Prevention of BSI and VAP Measuring Change in Outcomes Part II

1
Prevention of BSI and VAPMeasuring Change in
OutcomesPart II

• Ted Speroff, PhD

2
Using NNIS Rate Measures is a Problem for QI

• NNIS Rates are used in surveillance to detect
  outbreaks a rise in rates!
• Also, easier to make site comparisons
• And easier to pool data into single rate
• However, the goal of QI is to decrease the rate.
• The area you have to work with is between the
  mean rate and 0.
• It is very hard to show improvement using rate as
  your measure.

3
Control Chart for NNIS Rate
4
Solution g Chart

• Change your Measure
• The number of days between events
• Date 2 minus Date 1
• Goal to increase the number of days between
  events
• There is no upper boundary
• As the NNIS rate decreases, the number of days
  between events increases.
• G chart is sensitive for detecting a decrease in
  NNIS rate
• Dont need to know the census (denominator), just
  the dates of infections. Thus, not dependent on
  the number of line-days or vent-days of your ICU.

5
ExampleQual Saf Health Care 200514295-302

6
Additional Rules for Control ChartsStatistical
Significance

• Single point above the UCL
• 2 of 3 consecutive points between 2 and 3 sigma
• 6 consecutive points in an upward trend
• 9 consecutive points above or below the central
  line (mean)

7
Excel File TemplateTabs C-Line BSI VAP
8
ReminderChange Name of File andSave Often
9
First Entry MM/DD/YYYYDate of first BSI in your
ICU
10
Continue with Second EntryNote Days between
Events takes at least two entries
11
Completion of Second Entrystarts the Control
Chart
12
Baseline Data EntryFrom Row 20 to 42
13
Set Baseline Move Cursor to Cell F20Edit from
B45 to B42
14
Edit of Chart Title and Data EntryComplete
15
Print
16
VAP G Chart
17
VAP NNIS Worksheet
18
End of Part IIQuestions and Commentsso
far?Continue Part III