Understanding logistic regression analysis in clinical reports: an introduction Richard P Anderson, MD, Ruyun Jin, MD, Gary L Grunkemeier, PhD The Annals of Thoracic Surgery Volume 75, Issue 3, Pages 753-757 (March 2003) DOI: 10.1016/S0003-4975(02)04683-0
Fig 1 The percentage of papers appearing in three thoracic surgical journals, according to the year of publication. Numbers were obtained by searching the CTSNet Web site using the text string “logistic regression.” The Annals of Thoracic Surgery 2003 75, 753-757DOI: (10.1016/S0003-4975(02)04683-0)
Fig 2 Logistic regression model from the example in the text. The smooth curves are plots of the probability of operative mortality for patients with (solid line) and without (dashed line) renal failure, according to age. The numbered circles are the observed mortality by age group (4 = < 45 years, 5 = 45 to 55, 6 = 55 to 65, 7 = 65 to 75, 8 = > 75) for the patients with (black symbols) and without (white symbols) renal failure. The predicted mortalities for the 2 patients (A and B) discussed in the text are shown by the boxes. The Annals of Thoracic Surgery 2003 75, 753-757DOI: (10.1016/S0003-4975(02)04683-0)