作者: L. M. Katz
DOI: 10.1001/JAMA.1994.03520080030024
关键词: Randomization 、 Medicine 、 Surgery 、 Confidence interval 、 Exact test 、 Blinding 、 Internal medicine 、 Bacteriuria 、 CRANBERRY JUICE 、 Odds ratio 、 Pyuria
摘要: To the Editor. —While it is attractive in these high-tech times to embrace low-tech approaches disease prevention and control, Dr Avorn colleagues 1 have not made their case for cranberry juice. From data Table of article (page 753), apparent that randomization and/or blinding scheme failed two groups were comparable with regard previous UTI (Table). Analysis history prior 6-month 12-month periods demonstrates a significant excess placebo group during both ( P =.004, odds ratio [OR]=4.39, 95% confidence interval 1.55 12.43, using approximation Woolf, =.03, OR=2.50, 1.15 5.42, respectively [Fisher's exact test]), suggesting was more likely at baseline experience or bacteriuria/pyuria than treatment group.