Using Laboratory Automation as a Catalyst for Performance Improvement

Transcription

1 Using Laboratory Automation as a Catalyst for Performance Improvement James D. Peele, Ph.D. Director of Clinical Chemistry Baptist Health 2010 EXECUTIVE WAR COLLEGE Goals and Objectives Briefly review the evolution and benefits of automation in laboratory medicine. Understand how automation can be the catalyst for process changes. Review examples of how automation has impacted change at Baptist Medical Center.

2 Why Automate? Three reasons 1. Efficiency 2. Safety 3. Error reduction Chemistry Outpatient Test Volume and FTEs , , , , , , , ,000 50,000 OP Test Volume Total Paid FTEs

3 Medical Technologists are a dwindling resource Third largest group of health care professionals behind physicians and nurses Of 9300 positions needed annually MT schools graduate only 4800 leaving a shortfall of 4500

4 Preanalytic Analytic Postanalytic 1955 Effort / Activity Much Effort! Much Time! Time

5 Leonard Skeggs,, Ph.D. Inventor of the AutoAnalyzer Model II of the AutoAnalyzer Leonard T. Skeggs,, Jr. Persistence... And Prayer: from the Artificial Kidney to the AutoAnalyzer, Clin Chem,, 46 (2000)

6 The AutoAnalyzer platform allowed laboratories to construct their own menus and methods.

7 Preanalytic Analytic Postanalytic 1955 Effort / Activity Time Preanalytic Analytic Postanalytic 1955 Effort / Activity Much Effort! Much Time! Time

8 Preanalytic Analytic Postanalytic 1955 Effort / Activity Much less Effort! Much Less Time! Time

9 Preanalytic Analytic Postanalytic 1955 Effort / Activity 2010 Time In the past 40 years, quality has improved greatly, largely because of automation. Jan S. Krouwer,, Ph.D. Jan S. Krouwer,, Ph.D. Assay Development and Evaluation: A Manufacturer s Perspective, AACCPress 2002, p.1.

10 Preanalytic Analytic Postanalytic 1955 Effort / Activity 2010 What s Next? Time Automating a bad process only serves to speed-up problems. Ana Stankovic, MD,PhD,, MSPH Elizabeth DiLauri, MBA Ana Stankovic, Elizabeth DiLauri Quality Improvements in the Preanalytical Phase: Focus on Urine Specimen Workflow,, MLO, March 2010.

11 Human / Automation Integration There is significant preanalytical manual activity prior to samples entering automation. Automation success depends heavily on the success of preanalytical human processes. Successful automation has a integrated link with human tasks. Preanalytical Pre Automation Questions 1. How do samples arrive in the laboratory? 2. Do samples have to be relabeled? 3. How often are samples aliquoted? 4. To how many different workstations are samples routed? 5. Where are samples stored? 6. How often and how difficult is it to retrieve samples from storage for add-on testing?

12 Pre Automation Advice 1. Know your laboratory! Flow chart processes to determine areas for improvement. 2. Look for repetitive tasks and redundant processes. 3. Focus on the front end (preanalytical( preanalytical). This is the area where automation can have the biggest gains. 4. Eliminate as many inefficiencies as possible before automating (LEAN). 5. Outline goals for automation. Clinical Chemistry Laboratory Automation Baptist Medical Center Goals for Automation - Established 2007 Improve efficiency (LEAN) Consistency in quality and turn around time (6δ) Patient and technologist safety Reduction in medical errors Ability to continue expansion of outreach Ability to offset technologist shortage

13 Kim Torres OPI Project Laboratory/ED TAT Dec March 2009 Problem Statement In September 2008, 35% of the CBCs of the Adult ED Laboratory tests were not resulted within 32 minutes. In September 2008, 39% of the Chem 7 and PT of the Adult ED Laboratory tests were not resulted within 44 minutes. In September 2008, 34% of the CK, CKMB, Troponin, Liver Profile, and Lipase of the Adult ED Laboratory tests were not resulted within 60 minutes. Time is referenced from ORDER to VERIFY.

14 Goal Statement By March 17, % of the CBCs of the Adult ED Laboratory tests will be resulted within 32 minutes. By March 17, % of the Chem 7 and PT of the Adult ED Laboratory tests will be resulted within 44 minutes. By March 17, % of the CK, CKMB, Troponin, Liver Profile, and Lipase of the Adult ED Laboratory tests will be resulted within 60 minutes. Business Case To improve physician and ED Patient Satisfaction by contributing to improved ED throughput. Reduction of Laboratory Turnaround time will support reduction of o ED length of stay and LWBS (Left Without Being Seen) in the Adult Emergency Department. By reducing LWBS rate from 5.2% to 2%, assuming a historical admission rate of 20%, then we will see an additional contribution margin of $950,675.

