CHIMA Current Data Quality Initiatives

1
CHIMA Current Data Quality Initiatives

• OHIMA - SPRING INSTITUTE
• TORONTO, ONTARIO
• Gail Crook, May 7, 2004

2
Current Reality

• CIHI questions variability in the DAD and
  validity of complexity overlay develops
  National Grouper Redevelopment Committee
• Ontario MoHLTC questions variability of the DAD
  for planning and funding purposes- commissions
  pilot D.Q. Audit
• Ontario JPPC questions variability of DAD for
  rate based funding initiatives questions impact
  for multi-year funding

3
CHIMA Response To Issues With Variability Of The
Dad

• Position paper developed and sent to CIHI
  regarding reasons for variation in the DAD
• Met with Ontario MoHLTC to develop pilot
  Clinical D.Q. Audit proposal (Ontario)
• Met with Alberta Health and Wellness to develop
  D.Q. Audit proposal (Alberta)
• CPEs and national workshops on D.Q.
• Working with provincial health record
  associations for provincial and national
  strategies for D.Q.

4
Ontario MoH D.Q. Audit

• Request from MoHLTC to develop a proposal for a
  pilot clinical data quality audit project to
  assist the MoH to
• Increase understanding of the variances in
  hospital coding practices and case mix
• Look at a MoHLTC methodology for identifying
  hospitals to be audited
• Partnership of CHIMA, OHIMA and CIHI to develop
  proposal

5
Hire/Train Auditing Staff

• 6 expert coders to be recruited by CHRA/OHRA
• 5-day training provided by CHIMA/OHIMA
• 2.5 days refresher
• 0.5 day data collection application
• 2.0 days inter-rater reliability testing

6
Data Collection

• Original DAD data downloaded to laptops
• On site re-abstraction of charts
• Diagnoses and procedures re-abstracted blindly
• Non-clinical data reviewed/validated
• Identification of discrepancies and reasons for
• Reconciliation process
• Attempt to secure consensus among team members
• CIHI will rule on difficult cases

7
Discrepancies and Reasons
Reason Code
Discrepancy Code

• What did the re-abstractor identify as the reason
  for the difference

• What type of difference was there between
  original data submission and and re-abstracted
  data

8
Project Reports

• Provincial Report
• Aggregated data tables (including unweighted and
  weighted data)
• All facilities
• Teaching only
• Community only
• Summary findings and recommendations
• CHIMA/OHIMA
• Health Record/Hospitals
• CIHI
• MOH

9
5 CMGs Studied

• 013 Specific Cerebrovascular disorder (EX. -TIA)
• 143- Simple Pneumonia and Pleurisy
• 222 - Heart Failure
• 485 - Nutritional Miscellaneous Metabolic
  Disorders
• 751- Septicemia

10
Changes in Complexity Level
Complexity
Original DAD Data
Re-abstracted Data
Change
(PLx)
2
Sample

Sample



1
291
26.1
603
54.1
312
107.2
2
289
25.9
234
21.0
-55
-19.0
3
273
24.5
147
13.2
-126
-46.2
4
261
23.4
130
11.7
-131
-50.2
0
0.0
0
0.0
0
0.0
9
1,114
100.0
1,114
100.0
0
Total
11
Results by CMG
Original CMG
Change in
Change in
of Records
of Records
RIW
ELOS
that Switched
that Switched
CMGs
MCCs
013 - Spec cerebrovascular
-9.0
-22.4
11.0
4.5
disorders(xtia)
143 - Simple pneumonia
-18.3
-12.7
19.1
10.3
pleurisy
222 - Heart failure
-17.8
-18.3
9.5
3.6
485 -Nutritional/Miscellaneous
-18.4
-17.3
32.9
26.6
metabolic disorders
751 - Septicemia
-27.2
-21.4
29.3
24.8
Total
-17.5
-18.9
20.3
13.9
12
Diagnosis Discrepancies
Diagnosis Discrepancies
Discrepancies
Error
MRDx Discrepancies
MRDx different type
160
11.4
MRDX missing
55
3.9
Post-Admit as MRDx
3
0.2
Secondary as MRDx
24
1.7
Dx not coded, orig as MRDx
65
4.6
Total MRDx Discrepancies
307
22.0
CC Diagnosis Typing Discrepancies
CC diagnosis as Type 3
56
1.0
Pre-admit as post-admit
55
1.0
Post-admit as pre-admit
33
0.6
Secondary Dx as CC
1,596
29.3
Total CC Diagnosis Typing Discrepancies
1,740
32.0
CC diagnosis discrepancies
CC Dx missing
289
5.3
Dx not coded, orig as CC
1,432
26.3
Transfer Dx missing
11
0.2
Dx not coded, orig trans
5
0.1
Total CC Diagnosis Discrepancies
1,737
31.9
Diagnosis Prefix/Suffix Different Discrepancies
79
1.2
Any diagnosis different discrepancies
556
8.1
Mandatory type 3 missing
31
10.3
Total Diagnosis Discrepancies
4,450
65.0
13
Reasons for Discrepancies
Reason Code
Total


