Extrapolation of data from key population surveys and progr - PowerPoint Presentation

Slide1

Extrapolation of data from key population surveys and programs

Jessie K. Edwards

University of North Carolina at Chapel Hilljessedwards@unc.edu

1Slide2

PerspectiveTo improve public health, we must make a series of urgent decisions.

Resource allocationIntervention strategiesPrioritization Sometimes, we must act without perfect data.

2Slide3

PerspectiveBut, we rarely act from a position of complete ignorance.

To make decisions, we (often implicitly) summarize any existing information and our remaining uncertainty. This talk walks through some approaches to formalize this process in a specific setting.

3Slide4

Our SettingDecision makers often need

national estimates of indicators related to key populationsFor advocacy purposesTo set national targets

To inform funding allocation by international bodies4Slide5

What do we want to know?

Size estimates of key populationsFemale sex workersMen who have sex with menTransgender peoplePrevalence of HIV and other diseases

Distribution of risk factors for disease transmissionProgram coverage5Slide6

ChallengesUsually, information to describe a national HIV epidemic has gaps.

Subgroups of the population may be underrepresented

Geographic areas of the country may be excluded6Slide7

We need a principled approach to come up with national estimates when existing data are

incomplete.7Slide8

Ongoing efforts

WHO / UNAIDS meeting in March on strategic indicators for key pops MeSH

co-convened, UNC, LSHTM, JHSPH participants

8Slide9

Illustrative exampleWhat is the size of the population of female sex workers in the Dominican Republic?

9Slide10

Overview

Need for a national estimate: Specifically sizes of key populations, including female sex workersFunders request this information

National estimates guide MOH prioritiesInventory of existing data: Size estimates from 37 of 154 municipalities collected in priority provinces using a venue-based sampling approachStatistical approaches to compute national estimate: The focus of this talk.

Consensus building + decision making:

Project is ongoing – due to data security concerns, this example is illustrative only (data are hypothetical)

10Slide11

Parameter of interest

11

In simulations, I set

 Slide12

Missing data problem

But we only have measurements in a subset of the geographic areas from programmatic data collection in 2014 (let’s denote these regions by

)

12Slide13

Missing data problem

Our challenge is to use data from the sampled areas to make inference to the entire country

13Slide14

Approach

Original data:

Because data were collected for programmatic purses,

14

In simulated example,

 Slide15

Approach

Original data:

Expect that

, but did not collect data in all strata of

.

15

In simulated example,

 Slide16

Approach

Original data + targeted additional data collection

Expect that

,

and

16Slide17

Obtaining information on

The programmatic data collection activities only collected data on .But contextual data (

) can be found other placesCensus informationGeographic databasesOther population health surveys

DHS

17Slide18

Covariates in

for DR example 

From national data (publically available online): Age distributionMale/female ratioEmployment dataPopulation densityPovertyCountry of origin

From DHSAverage number of years of education among womenHIV prevalence

Adolescent pregnancy indicatorsFrom other sources (directly from stakeholders)

Indicator of tourist areaIndicator that the municipality was on a major transit corridor

18Slide19

Challenges using population-based surveysChallenges with DHS data

Indicators are generalizable to regional level, but we needed municipal estimatesRandom displacement of GPS coordinates for clusters19Slide20

Summarizing DHS data

Create

“surface” by interpolating values of between DHS clusters using a fine grid

Summarize values of by municipality

20Slide21

Summarizing DHS data

Create

“surface” by interpolating values of

between DHS clusters using a fine grid

Summarize values of

by municipality

21Slide22

Now a standard missing data problem

We have data on for 50 municipalities

We have data on for all 154 municipalitiesCould use any computation approach for estimating a mean in the presence of missing data. Up-weight existing data to represent distribution of

in entire countryPredict estimates in municipalities without data from a regression model containing

22

Here, we use the augmented inverse probability weighted (AIPW) estimator for

.

 Slide23

The AIPW estimator

The AIPW is a consistent,

semiparametric efficient, and doubly robust, estimator of

23Slide24

AIPW estimator is consistent, semiparametric efficient, and doubly robust

What does that mean and why should we care?Takes covariates

into accountSmaller variance than traditional weighting approachesTwo chances to correctly specify parametric modelsModel for inclusion in data collection (like weighting approaches) AND model for the outcome of interest (like regression-prediction approaches)In weighting only or regression model only approaches, if the parametric model is

misspecified, results are biased.Here, if

either of these models is correct, estimate of

will be correct

24Slide25

Data use/consensus building

In this example, stakeholders convened a workshop after the activity to understand the results and Suggest sensitivity analysesWhat if a different set of variables were included in

What if variables in were measured differently?What if variables in were modeled differently?

Discuss areas of uncertainty

25Slide26

Loose ends

Quantifying uncertaintyHow sure are we that assumptions hold?How much faith do we have in existing estimates of

How well is captured?Are we sure that we have measured all predictors of that differ between municipalities with and without data?Crucial for making decisions (i.e., maximizing expected utility)Best incorporated in a Bayesian framework

26Slide27

Loose ends

What if we had not been able to collect additional data?That is, what if we had had NO DATA within key strata of

?e.g., What if all data had been collected in urban areas?We could augment our existing knowledge/information using Parametric extrapolationBayesian methods

27Slide28

DiscussionDecisions can’t always wait for perfect and complete data.

We can use modern statistical tools to produce useful results from incomplete data.Inference may be updated as more information arises.

28Slide29

This presentation was produced with

partial support from the

United States Agency for International Development (USAID) under the terms of MEASURE Evaluation cooperative agreement AID-OAA-L-14-00004. MEASURE Evaluation is implemented by the Carolina Population Center, University of North Carolina at Chapel Hill in partnership with ICF International; John Snow, Inc.; Management Sciences for Health; Palladium; and Tulane University. Views expressed are not necessarily those of USAID or the United States government .

www.measureevaluation.orgSlide30

Extrapolation of data from key population surveys and programs

Jessie K. Edwards

University of North Carolina at Chapel Hilljessedwards@unc.edu

30