Using Six Sigma to Improve Cardiac Medication Administration and CAT Scan Capacity

Transcription

1 Using Six Sigma to Improve Cardiac Medication Administration and CAT Scan Capacity Harvard Quality Colloquium August 22, 2005 Susan McGann RN, BSN Adrienne Elberfeld

2 Virtua Health.Today Four hospital system in Southern New Jersey Two Long Term Care Facilities Two Home Health Agencies Two Free Standing Surgical Centers Ambulatory Care - Camden Fitness Center 8000 employees physicians 7,000 deliveries $650 million in revenues STAR Culture

3 Virtua Facilities

4 The Virtua STAR Excellent Service Resource Stewardship Outstanding Patient Experience Clinical Quality & Safety Caring Culture Best People

5 Virtua Health. The Future Change in HR Structure and Process Focus on Programs of Excellence Building a Greenfield site Potential consolidation of multiple sites Ambulatory Strategy Growth in the North Additional Strategic Partnerships

6 Define Project Title: Cardiac Medication: Indicators Six Sigma Project Sponsors: Jim Dwyer, Ann Campbell, Ellen Guarnieri, Adrienne Kirby, Mike Kotzen Champions: Pat Orchard & Jane Slaterbeck Master BB: Mark Van Kooy Black Belt: Adrienne Elberfeld Green Belt: Ted Gall Finance Approver: Gerry Lowe Project Start Date: July 22, 2002 Team Members: Jay Brewin, Darlene Euler, Christine Gerber, Val Torres, Kathy Halstead, Kathy Plumb, Cindy D Esterre, Lori Edell, Heather Scheckner, Angie Smolskis, Pat Quackenbush, Ronald Kieft, Michelle Weaks, Robert Singer, Vince Spagnuolo, Steve Fox R0 Cardiac Medication Indicators Project Description: Increase quality of patient care by use/non-use and appropriate documentation of aspirin, beta-blockers, and ACE inhibitors in CHF or AMI patients to achieve or exceed Virtua benchmark goals. Project Scope: To have all four acute care facilities, within all medical disciplines, meet the standards of Core/JCAHO guidelines Potential Benefits: To achieve improved outcomes for patients with AMI/CHF diagnosis by adhering to evidence based practice through education, documentation, and compliance while meeting regulatory standards and enhancing quality of patient care at Virtua. Alignment with Strategic Plan: IIA-Cardiology; Global MICP Goals for Virtua.

7 Measure QRA Chart Review Gage R&R Each Appraiser vs Standard Assessment Agreement Appraiser # Inspected # Matched Percent (%) 95.0% CI Appraiser A ( 9.9, 65.) Appraiser B ( 6.5, 99.8) Appraiser C ( 42.8, 94.5) Appraiser D ( 5.6, 97.9) # Matched: Appraiser's assessment across trials agrees with standard. Assessment Disagreement Appraiser # / 0 Percent (%) # 0/ Percent (%) # Mixed Percent (%) Appraiser A 0 * Appraiser B 0 * Appraiser C 0 * Appraiser D 0 * # /0: Assessments across trials = / standard = 0. # 0/: Assessments across trials = 0 / standard =. # Mixed: Assessments across trials are not identical. Between Appraisers Assessment Agreement # Inspected # Matched Percent (%) 95.0% CI ( 5.2, 72.3) # Matched: All appraisers' assessments agree with each other. Percentage of time QRA s agreed on assessment During this gage, it was determined that there was variation between the QRA s review of charts A Workout was held on September 8th with the QRA s and Case Management Directors to develop SOP s in reviewing of all CHF and AMI patients for core indicators

8 Analyze Root Cause Analysis Identified through Containment Issue Concurrent reviews of AMI & CHF patients Conclusion Between Case Management, Quality & Nursing charts needed to coordinate efforts in reviewing charts Solution Met with CCM s, Case Management & Quality to educate on core indicators Who Team members specific to campus, J. Slaterbeck, A.Elberfeld Ongoing information needed for medical staff and nursing staff of the core indicators Have team members develop a storyboard template with pathways and indicators to be available at key areas throughout the facility Identified key areas, (physician lounges, Cardiac specific units, nursing specific areas), and posted storyboards that are the same throughout the system Team members specific to campus Cardiac POE needs real time access to Clinical Care Advisor to review data Coordinate with IS accessibility to system Cardiac POE Director, AVP, and Black Belt access to system; able to review ongoing and provide feedback to Case Management C. Mullin, J. Slaterbeck, B. Rodin

