1	Point of Care Testing An Informatics Perspective
2	Outline Brief introduction to Point of Care Testing (POCT) Informatics/management issues in POCT Why these issues are important for the mission of the lab and the hospital Solutions available now Solutions available in the near future
3
4	Massachusetts General Hospital: Trends
5
6	The Expanding Menu Of Point Of Care Testing Chemistry Blood glucose testing Blood gases, fetal scalp pH, electrolytes, specific gravity Cardiac markers: troponin, CK-MB, myoglobin, BNP Dipstick urinalysis Pregnancy testing and ovulation assessment Fecal occult blood and gastric occult blood Cholesterol Intraoperative-PTH Neonatal bilirubin Hematology Coagulation testing, activated clotting time (ACT), platelet function Microbiology/Serology: Physician performed microscopy. HIV, Group A streptococcus, H. pylori serology and CLO testing, STD, other
7
8	The Informatics “Great White Whale” In future nearly all testing needs to be integrated in the EMR if the EMR is to be the source of truth for patient care The electronic patient record, evidence based medicine, and process improvement have become increasingly important The growth of POCT presents a challenge to the goal of one patient record as it is performed in a distributed fashion POCT programs need to be actively managed
9	POCT Management Structure At MGH
10	OK, so you’re in charge…Now what? POCT increasingly part of laboratory operations and under laboratory control Increasing utilization of POCT: 8% of all tests at MGH Problems in test performance replaced with challenges in information management
11	Information Management Principles
12	Evolution of POCT Connectivity First Generation
13	Evolution of POCT Connectivity Second Generation
14	Evolution of POCT Connectivity Third Generation
15	NCCLS POCT Interface Specification Developed from CIC consortium (2000) Attempts to support all POC diagnostic devices Handhelds Small benchtop analyzers Test modules incorporated into existing bedside patient monitors
16	NCCLS POCT1-A Interface
17	Device Interface Generally a simple, tightly coupled interface if device and DMS are from same vendor Governs the bidirectional flow of information between the POC device and the data manager Device status, location, events, patient results QC data, operator lists, patient lists
18	Observation Reporting Interface Governs communication between the data manager and the observation recipient (LIS or CDR) Leverages existing HL7 interface for laboratory instruments PROBLEM: LIS will typically not accept result unless an order is present
19	Data Manager to LIS/HIS Transmit Options Via scripted interface (terminal emulation) Emulates user logging into LIS, entering order, waiting for accession number, resulting test Less costly but must be updated regularly Less robust error capture (i.e. when LIS download fails) than interface, can be slow, especially with tests with multiple analytes
20	Enterprise Level Connectivity Advantages Foster cooperation, standardization ¯ Overall cost (equipment (servers), licensing fees, interfacing) Disadvantages Customizations, choice of vendors for each site may be limited
21	Wireless POCT Model
22	Connectivity to IS…why bother? Data management Operations Clinical value Billing
23	What is value of POCT information in the LIS? Currently 70-80% of POCT never makes it into the electronic patient record (LIS or HIS) “LIS is the source of truth” for lab test information Data mining, research, test volumes, operations, TAT, QA, QC, EBM The LIS is typically the conduit for laboratory information flowing to the HIS and billing system
24	What is value of POCT information in the electronic patient record? Hospital surveys indicate that less than 10% of hospitals currently transmit POCT data to EMR Most POCT leads to rapid response (give K, give insulin, etc.) so does having a record of the abnormal result add value? Growth of EMRs has created dependencies for clinical lab data to be available electronically How to get the information into the HIS? POCT à POCT DMS à LIS à HIS Attention must be paid to display of POCT in EMR Reference ranges, in labs section or separate section
25	Billing (and Payment) for POCT Historically most organizations have not billed for POCT Cost of testing included in the “facility charge” or “encounter charge” Lack of connectivity often cited as a reason not to bill Billing systems typically require that test is ordered, received and resulted in the LIS Billing requires a physician order to bill which may not exist Whether charges can be captured is determined by: Connectivity to hospital charge description master, payment criteria, CPT coding Patients Inpatients - DRG Hospital based clinics – Depends on payment system ED – Payment based on type of encounter Outpatients – Best opportunity for collection
26	Beyond Connectivity Systems to ensure clean data from the start are necessary if connectivity solutions are to automatically pass test data to the LIS Stopping errors at the bedside with automatic identification technologies is the optimal solution Bar coding Radiofrequency ID
27	Patient Identification in POCT
28	Why an improved patient wristband? Patient safety #1 of JCAHO 2003/2004 National Patient Safety goals is “To improve the accuracy of patient identification” Positive patient ID essential for medication administration. Med admin record (MAR) with patient bar coding successful at reducing error Positive patient ID essential for lab draws. Operations efficiency Automates charge capture, documentation for EKG machines, glucometers
29	Impact of Wristband Bar Coding
30	Inpatient Wristband
31	Radiofrequency Identification (RFID) RFID: hospital IDs, turnpike tolls, inventory control tags RFID tags can hold 100s of characters of data RFID tags are read-writable Wal-Mart, DOD are requiring RFID for top suppliers by 2005 Early adopters in retail & consumer packaged goods have achieved cost savings of 5% of sales
32	Wristband ROI Reduce liability exposure to sentinel events Improve charge capture Promote efficiencies (POCT) Reduce adverse drug events One major source of ROI for bar coding is avoidance of adverse events Adverse drug events on average result in 2.1 days added to LOS and $4600
33	Preventable Adverse Drug Events
34	Potential Cost Savings MGH Annual Admissions: ~45,500 projected for FY’04 Adverse Drug Reaction Rate (ADE): 6.1 per 100 admissions = 2,778 26% of ADE are Preventable = 722 Preventable administration errors = 38% x 722 = 274 Reduction of errors reported by bar coding averages 70% (range 59-86%). 70% x 274 = 192 errors avoided annually Cost Savings: $4600 per ADE x 192 = $883,200 /year LOS: (2.1 Days per ADE) x 192 = 403 days saved/year Backfill: (403 days/5.4 days per stay) x 0.85 = 63 stays
35	Barriers to Implementing POCT Informatics Solutions Infrastructure requirements Bar coded wristbands Interfaces Cost-benefit analysis is necessary The case for connectivity needs to be made on a variety of levels Regulatory/QA/QC Billing Clinical utility Limitations of current “multi vendor” connectivity solutions
36	What tests to connect? Based on test volume, utility in EMR/LIS, management issues, device type, connectivity Probably should be: Glucose Electrolytes Cardiac markers ACT Not so useful/difficult to implement FOBT Rapid strep test Manual urinalysis
37	Cost versus value Laboratory testing is cheap 4% of hospital budget Influences 65% of hospital costs
38	Connectivity Cost Assume interfacing for each device costs $ 15,000 POCT cardiac marker device with a three year useful life performing: 5 tests per day 30 tests per day POCT glucometers performing 1000 tests per day Cost per test to interface ? = Cost per year/number of tests $ 5,000 / (5 X 365) = $ 2.73/ test $ 5,000 / (30 X 365) = $ 0.46/ test $ 5,000 / (1000 X 365) = $ 0.01/ test
39	When to connect? POCT is not cheap with respect to consumables and devices (glucose strips ~$0.50 to 1.00 per test x 400,000 tests/year) Optimally, connectivity should be negotiated at time of contracting for the entire POCT system
40	POCT in the near future?
41	Summary Along with the increased capabilities and use of POCT the importance of connectivity is increasing The central laboratory is best suited to coordinate the hospital POCT management program Simple additions to the POCT program can improve systems Bar coding for data entry