15 Primary Process Flowchart Orders Placed Physician / RN initiates stat lab orders HUC / RN inputs order in computer Is Label in Triage? Label generated No in lab Is order An Add-on? No Yes RN / ACP reviews orders Labels in waiting box RN / ACP gets rainbow tubes No Is patient located? Yes Try calling patient again Yes 1 3 Lab drawn MLA draws blood Stick successful? No RN/Physician retries stick Lab called to ED to draw specimen Yes MLA labels blood with Cerner label Blood placed in tube MLA sends blood to lab Registration label placed on tube with date and time and RN / ACP initials 2 Flowchart transport/receive specimen 2 Is tube system working? No Specimen walked to lab Yes Sample sent in tube Specimen retrieved Orders in lab? No Specimen placed in waitingrack Yes 1 Match labels and specimen Do labels match specimen? No Call ED to clarify Specimen requirements met? No Call ED for recollection 3 Yes Receive specimen in computer Specimen labeled & moved to appropriate lab area 4 5 6

16 Multi-Voted Cause and Effect Diagram LABORATORY ED Turnaround Times - Multivote results Policy Material People Triage MLA under utilized (3) Too many recollects due to poor specimen (9) N o staff to sort orders & send to lab in Tube system (3) ED Lab Hotline calls transferred (4) Troponin repeated in excess (10) cancel/reorder on same accession is cumbersome Tech at Data Link in lab has too many responsibilities Communication of Add-On Labs (6) Recollects not drawn in timely manner by Triage Overwhelmed MLA s at ED processing station (3) Triage MLA unclear duties (3) N o Im pact Lab Te ch/mla (8) New Equipment in chemistry has downtime (7) No Tube System in Triage (7) Delayed Lab TAT Environment Procedures Equipment Critical Factors Identified Lab orders and recollects not communicated to MLA in ED Triage Inadequate assistance for HUCC or other RN MLA in Triage batching specimens to tube to laboratory Triage MLA under-utilized utilized

17 Rapid Cycle Process Improvements 1. Established a dedicated line for the ED staff to call the Laboratory. 2. Lab Automation in Chemistry went live the week of December 12th. 3. Indices automatically performed by Instrumentation. 4. Triage MLA flagging charts with RED tag so nurses know which patients have had blood drawn. 5. No batching of specimens from Triage (Tube station to be added) 6. Added Printer in Chemistry for Add-on Orders; changed process for ED nurses. Revised Flowchart Deliver to Chemistry 7 Specimen placed on Power Processor Specimen centrifuged 7 Does specimen have more than one label Yes Consolidate Orders No Specimen sorted Specimen placed on Power Processor 8 9 Does specimen have more than one label Yes Label placed in holding area for future tests No Specimen moved to appropriate machine Lab Automation now performs Specimen centrifuged Specimen moved to appropriate machine No Call ED for recollection Specimen requirements met? Document Hemolysis, etc. in Cerner Yes

18 Major Outcomes ER processing bench located next to automation inlet. Add on procedure simplified to sync with automation. Pneumatic tube station installed in ER Triage area linked to pneumatic tube at ER processing bench. Dedicated phone line for ER calls. Identified second OPI project for improving specimen quality.

19

20 Revised Flowchart Deliver to Chemistry 7 Specimen placed on Power Processor Specimen centrifuged 7 Does specimen have more than one label Yes Consolidate Orders No Specimen sorted Specimen placed on Power Processor 8 9 Does specimen have more than one label Yes Label placed in holding area for future tests No Specimen moved to appropriate machine Lab Automation now performs Specimen centrifuged Specimen moved to appropriate machine No Call ED for recollection Specimen requirements met? Document Hemolysis, etc. in Cerner Yes Symbols for Flow Charting Requires Human Handling Potential Time Delay Biohazard Potential Medical Error Potential

21 Specimen Arrives (Barcoded) Specimen Received in LIS Preanalytical Processing Manual Steps Specimen Racked for Centrifugation Centrifuge Loaded Centrifugation Centrifuge Unloaded Remove cap Preanalytical Processing Manual Steps Aliquot? Print labels Label aliquot tubes Aliquot Transport

22 Medical Tech Time Safety Error Physical Specimen Arrives Received in LIS X X Racked for Centrifugation X X Centrifuge Loaded X X Centrifugation Centrifuge Unloaded X X Remove Cap X X X Aliquot Necessary X X Print Labels X X Label Aliquot Tubes X X Aliquot X X X X Rack for transport X X X Transport to Chemistry X X X Transfer of Instrument Rack X X X Load Instrument X X X Analyze Unload Instrument X X Scan specimen for storage X X Place in storage rack X X Preanalytical Processing with Automation

23 Sample Introduced at INLET Barcode scanned and sample auto- received at INLET

24 Samples queued for Centrifugation CENTRIFUGE automatically loaded, balanced, and unloaded.

25 Caps removed and dropped into biohazard waste in DECAPPER Samples queried at instrument loading for routing instructions.

