Multi vehicular Pile ups on the EHDI Information Superhighway National EHDI Conference San Diego, CA March 15, 2015 Kathryn Aveni, RNC, MPH Disclosures Financial: Kathryn Aveni serves as a paid regional Technical Assistance Network consultant for the National Center for Hearing Assessment and Management. The EHDI Road Map 1
A New Jersey Road Map For The Trentonian/BRIAN McCARTHY Seven tractor trailers tangle on NJ Turnpike near Exit 8; thousands backed up for hours Published: Wednesday, May 11, 2011 Dump truck accident on NJ Turnpike will cause traffic delays for next two months. Times of Trenton 2
EHDI Database Inpatient Hearing Screening Outpatient Screening & Diagnostic Exams Data Entry Data Entry Birth Certificate System Immunization Registry System weekly downloads EHDI Database Microsoft Access EI System Annual automated match, quarterly manual matches Reports EHDI Database Accidents Happen Inpatient Hearing Screening Outpatient Screening & Diagnostic Exams Data Entry Data Entry Birth Certificate System Immunization Registry System weekly downloads EHDI Database Microsoft Access EI System Annual automated match, quarterly manual matches Reports CDC EHDI Functional Standards http://www.cdc.gov/ncbddd/hearingloss/ehdi is functional standards.html 3
Goal Statement 1: Document unduplicated individually identifiable data on the delivery of newborn hearing screening services for all infants born in the jurisdiction. Goal Statement 2: Support tracking and documentation of the delivery of follow up services for every infant/child who did not receive, complete or pass newborn hearing screening. 1.1 The EHDI IS SHALL provide a unique patient record for each newborn child born in the jurisdiction. 1.4 The EHDI IS SHALL receive and document all individual newborn hearing screening procedures and results, in a timely manner in accordance with scope of practice, organizational policy and jurisdictional law. 2.3 The EHDI IS SHALL provide the ability to generate and present a list of infants who did not receive or complete newborn hearing screening and are in need of recommended screening and/or diagnostic evaluation services. Goal Statement 5: Maintain data quality (accurate, complete, timely data) of individual newborn hearing screening, follow up screening and diagnosis, early intervention and demographic information in the EHDI IS. 5.5 The system SHALL provide the ability to analyze information with respect to data quality according to scope of practice, organizational policy, and/or jurisdictional law (see appendix B for guidance on essential data quality attributes). Essential Data Quality Elements Completeness: The proportion of stored data against the potential of 100% complete Uniqueness: Nothing will be recorded more than once based upon how that thing is identified. It is the inverse of an assessment of the level of duplication Timeliness: The degree to which data represent reality from the required point in time Validity: Data are valid if it conforms to the syntax (format, type, range) of its definition Accuracy: The degree to which data correctly describes the "real world" object or event being described. Consistency: The absence of difference, when comparing two or more representations of a thing against a definition 4
Crash 1 Fender Bender Stage: Inpatient screening results Data Quality Element: Accuracy In response to an EHDI fax to a pediatrician office for follow up information, they indicated baby passed initial screening and included their documentation from the hospital. Prevention and Mitigation Vehicular Accidents Prevention: Driver education, speed limits, sober & alert driving, lane departure & blind spot warnings Mitigation: Seat belts, air bags, crumple zones 5
Prevention and Mitigation EHDI IS Accidents Prevention: User education on data system, rules and regulations, eliminate data entry (i.e., HL7 messaging) Mitigation: Redundancies in identifying cases needing f/u Data reports and monitoring to quickly identify and correct problems Crash 1 Prevention & Mitigation Create written materials for parents & physicians that are not potentially confusing Screener training PCP education Results provided to parent and PCP and EHDI Crash 2 EHDI Data Point: Inpatient screening results Data Quality Elements: Completeness, Timeliness 6
Do you see a problem here? What happened? Primary birth certificate clerk was out on extended leave of absence. Birth records got completed through the legal fields, and the rest of the data was put off for later & never got caught up. OVSR only monitors timely completion of legal part. NJ EHDI administrative rules require hospitals to submit hearing results, via the birth certificate system, within 1 week of transfer or discharge. 7
With each weekly birth certificate file import, a report is run automatically. Rates out of expected ranges are highlighted. Mitigation Initially created to monitor screening and refer rate issues. Percent of expected births was added to monitor for missing data. Recent (April July 2015) conversion to new birth certificate system created lags for many hospitals. Caution Potholes ahead Spaces in names (Garcia Perez) Hyphens (Smith Jones) Apostrophes (Ja yonna) Special characters: ῆ, ë, é 2013 NJ birth records: 11% of last names had a hyphen (~3000 records) or a space (~7000) 8