A - Transcription error
9


0.2
B - Incomplete documentation
65


1.5
D - Lack of Code specifity
129


2.9
E - Code specificity not supported
128


2.9
F - Different interpretation - disagree
421


9.5
I - No significant impact on treatment
2,125


47.8
K - Other grey area coding
68


1.5
L - Inconsistent or conflicting information
155


3.5
M - Coding contrary to CIHI guidelines
691


15.5
N - Hospital Policy
21


0.5
O - Coding error
228


5.1
P - Information on chart missed
151


3.4
V - Other
33


0.7
W - No apparent reason
49


1.1
Z - Significant impact on treatment
177


4.0
Total
4,450


100.0
14
Summary of High-Level Results

• 18 decrease in RIW
• 19 decrease in ELOS
• 107 increase in the number of Plx 1 cases
• 20 of records switched CMG
• 14 of records switched MCC
• 48 of discrepancies occurred where re-abstractor
  found diagnoses had no significant impact on
  patients treatment or LOS
• 16 of discrepancies due to coding contrary to
  CIHI guidelines

15
Recommendation 1

• Standardization of Coding
• Work with key stakeholders to develop and
  implement comprehensive coding standards.
  Recognition and adoption of standards will
  support and promote uniform effective coding
  practices in Ontario acute care hospitals that
  will in turn, result in reliable data.
• The MoHLTC mandate all health record coders
  attend CIHIs Diagnosis Typing/Standards
  Workshop.
• CIHI to seek out stakeholder collaboration on
  redevelopment of Grade List Diagnosis
• CIHI to solicit increased consultation regarding
  the definition of diagnosis typing.

16
Recommendation 2

• Improve Clinical Documentation
• Facilitate the highest quality of health provider
  clinical documentation in Ontario acute care
  hospitals in order to improve the accuracy of
  health information coding.
• MoHLTC to review the Public Hospitals Act and
  ensure that all relevant regulations are
  implemented.
• College of Physicians and Surgeons of Ontario to
  lobby medical schools to teach the importance of
  high quality and timely clinical documentation
  in student training curriculum.
• CIHI to ensure that coding standards allow for
  capture of diagnosis based on documentation from
  all members of patients clinical team.

17
Recommendation 3

• Ensure Coding Expertise
• There is a need for continuing education on
  coding standards and data quality for health
  information professionals. It was noted there
  are varying levels of experience and expertise
  among Ontario Health Information Professionals.
• Ensure that the infrastructures and funding are
  in place for staff education in particular in
  the areas of Diagnosis Typing and CIHI Standards.
• Encourage Hospital Administrators promote
  continuing education for health
  record/information professionals.

18
Recommendation 4
Understand the Necessity of Data Quality
Accountability

• Clarify accountabilities for data quality amongst
  appropriate stakeholders.
• MoHLTC lead development of a data quality
  strategy including continued re-abstraction
  studies.
• MoHLTC investigate including provisions for data
  quality in Hospital Services Agreements.
• CIHI to consider findings of this study during
  redevelopment of CMG grouper.
• Hospitals to ensure that staff involved in coding
  health records are certified HIMPs and enforce
  continued professional education workshops for
  HIMPs
• Hospitals to be held accountable to MoHLTC for
  quality of the data they submit to CIHI

19
Data Quality Components

• Relevancy - data are relevant to purpose
• Timeliness - data are available when needed
• Currency - some data are obsolete after a period
  of time
• Consistency - data quality needs to be consistent
• Accuracy - data should be error-free

20
Elements of Data Quality Program

• Forum for discussion of coders challenges -
  meetings, e-mails, bulletin board.
• C.I.H.I.s coding standards - make sure that all
  coders access these, understand them and apply
  them - consider coder sign-off.
• C.I.H.I.s coders queries database - the best
  way to ensure that all coders are aware of the
  latest and greatest.
• Inter-rater reliability - this is the
  cornerstone, chief component, big cheese

21
D.Q. Related Issues and Protocols

• Policies and procedures - do we have them all?
• Does everybody know about them?
• Are they current?
• Do we follow them?
• Productivity
• Record completion - the nightmare returns!
• Coder education - students for hands-on training
  CIHI educational sessions. Everyone should go?
  How is that information shared with others? How
  do we leverage the emergent expertise?

22
Current National And Provincial D.Q. Initiatives

• CIHIs National Grouper Redevelopment T.T.
• CIHIs National D.Q. Steering Committee
• Provincial Ministries D.Q. Initiatives
• CHIMA National D.Q. Strategy
• Provincial H.I.M. D.Q. Strategies (ie) CQI
  Network
• Ontario D.Q. Task Force Report/Recommendations

23
Next Steps For Chima

• Develop a National Implementation Plan for the
  (4) D.Q. Recommendations
• Establish a forum for development and
  implementation of a national data quality
  strategy (discussions at CHIMA conference in June
  04)
• Finalize the Alberta Data Quality Study Report
  and Recommendations
• Secure Funding for a National D.Q. Strategy
• Work with Provincial Health Record Associations
  to ensure Provincial D.Q. Strategies