We prospectively enrolled Alaska Native children aged <10 years hospitalized with empyema or parapneumonic effusion at the Alaska Native Medical Center in Anchorage, Alaska during 2018-2023. The Alaska Native Medical Center is the tertiary care center for all Alaska Native people living in Alaska. Children were eligible to participate if they had a negative bacterial culture on a pleural fluid specimen.

Residual pleural fluid specimens were sent to the CDC Arctic Investigations Program, aliquoted into 0.5 mL, and stored at −70°C until batch-tested during 2024. DNA was extracted using the MagMAX viral/pathogen ultra nucleic acid isolation kit (Applied Biosystems) and tested with the TrueMark Respiratory III panel (Applied Biosystems) multiplex real-time PCR for S. pneumoniae, S. aureus, H. influenzae, Chlamydia pneumoniae, Klebsiella pneumoniae, Legionella pneumophila, Moraxella catarrhalis, and Mycoplasma pneumoniae. S. pyogenes was detected using species-specific primers and probes to amplify spy1258 by real-time PCR [6]. Specimens positive for H. influenzae were serotyped using two multiplex real-time PCR assays that detected serotypes a through f. Specimens positive for S. pneumoniae were serotyped at the CDC Streptococcus Laboratory in Atlanta, Georgia using PCR methods previously described [7]. Real-time PCR could not discriminate pneumococcal serotypes 9L and 9N or 15A and 15F, and these are reported as “9L/9N” or “15A/15F.”

Demographic and clinical data were collected by medical record review. The Alaska immunization registry was used to assess pneumococcal vaccination status. Study personnel administered a questionnaire to parents regarding household composition, environmental exposures, and recent illness.

We compared demographic and clinical characteristics and bacterial etiologies of study participants to those of children aged <10 years hospitalized at the Alaska Native Medical Center during 2018-2023 with empyema or parapneumonic effusion and a positive culture for S. pneumoniae, S. pyogenes, or H. influenzae from pleural fluid. The culture-positive comparison group was identified from Alaska statewide public health surveillance for invasive bacterial disease [8].

Data were analyzed using SAS v9.4. Categorical variables were compared using Fisher’s exact test; continuous variables were compared using the Wilcoxon rank-sum test. Urban residence was defined as Anchorage Municipality, Fairbanks North Star Borough, Juneau City/Borough, or Matanuska-Susitna Borough; all other areas were considered rural.

This study was approved by the CDC and Alaska Area Institutional Review Boards, the Alaska Native Tribal Health Consortium, and Southcentral Foundation Board of Directors. Parents provided written informed consent.

During April 2018-September 2023, 14 participants were enrolled in the study; 9 (64%) were male and 1 (7%) was an infant [Table 1]. Thirteen (93%) participants resided in a rural area, and the median household size was 6 persons (range, 2-14). Six (43%) participants’ households had woodstove heating, 3 (21%) had a smoker, 2 (14%) lacked piped water, and 1 (7%) lacked a flush toilet. Two (14%) participants had asthma, 6 (43%) reported a viral illness in the 90 days prior to hospitalization, and 13 (93%) were receiving antibiotics at the time of pleural fluid collection. The median time from antibiotic initiation to specimen collection was 4 days (range, 0-10). One (7%) participant had ≥1 positive culture: a blood culture with Staphylococcus hominis and Staphylococcus epidermidis collected from one lumen of a central venous catheter 13 days after admission. Twelve (86%) participants were hospitalized for ≥7 days; none died.

Bacterial DNA was identified in 11 (79%) of 14 pleural fluid specimens [Table 2]. Among these, 9 (64%) had S. pneumoniae, 3 (21%) had H. influenzae, and 2 (14%) had S. pyogenes. Three (21%) specimens had ≥1 organism detected.

Of the nine specimens with S. pneumoniae detected, six were a serotype contained in 13-valent pneumococcal conjugate vaccine (PCV13); all six were serotype 3. Five of the children with serotype 3 pneumococcal disease received four doses of PCV13 prior to specimen collection and one child received two doses. The median time between the last PCV13 dose and specimen collection was 1.9 years (range, 0.7-4.4).

During 2018-2023, there were 11 cases of culture-positive invasive disease due to S. pneumoniae, H. influenzae, or S. pyogenes among children aged <10 years hospitalized at the Alaska Native Medical Center with empyema or parapneumonic effusion [Table 1]. Culture-negative study participants were older (median age, 2.7 years; range, 0.6-6.7) than children with culture-positive disease (median age, 1.2 years; range, 0.6-3.0) (p=0.03). Study participants had similar lengths of hospitalization (median, 11 days; range, 2-29) compared to children with culture-positive disease (median, 9 days; range, 4-21) (p=0.12). One (9%) child with culture-positive disease died.

Among the 11 surveillance cases, 4 (36%) were caused by S. pneumoniae, 3 (27%) by H. influenzae, and 4 (36%) by S. pyogenes [Table 2]. Of the four pneumococcal disease cases, two were caused by serotypes contained in PCV13. One child with serotype 3 disease received three doses of PCV13, with the last dose given 50 days prior to specimen collection; one child with serotype 4 disease received four doses of PCV13, with the last dose given 13 months prior. Six (55%) of the 11 culture-negative, PCR-positive study participants had S. pneumoniae serotype 3 detected compared to 1 (9%) of the 11 children with culture-positive disease identified through statewide surveillance (p=0.06).