Introduction

Pediatric intensive care refers to a set of technologies and procedures delivered by highly specialized healthcare professionals, aimed at reducing mortality and morbidity among critically ill children and adolescents [1]. The effectiveness of these interventions, however, relies not only on clinical expertise but also on the proper organization of Pediatric Intensive Care Units (PICUs). To ensure optimal outcomes, PICUs must meet minimum standards of quality and technological capacity, and be strategically located and sufficiently available to address local healthcare needs [1–3].

In Brazil, the Brazilian Health Regulatory Agency (Anvisa) has defined minimum operational standards for all Intensive Care Units (ICUs), including pediatric units, to ensure safe and effective care delivery [4]. These units are designated as “critical areas intended for the hospitalization of seriously ill patients who require continuous specialized professional attention, specific materials, and technologies necessary for diagnosis, monitoring, and therapy.” Pediatric Intensive Care Units (PICUs) therefore play a crucial role in guaranteeing the right to health, as established by the Brazilian Federal Constitution of 1988 and Law No. 8,080 of September 19, 1990, which established the Unified Health System (SUS). Their importance is further highlighted in the context of Brazil’s epidemiological transition, which requires greater access to specialized pediatric care [5–8].

Specifically for PICUs, their characterization and operation have been regulated by specific laws and normative acts since 2000. These regulations stipulate that the calculation of the number of PICU beds must consider technical aspects such as: scientific evidence; clinical and therapeutic protocols related to care lines based on literature reviews; analysis of the structure and performance of the installed capacity of national hospitals; comparison between the actual situation of hospitals and ideal, recognized standards; parameter adjustments to be used in bed estimation, and the application of simulation models [9–10].

Despite regulations, access to pediatric intensive care in Brazil is difficult and complex. The national dimension of the country, socioeconomic and regional differences in health infrastructure contribute to an unequal distribution of services [2,11–14]. Locations with lower technological density often have difficulty providing highly complex care, resulting in patient transfers that can increase the risk of morbidity and mortality [15]. This study describes the national distribution and complexity of Pediatric Intensive Care Unit (PICU) beds in Brazil and examines their relationship with child mortality. Data were obtained from different official government sources. These databases together offer a comprehensive view of the Brazilian health system.

Materials and methods

Results

The analysis considered 5,570 municipalities across 450 health regions, encompassing a total population of 173,143,050 people. Of these, 19.8% were under 14 years old, with a mortality rate of 22.5 per 100,000 inhabitants during the study period. According to CNES data for July 2023, there were 708 service providers in the country (Table 1). These providers registered a total of 6,489 PICU beds, proportionally distributed by complexity level as follows: type I (16.6%), type II (64.7%), and type III (18.6%). Of these, 3,200 beds (approximately 50%) were indicated as being part of the SUS and were spread across 133 health regions in 303 municipalities. States such as São Paulo and Minas Gerais in the Southeast had the highest number of beds (both SUS and non-SUS), while smaller states like Roraima and Amapá had the lowest availability.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Distribution of PICU beds in Brazil, according to health regions, hospital typology, and sociodemographic and mortality characteristics.

https://doi.org/10.1371/journal.pone.0339186.t001

Table 2 presents the distribution of PICU beds in Brazil by region, hospital typology (PICU I, II, and III), and management type (SUS and non-SUS). The regional breakdown is as follows:

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Distribution of PICU beds in Brazil, according to regions, hospital typology, and management.

https://doi.org/10.1371/journal.pone.0339186.t002

• North Region – 493 PICU beds: 8.5% type I, 89.9% type II, and 1.6% type III. The region has a population of 17,354,884, with 25.2% under 14 years old. The mortality rate for this age group is 36.6 per 100,000 inhabitants.
• Northeast Region – 1,329 beds: 9.0% type I, 82.1% type II, and 9.0% type III. The population is 54,658,515, of which 21.1% are under 14. The under-14 mortality rate is 25.8.
• Southeast Region – 3,270 beds: 22.0% type I, 53.9% type II, and 24.0% type III. This region has 84,840,113 residents, with 18.0% under 14. The mortality rate for children under 14 is 18.8 per 100,000.
• South Region – 748 beds: 3.9% type I, 63.8% type II, and 32.4% type III. Total population is 29,937,706, with 18.5% under 14. Mortality rate for this group is 17.4 per 100,000.
• Central-West Region – 649 beds: 26.0% type I, 65.6% type II, and 8.3% type III. The population is 16,289,538, with 20.9% under 14 years. Mortality rate for this group is 25.1 per 100,000.

