The NEW ENGLAND JOURNAL Of MEDICINE 553m 225 Underwent randomization tto fo ad other illnesses (go 113 Analyzed 112 Analyzed Figure 1.Enrollment and Outcome. groups with the use of Student's ttest for con prostate volume as compared with those with a low prostate volume (dichotomized at 40 ml), ical va aldata obtained from men who did not com plete the study were included in the final analy- The funding organizations (the National Insti- ses.All reported I values are two-sided and have tute of Diab and Digestive and Kidney Dis not been adjuste esand the nter for Co on the basis of baseline data:an examination of palmetto had no role in the design or conduct changes in the primary outcome measures among of the study,the collection,management,analysis. men wit tion,review,and approval manuscrip 560 N ENGLJ MED 354:6 WWW.NEJM.ORG FEBRUARY 9.200 Downloaded from neim. s without permissionThe new england journal o f medicine 560 n engl j med 354;6 www.nejm.org february 9, 2006 groups with the use of Student’s t-test for con￾tinuous variables and chi-square tests for categor￾ical variables. All analyses were conducted ac￾cording to the intention-to-treat principle, so that all data obtained from men who did not com￾plete the study were included in the final analy￾ses. All reported P values are two-sided and have not been adjusted for multiple testing.31 Three subgroup analyses were planned a priori on the basis of baseline data: an examination of changes in the primary outcome measures among men with moderate symptoms as compared with men with severe symptoms, men with a high prostate volume as compared with those with a low prostate volume (dichotomized at 40 ml),32 and men with high PSA levels as compared with men with low levels (dichotomized at 1.4 ng per deciliter).32 The funding organizations (the National Insti￾tute of Diabetes and Digestive and Kidney Dis￾eases and the National Center for Complementary and Alternative Medicine) and the supplier of saw palmetto had no role in the design or conduct of the study, the collection, management, analysis, and interpretation of the data, or the prepara￾tion, review, and approval of the manuscript. 225 Underwent randomization 775 Men were assessed for eligibility 550 Were excluded 152 Declined to participate 106 Had an AUASI score <8 101 Were taking excluded medications 94 Had maximal urinary flow rates >15 or <4 29 Were taking saw palmetto 21 Had abnormal laboratory values 12 Had undergone prostate surgery 6 Had a history of prostate cancer 27 Met other exclusion criteria 2 Had a low adherence rate during placebo run-in period Recruitment source Letter from study, 553 men Poster, 110 men Newspaper ad, 42 men Friend, 26 men Other source, 44 men 113 Assigned to and received placebo 4 Lost to follow-up 1 Moved 1 Had urinary retention 1 Opted out of study 1 Could not be contacted 5 Discontinued intervention 2 Started an alpha-blocker 1 Started saw palmetto 2 Had other illnesses (testi￾cular pain and colon cancer) 113 Analyzed 112 Assigned to and received saw palmetto 5 Lost to follow-up 3 Moved 2 Opted out of study 5 Discontinued intervention 3 Had other illnesses (gout, rash, adrenomyeloneu￾ropathy) 1 Started an alpha-blocker 1 Purchased and started over- the-counter saw palmetto 112 Analyzed Figure 1. Enrollment and Outcome. The New England Journal of Medicine Downloaded from nejm.org on October 18, 2011. For personal use only. No other uses without permission. Copyright © 2006 Massachusetts Medical Society. All rights reserved