[Skip to content]

NHS
Loading
.

Matching Michigan

This project is finished. View the published journal paper.

Ninety seven per cent of acute trusts in England are participating in Matching Michigan. This is a quality improvement project based on a model developed in the United States which, over 18 months, saved around 1,500 patient lives.  It took place at Intensive Care Units (ICUs) in Michigan and introduced data definitions, technical interventions - changes in clinical practice - and non-technical interventions - linked to leadership, teamwork and culture change.  When these interventions are applied together these have been shown to reduce Central Venous Catheter Bloodstream Infections (CVC-BSIs).


Find out Why we're working to match Michigan from this interview with Professor Peter J. Pronovost of Johns Hopkins University School of Medicine, USA.

 

More information, including intervention resources and case studies, is available on www.patientsafetyfirst.nhs.uk

 

Data reporting

Reporting Infections in Critical Care (RICC) is the online infection data reporting tool for Matching Michigan. This is the first step in building a standardised reporting system to benchmark rates of infection. Monthly data needs to be reported by 15th of each month. 

 

 

 

Definitions

This is a quality improvement project starting with data collection to identify and measure infection rates. In many cases, it's the first time this type of data has been collected. There are two types of data that need to be collected:  CVC data and Central Venous Catheter bloodstream infections (CVC-BSI) data.
 
A Central Venous Catheter (CVC) is defined as:

 

  • An intravascular device terminating in one of the great veins or pulmonary artery, including those in, or near, the right atrium, and those inserted via a femoral vein.

  • Includes PICCs, haemodialysis catheters and parenteral nutrition catheters.

 

CVC data:  Data is collected daily.  Step-by-step guidance, see related documents for collecting denominator data that is the number of patients in the unit, number of CVC-patient days and total CVC-days and this sequence is recommended.
       
CVC BSI data:  Data definitions have been established for Adult patients and Paediatric patients. 

  

Guidance

In some cases, more clarity around CVC-BSI definitions is needed to meet the complexities of patients' conditions in the ICU environment. Here is further guidance for specific situations, this supports CVC-BSI definitions for Adult and Paediatric data reporting.

 

  • A second positive blood culture with same organism?

    • If less than, or equal to 7 days = same infection episode
    • If more than 7 days = new (second) infection episode

  • Two organisms at same time = 1 infection
  • Two organisms at different times and less than 7 days apart = use clinical judgement (intensive care & microbiology)
  • No peripheral venous access, only a CVC: interpret as for CABSI
  • ICU to ICU transfers: apply 48 hour rule. If BSI is positive less than 48 hours after transfer and ICU admission = pre-ICU infection
  • If missing data is found in CVC line day data boxes, take the mean between the two data entry points.

 

Please contact the project team if you need further information.

 

More information, including intervention resources and case studies, is available on www.patientsafetyfirst.nhs.uk

 

Register to participate in Matching Michigan

 

Please use this online registration form if you are interested in participating in Matching Michigan. 

 

More information, including intervention resources and case studies, is available at www.patientsafetyfirst.nhs.uk

Frequently Asked Questions

  1. When is the best time to collect the CVC data?  

  2. In a 10 bedded unit, with 100% patient utilisation of CVCs and 100% occupancy, what would be the total number of patients with one or more CVCs? 

  3. In a 10 bedded unit with 100% patient utilisation of two CVCs in every patient every day and 100% occupancy, what would be the total number of CVCs? 

  4. Do we need to record the different types of central venous catheters that we use, on a monthly data collection form? E.g. antibiotic coated, silver coated and the number of lumens on the line?   

  5. Does a patient with a CRBSI need a new line on the data collection form for each day that the CRBSI continues?

  6. Do we have to collect all data as stipulated on the data collection form? 

  7. How much scope is there for clinical judgement across CVC-BSI definitions?  

  8. Do PICCs count as CVCs? 

  9. Do we record data from patients up to 48 hours following ICU discharge? 

  10. When is the best time to collect the infection data? 

  11. How long is the interval between two or more blood cultures drawn on separate occasions?

  12. If a blood culture indicates a BSI, do we count the BSI as starting on the day the culture returned positive, or on the day the culture was taken? 

