Right-sizing expectations for hand hygiene observation collection.

Area of Science:

• Healthcare Quality Improvement
• Infection Prevention and Control
• Biostatistics in Healthcare

Background:

• Effective hand hygiene monitoring is crucial for patient and staff safety.
• Current hand hygiene observation standards (100-200/month/unit) may be excessive.
• The optimal number of observations for reliable data is not well-defined.

Purpose of the Study:

• To determine a statistically comparable number of hand hygiene observations below the current standard.
• To assess the impact of reduced observation numbers on data quality and statistical significance.
• To inform revised guidelines for hand hygiene monitoring in healthcare settings.

Main Methods:

• Retrospective analysis of 873,618 hand hygiene observations from 68 facilities.
• Stratification of data by facility, unit, and month.
• Resampling of data into sets of 25, 50, 100, and 150 observations for comparison with 200, including power analysis for adherence rates (50%-90%).

Main Results:

• Comparing reduced sample sizes (25-150) to 200 observations showed minimal significant differences (2.6%-4.3%) at P=0.05.
• Median confidence interval width differences ranged from 0.05% to 0.68%.
• Power analysis indicated that required percentage differences for significance varied from 7.8% (150 vs. 200 at 90% adherence) to 30% (25 vs. 200 at 50% adherence).

Conclusions:

• Reducing hand hygiene observations to 50 per unit per month is feasible without compromising data quality, even with lower adherence rates.
• Healthcare facilities can optimize resource allocation by decreasing the frequency of direct monitoring.
• A paradigm shift towards emphasizing training, education, culture, infrastructure, and feedback is recommended over extensive monitoring.