Improving child health globally through evidence-based programs George

1
Improving child health globally through
evidence-based programsGeorge Diana Sharpe
Perinatal LectureshipUniversity of Texas at
AustinSchool of Nursing

• Kirk Dearden 27 February 2009

2
Why the next 40 minutes matter

• Scope of the problem is large
• 27,000 children lt 5 y die every day
• 16,000 of those deaths have malnutrition as an
  underlying cause
• Solutions to the problem are well-known
• We are most effective in improving child survival
  when our efforts are evidence-based

3
Structure

• Evidence to improve programming
• Before programs begin
• During program implementation
• Innovative strategies to improve child survival
• After program completion

4
Our focus

• Who?
• Children lt 5 y old in developing countries
• Girls, marginalized, poverty-stricken
• What?
• Programs implemented by community-based NGOs
• Less on MOH activities
• Impact?
• On health systems, policies and populations

5
Our focus

• What evidence?
• Before programs begin
• Epidemiological evidence
• Formative research (usually qualitative)
• During program implementation
• Monitoring and evaluation
• After programs end
• Rigorous assessments of program impact (usually
  quantitative)

6
Speaking of evidence

• A quiz to start things off!
• Which country in each pair has twice the
  under-five mortality rate as the other?
• Ethiopia vs. Sierra Leone
• Mali vs. Benin
• Cambodia vs. Niger

7
Speaking of evidence

• Twice the under-five mortality rate
• Ethiopia vs. Sierra Leone
• Mali vs. Benin
• Cambodia vs. Niger

8
Speaking of evidence

• Twice the total number of deaths to children lt5 y
  old
• India vs. Nigeria
• Pakistan vs. Afghanistan
• Indonesia vs. Iraq

9
Speaking of evidence

• Twice the total number of deaths to children lt5 y
  old
• India vs. Nigeria
• Pakistan vs. Afghanistan
• Indonesia vs. Iraq

10
Our focus

• Quiz points to the need to understand where the
  numbers of deaths are greatest
• What evidence?
• Before programs begin
• Epidemiological evidence
• Formative research (usually qualitative)

11
The epidemiological evidence

• We MUST know the underlying epidemiology of child
  morbidity and mortality prior to initiating
  programs and policies designed to help children
  survive and thrive
• Sometimes we dont attempt to get this evidence
• Sometimes we are thorough in obtaining this
  evidence
• But evidence used poorly to champion a single,
  specific cause
• Or used effectively to bring attention to a
  neglected topic
• Deborah Maine/Allen Rosenfield Maternal
  mortality
• Lancet series on neo-natal survival

12
Epidemiology of child survival

• Who
• What
• When
• Where
• Why

13
Epidemiology of child survival

• Who
• 10 million children lt 5 y of age
• Poor and females at much greater risk

14
Epidemiology of child survival

• What
• Neonatal disorders 33
• Diarrhea 22
• Pneumonia 21
• Malaria 9
• Other causes 9
• AIDS 3
• Measles 1
• Undernutrition 60 of all deaths to children lt 5
  y old

15
Major causes of death, children lt 5 y
16
Epidemiology of child survival

• When
• 40 of all under-five deaths first 28 d of life
• 2/3rds of all IMR in first 28 d
• 2/3rds of all NMR in first week
• 2/3rds of all deaths in first week occur in the
  first d
• Most deterioration in nutritional status occurs
  in first 18 m of life

17
An example of using evidence effectively
Weight-for-age Z-score by age and region, Save
the Children, 1986-1997
18
Epidemiology of child survival

• Where
• Half of all deaths in just 6 countries
• India, Nigeria, China, Pakistan, DR Congo,
  Ethiopia
• 90 of all deaths to children lt 5 y old occur in
  42 countries

19
Epidemiology of child survival

• Where
• Half of all deaths India, Nigeria, China,
  Pakistan, DR Congo, Ethiopia
• 90 of all deaths to children lt 5 y old occur in
  42 countries

20
Epidemiology of child survival

• Why
• At a fundamental level, children die because
  those who have been entrusted to care for them
• Parents
• Family
• Health care providers
• Program planners and implementers
• Policy makers
• Donors
• Dont practice optimal behaviors

21
Epidemiology of child survival

• In most cases the technology to address these
  challenges exists
• What we dont know is why, for example
• Some mothers fail to exclusively breastfeed
• Some health care providers discourage exclusive
  breastfeeding
• Breastfeeding is not a priority for the MOH and
  for donors

