Randomization case study

Problem 1 : Side effects of Avandia

(a)Since more patients on pioglitazone had cardiovascular problems (5386 vs. 2593) we can conclude that the rate of cardiovascular problems for those on a pigolitazone treatment is higher.

# Cardiovascular problems
#                              Yes           No           total
#Treatment  Rosiglitazone     2593          65000         67593
#           Pioglitazone      5386          154592        159978
#           Total             7979          219592        227571


#Rosiglitazone
2593/67593
## [1] 0.03836196
#Pioglitazone 
5386/159978
## [1] 0.03366713
0.03836196-0.03366713
## [1] 0.00469483

Since the total number of data samples for those who take Rosiglitazone or Pioglitazone have different total, the first clause of the sentence is wrong assumption, thereby sentence (a) is incorrect.

Difference in the rate of cardiovascular problem of Rosiglitazone and Pioglitazone is 0.00469483.

  1. True

  2. True

  3. Each data sample for different kinds of treatment is independent, so the rates of incidence is not due to a relationship between the two variables; therefore the statement (b) is incorrect.

Problem 2 : Heart transplant

# Heart transplant
#                         Group 
#                   Control    Treatment       total
#Outcome    Alive         4        24           28
#           Dead          30       45           75  
#           Total         34      69           103

  1. 30 and 45, respectively.

  2. #Randomization technique

  1. H0; No significant difference in group whether they alive or dead group. H1; There is a significant difference in group.

  2. 28, 75, 34, 69, 0, lower

  3. There is no significant difference in group.

Problem 3 : Sinusitis and antibiotics

# Sinusitis and antibiotics
#                               Self reported improvememnt in symptons 
#                              Yes           No           total
#Treatment  ANTIBIOTIC         66           19            85
#           PLACEBO             65          16             81 
#           TOTAL                131          35            166
  1. Null hypothesis
  2. There is little failure in each group so it does no make use of data sample.
19/85
## [1] 0.2235294
16/81
## [1] 0.1975309
0.2235294-0.1975309
## [1] 0.0259985
0.0259985
## [1] 0.0259985
  1. Both of the treatments have 80% of effectiveness; from this percentage it shold be okay to say that it is effective.

  2. There is no significant difference in Antibiotic and Placebo. There is a significant difference in treatment.

  3. We can conclude that given p-value, we can reject that there is a significant difference of effectiveness between antibiotic versus placebo.