Latin American System Fragmentation by Country

Appendix 2 for “Real-World Data in Healthcare Technology Incorporation: A Strategic Opportunity for Latin America” 

Perhaps the most underestimated regional aspect of real-world evidence (RWE) in Latin America is that the region's systems are pluralistic and fragmented. Beyond availability of data, efforts to promote the use of RWE must also contend with biases that can result from institutional or epidemiological fragmentation and are difficult to adjust using standard statistical methods.

Brazil

In Brazil, the coexistence of the Unified Health System (SUS) and supplementary health plans—with approximately 25% of the population covered by private plans—produces patient flows, databases and coverage logics that operate in parallel. Mexico operates with the Mexican Institute of Social Security (IMSS), the Institute for Social Security and Services for State Workers (ISSSTE), IMSS-Bienestar, and the private sector. Colombia operates under the contributory and subsidized regime. This systemic fragmentation has been identified by ISPOR as a structural challenge for health technology assessment (HTA) in pluralistic systems, which requires methodological adaptations that frameworks originally designed for integrated systems do not consider. 

Peru

In Peru, system fragmentation is expressed in the coexistence of various health financing entities (such as SIS, EsSalud, the Armed Forces and National Police funds, and private insurers) and multiple networks of health service providers (managed by MINSA, regional governments, EsSalud, the armed forces, and the private sector). Each subsystem generates its own data, coverage standards, and care pathways. Therefore, an RWE study produced in a single subsystem may be valid for that population, but not necessarily transportable to the country. The Peruvian agenda must prioritize interoperability, population representativeness and intra-country external validity. 

Argentina

In Argentina, fragmentation takes on a particularly complex configuration due to the coexistence of 3 large subsectors with little interoperability: the public subsector, social security (national and provincial health insurance plans, including PAMI) and the prepaid medicine subsector. Each one has different financing mechanisms and coverage criteria, to which is added a strong provincial decentralization in health management. As a result, RWE generation usually occurs in highly segmented environments—for example, referral hospitals, specific funders, or highly complex private centers—and cannot always be extrapolated to the system as a whole. This creates a clear inefficiency in the use of resources, both budgetary and human and operational.

This high fragmentation, coupled with the low interoperability between the information systems of providers, funders, and jurisdictions, reduces the system’s capacity to produce integrated evidence and generates cumulative inefficiencies in the use of resources: duplication of diagnostic studies, repetition of administrative processes, unnecessary consumption of inputs, overload of human resources, and loss of care time. In this context, interoperability is not only a technical condition for improving the quality of RWE, but also a condition of efficiency for healthcare management. 

Central America and the Caribbean

In Central America and the Caribbean, systemic fragmentation acquires different configurations, but converges on the same methodological problem: the difficulty of constructing valid counterfactuals for causal inference from observational data.

In Panama and the Dominican Republic, the coexistence of multiple subsystems with ministries of health, public and private insurers, and heterogeneous provision networks, produces care trajectories and patterns of use that are not directly comparable to each other. This limits the possibility of defining control groups that adequately represent the target population of a coverage decision. Even though initiatives such as the functional integration of services or the institutionalization of EVTESA move toward a more ordered system, data production continues to be segmented by funder and level of care, resulting in observable populations defined more by administrative structures than homogeneous clinical definitions.

In contrast, Costa Rica, with a more integrated system under the Costa Rican Social Security Fund, partially reduces this problem by having more complete longitudinal trajectories within the same insured population. However, even in this case, the external validity of the estimates faces limits when policy decisions extend beyond the subgroups effectively represented in the records.

Taken together, these cases illustrate that the main challenge of RWE in the subregion is not only the availability of data, but the construction of credible counterfactuals in contexts where institutional or epidemiological fragmentation introduces structural biases that are difficult to adjust using standard statistical methods. As a consequence, the methodological agenda for RWE in Central America must emphasize not only interoperability and expansion of databases, but also the development of explicit approaches to transportability and emulation of assays that allow for the comparison of populations that, in practice, are generated by health systems that operate under different logics.