iSixSigma

Bloodstream infections

Six Sigma – iSixSigma Forums Old Forums Healthcare Bloodstream infections

Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • #24887

    Chris Dillinger
    Participant

    I am working on a project designed to reduce central catheter related bloodstream infections. I am struggling with financial validation and tracking of potential savings for this project. Various published articles suggest the financial value to be between $25K and $50K per infection. However, I am unable to validate that to be “real” dollars that hit the the bottom line. Does anyone have suggestions or supporting documentation to help? Thank you.

    0
    #59717

    Big Dave
    Participant

    Chris,
    I am a black belt who is also a clincial biologist.  I worked with Quest Laboratories and we also faced some of the same issues regarding savings on prevention.  The CDC has a great deal of information on mortality rates related to septicemia (blood infection).  Once you have the mortality numbers then you can check with the American Hospital Association to find out what the average cost of treatment is per day for septicemia.  This data should enable you to substantiate your projected savings on the project.  Let me know how it goes.  [email protected]

    0
    #59721

    [email protected]
    Participant

    If someone gets an infection what exactly occurs?  Does LOS increase, does it result in a readmission, does it result in more bloodwork, is it treated with meds, and etc.  Everything I listed above has expen$e written all over it.  Your job is to understand what happens when an infection occurs and quickly associate dollars to it.

    0
    #59718

    sai
    Member

    Dave, Are six sigma applications quite common in the area of clinical biology, any references or articles that i can look up.
    thanks,
    Sai

    0
    #59722

    Michael Schlueter
    Participant

    Hi Chris,
    I thought about using Taguchi’s quality loss function for your purpose. Here is an example, how you can use it.
    I assume the patients are already in hospital (so in contrast to home treatment this costs quite something). What you need is a reliable response Y, which is easy for you to measure and which reflects (or measures) infection correctly.
    Now the best way would probably be to measure bacteria level or whatever causes the infection. Temperature measurements from the patient may be sufficient, too, in case you don’t have those .
    Let’s assume that half of the patients can bear +2deg above their normal temperature, while the other 50 % can’t. Let’s assume treating a patient at that temperature (36.5+2) costs $1000/day on average (just my guess) in hospital: nursing, doctors, medical treatment, examinations etc.
    Now we can estimate the loss function:

    L = k * (T – Tnormal)^2
    => L = $1000/day = k * (2deg)^2
    => k = 250 $/day/deg^2
    So the average loss by different infection levels can be estimated as:

    T            (T-Tnormal)         L (average loss)
    36.5              0                     0 $/day
    37                 0.5                 63 $/day
    38                 1.5               563 $/day
    39                 2.5             1563 $/day
    40                 3.5             3063 $/day
    41                 4.5             5063 $/day
    You can quantify your initial and your improved situation using these estimates. Assuming average patients temperature was 40deg due to infection +-1deg in the old way of catheter treatment, then the initial quality loss is approximately:
        L1 = 250 $/day/deg^2 * (3.5^2 + 1^2) = 3312.5 $/day
    When your improvement changes this to 39+-0.5deg, loss drops down to:
       L2 = 250 $/day/deg^2 * (2.5^2 + 0.5^2) = 1625 $/day
    Which would be an average savings per patient of approximately
               L1 – L2 = 1687 $/day
    (I may be too conservative in my estimate; because when patients die at an average temperature of 45deg, loss would be just 18063 $/day, which does not reflect very well the value of a life. But anyway.)
    I’d like to add: you can (and should) add fixed cost to the loss function. When a patient is almost healthy, but in hospital, he/she will cost something anyway.
    The non-linear relationship takes into account that losses quickly accumulates, once deviation from the ideal target value increases (the square function is the least trivial approximation). E.g. 38deg is not very bothering. At 39 deg the hospital may take extra measures. At 40 deg care increases. At higher temperatures intense doctor-involvement may be required, such as surgery, intensive care or such. Not to forget about the extra loss imparted to relatives.
    Hope this helps you any further,
    Michael Schlueter

    0
    #59723

    Alan Villiers
    Participant

    I would suggest that you take a look at the literature via a MedLine search cross referencing central lines, infection rates, costs, and quality as partial search terms. If you work in a healthcare facility your medical librarian would run the search for you. I suspect that you might be overwhelmed with relevant citations.  The CDC as previously suggested is also good.  Also, you might want to do a Google search on Benneyan’s work with g-charts that are sensitive to low frequency events like these.
    Alan Villiers
     

    0
    #59724

    Alyce Carlson
    Participant

    I recently facilitated a PI Team related to Central Line Infections.  Some data that came our way was from the Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project, HCUPnet, 2000.  One quote was “Nationally, the average total charge for treating patients hospitalized with septicemia (except in childbirth) increased from $16,691 in 1993 to $24, 365 in 2000.”  A team member was able to correlate that with estimated costs for this medical center.

    0
    #59720

    Chris Dillinger
    Participant

    Thank you for your input. We are going to try calculating average cost of treatment per day. It is amazing at times how data poor we are in such a data rich environment as healthcare.

    0
    #59719

    Chris Dillinger
    Participant

    Sai,
    Six Sigma applications are becoming more common in clinical biology areas and healthcare. I work for Mount Carmel Health. We have deployed Six Sima as our formal toolbox to permanently solve the business processes that challenge our healthcare system. The opportunitinies are vast. BMG is the company that helped with our training and deployment. They may be able to provide applications and references for you as well.

    0
    #59725

    David Ross Scott
    Participant

    Chris,
    I work in the energy supply business – but face the same challenges as you do when it comes to identifying and validating benefits.
     
    The question I always ask is what is the cost of this defect to the business? If e.g. we issue incorrect energy bills, the cost to our business can be measured by uneccessary phone calls and customer complaints – these 2 activities can easily be measured by knowing the cost to handle a call / respond to a complaint.
     
    Having no medical experience at all (apart from being on the receiving end), I would imagine that the cost of a defect in your process is the cost of “putting right” the bloodstream infection e.g. the additional length of time the patient has to stay in hospital, the cost of the medical staff to deal with the infection and the cost of the drugs needed to treat the infection – these are the things which will hit your bottom line.
     
    I hope that the above helps you.
     

    0
    #60489

    Lois Yingling
    Participant

    I am also working on a SS project to reduce CVC related BSI. A finance team member ran a report using the confirmed CVC BSI patient account numbers by DRG, variable cost and LOS for those patients. A second report for the control group was run for the same time frame by the DRGs identified for for the CVC BSI accounts. The control group variable cost/case & LOS/case was used as the benchmark. The difference between the benchmark of non BSI patients with the same DRGs and the variable cost for the CVC BSI patients was used as potential savings ($16,699/case). LOS opportunity was calculated the same way (20.67 days/case).

    0
    #62402

    hassan
    Participant

    Would you send me the procedure from A to Z to reduce blood stream infection in health care
    Thanks in advance
     

    0
Viewing 12 posts - 1 through 12 (of 12 total)

The forum ‘Healthcare’ is closed to new topics and replies.