BILIS

An unified indicators database

Raphael Saldanha

LIS/ICICT/Fiocruz

2023-12-18

BILIS

  • Banco de Indicadores do Laboratório de Informação em Saúde, ICICT
  • A proposal to create an unified health indicators database

Motivation

  • Health indicators are necessary at several projects

  • Projects usually hires people to compute the indicators

  • Several projects uses a common group of basic indicators

  • The indicators may be computed with slight differences in methodology

  • Dependent on project lifecycle

Objectives

  • Provide a set of health indicators

  • Standardized methodology

  • Shared budget

  • Long term, independent initiative

Solutions

  • Primarly use already existing sources of secondary data

    • Death records, new born records, hospitalization records, etc

  • Easy to use R packages to compute health indicators

  • One package, one specific mission

Packages

  • brpop: yearly population estimates for Brazilian administrative levels, including age groups and sex (for standardized indicators)

  • rpcdas: query SIM, SINASC, SIH and other health information systems stored at PCDaS ElasticSearch stack, returning aggregated data per administrative level and time

  • recbilis: query other health information systems stored at ICICT Postgres instance

Packages

  • bilis: compute health indicators at several combinations of administrative levels (municipality, health region and state) and time units (weekly, monthly and yearly)

Overview

flowchart LR
sources_1(SIM, SINASC, SIH) --> pcdas_api(PCDaS API)
pcdas_api --> rpcdas[rpcdas]
sources_2(SINAN, \n Sivep-Malária) --> postgres(PostgreSQL Server)
postgres --> recbilis[recbilis]
pop(Population count) --> brpop[brpop]
rpcdas --> bilis[bilis]
recbilis --> bilis
brpop --> bilis
bilis --> csv(CSV files)
bilis --> parquet(Parquet files)
bilis --> db(Databases and destinations)

Initial results

  • 21 indicators based on RIPSA formulations

  • Crude rates

  • Age standardized rates is the next step

Thanks!