Question

Pine et al. [1983] followed patients with intra-abdominal sepsis(blood poisoning) severe enough to warrant surgery to determine the incidence of organ failure or death(from sepsis). Those outcomes were correlated with age and pre-existing conditions such as alcoholism and malnutrition. There were 106 patients that have been followed. For each patient, data were collected on death as well as five potential predictor variables: shock, malnutrition, alcoholism, age, and bowel infarction.

• Specify a multiple logistic regression model that can be used to test no association of malnutrition with death after controlling for age and shock. Identify and interpret the regression coefficient of interest.

Answer 1

• Those results were connected with age and previous condition, for example, liquor addiction and lack of healthy sustenance.
• The opportunity to death of end of study is named as time ( in days)
• Survival status is shown by paired variable x1( 0= alive and 1=death)
• Five potential indicator factors shock,malnutrition ,alcoholism,age and gut infraction were named X2,X3,X4,X5,X6 separately.
• The four factors X2,X3,X4,X6 were twofold variables,coded 1 if the manifestation was available and coded 0 if missing.
• The variable X5=age in years, was acquired as a consistent variable.