New Results for Computing Blaker’s Exact Confidence Interval for One Parameter Discrete Distributions

2015 ◽  
Vol 45 (3) ◽  
pp. 1041-1053 ◽  
Author(s):  
Bruno Lecoutre ◽  
Jacques Poitevineau
1996 ◽  
Vol 14 (9) ◽  
pp. 2546-2551 ◽  
Author(s):  
E Bajetta ◽  
A Di Leo ◽  
L Biganzoli ◽  
L Mariani ◽  
F Cappuzzo ◽  
...  

PURPOSE The aim of the study was to evaluate the activity of vinorelbine (VNLB) in a population of advanced ovarian cancer patients, with particular attention to defining its role in platinum-resistant disease. PATIENTS AND METHODS Thirty-three patients were recruited and treated with VNLB 25 mg/m2 intravenously (IV) weekly. the median age was 53 years, performance status 0 to 2, and number of previous chemotherapy regimens two (range, one to five). Twenty-four patients were platinum-resistant; the remaining nine either were platinum-sensitive (four cases) or had undetermined sensitivity (five cases). RESULTS The mean delivered dose-intensity of VNLB was 67% of the planned level, because 60% of the cycles were delayed due to neutropenia or anemia. Four partial responses (PRs) and one complete response (CR) were observed, for an overall response rate of 15% (95% exact confidence interval, 5.1% to 31.9%). All the responses occurred in the subgroup of 24 platinum-resistant cases, in whom the response rate was 21% (95% exact confidence interval, 7.1% to 42.1%). Seven patients became stabilized on VNLB, and 27% of the cases showed a reduction in serum cancer antigen 125 (CA 125) levels. G3/G4 side effects consisted of neutropenia, anemia, and worsening of preexisting peripheral neuropathy. No treatment-related deaths occurred. CONCLUSION VNLB led to a 21% response rate in the population of heavily pretreated and platinum-resistant ovarian cancer patients. Further studies of VNLB alone or in combination with taxanes are warranted in patients with less pretreatment.


2002 ◽  
Vol 27 (4) ◽  
pp. 335-340 ◽  
Author(s):  
Douglas G. Bonett

An approximate test and confidence interval for coefficient alpha are derived. The approximate test and confidence interval are then used to derive closed-form sample size formulas. The sample size formulas can be used to determine the sample size needed to test coefficient alpha with desired power or to estimate coefficient alpha with desired precision. The sample size formulas closely approximate the sample size requirements for an exact confidence interval or an exact test.


Sign in / Sign up

Export Citation Format

Share Document