gateway to the future: improving the national vital statistics system st. louis, mo june 6 th –...
TRANSCRIPT
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Interaction of Electronic Health Records and Vital Records Systems:
Vermont Comparisons of EBRS Data with Hospitals’ OBNet System Data
Cindy HooleyVermont Department of Health
Public Health Statistics
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Acknowledgements
Rich McCoyCenter for Health Statistics
Vermont Department of Health
Jason RobertsResearch & Statistics Section
Vermont Department of Health
Rachel Wallace-Brodeur Vermont Child Health Improvement Program University of Vermont, College of Medicine
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Background: VT EBRS and OBNet
2003 – 2005
Fletcher Allen Health Care (in partnership with Dartmouth Hitchcock Medical Center) developed a Web-based obstetrical delivery “registry” called OBNet. They planned to roll out the OBNet system to six hospitals and sell the application to the remaining VT hospitals.
OBNet application collected demographics, maternal and fetal risk factors, interventions, and outcomes.
Not a true “electronic health record,” but data entered real-time, point-of-service.
Users (physicians, nurses, medical records clerks, etc.) access screens and enter data based on their roles for patient care.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Background: VT EBRS and OBNet
2003 – 2005
Fletcher Allen approached us to propose a partnership:
1) Vital Records assist OBNet to modify the application to meet all national standards and requirements for birth reporting and registration.
2) Vital Records excuse the “OBNet hospitals” from using the VT EBRS (to avoid double data entry).
Health Department leadership agrees to collaborate with OBNet, resulting in a two year development cycle.
Utilized the NAPHSIS use cases and the NCHS guidelines and materials and incorporated the 2003 Revision of U.S. standard birth certificate.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
OBNet Hospitals VT EBRS Hospitals
6 Hospitals
57% of all births
6 Hospitals
40% of all births
Users: Medical Records Clerk
Users: Physician; Nurse; Medical Records Clerk; maybe others
1. Data sent from OBNet system to VT EBRS using PHIN-MS based on xml schema.
2. Batch file nightly.
3. Only sent to VT EBRS after the patient is discharged.
1. Data entered directly into VT EBRS.
2. Transmitted when record is completed (doesn’t depend on patient discharge or nightly batch).
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VT EBRS and OBNet: Differences in Data Collection
OBNet interface: Information for the birth data “extract” to Vital Records is pulled from different OBNet screens (e.g., medical conditions are not “grouped” as on NCHS birth worksheets).
Users at the OBNet hospitals do not use the NCHS worksheets.
OBNet screens allow access based on “roles” – for example, a medical records clerk will not have access to all data elements for entry or edit checks. Responsibilities for completion of OBNet screens are decentralized.
OBNet had to make significant changes to accommodate national standards:
Force the user to visit each data element and mark “unknown” (so we know they didn’t skip the question).
Changed several items from default responses to state-required options.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VT EBRS and OBNet: Implementation and Beyond
July 1, 2005
Six hospitals begin collecting and transmitting birth record data to Vital Records using OBNet. Other hospitals submitting births using VT EBRS.
2006 – 2007
Variety of defects and enhancements made to the OBNet. Many items found in the first year that did not meet our original requirements.
Collaboration was always positive, but OBNet was understaffed and therefore very slow to correct defects. Resulted in a lot of questionable data and extra “clean-up” of the births database by Vital Records.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
2008 – 2009 VDH changed font type and size for content (values) of all fields printed
on legal birth certificates and implemented new concatenation rules to accommodate long names.
Increase in the number of married teens giving birth. Learned that OBNet had 2 questions for marital status on their worksheet and on the interface – caused confusion for moms and clerks!
1) Marital Status: Single, Married, Divorced.2) Mother Married at time of Birth, Conception, or in Between?
Yes, No, Unknown (what VDH wants). Increase in the number of babies being transferred within 24 hours of
delivery. Learned that response to this was being lumped with additional question about any transfer before being submitted from OBNet to VDH.
VT EBRS and OBNet: Defects and Enhancements
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
The Big Question
Four and a half years later (July 1, 2005 – December 31, 2009) more data available for comparison to answer the question:
Is there a difference in the quality of birth data received directly from a hospital medical record compared to
what is entered into an EBRS?
Can’t fully answer the question since OBNet is not a true electronic medical record, but it is close enough to provide some guidance as we pursue HL-7 standards for birth / death collection by electronic medical/health record systems and the transfer of data to a state’s EBRS / EDRS.
We continue to examine the data (small numbers in VT – approx. 6,000 births / year).
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
The Big Question
Our hypothesis back in 2005 was that birth data from OBNet would be:
1) More accurate than EBRS (point of service rather than several days after the fact).
2) More complete than EBRS (less unknowns).
3) More timely than EBRS (more events reported w/in 10 days of the birth).