26 One sample routed to DXI for testing, the second sample routed down the automation track for the next instrument. Manual Specimen Arrives (Barcoded) Preanalytical Processing Automated Specimen Arrives (Barcoded) Specimen Received in LIS Specimen Loaded onto Automation Line Specimen Racked for Centrifugation Centrifuge Loaded Centrifugation Centrifuge Unloaded

27 Remove cap Aliquot? Print labels Label aliquot tubes Aliquot Transport Manual Preanalytical Processing Automated Preanalytical Processing Physical Specimen Arrives Received in LIS X X Racked for Centrifugation X X Centrifuge Loaded X X Centrifugation Centrifuge Unloaded X X Remove Cap X X X Aliquot Necessary X X Print Labels X X Label Aliquot Tubes X X Aliquot X X X X Rack for transport X X X Transport to Chemistry X X X Transfer of Instrument Rack X X X Load Instrument X X X Analyze Unload Instrument X X Scan specimen for storage X X Place in storage rack X X Physical Specimen Arrives Received in LIS Racked for Centrifugation Centrifuge Loaded Centrifugation Centrifuge Unloaded Remove Cap Aliquot Necessary Print Labels Label Aliquot Tubes Aliquot Rack for transport Transport to Chemistry Transfer of Instrument Rack Load Instrument Analyze Unload Instrument Scan specimen for storage Place in storage rack

28 Potential Steps, Delays, or Hazards Manual Automated Betsy Schifanella OPI Project ED Blood Specimen Quality Improvement April July 2009

29 Problem Statement Less than optimal quality specimens from the ED result in decreased quality results and increased TAT due to special handling or rework. From March 09 baseline data in the ED, the hemolysis rate was 14%, the short sample rate was 23%, the fibrin interference rate was 2.5%, the rework rate was 1%, and the blood culture contamination rate was 4%. Goal Statement By July 22, 2009 our team s s goal is to: Reduce hemolysis to 2% Reduce short samples by 50 % Reduce fibrin interference by 75% Reduce rework (clotted, gross hemolysis, QNS, mislabeled) by 50% Reduce blood culture contamination rate to 2%

30 Business Case Eliminating the gold tube from the collection requirements will save Downtown $4,987 annually Reducing rework costs will save $5,883 per year in Laboratory supplies and labor Decreasing rework will improve Laboratory turnaround time and will increase ED revenue $29,952 per year by increasing ED throughput and decreasing LWBS Chlorascrub Swabstick supply will save Downtown $15,222 per year and $17,363 system-wide Improvements Eliminated Gold Tube from collection requirements which results in Downtown savings of $4,987/yr. and reduction of short samples (low volume) by 72%. ED Skills Fair provided education on collection requirements based on best practice flowcharts. Standardized extension assembly Standardized IV blood collection (eliminated syringe/supplies for safety & proper tube filling) Standardized blood culture collection Reduced supply choices

31 Improvements Pre-packing IV Start (6 items) and Blood Culture (4 items) collection supplies which supported best practice flowcharts and education. Provided only the best practice supplies Facilitated gathering of supplies Improved compliance with safety goals Increased nurses satisfaction in the ED Replace Chloraprep Frepp Applicator with Chlorascrub Swabstick as a more effective blood culture site preparation solution for a system savings of $17,363. IV start with blood collection supplies consolidated to best practice items.

32 Supplies bagged to minimize searching and assembling items. Grab and Go! Laboratory tests Blood culture

33 Latest Results Occurrence of Less than Optimal Quality ED Specimens in DPMO 45.7% 1.63 Sigma Pre-Trial Trial % 19.6% 2.34 Sigma % % 6.7% % 0% 1.3% 1.6% 0 Hemolyzed Short Sample Fibrin Rework (clotted, hemolysis, % Decrease 28% 72% 100% QNS,mislabeled) -17% 4.0% 1.2% Blood Culture Contamination 70% Total defects 57%

34 Automation Performance

35 Distribution of Chemistry Tests by Priority - 5 Days (N=8274) TS 15% AM 14% EX 11% ST 40% RT 20% AutomationTurn Around Time by Priority AM EX RT ST EST TS

