Getting suitable data range of data sources surveys: sampling strategies, questionnaires reporting systems: forms, outputs sentinel sites: clinics, programmes,

Getting suitable data

• range of data sources• surveys: sampling strategies, questionnaires• reporting systems: forms, outputs• sentinel sites: clinics, programmes, survey clusters

Ctown4.ppt

Data Sources in Relation to Characteristics of Required Information.

Source/Characteristics

Administrative Householdsurveys

(vary by size,q’aire, etc)

Rapid Assessment

(RAP etc)Ad hoc(e.g. clinics)

Sentinel site School(census)

Timeliness(delay approx.)

1-3 months 1 month 6-12 months 3-6 months Variable

Associations/causality

(+) + + +++ ++

Fineness ofgeographicaltargeting

Medium(e.g. district)

Medium(e.g. district)

High(e.g. village)

Low (e.g.province) to

medium

Low

Externalvalidity

Low Medium Medium High Low

Comprehens-iveness ofvariablescovered

Low Low Low Medium High

Data quality Low High Medium High High

Source: ICN documents, FAO/WHO, 1992

Ad hoc(e.g. clinics)

Surveys• sampling strategy – general (see next slide)

o multi-stage, PPS, known probability of individual selectiono small-scale: 30*30 (or 33*6, or …); segmentation vs spin-a-bottle

• match earlier surveyso sample same populationo match age-bandso keep same measures/questions (don’t change questionnaire lightly)o match seasons

DHS anthropometric questionnaire module

Source: DHS Kenya report 1994

DHS feeding practices questionnaire module

Source: DHS Kenya report 1994

Example of reference standardsSource: WHO ‘Measuring Change in Nutritional Status’ (1983)

FSAU nutritionquestion-naire module (1)

Source: FSAU

FSAU nutritionquestion-naire module(2)

Source: FSAU

UNICEF modelquestion-naire module

Source: UNICEF MICS Manual (1995)

UNICEF modelquestion-naire module

Source: UNICEF MICS Manual (1995)

Reporting systems (clinics, programmes)

• need to be useful at all levels• provide information on trends not levels• use all, or select by convenience• stepwise aggregation (district, province …)• preferably should have validation surveys/capacity

Sentinel sites (clinics, programmes)• same principles plus:• select sites usually for early change• focus on good data quality, training, data flow, supervision• use as signal of change in that area (but note not representative, by design)• capacity to follow up, validate, important

Sentinel sites – survey• resample same clusters at regular intervals• issue of if same hhds/kids, replacement, etc• otherwise much the same as for sentinel clinics etc• see examples of Zimbabwe, ALRMP-Kenya, Namibia plan.

Reporting form

Source: FSAU

Reporting form. Source: EOS, Ethiopia

FSAUIntegratedPhase Classification(IPC).

Source: FSAU, Tech Manual V 1, Table 1, May 2006

FSAUIPC:General Interpretation

Source: FSAU, Tech Manual V 1, Table 2, May 2006

FSAUIPC:Wasting

Source: FSAU, Tech Manual V 1, Table 4, May 2006

FSAUIPC:Responseframework

Source: FSAU, Tech Manual V 1, Table 16, May 2006

Equivalent wasting level

Uganda 10%Somalia 15%Ethiopia 20%Kenya & Sudan pastoralists 25%

Eh?

1.1 Malnutrition by areaWasting was highest in Mudzi (9%). A verification exercise using clinic data was done for Mudzi and there was an indication of sharp increase in malnutrition in January 2005. Results from the 2005 vulnerability assessments done in May 2005 revealed that Mudzi district was among the districts that were food insecure. Comparison with data collected in November 2004 shows that wasting rates are higher in all the 10 sites. This is an indication of worsening of nutritional situation as it is expected that nutrition should improve during this time (March) as people start eating food from their agricultural produce.

Comparison of wasting rates, Nov 2004 and Mar 2005

0123456789

10

% c

hild

ren

w asting Mar 2005 w asting Nov 2004

Source:Zimbabwe

Pilot Food and Nutrition Sentinel Site Surveillance Report

March 2005

Food and Nutrition Council in collaboration with Epidemiology Dept, Nutrition Unit, Ministry of Health and

Child Welfare

Region 2002 2003 2004 2005 2006Turkana Kaleng, Kibish, Lapur, Lokitaung

11%(9-13.3)

27.6%(23.8-29.8)

34.4%(31.3-37.4)

22.1%(18.5-26.2)

24%(20-27.9)

Kakuma, Oropoi, Lokichoggio

11.4%(9.4-13.7)

18.9%(15.8-21)

23.3%(20.7-26.2)

19.2%(15.8-23.1)

26.6%(22.4-30.7)

Kalanuk, Katilu 12.7%(10.6-15.1)

24%(21.2-27.1)

20.1%(17.6-22.9)

21.3%(18.8-24.1)

1.2%(17.3-25.1)

Loima, Turkwell 11.8%(9.8-14.4)

22.4%(19.7-25.3)

23.3%(20.7-26.2)

21.4%(18.8-24.1)

23.6%(19.6-27.7)

Lokichar, Lokori 19.4%(16.9-22.2)

32.8%(30-35.7)

25.5%(22.9-28.3)

25.9%(21.7-30)

Central, Kerio, Kalokol

21.3%(18.7-24.2)

37.3%(34.3-40.3)

25%(22.3-27.8)

26.6%(21.7-30)

Isiolo Merti, Sericho

15.6%(13.5-18)

28.5%(25.6-31.6)

Kina, Garbatulla, Oldoniyro and Central

13.2%(11.2-15.6)

Kwale 5.8%(4.48-7.28)

5.9%(4.6-7.4)

West Pokot 10.9%(9.1-13.1)

Makeuni 2.3%(1.4-3.7)

4%(2.8-5.4)

Taita Taveta Wundanvi Mwambi

3% (1.7-4.1)

Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov DecPastoralistsAgro-Pastoralists

= Hunger Season

= Post Rains/Harvest

= Moderate

Source: Small scale survey dataset SEMmrge10_21B.sav

Results of area-level surveys, Kenya

Source: Small scale survey dataset SEMmrge10_21B.sav

Source: CHANIS Report, Oct 2006

Source: CHANIS Report, Oct 2006

Source: FSAU Nutrition Update September 2006

IPC Survey Results, Sool Plateau, Somalia June 2006

Sentinel site surveillance results, Bakool, SomaliaAugust 2006

Source: FSAU Nutrition Update September 2006

Source: Small scale survey dataset SEMmrge10_21B.sav

Program data