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Abnormal Conditions of the Newborn
Assisted Ventilation Immediately following delivery
Assisted Ventilation for more than 6 hours
NICU Admission Newborn Given Surfactant Antibiotics Received for Suspected
Sepsis Seizure/Neurologic Dysfunction Significant Birth Injury
Congenital Anomalies
Anencephaly Meningomyelocele/Spina Bifida Congenital Heart Disease Congenital Diaphragmatic Hernia Omphalocele Gastroschisis Limb Reduction Defect Cleft Lip and/or Cleft Palate Down Syndrome Suspected Chromosomal Disorder Hypospadias
A significant difference was identified in the reporting of abnormal newborn outcomes and also for congenital anomalies when comparing hospitals that use VT EBRS compared to OBNet hospitals.
VT EBRS versus OBNet: Accuracy?
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Number of Infants Requiring Any Assisted Ventilation: Level 1 Hospitals 2005
Jan. - June 2005
15
13
EBRS
OBNet
July - Dec. 2005
41
1
EBRS
OBNet
Before OBNet Implemented* After OBNet Implemented
*The “OBNet” hospitals are those that were using Vermont’s vital records’ birth system, but converted to OBNet after July 1st, 2005.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Number of Infants Requiring Any Assisted Ventilation: Level 1 Hospitals 2006-2009
2006-2009
566
13
EBRS
OBNet
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Infants who received Surfactant: Level 1 Hospitals 2005 - 2009
0
1
2
3
4
5
6
7
July-Dec2005
2006 2007 2008 2009
Num
ber
of In
fant
s
EBRS
OBNet
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Seizure / Neurologic Dysfunction: Level 1 Hospitals 2005 - 2009
Seizure/Serious Neurological Dysfunction
0
1
2
3
4
5
July-Dec2005
2006 2007 2008 2009
Num
ber o
f Inf
ants
EBRS OBNet
Seizure/Serious Neurological Dysfunction
July 2005 - December 2009
10
1
EBRSOBNet
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Significant Birth Injury: Level 1 Hospitals 2005 - 2009
Significant Birth Injury
0
5
10
15
20
25
30
35
40
45
July-Dec2005
2006 2007 2008 2009
Num
ber o
f Inf
ants
EBRS OBNet
Significant Birth Injury July 2005 - December 2009
EBRSOBNet
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Any Congenital Anomaly: Level 1 Hospitals 2005 - 2009
Any Congenital Anomaly July 2005 - December 2009
56
6
EBRS
OBNet
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VRPHP volunteered to do a small study from 2005 - 2007.
VRPHP asked three (3) OBNet hospitals to submit a small portion of the Birth Certificate data to them manually (on a worksheet) for an annual review of perinatal statistics.
The data was collected by a nurse using a worksheet provided by VRPHP.
This allowed for a comparison of the birth data reported to VRPHP on paper to the OBNet hospitals’ birth data submitted electronically to VDH.
The study substantiated the differences being noticed in the post-OBNet implementation.
Vermont Regional Perinatal Health Project (VRPHP) Study
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Total Transmitted by OBNet to VT EBRS: 2
Total Reported on Paper to VRPHP: 96
Total Transmitted by OBNet to VT EBRS: 15
Total Reported on Paper to VRPHP: 39
NICU Admissions
0
5
10
15
20
25
July-Dec2005
2006 2007
# o
f In
fan
ts
OBNet Paper
Assisted Ventilation
0
10
20
30
40
50
July-Dec 2005 2006 2007
# o
f in
fan
ts
OBNet Paper
Vermont Regional Perinatal Health Project (VRPHP) Study
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VRPHP Conclusions: Reasons for OBNet Underreporting
1. Abnormal Conditions of Newborn and Congenital Anomalies are rarely entered into the OBNet application by the physician.
Physician may not consider it as “their role” and it has not been assigned or clarified for the nurse, nurse manager, or other staff to enter the information.
2. OBNet hospitals are not using the NCHS worksheet.
Using their own worksheets and guidance documents, which may result in different interpretations for what qualifies as a confirmed condition, per the birth certificate standards.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VRPHP Conclusions: Reasons for OBNet Underreporting
3. Birth data items in OBNet that are incomplete are left to the medical records clerk to obtain and enter. An assumption is made that the medical records clerk will take care of it.
Medical and Health Information items on OBNet screens that the medical records clerk cannot access (re: not their assigned “role”).
VRPHP conclusion after meetings with the OBNet hospital staff:
“…clerks are entering just the information from the parent worksheet….they assumed that the other information was being entered by OBs, thus they did not have to enter it. When we presented this to the nurse managers, they had no idea that those fields were not being completed.”
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
NCHS questioning the underreporting of “Non-vertex presentation”, a new checkbox item under Characteristics of Labor and Delivery when compared to responses for Method of Delivery.
While we found this to be true in Vermont, it appears to be an issue only at hospitals using VT EBRS.
