A method of sample size determination to estimate the fatigue life at desired probability of survival, confidence level, and maximum acceptable error is presented.
This is a more sensible use for the AD, as it would be up to the end user to select the maximum acceptable error rate level.
The maximum acceptable error rate was 0.5% for deficiencies in first name, last name and gender.
The maximum acceptable error rate was 5% for deficiencies in father's name, demographic information and cancer encoding.
We choose an error of five minutes as the maximum acceptable error in order to precisely identify the start of mechanical ventilation.
However, with regard to father's name and cancer encoding, the performance of the registry failed to meet our standard for maximum acceptable error rate.
From 2006 to 2009 the difference between the real and maximum acceptable error rate decreased; however, by the end of the study period in 2009 the error rate (7.87%) remained above the 5% threshold.
The percentage of complete data registration (patient's name, age, gender, address, phone number and father's name) and correct cancer encoding was calculated for each year and compared to the maximum acceptable error rate for each item.
For 2007, we adjusted the a priori estimated seroprevalence of BTV-8 infected locations with goats to 30% with a maximum acceptable error in the estimated prevalence of 10% and a 95% confidence level required for the estimated prevalence, resulting in a sample size of approximately 81 locations with goats for 2007.
For the locations with sheep, we used an a priori estimated prevalence of BTV8-infected locations with sheep of 10% and a maximum acceptable error in the estimated prevalence of 5%, resulting in a calculated sample size of approximately 140 locations with sheep for 2006.
