Disclaimer The Views Expressed In This Presentation Are Those Of The

Transcript

3/14/2016 Laboratory Report Formats and Reference Ranges: How Labs Can Facilitate the Medical Decision Process CPT Joshua G. Martinez, MT (AMT) CPT Jeffrey Baker, MT (ASCP) Disclaimer The views expressed in this presentation are those of the speaker and authors, and The views expressed in this presentation are those of the speaker and authors, the do not reflect the official policy and do not reflect the official policyof of the Department of Army, Department of of Department oforArmy, Department Defense, U.S. Government Defense, or U.S. Government Learning Objectives • Understand that proper reference are required to make accurate medical decisions • Understand how to properly validate a reference range and establish a range if one cannot be validated • Understand the components of a laboratory result as outlined by the CAP standards and identify common mistakes in the laboratory report format. • Provide tips on how to make lab reports easier for the various stakeholders to understand. 1 3/14/2016 Why do we care? 42CFR§493.1253 Standard: Establishment and verification of performance specifications. (ii) Verify that the manufacturer's reference intervals (normal values) are appropriate for the laboratory's patient population. We care about you! 2 3/14/2016 What is a reference range? • An interval between and including two points • Upper and lower limits • “Normal” values/”normal” limits • 95% of healthy individuals Where did we get these ranges? 3 3/14/2016 How do I know if our reference ranges are appropriate? • Ask the question “where/how did we get these ranges?” • If unknown, then need to validate or establish ranges • If known, ask to see the study > Was a study performed? > Does it represent the population? > Are they outdated? What is normal? • Two variables that affect reference ranges • Analytical System • Population Do all tests need verification? NO, if: • Decision limits • • • • • National/International consensus i.e. Cholesterol, HbA1c, TDM, etc. Specialized physician provides a cutoff (i.e. endocrinologist) No need to establish de novo ranges Transference 4 3/14/2016 Transference • Transferring of ranges from a valid study • Only works if population and testing methodology are the same • Comparability of the analytical system > Population stays the same; change in analytical system > Must perform method comparison study • Comparability of the test subject population > Transfer of range from another lab or manufacturer > Same analytical system; change in population • Recommend performing range validation (n=20) When should we validate vs. establish? VALIDATE (n=20) When transferring ranges from previous study (transference) When using manufacturer ranges When changing analytical methods ESTABLISH (n=120) When validation fails When population changes When appropriate and/or necessary Validating a Reference Range • Most feasible/inexpensive • Manufacturer’s range* • Published literature* (e.g. Tietz, Henry’s) • *CAUTION! Can be OUTDATED!! • 20 healthy individuals • • • • Consent form >10% outside proposed range = fail Obtain another 20 specimens from healthy individuals >10% outside proposed range = fail; consider establishing a range 5 3/14/2016 Establishing a Reference Range • Develop a plan – “garbage in, garbage out” • A priori vs. a posteriori • Direct vs. indirect sampling • Exclusion criteria > Pre-analytical > Work w/ your medical director • Consent form • Minimum of 120 healthy individuals for each population- the more the better > 95% reference interval- Gaussian > 90% confidence limits • Partitioning: Age/sex/race/sample type etc. • Non-parametric analysis- recommended • Break up the burden! > Consider a multicenter approach- pooling data Confidence Intervals (CI) • Range of value that includes the “true” percentile w/ a specified probability i.e. 90% • Narrows as sample size increases • Measures variability • Minimum 120 sample size to achieve 90% CI Source: CLSI EP28-A3C: Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory 6 3/14/2016 Who Really is Normal? Post Analysis • Laboratory Notification • Involve all stakeholders • Memorandum signed by Medical Director > Provide synopsis of study- 5 Ws > Highlight major changes > Provide old and newly adopted range • Ensure instrumentation accurately reflect reference ranges • In case of LIS downtime, raw instrument reports may needed to continue patient care – incorrect ranges can have consequences 7 3/14/2016 Case Study #1 A primary care provider from outlying clinic calls laboratory management concerned with a sudden increase in K results (4 w/in 72 hours)… • Day 1- Lab management asks clinic lab for calibration and QC documentation- no issues • Day 7- physician calls back w/ more patients presenting w/ high K.- examined possible pre-analytical interferences, also pulled last calibration verification & PT survey- no issues • Day 14- Multiple clinic staff now concerned Pulled all results from past 30 days. Calculated % high K = 25%!! Case Study # 1 Discussion • Current reference range for K = 3.4-4.4 mmol/L • “Where did we get this reference range?” • Study performed upon analyzer implementation; not signed off! • Study was pulled and revalidated 8 3/14/2016 Case Study #2 5 year old Caucasian male undergoing chelation therapy for elevated lead levels. Lab monitoring for hepatotoxicity… • OUTSIDE LAB on 3 JUN 15 • Alk Phos 295 (60-415) U/L • Remaining CMP and GGT are normal • Family PCS • DIFFERENT LAB on 1 SEP 15 • Alk Phos 323 H (38-126) U/L • Remaining CMP and GGT are normal 9 3/14/2016 Case Study #2 Discussion • Are the results statistically different? • Is 295 = 323? Analytic error • Variability (CVs or Random Error) • Methodology • Age appropriate ranges!! • Lab #2 was using adult ranges Bottom Line • Know where your ranges come from • Ensure a study was performed • “Garbage in, garbage out” • Notify the stakeholders • Listen to your customers 10 3/14/2016 What Does CAP Say Required Components 1. Name and address of testing laboratory 2. Patient name and identification number, or unique patient identifier 3. Name of physician of record, or legally authorized person ordering test, as appropriate 4. Date and time of specimen collection, when appropriate 5. Date of release of report (if not on the report, this information should be readily accessible) 6. Time of release of report, if applicable (if not on the report, this information should be readily accessible) 7. Specimen source, when applicable 8. Test result(s) (and units of measurement, when applicable) 9. Reference intervals, as applicable 10. Conditions of specimen that may limit adequacy of testing Find the Required Components 11 3/14/2016 The Provider may not see it the way you do Common Mistakes • Different systems • Not just for providers • Viewing Method • Impact of changes Report Format Tips • View through the Providers Eyes • Lab Access to the provider system • Solicit provider feedback • Review printed report and on screen view • Understand the limits of your LIS • Get creative • Draw attention to abnormal results • If you can not find the required components, no one can 12 3/14/2016 Additional Tips • Not just about requirements • What does your policy say • Special Cases • Standardizing all locations Name the Positive Analyte and Detection Level Blank Slide – Do Not Remove 13 3/14/2016 Improved Format Same Results Different Systems Another Example 14 3/14/2016 References • CLSI EP28-A3C: Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory • College of American Pathologists All Common Checklist dated 7/28/2015 • Code of Federal Regulations (CFR) Title 42 Part 493Laboratory Requirements Acknowledgments USAMEDDAC Fort Stewart, GA Department of Pathology • • • • COL Donald Taillon, C, Department of Pathology Mr. Lennox Wildman, Quality Assurance Coordinator Ms. Maria Pagan-Collazo, Assistant QA Coordinator Ms. Deanne Nomura, LIS Manager Questions? Thank you for your attention! 15