2005 2006 2007 2008 2009EBRS 5 3.9 3.2 3.8 2.6OBNet 0 0 0 0 0
VT EBRS versus OBNet: Accuracy, continued
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VT EBRS versus OBNet: Accuracy
NCHS concerned about inconsistency in reporting “Was delivery w/ forceps attempted by unsuccessful” under Method of Delivery.
Small numbers (39) in Vermont but significant differences found between VT EBRS records and OBNet records:
For EBRS records, 7/1/05 – 12/31/09:7.1% of records w/ failed forceps attempts were also coded with “forceps” as the final route of delivery.
Further, 42.9% of failed forceps attempts were marked as “spontaneous” vaginal deliveries, a highly unlikely, if not inconsistent result.
So, 50% of all records marked as “attempt at forceps delivery failed were miscoded.
For OBNet records, NONE were miscoded.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VT EBRS versus OBNet: Accuracy
NCHS concerned about inconsistency in reporting “Was delivery w/ vacuum attempted by unsuccessful” under Method of Delivery.
Small numbers (181)in Vermont but differences found between VT EBRS records and OBNet records:
8.6% of EBRS records and 6.3% of OBNet records coded w/ failed vacuum attempt were coded as “vacuum” deliveries.
28.6% of EBRS records and 9.0% of OBNet records coded w/ failed vacuum attempt were coded as “spontaneous” vaginal deliveries.
So, total of 37.1% EBRS and 15.3% OBNet records miscoded.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VT EBRS versus OBNet: Completeness?
Following is a chart showing rates of “unknowns” for selected items including those used in calculating bmi, weight gain, and smoking rates.
10,534 hospital births reported via EBRS for 7/1/05 – 12/31/2009.
15,953 hospital births reported via OBNet for same period.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VT EBRS versus OBNet: Completeness?
Fields containing Unknown values% of total EBRS
% of total OBNet
HEIGHT 0.1 0.7PWEIGHT 0.9 1.2DWEIGHT 0.4 1
WIC 1.4 0.7PAYER 0.1 0.2
BFED 0.3 0.5
CIGB4 1.8 0CIGT1 1.8 0CIGT2 1.8 0CIGT3 1.8 0
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
VT EBRS versus OBNet: Completeness?
Fields containing Unknown values% of total EBRS
% of total OBNet
Month first prenatal visit 0.3 0.5Day first prenatal visit 0.4 0.6Year first prenatal visit 0.3 0.4
Month last prenatal visit 0.4 0.5Day last prenatal visit 0.6 1.2Yearl last prenatal visit 0.4 0.5
Total # prenatal visits 0.6 0.9
Month last menses 8.3 7Day last menses 11.4 10Year last menses 8.1 6.1
Born alive now living 0.1 0Born alive now dead 0.1 0Terminations 0.1 0.1
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
We have observed that transmission of the birth certificate data from the hospital to the Vital Records’ birth reporting system occurs more often within 10 days of the event for OBNet hospitals than EBRS hospitals.
VT EBRS versus OBNet: Timeliness?
76% 76%
79%
85%
79%80%
85%
89%90%
83%
65%
70%
75%
80%
85%
90%
95%
Jul - Dec2005
2006 2007 2008 2009
EBRS Hospitals
OBNet Hospitals
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Conclusions
OBNET • Receive more complete data (less unknowns) with OBNet for birth outcomes,
complications, etc. because it is physician or nurse entering those fields.• Receive more accurate (consistent) data for method of delivery information
because medical personnel enters those fields.• Receive less accurate data for abnormal conditions and congenital anomalies.
Trend confirmed in VRPHP study appears to have continued through 2009; VDH may need to intervene with hospital/OBnet procedures.
• Receive the data in a more timely manner (higher % transmitted within 10 days of the birth). Doesn’t take as long to complete records
EBRS • Receive more accurate data for abnormal conditions and congenital
anomalies. If it’s there, clerk will find it?!• Receive less complete data (more unknowns overall for more fields).• Receive the data in less timely manner.
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Future Discussion / Topics
As software vendors prepare hospital or practice-based EHR systems to include data collection for states’ EBRS or EDRS reporting, those systems need to account for:
1. What (who) is the best source of the information?
2. Are the required data fields being populated with the best data source (e.g., data from mother’s worksheet)?
3. Have appropriate edit checks been put into place to ensure review of all data fields and minimize “unknowns” or blank/null fields?
4. Does the system integrate the use cases, standards and guidance from the National Association of Public Health Statistics and Information Systems (NAPHSIS) and the National Center for Health Statistics (NCHS)?
5. Is there capacity within the EHR to meet any state-specific requirements, depending on the jurisdiction’s laws, rules, and policies? (e.g., same sex parents; additional birth defects; hospice care question; etc.)
Gateway to the Future:Improving the National Vital Statistics System
St. Louis, MO June 6th – June 10th, 2010
Questions / Comments?
Cindy HooleyVital Statistics Information ManagerPhone (802) 651-1636Email: [email protected]
Richard McCoyDirector, Vermont Center for Health StatisticsPhone (802) 651-1862Email: [email protected]