9 Analyze Root Cause Analysis Identified through Containment (continued) Issue Conclusion Solution Who Who is going to perform the task of daily chart reviews concurrent with care? Communication with physicians per need for documentation Nursing, case management and quality are all reviewing charts; need to coordinate efforts in regard to the indicators Need one point person to communicate directly with physicians in a timely manner Case Management to take the lead on chart reviews for patients with AMI, CHF & related diagnosis. Support from quality & nursing If nursing and/or case mgt has direct contact with physician, they give necessary feedback. Next step is the facility QRA and physician champion Case Mtg Directors, Quality Directors, CCM s Case Mgt, QRA s, B. Singer, V. Spagnuolo, S. Fox Coordination of ongoing chart reviews, documentation completion, and data information Need to appoint point people within the facility to ensure that activities are being completed and coordinated Case Management to coordinate with nursing & quality; all paperwork forwarded to Black Belt & VP Quality Case Mgt, QRA s, C. Mullin, A. Elberfeld

10 Improve Root Cause Analysis Factor MICU run sheets not available on charts Inconsistent availability of patient census with diagnosis for Nursing and Case Management Physician compliance in completion of discharge instructions Consistent practice of multi-disciplinary care of the patient across Virtua Root Cause Medics unable to complete; shortened documentation not part of permanent chart IS integration with Canopy system; initial information input by ICD-9 code, not description Inconsistent followthrough Need for champion at each campus to lead initiatives of the Cardiac Programs of Excellence Proposed Solutions Sponsor to work with Ambulatory Quality Director to have MICU run sheets completed & submitted concurrent with care Work order placed with Information Services with actual cases to research and advise on proper input process Directive from Medical Staff leadership to complete discharge instructions; two week trial period in April, 2003 by HIM to tag all charts without discharge instructions Appointment of Nurse Leader within each facility to coordinate activities of Cardiac Programs of Excellence at local level

11 Control Realized Results of Implemented Solutions Improvement Y Benefit Quality Benefit MICU run sheets on patient charts within 24 hours of admission Increased compliance for aspirin given with 24 hours Compliance with PRO indicators for aspirin given within 24 hours of admission; DOH regulations for transfer of patient care Physician completion of written discharge instructions specific to medications for cardiac patients Standard Operating Procedures by Nursing and Case Management in chart review, stickie reminders for physicians, and availability of discharge instructions Consistent education of nursing per cardiac medication indicators Accurate daily census with diagnosis available through OAS Gold and Canopy Appointment of a Process Owner at each hospital to coordinate care with directives from Cardiac Programs of Excellence Compliance and proper documentation of care for discharge medication indicators Increased compliance in care and documentation for all indicators Increased compliance for medications given within time frames Increased compliance in care and documentation for all indicators Sustained improvement in all indicators Quality of care documented Coordination of care for the cardiac patient by the multidisciplinary team Increased knowledge base of the nursing staff of the cardiac medications for AMI and CHF patients Timeliness of care improved Sustained results maintained and reported to CMS and public; appropriate recognition and

12 Control P Chart Virtua Health Control Chart for Aspirin Within 24 Hrs 0.0 UCL= Proportion 0.05 Project Started June 03 Feb 05 Goal=95% Compliance P= LCL= Sample Number

13 Define R0 CT Scan Capacity Project Title: CT Scan Six Sigma Project Sponsors: Ellen Master BB: Adrienne Elberfeld Black Belt: Kathy Eichlin Green Belt: John Graydon, Wendy Seiler Finance Approver: Rex Rueblinger Project Start Date: July 28, 2004 Team Members: Beverly Crawford, Melody DeLaurentis, JoAnn Domingo, Audrey Fley, Darryl Fussell, Cynthia Koller, Jo Nast, Heather Pierce, Donna Rapp, Elizabeth Zadsielski Project Description: Increase capacity by reducing in and out of room times for the CT Scan to adhere to GE industry benchmarks of 5 minutes without contrast and 25 minutes of with contrast. Project Scope: Marlton CT Scan department Potential Benefits: A more efficient process will lead to increased capacity thereby increasing opportunities for increased volumes. Alignment with Strategic Plan: Resource Stewardship Patient Satisfaction

14 Measure Descriptive Statistics Descriptive Statistics Y-CT Abdomen/Pelvis Without Contrast Updated /0/04 Variable: Avg Time Y2-Abdomen/Pelvis With Contrast Descriptive Statistics Variable: Avg Time Anderson-Darling Normality Test A-Squared: P-Value: Anderson-Darling Normality Test A-Squared: P-Value: % Confidence Interval for Mu Mean StDev Variance Skew ness Kurtosis N Minimum st Quartile Median 3rd Quartile Maximum % Confidence Interval for Mu % Confidence Interval for Sigma % Confidence Interval for Mu Mean StDev Variance Skew ness Kurtosis N Minimum st Quartile Median 3rd Quartile Maximum E % Confidence Interval for Mu % Confidence Interval for Sigma % Confidence Interval for Median 95% Confidence Interval for Median % Confidence Interval for Median 95% Confidence Interval for Median Y Mean = Standard Deviation = Z Score = 2.78 Mode = 9 Percent of Defects =.% Y2 Mean = Standard Deviation = Z Score =.90 Mode = 20, 2 and 24 Percent of Defects = 34.4%

15 Measure Descriptive Statistics Y3-CT Brain Without Contrast Descriptive Statistics % Confidence Interval for Mu % Confidence Interval for Median Variable: Avg Time Anderson-Darling Normality Test A-Squared: P-Value: Mean StDev Variance Skew ness Kurtosis N Minimum st Quartile Median 3rd Quartile Maximum % Confidence Interval for Mu % Confidence Interval for Sigma % Confidence Interval for Median Y3 Mean =.367 Standard Deviation = Z Score = 2.58 Mode = 7 Percent of Defects = 3.98% The problem is too much standard deviation/ variation in the process!!

