Presentacin de PowerPoint

1
The Johns Hopkins Universitys 2007 European
Conference
ACG as a tool for primary care improvement in a
publich health system environment Antoni Arias,
M.D., PhD, MsC IASIST S.A.
Karlskrona, Landstinget Blekinge, Sweden
September 18th-19th, 2007
2
Our acknowledgments
This presentation has been prepared with the
collaboration of the Basque Health Services
Agency (Osakidetza)
ACG applications for clinical management
developed in collaboration with
3
Contents

• The health system in Spain
• The contribution of case-mix systems to health
  care. The Spanish situation.
• ACG in Spain
• Applications in Primary Care (PC)
• The Basque experience using ACG in a public PC
  network
• The use for clinical management
• Applications in Integrated Care. Preliminary
  results.
• Some difficulties and limitations
• Expectations

4
Who are we?
Healthcare Knowledge International
Companies of HKI

• CHKS, UK
• IASIST, Spain

Countries where we operate
Leader in health care information delivery in
Europe, with extensive and exhaustive databases
for benchmarking IASIST, distributor of ACG in
Spain and Portugal
5
The Health System in Spain

• Public health system providing universal health
  care coverage
• Highly decentralized political structure 17
  regional governments with power to establish
  policies and regulations

Financing from state taxes Purchasing 17
governments Provision mainly public through 17
regional agencies
6
Organization of health care delivery

• Splitted health care divisions
• Primary Health Care
• Specialized Care (Hospitals)
• Specific budgets and management structure
• Primary care geographic structure (Basic Health
  Areas)
• Each BHA is served by a Primary Care Team (PCT)
• Coordination between Primary Care and Specialized
  Care never achieved
• Current trends Integrated Health Care
• Unification of primary and specialized care
  organizations into a single management agency and
  budget

7
Introduction of case-mix measures in Spain
Minimum Basic Data Set
Data
Goverment decision
DRG Implementation
DRG Research
Hospitals
1980
2000
1990
ACG Patchy Implem.
ACG Research
Primary Care
Data
Minimum Basic Data Set
Goverment decision
8
Current ACG situation in Spain
BASQUE COUNTRY (2004) 2,1 M 100
ARAGÓN (2007) 1,2 M (100)
CATALONIA (2005) 1,2 M (17)
Pilot projects 2 milion pat.
9
The Basque experience

• Aim To set up ACG system as a tool to analyze
  morbidity of population
• Objectives
• To increase the knowledge of health care
  professionals about ACG and share with them the
  results
• To detect organizational factors that affect
  validity of ACG in real life practice
  environment
• To set up practical applications
• Up to now two years of validation and
  applicability studies (October, 2004-September,
  2006)

10
The Basque experience
Percentage of visits registered in the electronic
health record Osakidetza (2004-2007)
80,6
Abril 2007
2006
2005
2004
Common information system 3S-Osabide
11
The Basque experience

• Criteria of data quality to include a doctor into
  the study group
• Level of use of the health electronic record
  system (gt 4 diagnostic registration/patient)
• Diagnostic registration coded (lt 22 of
  diagnostics as literal quotations)
• Valid diagnostic coding (lt10 of errors in
  ICD-9-CM coding, based on a sample of 250
  patients for each doctor)

12
The Basque experience

• Population and medical units included into the
  study have substantially increased

13
The Basque experience
63
Distribution of patients by no. of ADG
25
12
0,2
14
The Basque experience
Distribution of patients by ACG, Osakidetza
2005-06
4100 2-3 ADG combined, age gt34
003Acute minor, age 6
15
The Basque experience

• Challenges for the near future
• To reduce the time span between data collection
  and delivery of results
• To achieve the inclusion of 70 of units into the
  study
• To deliver information to 100 of medical units
• To improve the coding process
• To improve the electronic health recording
  systems
• To develop applications of data for clinical
  management

16
ACG for clinical management Our experience

• Delivery of useful information to improve
  clinical care through performance comparison with
  peers
• Adressed to action

17
Information for clinical management
Levels of analysis
Dimensions of analysis
Demand

• Population coverage
• No. of patients/year
• Age and gender profile
• Analysis of overusers
• Attended non-assigned population

External standard
Organization
Primary Care Team (PCT)
Case-mix

• Relative morbidity index
• Frequency of ACG and EDC
• Resource utilization bands (RUB)

Professional
Efficiency

• Visits
• Pharmacy cost
• Referred specialist consultation
• Volum/cost lab tests
• Volum/cost X-ray
• Total cost

