Project Outcomes
Sepsis mortality rate declined by 44 percent overall:
2008 sepsis rate = 28%, reduced to 15% in April 2011
*Data displayed is for the 28 measurement months
Medication Administration
Six hospitals were included in the final analysis of the Medication Administration project. Improvement was shown in medication administration accuracy at five participating hospitals. Findings for the total group over the three time periods measured were statistically significant (p < .05), representing substantial clinical improvements.
Sepsis Project
Summary of Findings
-
Through the sepsis improvement program, sepsis cases increased by 67 percent and the sepsis mortality rate was reduced by 44 percent
-
Total of 9 hospitals completed the final measurement with 8 of the hospitals completing the full program
-
Contra Costa joined in 2010 and did not have Year 1 and Year 2 data for comparison
-
The total program investment was approximately $2.5M
-
Estimated savings is $1.1M for a 56 percent ROI
-
The 67 percent increase in cases of sepsis along with the 46 percent decrease in the mortality rate demonstrates an overall successful implementation of improved identification and treatment of sepsis
*Based on the IOM proxy cost (see assumptions)
Sepsis Bundle Compliance
The INLP cohort improved compliance with nearly every EGDT bundle element.
-
Cohort-wide sepsis screening compliance increased from 89.7% in January 2010 to 91.7% in June 2011
Between January 2010 and June 2011:
-
The proportion of patients screening positive for sepsis who received a lactic acid blood test improved from 58% to 77%
-
The percentage of severe sepsis/septic shock patients who received antibiotics improved from 48% to 67%
-
The percentage of severe sepsis/septic shock patients who received fluid resuscitation improved from and 34% to 72%
-
Cohort-wide blood culture compliance remained unchanged at 89%
-
The proportion of severe sepsis/septic shock patients receiving all bundle elements had a relative increase of 172%, from 18% to 49%
Summary of overall findings:
Medication administration accuracy (as measured by the percent of correct doses administered) improved in both the initial pilot units, as well in the spread units.
Spread units improved from:
85% (at baseline)
92% (6 months post-intervention)
96% (as measured at 18 months post-intervention)
Six safety processes score (scale of 0-6) improved from:
4.8 (baseline)
5.6 (6 months post-intervention)
5.75 (18 months post-intervention)
Initial pilot units showed continued improvement: accuracy from 94.0% to 96.8% and safety processes overall from 95.3% to 97.2%.
Medication Doses
1,841 medication doses were observed on seven units in the six hospitals with comparable data:
-
604 (baseline)
-
623 (6 months)
-
614 (18 months)
Types of Medication Errors
Over the course of the project, the prevalence of different types of medication errors changed.
-
At baseline, the most common type of error was wrong technique, which showed the largest decrease—from 41 errors (baseline) to just 5 errors (18 months). (E.g. A medication with food that should be taken on an empty stomach.)
-
Meanwhile, the percentage of wrong-time errors increased from 38% to 61% of all errors, However, the actual number of wrong-time errors decreased from 33 errors (baseline) to 13 (6 months) and then rose again to 16 errors (18 months). The fluctuation, and on-going presence of this type of error was assumed to be attributed to the challenge of administering medications during a specified one-hour window.
-
Other categories of medication errors were less frequent throughout the project.