s are changed. True scenarios NJIIS lookup Registry DOB Sex Address ID 198 Jones, Unique 5/30/14 F 123 Main St, Newark NJ 723 Jones, Unique Maria 5/30/14 F 988 Elm St, Paterson NJ EBC ID First Middle Last DOB Mom 545454 Unique Marie Jones 5/30/14 General Hospital 223344 Un ique Ursula Smith Jones 5/30/14 St. Elsewhere Crash 3 Stage: Inpatient screening & outpatient diagnostic results Data Quality Element: Uniqueness Twins Similarly named children Baby Blues June 14, 2004 9
DOB s are changed. True scenarios Child First Child Middle Child Last Suffix Plurality Birth Order 1/1/2015 John Daniel Rodriguez 2 1 1/1/2015 John Dariel Rodriguez 2 2 DOB Child First Child Middle Child Last Suffix Plurality Birth Order 2/1/2015 George David Forman III 2 1 2/1/2015 George David Forman IV 2 2 s are changed. True scenarios on report submitted: Sai Srihan Madhu NJIIS lookup Registry DOB Address ID 3322 Madhu, Sai 5/30/14 456 Main St, Edison NJ EBC ID First Middle Last Records were inappropriately merged in NJIIS DOB Plurality Birth Order 667111 Sai Srihan Madhu 5/30/14 2 1 677112 Sai Srihith Madhu 5/30/14 2 2 Adam and Zachary (names changed) Birth certificate data indicated Adam passed and Zachary referred screen (and was in NICU x 4 d) Audiologist sent a f/u report for Adam. No report sent for Zachary. Call to audiologist. She called Mom, who was sure it was Adam that needed follow up (stated he was first born, he was not in NICU) Review of medical records verified the WBN (not NICU) baby referred. EBC system & worksheet review showed Adam noted as born at 7:02 am and Zachary noted as born at 7:03 am But, Adam s worksheet said birth order = 2 and Zachary s said birth order = 1. Hearing results entered based on birth order, thus swapped. Mom instructed to go to local registrar for birth certificate correction to birth order. 10
s are changed. True scenarios NJIIS lookup Registry DOB Sex Address ID 336 Santos, Sabrina 4/30/14 F 123 Main St, Anytown NJ 445 Santos, Sabrina 4/30/14 F 988 Oak St, Nearby NJ Anytown Hospital audiology department saw Sabrina Santos and diagnosed her with bilateral SNHL. Form done on ID 336. ID 336 really passed at birth. ID 445 referred at birth and is the child with hearing loss. Nearby hospital does not have audiology department. Prevention User training and retraining and retraining. Users must verify ALL fields! Be cautious with the use of Soundex matching. Carefully review all non exact matches. Be cautious with merging duplicates. Cross reference with other datasets. Data entry system validation checks may prevent some errors. Mitigation EHDI Quarterly Report includes Possible Inpatient Result Errors Selects cases where inpatient result = pass/pass and reason for test on OP follow up notes refer prior screen or not screened previously Follow up form: Report: 11
Crash with Injury Stage: Inpatient screening & outpatient testing results Data Quality Element: Accuracy Crash with Injury DOB 6/10/14. Inpatient bilateral refer on 6/13/14 EBC results incorrectly entered as pass/pass Hospital f/u contact attempt documented 9/22/14 Appointment scheduled. No show documented 9/24/14 2/4/15 another missed appointment documented No EHDI program follow up (believe passed) 6/12/15 (age 1 yr), diagnostic test: bilateral moderate SNHL, now showed up on possible errors report Exceeded goal of diagnosis by 3 month of age, with data entry error contributing to lack of additional f/u efforts EHDI reports modified to include lost to f/u forms in check for possible data entry errors 12
Flaming Multi Vehicular Pileups Stage: Inpatient screening & outpatient testing results Data Quality Element: Completeness, Uniqueness, Timeliness, Accuracy s are changed. True scenarios Natia Dobrcyzka born 8/14/13, referred ABR on R ear inpatient at St. Elsewhere EBC record not completed timely (finished in January 2014) Outpatient rescreening was done for Natia (refer right ear) on 8/29/13 with documentation submitted Outpatient f/u form not automatically matched to birth record because birth record not in system yet Manual match incorrectly matched follow up to Esmerelda Dobrcyzka, dob 8/13/13 born at St. Elsewhere on a close match review. The story continues. Audiologic testing 10/18/13. Pass L, could not test R (baby would not sleep) Not flagged for further f/u because EHDI not aware of inpatient refer yet AND because f/u linked to the wrong baby based on prior match Mismatch identified after report reconciliation and final late data submission 1/16/14 EHDI outreach to pediatrician to refer for more testing 4/17/14 diagnostic eval: TCHL R (type B tymp), normal hearing L. Pass OAE bilaterally, ABR not done (8 mo old). Recommended for f/u at 24 30 mo. No additional testing documented. 13
Questions? Kathryn.Aveni@doh.state.nj.us 609 292 5676 14