I want to conduct a parallel trial study in which 3 care bundles will be implemented aimed to decrease neonatal sepsis.How is sample size calculated?
Data to be collected from patients will be clinical signs of sepsis and blood samples for culture to determine the presence of infection. The outcome variable will be incidence of neonatal sepsis among participants.The study aims to evaluate the impact of the intervention on incidence of sepsis. What will be the best method to analyze this data
Hi
For sample size calculation you will need what is the current incidence of sepsis and what amount of changes you are expecting from the intervention. You can go for both randomized and non-randomized design. In the sample size calculation, you have decide how much absolute precision (1-alfa) and power (1-beta) you want. Usually, absolute precision of 5% and power of 80% is acceptable lower limit. If your trial is non-randomized, it is better to put a design effect. You can use online free statistical site www.openepi.com for this.
For analysis, it depends on the data. but from your objective, it appears that you will require unpaired t-test, Fisher's exact/Chi-square test. You can also calculate the sensitivity and specificity if the intervention is diagnostic and calculate the area under the curve for all interventions. For this you have to have one gold standard (these tests will be required if intervention is diagnostic only). At present statistical software also prompts what tests are appropriate provided you enter the data type correctly.
Hope this helps you
The STATA or Epi info software are very important to calculate a sample size for such study.
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Pharmaceuticals
- Drugs