Vermont Explor Annual Meeting

1
Vermont Explor Annual Meeting

• April 29, 2003
• Randolph, VT

2
Welcome Introductions
3
Agenda

• Highlights Issues in 2002
• Data Quality in 2003
• Presentations by BISHCA VDH
• Error Thresholds (Steve Reynolds)
• Personal Identifiers (Ken Kuebler)
• HIDI web tool for corrections
• E Codes
• Outpatient Procedure Coding

4
Data Quality in 2002

• Highlights Error Rates
• Southwestern (lt0.33)
• Porter Medical Center (lt0.75)
• Fletcher Allen (lt3.0)
• Northeastern (Some 0 months)

5
Data Quality in 2002

• Not so Highlights
• HCPCS (Missing)
• Operating Physician (Missing)
• Revenue Code (Duplicate)

6
Data Quality in 2003

• New error thresholds (0 and 5)
• Validation reports are important!
• Improvements needed?
• Correct errors before next submission!
• New web-based correction tool

7
BISHCA Vermont Dept of Health

• Dian Kahn, BISHCA
• Charles Bennett, Epidemiological Surveillance
  Chief
• Richard H. McCoy, Public Health Statistics Chief
• Peggy Brozicevic, Public Health Statistics
• Pat Worcester, Caroline Dawson, Laurel Decher,
  Annette Rexroad

8
Data Corrections

• How much is enough?
• Hospital work load vs. data quality
• Statistical significance

9
Personal Identifiers

• Ken Kuebler, Executive VP, Hospital Industry Data
  Institute, Missouri Hospital Association

10
Personal Identifiers

• Why collect a personal identifier?
• What identifiers are normally used?
• How are identifiers kept secure and HIPAA
  compliant?

11
Personal identifiers can be used to

• eliminate the need for re-admit flag
• reduce E-code messages for accident patient
  follow-up visits
• allow tracking of patients across hospitals and
  services for research and analysis

12
Types of Identifiers

• The best identifier is one that a person uses
  consistently for identification purposes (e.g.,
  Social Security number, drivers license number).
• An identifier built from a combination of
  existing data (e.g., date of birth, parts of
  name, ZIP code, sex).
• An identifier built from a unique identifier
  (e.g., encrypted or sequential number). This
  translation would have to be performed in one
  office.

13
Keeping Identifiers Secure and HIPAA Compliant

• Obtain confidentiality agreements from all users
  of the data.
• Encrypt original I.D. and then destroy original.
  Keep encryption key at one location only.
• Allow a minimum number of persons access to
  personal identifiers.

14
Breaks Are Good

• Take 5!

15
HIDI Web Tool

• Ken Kuebler, Executive VP, Hospital Industry Data
  Institute, Missouri Hospital Association

16
E Codes

• Correct coding
• Issues with error messages
• Value to epidemiology surveillance

17
Outpatient Procedure Codes

• ICD-9 vs. CPT-4
• Coverage of the two code sets
• Variation across settings
• Precision
• Grouping
• Future trends

18
Conclusions Lunch

• 2002 was a much better year than 2001
• Timeliness has improved
• Most of the important areas have few errors (dx,
  px, demographics)
• Future focus logical consistency from quarter to
  quarter