Blood was obtained by heel puncture and applied on filter paper Schleicher and Schuell 903 (Keene, NH, USA), between day 3 and day 20 of life (mean day 10). The samples were collected in areas throughout Brazil and sent by mail to Porto Alegre, South Brazil, where the tests were performed.

An indirect enzymatic immunoassay for IgM anti–Toxoplasma gondii prepared in-house (21) was used to test the first 78,350 samples. A fluorometric capture enzymatic immunoassay (FEIA) (Neonatal Toxoplasma gondii, AniLabsystems, Helsinki, Finland) was used in 285,780 samples. A kit produced by Wiener Laboratory (Rosario, Argentina) to detect IgG against Trypanosoma cruzi in human serum was adapted. In brief, a 3.2-mm DBS paper disk was placed in microtiter plates precoated with T. cruzi. The serum was eluted with 200 µL of phosphate-buffered saline (PBS)/bovine serum albumin (BSA) buffer on an orbital shaker set at 100 rpm for 60 min and incubated for 14–16 h at room temperature. After washing with PBS/BSA buffer, the protocol was followed according to the manufacturer's instructions with two modifications: the reactions occurred at room temperature, and the incubation times were duplicated, except after the addition of the color reagent. Also, two kits produced by Diesse Diagnostica Senese (Monteriggioni, Italy) were adapted to detect IgM against cytomegalovirus and IgM against rubella in human serum eluted from filter paper. A 3.2-mm DBS paper disk was placed in microtiter plates precoated with an anti-human IgM monoclonal antibody. The serum was eluted with 150 µL of PBS/BSA buffer on an orbital shaker set at 100 rpm for 2 h at room temperature. Afterwards, the protocol was followed according to the manufacturer's instructions with the same modifications made with the IgG Chagas test.

The cutoff for each test was obtained by testing 97 whole blood samples negative for IgM Toxoplasma gondii antibodies, 95 whole blood samples negative for IgG Trypanosoma cruzi antibodies, and 86 whole blood samples negative for IgM cytomegalovirus and rubella virus antibodies. The cutoff was established as three times the mean optical density of the negative samples. Negative, cutoff, and positive control samples were prepared in DBS for each test. The sensitivity of the methods was tested with 55 positive IgM Toxoplasma gondii samples, 43 IgG positive Trypanosoma samples, and 40 positive IgM cytomegalovirus and rubella samples. All samples were over the cutoff point for a preliminary analytical sensitivity of 100%. The presumptive positive samples were confirmed in a new duplicate run.

Serum tests were performed on samples from the mothers and neonates. For the first 202 case-patients with possible congenital toxoplasmosis and Chagas disease, an indirect immufluorescence test (Biolab-Meriéux Diagnóstica, Rio de Janeiro, Brazil) was used. Confirmatory serum tests for toxoplasmosis, cytomegalovirus, and rubella (IgM and IgG) were run by microparticle enzyme immunoassay (MEIA) in the Axsym (Abbott Laboratories, Chicago, IL). The FEIA method was used for serum tests and run in parallel with the Axsym, which showed good agreement.

Patients suspected to have congenital toxoplasmosis and cytomegalovirus were given a skull ultrasound, tomography, or x-ray and ophthalmoscopic and audiologic exams. Patients suspected to have congenital Chagas disease and their mothers were evaluated for cardiac and esophageal malformations. Patients suspected to have congenital rubella were evaluated for hearing loss and eye lesions. When the samples were above or maximally 20% below the cutoff value, serum samples from the infant and the mother were requested. All clinical and follow-up information was obtained by contacting the pediatricians or, in rare cases, the families.

A neonate was followed and classified as infected by meeting one of the following criteria: antigen-specific IgM and IgG in the neonate and in the mother, antigen-specific IgM in the neonate only, antigen-specific IgM in the mother only, or increased amount of antigen-specific IgG in the neonate. An increase in the neonate's IgG antibodies excluded maternal origin.

We analyzed 364,130 DBS samples for IgM against Toxoplasma gondii, and 699 samples were positive; all were recalled for serum confirmation. Serum samples from 594 neonates and 576 mothers were received, and of these, 202 suspected cases were tested by indirect immunofluorescence (IIF) (17 diagnosed with congenital toxoplasmosis), and 497 were tested by MEIA and FEIA (178 were diagnosed with congenital toxoplasmosis). A total of 195 neonates (1 in 1,867) were confirmed to have congenital toxoplasmosis. The laboratory findings are presented in Table 1, and the clinical findings are summarized in Table 2. The false-positive percentage was 0.16%. All patients with confirmed diagnoses were given sulphadiazine, pyrimethamine, and folinic acid.

Of the 195 patients with congenital toxoplasmosis, 138 (70.7%) were asymptomatic until 7 years of age. One IgM-positive asymptomatic infant also had HIV, and six patients with sequelae received late treatment (6–14 months after diagnosis) and could be asymptomatic if treated early. The follow-up was 1–84 months (mean 30 months).

We analyzed 15,873 DBS samples for Chagas disease and had 36 positive results. Serum samples from 31 neonates and 30 mothers were received for confirmatory tests. Results are shown in Table 1. The prevalence of specific Trypanosoma cruzi–specific IgG was estimated in 1 in 756 mothers (false-positive rate of 0.08%). All mothers and neonates had x-rays and echocardiographys. One mother (age 41) had an expanded heart and had a brother with Chagas disease. All others were asymptomatic and are under clinical observation. The antibodies observed in the neonates disappeared in time. The follow-up for positive children was 1–24 months (mean 15 months).

A total of 15,873 DBS samples for IgM against cytomegalovirus were analyzed, and 39 were positive. Thirty-two serum samples from neonates and 30 from mothers were received for confirmatory testing, and 16 cases were confirmed (Table 1). In 8 case-patients, the IgG levels decreased in the neonates, and in 15 case-patients, IgG in the neonates' serum only was interpreted by the clinicians as being of maternal origin. Likewise, in one case, antigen-specific IgM was detected only in the mother's serum and antigen-specific IgG only in the neonate's serum. This neonate was not followed. The prevalence of cytomegalovirus was estimated at 1 in 992 live births, and the false-positive was 0.11%. The follow-up of the infected patients showed that 11 were asymptomatic until 2 years of age, including a premature baby (36 weeks' gestation). Laboratory results and clinical findings are presented in Tables 1 and 3. The follow-up was 1–24 months (mean 15 months).

A total of 15,873 DBS samples for IgM against rubella virus were analyzed, and positive results were obtained in 67. Serum samples from 55 neonates and 52 mothers were received, and 16 were positive. Four mothers and one neonate were vaccinated against rubella and were excluded from the sample. The prevalence estimated was at 1 in 1,443, and the false-positive rate was 0.30%. Of 49 neonates and mothers without detectable IgM, 30 were followed until the specific IgG levels decreased. In 19 cases, the infant's IgG levels were interpreted by the pediatrician as being of maternal origin and were not followed. From these cases, three mothers received rubella vaccine before pregnancy. The follow-up of the 11 positive cases is shown in Table 4. The follow-up was 1–24 months (mean 15 months).