16 Analyze T Test for Equal Variances Test for Equal Variances for multiple 95% Confidence Intervals for Sigmas Factor Levels CT Tech 2 CT Techs Bartlett's Test Test Statistic: P-Value : Levene's Test Test Statistic: P-Value : Levene s test Test for equal variances for continuous data that is not normally distributed. There is a statistical difference in the variance! 3 CT Techs

17 Analyze Pareto Chart CAT Scan Delays Count Percent Defect Count Percent Cum % 0 Phone Door interruptions Assistance for pt Tech Delays Other Patient Delays IV Started in CT Rm MD Interuption Radiologist Delay Pt Rescan 3 Pt Uncooperative Equipment Failure Others A Pareto Chart shows where within the process the greatest opportunity exists for improvement. Here we see opportunities for the need for improvement with interruptions caused by the phone, door interruptions and assistance needed to move a patient resulting in 59 % of CAT Scan Delays. Use LEAN opportunities to streamline process.

18 Improve 2 Sample T Test & ANOVA Y 60 Boxplots of Before-A and After-Av (means are indicated by solid circles) Y-CAT Scan of Abdomen/Pelvis Without Contrast Y-Abdomen-Pelvis Without Contrast One-way ANOVA: Before-Avg. Time, After-Avg. Time Before-A After-Av Two-sample T for Before-Avg. Time vs After-Avg. Time Analysis of Variance Source DF SS MS F P Factor Error Total Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev Before-A ( * ) After-Av (------*------) Pooled StDev = N Mean StDev SE Mean Before-A After-Av Difference = mu Before-Avg. Time - mu After-Avg. Time Estimate for difference: % CI for difference: (0.6, 5.99) T-Test of difference = 0 (vs not =): T-Value = 2.44 P-Value = 0.07 DF = 8 P-value was less than.05, therefore, there is a statistical difference!

19 Improve 2 Sample T Test & ANOVA Y 40 Boxplots of Before-A and After-Av (means are indicated by solid circles) Y2-CAT Scan of Abdomen/Pelvis With Contrast Y2-Abdomen-Pelvis With Contrast One-way ANOVA: Before-Avg. Time, After-Avg. Time Before-A After-Av Two-sample T for Before-Avg. Time vs After-Avg. Time N Mean StDev SE Mean Before-A After-Av Analysis of Variance Source DF SS MS F P Factor Error Total Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev Before-A (------* ) After-Av ( * ) Pooled StDev = Difference = mu Before-Avg. Time - mu After-Avg. Time Estimate for difference: % CI for difference: (2.09, 8.74) T-Test of difference = 0 (vs not =): T-Value = 3.27 P-Value = DF = 49 P-value was less than.05, therefore, there is a statistical difference!

20 Improve Mood s Median/Non-Normal Data P-value was less than.05, therefore, there is a statistical difference! Mood median test for CT Scan Chi-Square = 6.76 DF = P = Individual 95.0% CIs Subscrip N<= N> Median Q3-Q After ( ) Before-A ( ) Overall median = 9.00 A 95.0% CI for median(after -) - median(before-a): (-3.2,-.00)

21 Control I & MR Control Chart I and MR Chart for Y-Avg Time Can Can we we see see the the improvement on on the the chart chart post post SOP SOP implementation? Individual Value Y-CT Scan Abdomen-Pelvis Without Contrast Subgroup UCL=29.70 Mean=2.87 LCL= Moving Range UCL=20.68 R=6.329 LCL=0 Take Take away: away: Process is is capable and and in in control.

22 Control I & MR Control Chart Can Can we we see see the the improvement on on the the chart chart post post SOP SOP implementation? I and MR Chart for Y2 Avg Time Individual Value Y2-CAT Scan of Abdomen-Pelvis With Contrast 0 Subgroup UCL=36.04 Mean=2.38 LCL=6.73 Moving Range UCL=8.00 R=5.50 LCL=0 Take Take away: away: Process is is capable and and in in control.

23 Control I & MR Control Chart Can Can we we see see the the improvement on on the the chart chart post post SOP SOP implementation? I and MR Chart for CT Scan Time Individual Value Y3-CT Brain Without Contrast UCL=20.9 Mean=0.43 LCL=0.667 Subgroup Moving Range UCL=.99 R=3.669 LCL=0 Take Take away: away: Process is is capable and and in in control.

24 The other results Ahead of the hospital curve Data driven organization The dots are connected: Strategy, Operations, Quality, Finance, People Financial up-spin Leadership Development The Results Go Well Beyond the Project!

25