Brazil’s Atlas of Variation in Healthcare: Is Geography Destiny?

Published Aug 16, 2022

Marcia Makdisse, MD, PhD, MSc in Care Transformation, Academia VBHC, Sao Paulo, Brazil.

Henrique Diegoli, MD, MSc, Academia VBHC, Sao Paulo, Brazil.

Pedro Magalhaes, MD, Academia VBHC, Sao Paulo, Brazil.

Muir Gray, MD, DSc, Academia VBHC & Oxford Value and Stewardship Programme, Oxford, United Kingdom

Variation in healthcare is one of the biggest challenges faced by healthcare organizations around the globe and a major driver of wastage and poor health outcomes. Distinguishing between warranted and unwarranted variation is paramount to guide improvements in the quality of care delivery. Warranted variation reflects the accurate application of different levels of resources because of different levels of need, or because an explicit decision has been made relating the allocation of resources to meet the particular needs and priorities of a certain population. Unwarranted variation, on the other hand, is the variation that cannot be explained by discrepancy in need or patient preference, as defined by Professor John Wennberg, creator of the Darmouth Atlas of Variation published in 19961. Much of the variation seen when comparing different populations is unwarranted and it reveals two other major problems: overuse of low-value interventions and underuse of high-value interventions.

The Darmouth Atlas of Variation was the first of a series of atlases that have contributed to accelerate the understanding of the causes of variation and led to improvements in healthcare in many developed countries. Brazil’s Atlas of Variation, launched in April of 2022 in Luzern, Switzerland,  during the Wennberg International Collaborative & Swiss National Science Foundation Spring Synthesis Conference 2022, is the first to be published in a middle-income country, as part of the Wennberg International Collaborative2.

Brazil’s Atlas comprises 30 maps built upon official data from different information systems linked to the DATASUS, Brazil’s public and open-access medical claims database, from 2016 to 2019, therefore, the analyses eliminated the impact of the COVID-19 pandemic on care variation in healthcare, establishing a reference period for future analyses4.

Substantial variation was seen across all geographic regions. Among health regions, for example, C-section rates ranged from 19.5% to 84.0% (a 4.3-fold variation), all above the 10-15% rates recommended by the World Health Organization, suggesting widespread overutilization.5 The proportion of strokes treated with intravenous thrombolysis, on the other hand, ranged from 0% to 33.1% and in-hospital stroke case-fatality ranged from 5% to 39% (a 7.9-fold variation).

The magnitude of variation seen is unlikely to be caused primarily by differences in patient needs and preferences and further analyses are needed to identify causes, gaps in care and challenges around care variation.

  1. Wennberg JE. Unwarranted variations in healthcare delivery: implications for academic medical centres. BMJ. 2002 Oct 26;325(7370):961-4. doi: 10.1136/bmj.325.7370.961. PMID: 12399352; PMCID: PMC1124450.
  3. Ministério da Saúde. DATASUS.
  4. Diegoli H, Makdisse M, Magalhaes P, Safanelli J, Gray JAM, em nome da Academia VBHC. Atlas de Variação em Saúde Brasil. S√£o Paulo, Brasil, 2022.  (English version); (Portuguese version).
  5. WHO Statement on Caesarean Section, 2015.
Your browser is out-of-date

ISPOR recommends that you update your browser for more security, speed and the best experience on Update my browser now