22
The most effective preventive and treatment
services and their impact

• The following preventive interventions would do
  the most to reduce U5MR
• Breastfeeding 13
• Insecticide treated materials 7
• Complementary feeding 6
• Zinc 6
• Treatment interventions
• ORT 15
• Antibiotics for sepsis 6
• Antibiotics for pneumonia 6
• Antimalarials 5

23
What happens when we dont use epidemiological
evidence to guide programs?

• Fail to address the greatest causes of morbidity
  and mortality
• Ignore country- and region-specific disease
  patterns
• e.g., malaria in Africa
• Misdirect scarce resources

24
Our focus

• What evidence?
• Before programs begin
• Epidemiological evidence
• Formative research (usually qualitative)

25
What happens when we dont conduct formative
research?

• Programs targeted at the wrong populations
• No community buy-in
• Poor understanding of the facilitators and
  barriers to engaging in optimal behaviors
• Inappropriate/ineffective programs

26
An example of a program that didnt use formative
research

• Peru no needs assessment, no clear
  understanding of the underlying epidemiological
  profile
• Misunderstanding of what was needed
• Community largely uninvolved
• Focus on a very small town ? public health
  impact?
• Potential for public health impact is doubtful
  despite massive resources

27
An example of successful use of formative research

• Multivitamins for Women of Reproductive Age in
  Bolivia

28
Background

• Hypothesis social marketing improves womens
  awareness and consumption of multiple vitamin and
  mineral supplements, especially among low-income
  women
• Design formative research and baseline and final
  surveys

29
Intervention

• Formative research to inform micronutrient
  product and marketing strategy
• Product name
• Location of manufacture
• Appearance
• Cost
• Packaging
• Advertising including appropriate media

30
Intervention

• Commercial distributors and medical staff to work
  with doctors and pharmacists
• Department-wide events for govt, NGO and other
  leaders
• 6 months of media advertising
• Poster, dangling product shots
• 148,000 brochures distributed thru pharmacies
• 900 TV spots

31
Percent of Women Who Had Ever Taken Multiple
Supplements, by Years of Formal Schooling
32
Structure

• Evidence to improve programming
• During program implementation
• Innovative strategies to improve child survival

33
Structure

• Evidence to improve programming
• During program implementation
• Monitoring and evaluation critical
• Sometimes we incorrectly conclude that a program
  is ineffective when in fact, the program wasnt
  implemented as designed
• Example Positive Deviance Initiative in Vietnam

34
Structure

• Evidence to improve programming
• During program implementation
• Innovative strategies to improve child survival

35
Behavior change needs to occur at a variety of
levels
Advocacy for policy change
Policy
Training, advocacy
Health Care Providers - Services Provision
Groups to change community norms
Norms of Surrounding Society
Secondary Target Audience
Groups to create support
Primary Target Audience
Individual Behavior Change
36
Behavior Change Strategies

• Policy
• Advocacy including the use of data-driven models
  to inform decisions
• REDUCE Maternal health
• ALIVE Neonatal mortality
• PROFILES infant nutrition
• Simulated models to estimate the relative
  advantages of exclusive breastfeeding over
  replacement feeding and vice versa
• IMR lt 25/1000 live births exclusive replacement
  feeding

37
Behavior Change Strategies

• Health care providers
• Assessments of existing policies, health care
  provider knowledge
• Changes to national guidelines
• JHU reproductive health
• Pre-service and in-service reform and training
• Vietnam training in breastfeeding for clinicians

38
Behavior Change Strategies

• Norms of surrounding society
• Information, education and communications (IEC)
  strategies including social marketing
• VitalDía in Bolivia

39
Behavior Change Strategies

• Secondary target audience
• Inclusion of husbands and in-laws
• The Grandmother Project
• Positive Deviance
• Other efforts

40
Behavior Change Strategies

• Primary target audience
• 3 strategies
• Negotiation
• ORPA
• Positive Deviance
• ALL involve collection of data to inform
  programming

41
Negotiation

• Negotiation
• ASK
• RECOMMEND
• AGREE
• REMIND
• APPOINTMENT

42
Example of Negotiation reduction of indoor air
pollution

• ASK the mother about current use of the stove to
  identify any problems
• RECOMMEND options to the mother and help her to
  select one she can try
• AGREEMENT on a behavior that the mother will try
• REMIND mother of optimal practice and help
  overcome obstacles
• Make an APPOINTMENT for a follow-up visit

