Summarizing benchmarks in the national database of nursing quality indicators using bootstrap confidence intervals

Area of Science:

• Nursing Quality Improvement
• Biostatistics
• Healthcare Management

Background:

• Summarizing nursing quality indicators often involves confidence intervals around the mean.
• Out-of-range confidence interval bounds can occur due to small sample sizes or non-normal data distributions.
• Existing methods like data transformation or interval truncation have limitations.

Purpose of the Study:

• To address the challenge of inaccurate confidence intervals in nursing quality indicators.
• To evaluate the percentile bootstrap method as a viable alternative for calculating these intervals.

Main Methods:

• The study discusses the application of the percentile bootstrap method.
• This statistical technique is compared to traditional methods like data transformation and interval truncation.
• Focus is on achieving realistic confidence interval bounds without normality assumptions.

Main Results:

• The percentile bootstrap method provides realistic confidence interval bounds for nursing quality indicators.
• It does not rely on the assumption of data normality, overcoming a key limitation of other methods.
• This approach offers a convenient way to obtain appropriate confidence intervals.

Conclusions:

• The percentile bootstrap is a valuable tool for calculating confidence intervals for nursing quality indicators.
• It effectively handles issues arising from small sample sizes and non-normal data.
• This method enhances the reliability of nursing quality benchmark summaries.