Question

For each of the following scenarios, (a) state the study design, (b) calculate the risk of the outcome among the exposed relative to the unexposed being sure to use the appropriate measure of risk, and (c) interpret your findings in a full sentence in the context of the study.

Hint: First, identify the exposure and outcome.  Then, set up the 2x2 table and calculate the risk.

1. The incidence rate of stomach cancer in a population of middle-aged men who smoke is observed to be 5.3 per 1000 person-years.  In an equivalent population of middle-aged men who do not smoke, the incidence rate is 1.7 per 1000 person-years.
2. 150 Charlotte residents were identified via medical records as having had the flu during the 2018 winter season. When interviewed, 24 of these residents reported having had surgery during the previous year.  200 other Charlotte residents were identified who did not have the flu during the winter of 2018.  When interviewed, 24 of these individuals reported having had surgery during the previous year.
3. 1,700 bone fracture patients participated in a trial of a new topical pain-killer.  700 patients were randomized to receive the new topical pain-killer, and 329 reported good or very good pain relief during the study period.  The remaining 1000 patients received the standard oral pain-killer and 692 reported good or very good pain relief during the study period.

Answer 1

Given data:

(a) Study design:

Study design is processes wherein the trial methodology and statistical analysis are organized to ensure that the null hypothesis is either accepted or rejected and the conclusions arrived at reflect the truth.

(b) This is a retrospective study design.

· The exposure is the occurrence of flu during the 2018.

· The outcome is whether the residents have been hospitalised during the previous year.

X = 0 à charlotte residents who were identified via medical records as having had flu during winter of 2018.

X = 1 à charlotte residents not having flu in winter of 2018.

Y = 1à hospitalized during previous year

Y = 1 à Not-hospitalized during previous year.

X	Y = 0	Y = 1
X = 0	24	126	150
X = 1	24	176	200
	48	302	350

· Odds of being hospitalized with flu Ω₁ = 24/126

· Odds of being hospitalized without flu Ω₂ = 24/176

Odds ratio = Ω₁/ Ω_{2 =} 1.3968

(c) This is a prospective convert study design, where the exposure is the receipt of new topical pain-killer.

· The outcome in the pain relief sensation:

X = 0 à Received new topical pain killer

X = 1 à Received standard oral pain killer

Y = 0 à Reported good (or) very good pain relief

Y = 1 à Did not report good (or) very good pain relief

X	Y = 0	Y = 1
X = 0	323	371	700
X = 1	692	308	1000
	1021	679	1700

P[Y = 0/X = 0] = 329/700

P[Y = 0/X = 1] = 692/1000

            Relative risk = P[Y = 0/X = 0] / P[Y = 0/X = 1] =0.809