43
What might you recommend?

• Unblock/properly seal chimney
• Make sure door has hinges
• Repair holes and missing/broken plates
• Keep at least 2 windows/doors open during burning
• Open long enough to ventilate house
• Keep child away from stove/outside during
  ignition, morning hours, and burning
• Put out fire when burning is finished

44
ORPA

• Observe
• Reflect
• Personalize
• Act
• Case study from West Africa (feeding sick
  children)
• Feeding as much or more during and after illness
• Feeding patiently
• Feeding special foods (enriched broth, fish soup,
  mashed banana or other fruit)

45
How are Negotiation and ORPA different from
education?

• Give individuals options
• Individuals choose options that are most
  feasible/do-able given their own culture, social
  environment, etc.
• Put the health promoter and the individual on an
  equal footing
• Require two-way communication
• intense listening by the health promoter followed
  by tailor made recommendations
• Require reflection

46
PD/Hearth
47
PD/Hearth
48
PD/Hearth
49
PD/Hearth
50
PD in Vietnam

• Some children from poor houses well-nourished.
  How did they do it?
• Answers vary by setting but include crabs,
  shrimps and greens from rice paddies
• PD hearth involves
• Discovering local solutions (evidence-based)
• Sharing those solutions
• Designing hearth sessions for malnourished
  children
• 2 weeks, 6 days per week
• Parents of malnourished children practice the
  practice
• Example contributing a handful of PD foods as
  the price of admission to a hearth session
• PD/Hearth requires evidence anthropometry before
  and after 2-week session

51
PD/Hearth

• Turn to neighbor and identify one PD outcome and
  risk factor
• She/he does the same
• Outcome inner city youth who get a college
  education
• Risk poor schools in inner city (or parental
  disinterest or lack of resources or)
• Well share 2 or 3 examples in plenary

52
What are the benefits of applying a PD framework
to development?

• PD behaviors are affordable, acceptable, and
  sustainable
• already practiced by those at-risk, do not
  conflict with local culture, and they work
• PD introduces a generic model for local
  problem-solving
• PD provides solutions today to challenges that
  cannot await long-term development
• Focus on whats right not prescriptive, top
  down, or donor-driven
• Easier to sustain without on-going external
  resources

53
What are the limitations of using PD?

• limited generalizability of findings
• labor- and cost intensive
• potential for scale uncertain

54
What is the impact of PD?

• PD study in Vietnamdisappointing results.
  However
• Trinh MacIntosh study on sustainability was quite
  encouraging

55
Positive Deviance and Neonatal Health A Case
Study from Pakistan

• How do you find PDs?
• Situation analysis to discover norms
• Community/clinic investigation to find PDs
• Positive Deviant Inquiries uncommon behaviors
  among
• Surviving asphyxiated newborns
• Thriving LBW babies
• Surviving newborns who had danger signs
• Normal newborns

56
Marsh, Pakistan

• Both groups, weak practice of
• Clean delivery
• Thermal control
• Immediate/exclusive breastfeeding
• Fathers involvement

57
Marsh, Pakistan

• PD behaviors (Afghani refugees)
• Mother prepared own delivery kit
• Mother given diet of chicken and eggs
  before/after birth
• Mother-in-law washed hands with soap before and
  after cutting cord
• Room kept warm at all times
• Dai used mouth-to-nose resuscitation

58
Our focus

• Evidence to improve programming
• Before programs begin
• During program implementation
• Innovative strategies to improve child survival
• After program completion

59
A variety of sources that examine the impact of
specific interventions

• Lancet series on
• child survival
• maternal health
• neonatal health
• Adolescents
• Undernutrition
• Alma Ata, etc.
• Perry H, Freeman (2008). How effective is
  community-based primary health care in improving
  the health of children? a review of the evidence.
  Report to the Expert Review Panel, the World
  Health Organization, UNICEF, and the World Bank

60
UNICEF

• more than enough information to act

61
Community-based Primary Health Care

• any activity which directly or indirectly has a
  positive influence on health, and does not take
  place exclusively in a health center or hospital

62
A review of the evidence
63
Extensive evidence that interventions are
effective and should receive priority

• Immunizations for mothers and children (TT for
  mothers and measles for children
• Supplemental vitamin A
• Exclusive breastfeeding during the first 6 months
  of life and continued breastfeeding thereafter
• Hygiene, safe water, and sanitation
• Oral rehydration therapy and zinc supplementation
  for children with diarrhea
• Handwashing