18
Information for clinical managementAn example
Complexity / Morbidity burden
Relative weight, by PCT
Lowest complexity Relative W 0,46 54 lower
morbidity burden
Highest complexity Relative W 1,74 74 higher
morbidity burden
19
Information for clinical managementAn example
Efficiency
Pharmacy cost x patient observed ( ) and
expected ( )
Efficiency Index 0,79 21 undercost 943.000
Efficiency Index 1,27 27 overcost 737.000
400
350
300
250
200
150
100
50
0
001
002
003
004
005
006
007
008
009
Estándar
182,58
291,57
274,75
212,19
337,71
289,03
328,99
287,14
196,36
270,49
Mean Cost ()
Mean cost () expected
231,02
271,59
293,94
243,63
296,59
295,57
258,10
280,21
241,01
270,49
Overcost or undercost, related to standard
Efficiency Index
1,07
0,97
0,87
1,14
0,98
1,27
1,02
0,81
0,79
Impact ()
510.658
943.068
280.254
481.278
715.386
121.540
736.869
144.487
281.209
20
Information for clinical managementAn example
Medical practice, clinical conditions (EDC)
PCT with lower prevalence Hypertensive patients
are grouped in several ACG with different
morbidity burden
CAR14 Hypertension, without severe complications
Patients, by acg
1.711 Patients
Prevalence x 1.000 Patients
PCT
Standard
PCT 104,1 Standard 122,9
18,7
8,8
12,0
13,6
15,6
Visits x patient (PCT Standard)
25,3
13,6
17,7
19,4
21,0
976
511
549
797
686
Pharmacy x patient (PCT Standard)
913
494
557
840
570
21
Information for clinical managementAn example
Medical practice, clinical conditions (EDC)
CAR14 Hypertension, without severe complications
Associated EDC
Hypertensive patients have other associated EDC
with impact in their morbidity Higher prevalence
of respiratory infections, musc.-esk. symptoms
and lipid disorders
Standard
Hypertensive patients are attended with less
visits (19 less) and higher pharmacy cost (5)
22
Integrated care ongoing research
Disease burden and coding relationship between
no. of diagnostics and disease burden
Adding hospital-based diagnostics increase the
disease burden (weight) of populations Increase
is lower as higher is the no. of diagnostics in
primary care
0,48
Mean weight
1,46
Primary Care
Primary Hospital Care
No. diagnostics/patient
23
Integrated care ongoing research
Total Cost Explanatory Power Analysis

• Goal to test wether acg can explain individual
  healthcare costs

All costs included Inpatient, outpatient,
Primary Health Care, Drugs, Skilled Nursing
Facilities
Data year 2005 Population 151.542 inhabitants
(Inner Area of Catalonia) Providers 10 Primary
Care 1 hospital Healthcare users 124.943 users
(82) Avg. Cost per inhabitant 643 Avg. Cost
per user 780 (Max 202.700 - Min 0,44)
24
Integrated care ongoing research
Total Cost Explanatory Power Analysis

• ACG explanatory power on individual patient costs
  range from 23 (costs untransformed nor
  untrimmed) to 52 (costs untransformed but
  trimmed)
• Age Sex highest explanatory power reaches 10
  (costs untransformed but trimmed)

25
Integrated care ongoing research
Pharmacy Cost Explanatory Power
(Cost in )
Analysed 72 of pharmacy cost, more
representative in primary care (78) than in
specialized care (45)
aWithout diagnostic data
bEstimated population
cExcept cost of antineoplastic drugs
dNo data available at the moment
26
Integrated care ongoing research
Pharmacy Cost Explanatory Power
Explanatory variables
Method used
Explanatory power
All
Inliers
Pharmacy Pcare ()

• Age sex
• ACG

• Multiple regression
• ANOVA

18,5 35,4
16,1 41,0
78 cost
() All explanatory variables analysed with
logarithmic transformation
27
Up to now

• Certainty about the potential of ACG to better
  know the subject of the health care business THE
  PATIENT
• The resulting information raises a lot of
  questions
• THE CHALLENGE to develop a way of analysis
  useful to answer those questions and take actions
  to improve
• The use of such information to improve practice
  and management is a learning process that needs
  to be worked out

28
Up to now

• Main focus of interest of our clients in ACG
• To measure differences in disease burden among
  populations attended by different PCT
• CASE MANAGEMENT to identify overusers (e.g.,
  visits) and their clinical profile.
• DISEASE MANAGEMENT to adjust performance
  indicators in relevant PC conditions (diabetes,
  hypertension, )
• OVERALL PERFORMANCE IMPROVEMENT introducing
  objectives in contracts between providers and
  purchasers (quality of data, efficiency
  indicators)

29
Difficulties and limitations

• No agreement of minimun data set on patients in
  primary health care at national level
• Critical issue to facilitate diagnostic
  registration by doctors in their clinics
• There is high variability in exhaustivity of
  diagnostics recording data will improve as we
  used it
• Quality of data is essential for research.
  However, it must NOT be a restrictive factor to
  use it for clinical management with no data use
  and feed-back , professionals will not care about
  ACG usefulness and will not register the
  necessary data

30
Expectations

• To get better recording of clinical data
• Which will be the added value of the RX-Model to
  the current ACG ?
• To link quality indicators to efficiency
  indicators
• To link specialized care data to primary care
  data to get an integrated appraisal of morbidity
  and health care use.
• To develop further the application to clinical
  management
• To apply ACG for case/disease management
• To extend ACG use to capitative payment or
  budgeting

31
Thank you very much! aarias_at_iasist.com www.iasist.
com