A multivariate analysis based on the number of pediatric intensive care beds and the mortality of children under 14 identified six groups (clusters), revealing important regional patterns and disparities (Table 3). These groups were defined based on the proportionality of registered PICU beds in each health region and observed child mortality rates. They are described as follows:

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Distribution of PICU beds in Brazil, according to health regions, hospital typology, and sociodemographic and mortality characteristics (multivariate analysis).

https://doi.org/10.1371/journal.pone.0339186.t003

• Cluster 1: Highest number of PICU beds (2,676) across 334 providers, mainly in the Southeast and Northeast. Predominance of type II beds (92.3%). Mortality rate under 14: 23.9 per 100,000.
• Cluster 2: 48 PICU beds from five providers, mostly type I (85.4%). Mortality rate under 14: 21.1 per 100,000.
• Cluster 3: 1,329 beds in 20 health regions, with 58% type II and 33.6% type I. Providers concentrated in São Paulo and the Federal District (53 and 14 of the 135 providers). Mortality rate under 14: 20.4 per 100,000.
• Cluster 4: Second-largest number of beds (2,160), with 35.8% type III. States of Rio de Janeiro and São Paulo host 164 of the 204 health providers in this group. Mortality rate under 14: 19.1 per 100,000.
• Cluster 5: 276 beds in 153 grouped health regions. High frequency of type III beds (76.4%), mostly in the South and Southeast. Lowest mortality rate under 14: 16.5 per 100,000.
• Cluster 6: Comprises health regions without any registered PICU beds. Mortality rate under 14: 25.0 per 100,000.

The spatial distribution of pediatric intensive care resources across Brazil is illustrated in Fig 1. The map on the left displays the absolute number of PICU beds per Health Region in 2023, evidencing a marked territorial imbalance, with large areas—especially in the North, Northeast, and parts of the Midwest—holding few or no beds, while higher concentrations are found in the Southeast and Southern regions. The map on the right presents the regional classification derived from the clustering analysis, which groups Health Regions according to their PICU availability and service characteristics. Regions classified in Cluster 1 occupy the largest national territory, whereas higher-performing clusters are restricted to a small number of Health Regions, primarily in more developed areas of the country.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Geographic distribution of Pediatric Intensive Care Unit (PICU) beds and service clusters across Brazilian health regions (2023).

Left: Number of PICU beds per health region according to the National Registry of Health Establishments (CNES), demonstrating substantial variation in availability. Right: Spatial distribution of hierarchical PICU service clusters identified through spatial analysis. Colors represent cluster levels (1–6).Together, the maps illustrate marked regional disparities in the organization, concentration, and territorial accessibility of PICU services in Brazil. Base map derived from public-domain cartographic data provided by the Brazilian Institute of Geography and Statistics (IBGE). No copyright restrictions apply.

https://doi.org/10.1371/journal.pone.0339186.g001

Discussion

The results of this study achieved the intended objective and corroborate other studies that have already demonstrated an unequal distribution of PICU beds across the evaluated health regions [12–15]. However, this study advanced the analysis through a multivariate approach that allowed for a deeper exploration of inequality patterns, revealing, for example, regions with a higher or lower proportion of PICU beds by typology. Moreover, although causal relationships cannot be inferred, the under-14 mortality rate observed in each cluster appears to be correlated in some way with the distribution of these beds.

One of the striking findings is that, of the approximately 6,489 PICU beds registered in the CNES database, only 50% are available to the SUS. According to official statistics, more than 80% of the Brazilian population depends exclusively on SUS services due to lack of access to private health plans or supplemental providers [19–21]. Another critical point is that 259 identified health regions—covering 2,896 municipalities and home to more than 53 million people (25% of whom are under 14 years old)—do not have any PICU beds registered in the CNES database.

To confirm and fully understand these results, audits must be conducted with healthcare providers and local regulation centers, since underreporting in CNES cannot be ruled out. If confirmed, the lack of PICU beds represents a major challenge for health managers in these regions to expand services and ensure population access to intensive pediatric care. The Ministry of Health must ultimately guarantee resource flows that support the creation or maintenance of these beds.

As previously mentioned, the presented data show significant regional differences in the distribution of PICU beds by type (I, II, III), and in child mortality rates. Regions with a higher proportion of type III beds, such as the Southeast and South, show lower pediatric mortality. Meanwhile, the North region, with fewer beds and a higher concentration of type II PICU beds, faces greater challenges in reducing child mortality.

