iSixSigma

Decreasing LOS

Six Sigma – iSixSigma Forums Old Forums Healthcare Decreasing LOS

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #25491

    GeJay BB
    Participant

    Hello All
    My first post so bear with me! I am working on a BB project to decrease LOS of acute resp failure pts. I have come up with numerous inputs that can potentially increase LOS. What I have found from an intial pull of data is that I am looking at 2 populations of people COPD and Pneumonia (based on admit dx). I am struggling on types of analysis test I can run to begin to start pin pointing significant inputs. Any suggestions?

    0
    #61495

    BritW
    Participant

    Could run a t-test to see if the data you are looking at betweenthe two different modalities are really different.  If they are not, then you might be able to treat the process issues the same for both.  My guess is you will find differences because there are issues around both treatments that affect LOS.

    0
    #61496

    Bob Sutter
    Participant

    There are two generic factors that impact a patient’s length of stay: severity of illness and processes. In order to discern which of the factors has a statistically signficant impact on LOS a multivariate regression analysis would be helfpul. Severity of illness variables include: age, comorbid conditions, lab values, etc. Process variables inlcude: medications, procedures, time to discharge, etc. Hope this helps.
    Bob 

    0
    #61497

    BritW
    Participant

    Policies also play a role.

    0
    #61498

    GeJay BB
    Participant

    Thanks Brit I will do a t test on the two populations that is a good starting point to validating if the populations really are different.

    0
    #61499

    GeJay BB
    Participant

    Great suggestion Bob I never thought about it that way. I am pulling information out fo the chart that is telling me about how aggressive we our with medications and treatments (ABX, Diuretic, Resp Tx, ABG, PT OT, MRI, CT, Cardiac Test) How aggressive we are with consultations (pulmonary, cardiology). I have data on co-morbidities, age. To do a multi variate analysis the Y would be LOS. I can capture all these other variable as X’s (inputs). Do you think it would be wise to severity of Illness and Processes seperate or together or does it even make a difference- GJ

    0
    #61503

    tparadise MBB
    Member

    There are a lot of multiple factors that influence the care for both patient populations.  There are a lot of ways to do a factor analysis.Which software are you using?  Excel is very limited on its statistics capability. Some of the anlaysis would include: Logistic Regression, Principal Component Analysis, Multivariate Study….
    I have been working with our CNS Group on these projects at our hospital and would be glad to share the information with you.  We are currently the benchmark level for COPD and Pneumonia care.
    Please contact me at:
    [email protected]
    Thanks,
    Terrill Paradise
    CSSBB,CQE,CQT,CMI
    Columbus Regional Hospital

    0
    #61518

    Bob Sutter
    Participant

    I think combining severity and process together will provide a more complete picture. With severity alone the analysis will not reveal the impact of processes and with processes alone the patient’s severity of illness is not considered. However, if you have the time conducting all three analysis could be insightful as well.
    Bob

    0
    #61519

    GeJay BB
    Participant

    Thanks Bob
    Statistics is not a strong point fo rme. Do you have any suggestion for how I can go about running the anlaysis to determine how processes and severity affect LOS? What type of test.

    0
    #61520

    Bob Sutter
    Participant

    There are various types of linear regression analysis that can be performed. The key to using them correctly is to assess the regression diagnostics after the analysis is run in order to ensure that the assumptions of the analysis are met. If the assumptions are not met then the output from the analysis can be questionable. If you are not that familiar with statistics then you may want to refer to a statistics textbook or use the services of a statistician. Hope this helps.
    Bob

    0
Viewing 10 posts - 1 through 10 (of 10 total)

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