#5

Overview:

-Location: Hospital wide MDACC

-Reason change is needed: Several studies have shown that early and adequate antimicrobial therapy improves the outcome of patients with multidrug-resistant (MDR) gram negative rod (GNR) bloodstream infections (BSI). Moreover, currently there are limited antimicrobial options for treatment, and there are no novel antimicrobial agents under development in the pharmaceutical industry against GNR infections.

-Groups involved: All healthcare providers

-Alignment to organizational goals: Safety and Effectiveness

-Our CS&E team decided to continue this project, with the objective to show the significant impact and the sustainability of this new QI process.

Aim Statement:

-To improve the outcome of patients with MDR-GNR BSI by starting adequate, prompt and early antimicrobial therapy. This will be achieved by:

  1. Increasing the number of Infectious Diseases consultations, recommending the best possible active antimicrobial therapy (based on in-vitro susceptibility data)
  2. Reducing the timeframe from the identification of each new episode of MDR-GNR BSI at MDACC microbiology laboratory to the initiation of active antimicrobial therapy

Measures of Success:

-Increase in active antimicrobial therapy against each specific strain of MDR-GNR causing BSI

-Reduction in the time between the final susceptibility to the start of active antimicrobial therapy

-Reduction in the infection-related ICU admissions

-Reduction in the infection-related mortality (associated to MDR-GNR BSI) and overall mortality

Use of Quality Tools:

-Data collection sheets, fishbone diagrams, control charts, process flow maps and the lean process concept. Two tools, a control chart and 3process maps (flowcharts), are shown below.

-Figure 1 (control chart) showed a significant increase in the prevalence of MDR-GNR BSI at MDACC, comparing 2012 with 2009 data (based on binomial probability distribution). Even though we do not have complete data for 2013, there is an even greater rise in the frequency of this type of infection during the first months of this year.

Figure 1: Prevalence of MDR-GNR BSI at MDACC from January 2009 to July 15, 2013

Figure 2: Pre-CS&E Process (Process Flow Map A)

Interventions:

-In an attempt to make the process more “lean”, we developed the following steps to reduce the timeframe between the identification of each new episode of MDR-GNR BSI in the microbiology laboratory to the initiation of active antimicrobial therapy (based on in-vitro antimicrobial susceptibility data).

-Starting on July 16, 2012, after approval from the Quality Improvement Assessment Board, the following new process was developed:

  1. All GNR BSI results reported by the microbiology laboratory were checked daily by an Infection Preventionist [IP] (Mrs. Polly Williams).
  2. The IP notified ID team members by email when a patient with a MDR-GNR BSI was identified
  3. An ID physician/team member notified the primary team of the MDR-GNR BSI by email (Dr. Adachi or on behalf of Dr. Adachi).
  4. The primary team made the decision whether to place an online ID consult request
  5. A system was developed to prospectively collect data from all patients with MDR-GNR BSI

Figure 3: Current CS&E Process (Process Flow Map B)

-This QI process was executed every day by our team, including weekends and holidays.

-This QI process was started as a CS&E project (Session 17). However, our team decided to continue, taking it as an additional and extra-curricular responsibility for a full year, from July 16, 2012 to July 15, 2013. The objective was to show the significant impact of this intervention and the sustainabilityof this new QI process.

Results:

Baseline Data pre-CS&E Process (January 2010 to July 15, 2012)

-Of 193 patients, Infectious Diseases was consulted in 117 cases (60.6%), resulting in prompt and better active antimicrobial therapy in those patients with multidrug-resistant gram negative rod bloodstream infections.

-However, when comparing patients with an ID consult to those without an ID consult, there were no significant differences in infection-related ICU admissions, infection-related mortality or overall mortality.

Table 1:Impact of the QI (CS&E) Process in the Outcome of Patients with MDR-GNR BSI at MDACC (July 16, 2012 to July 15, 2013)

*p value, comparing pre-CS&E data with post-CS&E process,
and using Chi-square or Fisher Exact Test, when indicated.

Novel and SustainedCS&E Process (July 16, 2012 to July 15, 2013)

-A total of 135 cases of MDR-GNR BSI have been identified during this period of 1 year, representing a significant increase compared to the previous baseline data (over 2.5 years)

-The significant rise in patients followed with an ID consult (from 60.6% to 76.3%) was followed by the concomitant initiation of early (from 11.00h to only 7.53h from the report of in-vitro susceptibility to the start of active therapy) and adequate active therapy (from 95.7% to 98.3%).

-These improvements in the process have had a significant impact in the outcome of cancer patients with MDR-GNR BSI, with:

1. Reduction in the infection-related ICU admissions (from 16.2% to 3.9%, p=0.018)

2. Reduction in infection-related mortality (from 15.4% to 4.9%, p=0.041) and in overall mortality (37.6% to 10.7%, p=0.009)

-Even though this new process had a significant impact, there was no significant reduction in the length of stay (LOS) of those cancer patients, mainly because of the severity and complexity of our cancer patients and because of the need for adequate therapy with at least 14 days of IV antimicrobial therapy.

-The most important outcome of this QI process was the significant reduction in infection-related mortality, helping to “save” at least 14 cancer patients with MDR-GNR BSI at MDACC.

Revenue Enhancement /Cost Avoidance / Generalizability:

-The reduction in infection-related ICU admissions from 16.2% to only 3.9% represents at least 16 patients who did not require transfer to ICU during our current process, making more ICU beds available for other patients.

-From this process we also identified that the average LOS of cancer patients with MDR-GNR BSI was 6.5 days, both inside and outside the ICU.

-There was a significant cost avoidance, considering only the cost of ICU bed per day ($19,295, average data cost per day for FY11 at MDACC). This represents a total cost avoidance of at least $2,006,680.00, with average cost avoidance per patient of $125,417.50.

Conclusions and Next Steps:

-The lean process had a significant impact in our cancer patients with MDR-GNR BSI, with more prompt and adequate active therapy through Infectious Diseases consultation, decreased number of transfers to ICU, and more importantly, decreased infection-related mortality and overall mortality among those patients.

-The main barrier to successful implementation of this new process is the lack of an electronic, automatic notification process.

Next Steps:

Figure 4:Ideal New Process (Process Flow Map C)

Our team is currently in active interaction with our institutional leadershipfor thedevelopment and implementation of an automatic electronic notification process for the management and treatment of all cancer patients infected with MDR organisms, including the possibility of an automatic infectious diseases consultation.