Dengue Cost of Illness Studies: Status Report

1
Dengue Cost of Illness Studies Status Report
Presented at Spring Meeting of Board of
Counselors  Donald S. Shepard, Ph.D. Jose
Suaya, M.D., Ph.D. Mariana Caram, M.A. Site
colleagues   Program Management Team, Dengue
Burden Studies Schneider Institute for Health
Policy, Heller School, MS 035 Brandeis
University, Waltham, MA 02454-9110 USA Tel
781-736-3975 Fax [phone] Web
http//www.sihp.brandeis.edu/shepard E-mail
Shepard_at_Brandeis.edu April 25-26, 2006, Seoul,
Korea
2
Site Colleagues

• Prof. Lucy Lum Chai See, MD
• Malaysia
• Dr. Sukhontha Kongsin, PhD
• Thailand

3
Overview

• Objectives, Background
• Methods
• Preliminary Results
• Conclusions and Next Steps

4
Objectives

• Examine economic burden of dengue on households,
  governments, and private sector (employers,
  insurers).
• Build capacity for health services and policy
  research on dengue by involving policy makers and
  practitioners from many countries.

5
Background

• 2001 Ho Chi Minh City, Vietnam stakeholder
  dengue meeting included economic topics.
• 2002 Burden of Illness (BOI) included in PDVIs
  grant application to Gates Foundation and
  countries invited to submit concept papers.
• 2003 PDVI issued a Request for Proposals for
  the BOI Program Management Team (PMT) and
  selected Brandeis University
• 2004 PMT recommended and PDVI approved and
  funded studies in 8 countries.
• 2005 Data collection started.
• 2006 Data are being cleaned and analyzed.
  Reports of results and international workshops
  are being planned.

6
Definitions

• Cost of illness Economic cost imposed by a
  disease due to prevention, treatment, and lost
  economic output (major focus of work to date).
  Key elements
• Cost to health sector and lost economic output
• Costs of treatment and prevention (vector
  control)
• Treatment costs calculated per case and aggregate
  to the country
• Age breakdown of treatment costs children and
  adults
• Burden of illness Loss of good health, generally
  measured in DALYs, due to morbidity and mortality
  imposed by the disease (minor focus of work to
  date).

7
Overview

• Objectives
• Methods
• Preliminary Results
• Conclusions and Next Steps

8
Methods and Data

• Conceptual frameworks National Health Accounts
  (NHA) and Cost of Illness (COI).
• Data collection and analysis
• Collaborating institutions (research and patient
  care), subjects, and labs.

9
Frameworks NHA

• Describes the spending on health for a country or
  region.
• Uses a consistent set of definitions limited to
  the organized health sector to compare data
  across countries
• Organizes results in a matrix structure that
  classifies financing agent (e.g. government or
  out-of-pocket) against type of care.

10
Disease-Specific NHA

• Uses
• Shows relative importance of that disease.
• Informs about treatment choices and settings.
• Informs about prevention choices including the
  development of new approaches (e.g. vaccine).
• Challenge data needs
• Consistent definitions
• Cost per treatment episode or per year by stage
  or type
• Incidence (acute conditions or complications) or
  prevalence (chronic conditions), often by stage
  or type
• Current and potential prevention options, costs,
  and effectiveness

11
Frameworks COI

• Describes the overall economic cost of a disease
  of a country or region.
• Determines direct and indirect costs.
• Direct costs are economic costs within the health
  sector for prevention and treatment.
• Indirect costs are economic costs of lost time,
  lost quality of life, and lost wages associated
  with the disease.
• Based on welfare economics collateral treatment
  expenses economic losses.

12
Frameworks combined
13
Data collection

• Subjects (or caretakers) interviewed once or
  twice about illness, treatment, and caretaker
  visits.
• Medical records abstracted.
• Lab data collected from records or tests during
  the study.
• Information entered into customized Microsoft
  Access database.
• Data converted into SPSS, cleaned, aggregated by
  subject.

14
Data analysis

• Key formula
• Total cost of a resource Quantity x Unit cost
• Quantity and unit cost can be obtained from
  separate data sources.
• Example 1 Hospital costs
• Hospitalization cost length of stay x cost per
  day
• Example 2 Lost time from self-employment or
  household activities
• Cost of lost time days lost x value per day

15
Collaborating Institutions 1
16
Collaborating Institutions 2
17
Types of Facilities
18
Number of Participants(as of 4/20/06)
19
Serology Gates Foundation Question

• Asked at midterm review What is status of
  serology testing?
• Studies were designed to follow current
  laboratory test practices at study sites, so our
  samples and costs are representative of the
  current laboratory capacity in the countries
  studied.
• Tests were performed at national or regional
  reference laboratories.
• Depending on the timing and number of blood
  samples, laboratories performed serology tests,
  viral isolation, and/or PCR.
• Fraction of patients receiving serology tests
  varied with care patterns in the countries.
• In six countries, all patients tested.
• In three countries, about half tested.

20
Serology Results
21
Overview

• Objectives
• Methods
• Preliminary Results
• Conclusions and Next Steps

22
Case Study Countries
23
Direct cost per hospitalized case
24
Cost per hospitalized case, US

• Data collected by Prof. Lucy Lum from 654
  dengue hospitalizations in 2 public and 2 private
  hospitals around Kuala Lumpur in 2003.

25
Financing of a dengue hospitalized case, US
26
Financing of dengue care by level and setting
27
Preliminary estimate of national cost

• Needs national estimate of cost per case (assumed
  data from Malaysia were representative)
• Needs national number of cases (derived for
  hospital cases from expansion factor)

28
Expansion
FactorRelation to Notification Rate
29
Projected number of hospitalized dengue cases in
Malaysia
30
Aggregate NHA cost of dengue cases in Malaysia
for hospitalization (US million)
31
NHA Overall cost of dengue in Malaysia

• Preliminary total 12.79 million
• Excludes ambulatory-only cases (not yet analyzed)
• Equivalent to 940,000 workdays of output

32
Derivation of indirect costs
33
COI in Malaysia other costs
34
Quantities of key activities, Thailand
35
Estimated cost per hospitalized case
36
Overview

• Objectives
• Methods
• Preliminary Results
• Conclusions and Next Steps

37
Country conclusions

• The cost of dengue treatment per hospitalized
  case in Malaysia (718) is equivalent to 53 days
  of lost output.
• These treatment costs are financed by government
  (70), employers and insurers (13), and
  households (17).
• Development and adoption of a dengue vaccine
  would offset 87 of treatment costs for all of
  Southeast Asia and may permit reductions in
  spending on vector control.
• The National Health Accounts analysis shows the
  value of investing in a dengue vaccine to
  multiple stakeholders.

38
Project status conclusions

• As most investigators were new to health services
  research, we are proud of their work.
• Data should allow a good estimate of dengue cost
  per case at each site.
• Burden of Illness Studies are possible to conduct
  in several countries.
• Existing data should support estimates of
  population-based costs in two countries (Brazil
  and Malaysia).
• Numerous extensions under discussion with PDVI to
  strengthen population-based studies of treatment
  and costs of vector control.