Impact of a Telemedicine Network on Neonatal Mortality in a State in Northeast Brazil

Infant mortality is a significant health indicator. It can be divided into neonatal and post-neonatal mortality. Conversely, maternal–child perinatal diseases, infections, and congenital anomalies directly influence neonatal mortality (NM). Among the causes of NM, congenital heart disease (CHD) has a leading role. However, CHD is underdiagnosed in Brazil, ¹ leading to delays in treatment and further clinical deterioration.

Paraíba is a State in Northeast Brazil. Its population is close to 4 million, and its area is divided into 4 mesoregions (Sertão, Mata, Borborema, and Agreste). Nearly 80% of the population use public health services, and many live in rural areas. In 2012, a telemedicine pediatric cardiology network was established to facilitate early diagnosis and treatment of CHD in public health units throughout Paraíba. ² The network increased diagnoses of CHD, bringing detection levels close to those of developed countries, ² with a probable impact on NM.

To analyze this hypothesis, we conducted a retrospective study based on data from Brazilian public databases. ³ We calculated the early (ENM) and late neonatal mortality (LNM) rates for Paraíba, other states from northeast Brazil, and all Paraiba's mesoregions. Mortality rates were compared between 2007–2011 and 2012–2016 (periods before and after the network was established). Linear regression models were adjusted to verify the relationship between the decline in NM rates and years of network operation (dependent variables were ENM and LNM rates; the independent variable was the number of municipalities in which the network was present in a given year).

NM decreased in Paraíba throughout the years, as well as in all northeast States. An analysis of ENM and LNM rates in the periods before (2007–2011) and after (2012–2016) the network demonstrates that these rates decline, especially ENM. The rates in other Northeast States also demonstrated a tendency to decline; however, Paraíba depicted the greatest decline for ENM and the second greatest decline for the LNM group. The linear regression model demonstrated a relationship between the fall in ENM (P = 0.037) and LNM (P < 0.001) with the years of network operation. However, in the analysis by mesoregions, only Sertão showed a significant reduction in ENM (P < 0.001), whereas the Borborema and Agreste mesoregions showed a substantial decrease in LNM (P < 0.002 and P = 0.036, respectively). Sertão and Borborema are the 2 poorest mesoregions and have the most difficult access to health centers.

Therefore, it can be concluded that the establishment of a telemedicine network led to a reduction in NM in Paraíba. Two factors contributed to that: greater detection of CHD ² and better access to health care – a critical factor that improves all components of infant mortality ⁴ – in the poorest mesoregions of Paraíba. Additionally, it is the authors' view that this experience need not remain geographically or culturally confined. Instead, it exemplifies the application of a work model that has the potential to be adapted and disseminated to other areas.