  13. Should we count the PA catheter and the sheath as two CVCs?

  14. What should we do if we are unable to collect CVC data for every day of the month, or at the end of the month we find we are missing CVC data for a few days? 

 

  1. When is the best time to collect the CVC data? 
    Ensure you collect CVC-patient days and CVC-days data at the same time in every 24 hour period, this should be agreed at an individual unit level.
  2. In a 10 bedded unit, with 100% patient utilisation of CVCs and 100% occupancy, what would be the total number of patients with 1 or more CVCs?
    This would range from 280-310 per month depending on how many days in the month; thus, a 31 day month would be 310.
  3. In a 10 bedded unit with 100% patient utilisation of 2 CVCs in every patient every day and 100% occupancy, what would be the total number of CVCs? 
    This would range from 560-620 per month depending on how many days in the month; a 31 day month would be 620.
  4. Do we need to record the different types of central venous catheters that we use, on a monthly data collection form? E.g. antibiotic coated, silver coated and the number of lumens on the line?
    No, this information is not required on the monthly data collection form. However, at a local level you may wish to collect this information for your own interest. 
  5. Does a patient with a CRBSI need a new line on the data collection form for each day that the CRBSI continues? 
    No. Once it is identified that a CVC is the source of a BSI, the CVC-BSI is only counted once.  The exception is if the same organism is grown from a blood culture sample taken more than seven days after the first positive blood culture. This will count as a second, assuming that there is, or has been, another CVC in place within the 48 hour window of the sample being taken.
  6. Do we have to collect all data as stipulated on the data collection form?
    Yes, all data should be collected. Without robust data, rates of infection cannot be determined reliably. As a consequence, ICUs will not be able to benchmark their data against the national dataset.
  7. How much scope is there for clinical judgement across CVC-BSI definitions?  
    Taking the 3 definitions:
    CRBSI is free of clinical judgement,
    CABSI requires that you exercise clinical judgement.
    CSBSI are all clinical judgement.
  8. Do PICCs count as CVCs? 
    Yes. Any catheter the tip of which ends in a central vessel near or in the heart, including those inserted via the femoral route and including CVVH catheters, are defined as CVCs
  9. Do we record data from patients up to 48 hours following ICU discharge? 
    Yes. Infections occurring up to 48 hours after ICU discharge count as ICU-acquired infections. This is ‘counterbalanced’ by infections appearing in the first 48 hours after ICU admission being defined as pre-ICU infections.
  10. When is the best time to collect the infection data? 
    The best time to collect infection data is when the infections are identified. Often this can be on the ward round with the ICU consultant and microbiologist, but the time may vary from one day to the next. CVC-patient days and CVC-days however, should be recorded at the same time every 24 hours.
  11. How long is the interval between two or more blood cultures drawn on separate occasions?  
    There is no specific interval; it could be a few minutes apart. Use your local policy.
  12. If a blood culture indicates a BSI, do we count the BSI as starting on the day the culture returned positive, or on the day the culture was taken? 
    The day the blood culture was taken.
  13. Should we count the PA catheter and the sheath as two CVCs?
    There are two issues to consider in response to this question.  Firstly, you can count the PA catheter placed through a sheath as one single CVC. The rationale for this is that if a PA catheter were infected, the sheath would also be infected and both would have to be removed.  They would also share the same insertion site. An alternative instance is if the PA catheter were removed but the sheath was retained in situ, the sheath would still count as a CVC and continue to be counted in the denominator. The removal of the PA catheter would not alter the number of CVC patient days.
  14. What should we do if we are unable to collect CVC data for every day of the month, or at the end of the month we find we are missing CVC data for a few days? 
    If missing data is found in CVC line day data boxes, take the mean between the two data entry points. For example:  If one day is missing and the total CVC patients the day before and the day after was 10, then enter 10 for that day.  If two days are missing and the day prior has total CVC patients of 12,and the day after the missing days 14, then enter 13 for the two missing days. 

 

 

TitleIssue dateType
Matching Michigan | Data collection tools14 December 2009Guidance