64
Extensive evidence

• Clean deliveries when births are at home and
  where hygiene is poor
• Home-based neonatal care (immediate/exclusive
  breastfeeding, cleanliness and prevention of
  hypothermia)
• Community-based treatment of childhood pneumonia
• Insecticide-treated bednets
• Detection and treatment of syphilis in pregnant
  women, and
• Iodine supplementation

65
Efficacious interventions that need more
evaluation in routine settings

• Community-based treatment of malaria
• Community-based rehabilitation of malnourished
  children through Positive Deviance/Hearth or
  through ready-to-use dry therapeutic foods
• Prophylactic supplemental zinc
• Complementary feeding from 6-9 months of age
• Prenatal calcium for prevention of pre-eclampsia
  and eclampsia
• Intermittent preventive treatment of malaria
  during pregnancy
• Detection and treatment of asymptomatic
  bacteriuria

66
Need more evaluation

• Application of a topical antiseptic to the
  umbilical cord of neonates
• Skin cleansing of newborns with a topical
  antiseptic soon after birth
• Improved airway management and resuscitation in
  neonates by trained community health workers
• Detection and treatment of neonatal sepsis by
  trained community health workers
• Improved cooking stoves through improved stoves
  (to reduce childhood pneumonia)
• Participatory womens groups for empowerment and
  education about maternal and neonatal health
  issues
• Non-health interventions, including micro-credit
  and conditional cash transfers to women
• Improved socio-political environments which
  support maternal and child health and allow
  access to high-quality basic services

67
Do not appear to have a beneficial effect on the
health of children

• Supplementary feeding programs in non-emergency
  situations
• De-worming medication for children (on growth or
  on cognition/school performance)

68
Havent had sufficiently rigorous evaluations

• Growth monitoring?
• Antenatal care
• Large-scale integrated programs to reduce
  stunting and wasting
• Birthing homes

69
Adverse effects

• Iron supplementation in malaria-endemic areas
• Micronutrient mix of iron, other minerals
  including zinc, and riboflavin

70
Successful programswhat do they have in common?

• Perry and Freeman the most successful integrated
  programs with a sustained and documented impact
  on child health
• Jamkhed Comprehensive Health Project in Jamkhed,
  India
• SEARCH (Society for Education, Action and
  Research in Community Health) in Gadchiroli,
  India
• Matlab MCH-FP field site in Bangladesh, and
• Hospital Albert Schweitzer in Haiti
• Common characteristics
• in operation for 20-50 years
• published, documented mortality impacts, and
• BRAC also worthy of attention but no published
  mortality impact

71
Common characteristics
72
Whats missing

• Geographic
• Info on program effectiveness outside S Asia,
  especially Sub-saharan Africa
• Content
• Urban health
• Health systemsnot simply health programs
• Methods
• Formative research
• Small-scale research to test elements of
  successful program strategies
• Operations research neededeffective relative to
  what?
• All
• Honest assessment of what does and doesnt work
• Tendency toward PR means that there are few
  unsuccessful experiences documented

73
The way forward

• Program planners
• Implement effective packages first on a pilot
  basis then at scale
• Donors
• Divest of requirements that every projectlarge
  or smalltrack progress on a host of indicators
• Provide broader support to integrated packages
  described by Perry and Freeman

74
The way forward

• Program evaluators
• Rigorously assess packages to judge
  effectiveness/make adjustments to programs as the
  scale expands
• Develop innovative methods for assessing impact
• Assess packages of interventions in routine field
  settings at scale over long time periods
• Bhutta et al. (2005) reviewed 740 studies of the
  effectiveness of community-based interventions
  for improving perinatal and neonatal health
  outcomes
• only 10 carried out in routine field settings
  that could be considered effectiveness trials
• Haws et al. (2007) looked at packages to improve
  neonatal health
• no studies at scale in routine settings

75
The way forward

• Program evaluators
• More info needed on program context and extent to
  which programs are implemented as planned
• PD in Vietnam is one example
• More cost-effectiveness studies
• More community empowerment studies
• More on service delivery mechanisms including
  Behavior Change Communications strategies
• Which approaches work best? In which contexts?

76
The way forward

• If we are to effectively address child survival,
  we need an evidence base
• Prior to beginning programs
• Must answer what, when, where, how and why?
• Must examine feasibility (formative research)
• During program implementation
• After program completion
• Rigorous testing of the most promising
  strategiesduring pilot phase and at scaleis
  absolutely essential
• No justification for allowing 10 million children
  to die every year
• Our obligation as practitioners of public health
  is to ensure that the programs and policies we
  implement do the most to help children survive
  and thrive

77
Thank you!