36 Average Turn Around Time (+SD) (Receive to Verify) ER-STAT Manual ER-STAT Automated Automation reduces variation by elimination of manual steps and by consistently and reproducibly handling specimens. STAT Manual STAT Automated Time (Minutes) REFERENCE ACCOUNT CHEM PANEL COMPARISON TAT - RECEIVE TO RELEASE OF RESULTS 70% 60% 50% 40% 30% 20% 10% 0% minutes More

37 160 Histogram Inclusion criteria: TAT <100 from MAR09 Distribution COMMENTS.svd 140 Automated 120 Manual 100 Count Automated Problem Specimen Manual Problem Specimen TAT Impact of Automation Automation TAT is consistent and independent of priority. Variation in TAT is less than non-automated results Automation performance is on par with other facilities with similar configuration. Good samples = good results = good TAT

38 Betsy Schifanella Kim Torres OPI Project Lab / Triage ER TAT Sept Dec 2009 Problem & Goal Process Capability - Laboratory tests Order to Receive Calculations Based on Loglogistic Distribution Model ED Triage - CBC & Chem 7 Process Data LS L 0 Target * USL 20 Sample Mean Sample N 4212 Location Scale O bserv ed Performance PPM < LSL 0.00 PPM > USL PPM Total LS LU S L 89% > than 20 minutes 89% Opportunity Sigma Level

39 Problem & Goal 250 July Triage ED (Cbc & Chem 7) - Order to Receive 200 UCL=207.8 Minutes _ X= Lab Tests Meeting Lab turnround time of 20 minutes - 11% of the time LCL=-76.7 Business Case Project could contribute to reduce Left without Being Seen patients in the Emergency Department. If LWBS is reduced to 2 percent (the benchmark), the Hospital Contribution Margin would be between $1.5 and $2.5 million dollars.

40 Improvements Create Triage RN-MLA phlebotomist teams; one MLA assigned to each nurse; laboratory specimens drawn while nurse is triaging patient Dedicate Processing Station in laboratory to Emergency Department specimens only Triage rooms modified to a multi-functional standard Patient taken to triage room and Triage Teams move between rooms Rescue Patients triaged by Triage RN-MLA team; labs ordered and drawn before patient sent to Main ED bed Latest Results Control Chart - ED Triage Lab Tests (Chem 7 & CBC) Order to Receive time in minutes (GOAL = 20 minutes) July 2009 Baseline & Triage RN-MLA Team Dedicated ED lab processing Mobile Team Triage RN-MLA 100 Minutes AVG=65 min AVG=11 min _ UCL=42.7 X=17.2 LCL= Lab Tests July Baseline - meeting 20 minute goal 11% of the time meeting 20 minute goal 92% of the time & meeting 20 minute goal 80% of the time

41 Latest Results Control Chart - Rescue Patients - Arrival to First Order (CHEM 7 & CBC) Baseline & Meeting 20 min. goal 100% of the time Meeting 10 min. goal 91.4% of the time Minutes minutes 0 Avg=40 min Avg=12.3 min Avg=5.5 min UCL=13.0 _ X=5.5 LCL= Patients Baseline 9-21 meeting 20 min. goal 31% of the time Awards Greatest staff satisfier Team Concept Greatest patient satisfier Expedited patient care Most innovative improvement Team comes to the patient Most successful project - Arrival to first order 40 to 5.5 min. (100%) - Order to received in Lab 65 to 11 min. (92%)

42 Post-analytical Data Review and Monitoring Station Clinical Laboratory Automation Baptist Medical Center - Downtown Goals for Automation - Established 2007 Improve efficiency (LEAN) Consistency in quality and turn around time (6δ) Patient and technologist safety Reduction in medical errors Ability to continue expansion of outreach Ability to offset technologist shortage Goals for Automation - Accomplished 2009

43 Lessons Learned (OPI Team Perspective) Little changes can make a big difference! Well designed changes improve outcomes for patients and staff. Education alone had minimal impact on changing a process. When prep re- packed supplies were provided the new process was enforced and positive change resulted. The power and momentum of a dedicated team is awesome! Lessons Learned (OPI Team Perspective) Different perspectives from multiple departments give better insight into improving how processes in one department affect another department. Improved relationships between departments contributes to improved processes and better outcomes for patients. Appreciated their ideas being heard and incorporated into possible solutions. We are contributing to better patient care!

44 Successful LEAN processes = Successful Automation MLA / RN Triage Team Prepackaged Collection Sets for ER Pneumatic tube station in Triage ER processing bench next to automation inlet Automation Line Add-on procedures utilize automated storage ER Tracker board in Chemistry Summary Automated laboratories today are a reflection and continuation of clinical laboratory automation that begun with the invention of the AutoAnalyzer. Automation is efficient, safe, and accurate... but only when successfully integrated with well implemented LEAN human processes.