The correlation between the number of beds and the mortality rate of children under 14 is plausible and consistent with earlier studies, both in adult critical care [22] and neonatal care [23], that emphasize the importance of health infrastructure in reducing mortality. Based on these studies, it is reasonable to assume that the availability of PICU beds with appropriate infrastructure and in sufficient quantity for the local morbidity profile can positively impact survival among critically ill children. However, it is also essential to ensure qualified staffing and continuous resource flows for complementary actions.

The cluster analysis revealed six distinct profiles based on the distribution of PICU bed types. These clusters underscore the unequal organization and availability of pediatric intensive care services in Brazil. None of the clusters reflect the national average distribution, emphasizing the need for region-specific evaluations in public policy planning.

The clusters demonstrate that the predominant types of PICU beds vary across health regions, and factors such as the concentration of more complex beds (type III) are associated with lower child mortality rates. Clusters with a predominance of lower-complexity beds (type I) or no PICU beds tend to have higher mortality rates. For example, Cluster 5—with the highest proportion of type III beds—had the lowest child mortality rate, while Cluster 6—without any PICU beds—had the highest.

This discrepancy reinforces that simply analyzing the total number of beds is insufficient. The complexity and quality of the infrastructure are key determinants of the ability to provide care and reduce child mortality. Local factors such as population distribution and technological density must also be considered, as more developed regions tend to concentrate both a greater number of beds and better outcomes—often to the detriment of peripheral or underserved areas.

The findings highlight the importance of regionalized strategies in planning the allocation of healthcare resources. Adopting approaches that consider local specificities can promote greater equity in access to pediatric intensive care beds. Additionally, the absence of a distribution pattern reflecting the national average underscores the need for decentralized, yet coordinated planning to effectively address regional healthcare demands.

Although this study suggests an association between the distribution of PICU beds and regional child mortality, these findings should be interpreted with caution. Several potential confounding factors may influence mortality rates independently of PICU availability, such as differences in socioeconomic conditions, prevalence of comorbidities, access to primary and emergency care, quality of neonatal and perinatal services, and timeliness of patient transfers. Regional variations in data reporting and health system performance may also contribute to apparent disparities. Future studies integrating these contextual variables could help clarify the complex relationship between pediatric intensive care capacity and child health outcomes.

The main limitations of this study relate to the use of secondary data from the CNES, which may not fully reflect the actual availability of PICU beds due to underreporting and inconsistencies in data entry. For example, the classification of PICU beds (Types I, II, and III) adopted in this study is based on the standards established by ANVISA Resolution RDC No. 7/2010 [4], which defines the minimum requirements for the structure, staffing, and technology of intensive care units in Brazil. These categories are consistently applied within the CNES and are used nationwide for regulatory and planning purposes. However, as CNES data depend on self-reporting by healthcare providers and periodic updates by local managers, a degree of misclassification or delayed reporting cannot be excluded. Previous evaluations have identified variability in the completeness and reliability of CNES data, particularly regarding the number of registered beds and equipment, underscoring the need for continuous validation and improvement of these systems [24,25].

Additionally, it was not possible to verify the results directly with regional regulatory centers, as access to such data requires specific authorization from local health authorities and, in some cases, ethics committee approval due to their sensitive nature.

Despite these limitations, the findings are considered robust, as they are consistent with previous studies that examined PICU availability using provider or regulatory data [26]. Moreover, this study advances beyond descriptive analyses by employing cluster analysis, which provides a nuanced understanding of regional inequalities and supports the formulation of targeted public policies to address the unequal distribution of pediatric intensive care resources in Brazil.

Finally, despite areas that could be improved, the results underscore the importance of regionalized strategies in health resource planning. Approaches that take local specificities into account may lead to more equitable access to pediatric intensive care. Moreover, the absence of a national distribution pattern highlights the necessity for decentralized—but coordinated—planning to meet regional demands effectively.

Conclusion

The results highlight an unequal distribution of PICU beds in Brazil, with some municipalities exhibiting a desirable service supply profile, while others lack any infrastructure at all. This scenario contradicts the principles of the Unified Health System (SUS), especially the principle of equity. Although some progress has been made, there is still a long way to go to ensure fair and efficient coverage for the entire population.

The challenge is complex and requires further studies on this topic within the context of public health. It is hoped that this study will contribute to expanding the discussion on care for critically ill children in Brazil—not only focusing on clinical care protocols but also supporting SUS management and planning efforts.

Supporting information