Methodologies for

1

• Methodologies for
• District Surveys of health status

P Jha, DK Sikri, R Laxminarayan, S Rao-Seshadri,
D Jamison, GA Alleyne, R Kumar, S Darley, J
Chow, A Kurpad, N Dhingra, R Jotkar, S Jha, R
Poornalingam, A Singh, L Menezes, P Brown, S
Chandra, R Glass, P Laxmi on behalf of the DCP-2
India Study Team www.cghr.org prabhat.jha_at_utoron
to.ca
2

• Outline
• Rationale of District Indicators
• Background of District survey
• Consultative Process for District survey
  conforming to DISHA
• Output of Consultative Process
• DISHA for reducing MMR
• DISHA for reducing U5MR, especially that of
  female
• DISHA for reducing Premature Adult mortality
  (16-59 yrs)
• DISHA for Disease specific reform
• DISHA for System wide reform
• Indicators for outcome, service use, household
  determinants, financial risk protection, public
  responsiveness
• Revised Survey Tool envisaging the indicators
  conforming to the DISHA
• Possible Future course

3

• District Level Health Indicators
• Rationale
• For monitoring district wise outcome and coverage
  of various health programs by Central/State
  Governments. Pre-requisite in the era of
  de-centralized planning.
• Honble Prime Ministers desire to undertake and
  publish District Level Annual Health Indicators
  regularly and compare against benchmarks.
• ORGI plan to cover all Indian districts (593) in
  3 phases in order to restrict effective sample
  size to 39000 population or 7200 eligible women
  or 7000 households per district. Thus each
  district would be covered once in 3 years.

4

• Sample Size Criteria
• The sample size would depend upon
• Choice of the decisive indicator
• Birth Rate/Death Rate/Infant Mortality Rate
• Permissible Level of Error
• 5/10/15 percentage relative standard error
• Level of Aggregation
• National/State/District
• Periodicity and Reference Period
• 1/2/3 years

5
Sample Size (Population) at varying rates of
Infant Mortality Rates and Birth Rates at 10
percent relative standard error (prse)
Population in 000
6

• Periodicity of events Vs
• Population covered
• (Level of error 10 prse of IMR)

Note Sample size estimated for a district with
birth rate of 25 and infant mortality rate of 60.
7
Expected District Indicator Outcome

• -

8
Expected District Indicator Output

• -

9
Expected District Indicator Output

• -

10
Expected District Indicator Output

• -

11
Expected Health determinant inOutcome Indicator

• -

12
Possible Concerns regarding District Health
Indicators

• Experience with SRS health check up reveals
  under-reporting of adult deaths by about 13 and
  it was marginally higher for female adults. Newer
  SRS sample frame of 2001 might resolve this
  partly.
• Exclusion of death of in-migrated could be
  feasible option in view of earlier experience.
• Seasonal Out migration of impoverished people
  could reduce validity of adult deaths.
• Matching of response rate of tobacco use in
  Specific Fertility and Mortality Survey (SFMS)
  with NFHS revealed that SFMS underreports the
  risk factor.
• Matching of response rate of alcohol use in
  Specific Fertility and Mortality Survey (SFMS)
  with NFHS revealed high correlation of the risk
  factor (r .0.92)

13
Deliverables within 5 years

• GOAL Increase use of effective health services
  (by year 3) and show reductions in mortality (by
  year 5)
• Prioritization and focus of 5-10 districts on
  most effective interventions
• Increased allocation of state budgets for health
  in intervention states
• Accelerated reform (focusing on actions against
  specific constraints) in intervention states
• Training materials and critical mass of DCP-2
  implementers in place in various states and
  levels
• Rigorous evaluation of new interventions measles
  plus antigen program maternal mortality
  polypill for CVD
• Quantity of cost information dramatically
  improves
• Put in place simple monitoring system with RGI
• Report cards for each Member of Parliament
  produced

Notes Logical framework has been developed
with indicators for specific diseases
14
Thanks