NCHARM Pilot Project

1
Utah Department of Health
2
UNIQUELY IDENTIFYING NEWBORNS IN STATEWIDE
DATABASES

• John Eichwald, M.A.
• Utah Department of Health
• Children with Special Health Care Needs
• Barry Nangle, PhD, Jan Bagley, RN, BS,
  Martee Hawkins, RN, Jane Johnson, BS, and
  Nita Owens

3
Statutory Authority

• The Utah Department of Health (UDOH) is required
  to collect three distinct data sets for each
  birth that occurs in Utah
• Birth Certificates
• Newborn Heelstick Screening
• Newborn Hearing Screening

4
The UDOH Mission

• to coordinate the development of a Information
  System (IS) that will allow customers and staff
  seeking health information to obtain the
  necessary information they need, regardless of
  where it is collected.

5
The IS Goals

• 1) public health data are complete, uniform and
  accurate
• 2) stakeholders are satisfied with the
  availability and usefulness of public health
  data and
• 3) public health data are entered only once, and
  are readily retrievable by all authorized people.

6
The CHARM Project

• CHARM stands for Child Health Advanced Record
  Management
• The primary goal of CHARM is to create an
  electronic child health profile containing
  public health information.
• One of the first objectives is to link newborn
  hearing screening records, heelstick screening
  records, and birth certificates.

7
Improved Program Management

• If newborn data can be reliably shared it
  increases the potential for

• decreasing redundant data collection and
  reporting in Utah hospitals
• increasing the accuracy of the data being
  collected
• locating infants missed in screening
• identifying infants lost to follow-up
• not contacting families with infants who have
  died
• Improving the health status of newborns

8
Informed Decision Making

• Challenges of collecting newborn data
• Naming Issues
• first names unassigned baby boy
• last name paternity
• adoptions
• maternal name
• Nursery Issues
• pre-term delivery
• intensive care
• transfers
• early discharges

9
Pilot Project

• The Problem How do we uniquely identify newborn
  babies in these three databases?
• Do we use a deterministic or probabilistic
  matching paradigm?
• The primary goal of the Pilot Project was to find
  out if one number could be used to uniquely
  identify newborns.

10
Proposed Solution

• Since the process to order and manufacture the
  uniquely numbered heelstick kits occurs at least
  six months prior to the births in any calendar
  year, it was decided to see if this number could
  be used to uniquely identify newborns in the
  other two data systems.

11
Pilot Description

• The Pilot Project was funded in part by the State
  Systems Development Initiative (SSDI) grant
  through HRSAs Maternal and Child Health Bureau

12
Pilot Description

• Two pilot hospitals selected
• Salt Lake Regional Hospital
• large urban
• (1,667 births/year)
• Brigham City Community Hospital
• smaller rural
• (515 births/year)
• Pilot Timeframe
• July 1 through September 30, 2000

1998 Vital Statistics
13
Pilot Description

• Created new Birth Record Number (BRN) field
  within Hearing Screening and Birth Certificate
  databases

14
Alternate Record Number
15
Pilot Description
Created new Birth Record Number (BRN) field
within Hearing Screening and Birth Certificate
databases

• Designed and printed 1,000 label sheets with six
  (6) peel-off pre-numbered labels which were
  included with their respective heelstick kit

16
Pilot Description

• Size 2½ x 5 inches

17
Pilot Description

• INSTRUCTIONS
• Place these labels as listed below
• 1. Place on Newborn HEARING Screening
  record
• 2. Place on the Delivery Record
• 3. Place in Medical Records chart
• 4. Place in Newborn Screening Log
• 5. May be used for Immunization

18
Pilot Description
Created new Birth Record Number (BRN) field
within Hearing Screening and Birth Certificate
databases Designed and printed 1,000 label sheets
with six (6) peel-off pre-numbered labels which
were included with their respective heelstick kit

• Hospital staff chose their own procedures for
  distributing the labels to the birth certificate
  and newborn hearing screening staff
• This number was then entered into their
  respective computer database systems

19
Pilot Results
597 Heelstick Screens 594 BRN (99.5)
598 Birth Certificates 594 BRN (99.3)
597 Hearing Screens 592 BRN (99.2)
Birth Record Number (BRN) propagation results
20
Pilot Results

• Potential reasons why BRN would not match between
  databases
• record not in database
• missing BRN
• invalid BRN
• duplicate BRN
• different BRN between databases

21
Pilot Results
Linked Records 591 (99.5)
594 Birth Certificates
594 Heelstick Screens
Linked Records 585 (98.8)
Linked Records 585 (98.8)
592 Hearing Screens
22
Pilot Results
Heelstick Screens
Birth Certificates
Linked Records 583 (97.5)
Hearing Screens
Total Records Linked (598 Births)
23
UDOH Conclusions

• Pilot hospital did an outstanding job!
• Other hospitals not notified of BRN field
• Need to define data procedures when kits are
  contaminated
• Procedures not defined for missing screens
• Three way matching is complex
• Recognize processes to be integrated, not just
  data elements

24
Hospital Conclusions

• Implementation process was individualized for
  each site
• Implementation was felt to be reasonably simple
• Staff felt the label was helpful
• did not have to manually enter heelstick number
  into birth log
• The BRN was used for other tracking
• Staff felt UDOH should implement BRN

25
Outcome of the Pilot

• Due to the apparent success of the pilot project
  the use of the Birth Record Numbers will be
  phased in for all 44 Utah birthing hospitals
  beginning in January 2002

26
Thank you!
jeichwal_at_doh.state.ut.us