends were higher than rates for patients admitted during the week, University of Toronto researchers report. There were no conditions in which death rates were lower on the weekends. Researchers defined weekends as the period from midnight Friday to midnight Sunday. Fewer staffers work then and those who do tend to have less seniority and experience. Moreover there are fewer supervisors on weekends and they are often in charge of staffers they don’t know well. Dr. Bell and Dr. Redelmeier of The University of Toronto report in today’s New England Journal of Medicine.
Researchers analyzed 3.8 million admissions from emergency rooms in Ontario, Canada. They found, for example, that 42 percent of patients admitted on weekends for ruptured abdominal aortic aneurysms died compared with 35 percent of patients admitted during the week.
Police and fire departments are fully staffed on weekends, and hospitals should try to follow the pattern of these essential services. Hospitals could entice doctors and nurses by paying more for weekend duty. Hospitals could save money because it would be more efficient to fully use expensive equipment seven days a week, Redelmeier said.
Data Set 1: Ruptured abdominal aortic aneurysms*
Weekend = 1,559 cases, 655 deaths
Weekday = 3,895 cases, 1,403 deaths.
* An aneurysm is when a blood vessel becomes abnormally large or balloons outward. The abdominal aorta is a
large blood vessel that supplies blood to your abdomen, the pelvis, and legs.
Test the claim that the proportion of deaths (mortality rates) is greater on the weekend then weekday for this disease,” at the 0.05 level of significance.
Part II:
Is there a need for Public Health Care?
Point: Emergency rooms by law have to treat all patients. It does not matter if they do or do not have medical insurance. Thus there is no need for public health care. The uninsured can get medical treatment.
Counterpoint: There is a difference in the mortality rates for patients who do and do not have medical insurance. The percentage of people dying is greater if they do not have medical insurance than the percentage of people who do have medical insurance. Thus there is a need for public health insurance.
To answer this question and resolve this debate consider the following article about medical insurance.
No Insurance? That’s a killer!
What insurance (and the lack of it) often represents, as numerous studies have shown, is the difference between care and no care, between an early cancer diagnosis and a late diagnosis, between properly managing a chronic condition like asthma and waiting until a dangerous attack occurs. For some of the patients in the Archives of Surgery study, which was led by Johns Hopkins trauma surgeon Adil Haider, what insurance represented was nothing less than the difference between life and death.
Drawing on the National Trauma Data Bank, which collects information from approximately 700 U.S. trauma centers and hospital emergency departments, Haider and his colleagues analyzed almost 430,000 moderate to severe cases of traumatic injury (from auto accidents, gunshots and other causes) treated between 2001 and 2005. Controlling for age, gender, type and severity of injury, they found that, among white patients, the mortality rate for those with insurance was 4.2 percent, compared with 7.9 percent for the uninsured. The findings by Haider and his colleagues erase any illusion that emergency care is the great equalizer in our health-care system that our differences get left behind when we are rolled through those double doors, injured and in danger of dying.
Based on a study that appeared in the October issue of Archives of Surgery.
Data Set 2: White Patients
Insured Patients: n = 162,984 Insured Mortality Rate: 4.2%
Uninsured Patients: n = 99,894 Uninsured Mortality Rate: 7.9%
Data Set 3: African-American Patients
Insured Patients: n = 22,397 Insured Mortality Rate: 4.9%
Uninsured Patients: n = 49,852 Uninsured Mortality Rate: 11.4%
Data Set 4: Hispanic Patients
Insured Patients: n = 15,873 Insured Mortality Rate: 6.3%
Uninsured Patients: n = 25,897 Uninsured Mortality Rate: 11.3%
For each data set test the claim that the proportion of deaths (mortality rates) is greater for uninsured patient than the insured patient,” at a 0.05 level of significance.
Based on the results of your hypothesis test, is there a need for public health care?
Data for Part I:
Make weekend your sample 1 and weekday your sample 2
Part I claim: The proportion of deaths (mortality rates) is greater on the weekend then weekday for this disease,” at the 0.05 level of significance. a.) State the null and alternative hypotheses (as a mathematical statements) and indicate where the claim is (in the null or alternative).
Circle or highlight the correct sign, indicate where the claim is.
b.) Perform the test by finding the p-value (use your calculator, 2-PropZTest). Use data set 1.
c.) Make the appropriate decision to reject or fail to reject, Indicate why you made this decision.
d.) State the conclusion about the claim for this hypothesis test.
Data for Part II: Based on a study that appeared in an issue of Archives of Surgery.
Data Set 2: White Patients
Insured Patients: n = 162,984 Insured Mortality Rate: 4.2%
Uninsured Patients: n = 99,894 Uninsured Mortality Rate: 7.9%
Data Set 3: African-American Patients
Insured Patients: n = 22,397 Insured Mortality Rate: 4.9%
Uninsured Patients: n = 49,852 Uninsured Mortality Rate: 11.4%
Data Set 4: Hispanic Patients
Insured Patients: n = 15,873 Insured Mortality Rate: 6.3%
Uninsured Patients: n = 25,897 Uninsured Mortality Rate: 11.3%
You will need to calculate X in each case (for each race), for example for white patients x=(.042)(162984), round this to the nearest whole number
Make uninsured sample 1 and insured sample 2.
Notice in here they decided to separate by race, because there could be some hereditary predisposition to certain illness among different races, for example some races have a higher rate or heart decease, it makes sense to control for that, this way the only difference between the two groups you are comparing is insurance (whether they are insured or not, here insurance was isolated).
You can do the test first for the 3 races, make the decision to reject or fail to reject, then to make it easier you can just write one conclusion in the end for the three races. Just mention what the evidence indicates about the claim for White patients, African-American patients and Hispanic patients.
Part II claim: The proportion of deaths (mortality rates) is greater for uninsured patient than the insured patient, at a 0.05 level of significance.
Data Set 2 (white patients)
a.) State the null and alternative hypotheses (as mathematical statements) and indicate where the claim is (in the null or alternative).
Circle or highlight the correct sign, indicate where the claim is.
b.) Perform the test by finding the p-value (use your calculator, 2-PropZTest).
c.) Make the appropriate decision to reject or fail to reject, Indicate why you made this decision.
Data Set 3 (African-American patients) a.) State the null and alternative hypotheses (as mathematical statements) and indicate where the claim is (in the null or alternative) .
Circle or highlight the correct sign, indicate where the claim is.
b.) Perform the test by finding the p-value (use your calculator, 2-PropZTest).
c.) Make the appropriate decision to reject or fail to reject, Indicate why you made this decision.
Data Set 4 (Hispanic patients) a.) State the null and alternative hypotheses (as mathematical statements) and indicate where the claim is (in the null or alternative) .
Circle or highlight the correct sign, indicate where the claim is.
b.) Perform the test by finding the p-value (use your calculator, 2-PropZTest).
c.) Make the appropriate decision to reject or fail to reject, Indicate why you made this decision.
Conclusion for Part II: State the conclusion about the claim of this hypothesis test. You can summarize the conclusion about the claim mentioning white patients, African-American patients and Hispanic patients.