Pneumococcal carriage and antibiotic susceptibility patterns in mother-baby pairs in a rural community in Eastern Uganda: a cross-sectional study [version 3; peer review: 1 approved, 1 approved with reservations]

Background: Pneumococcal carriage predisposes children to pneumonia. Pneumonia poses a significant threat to the lives of children below five years old worldwide, contributing to a high number of hospitalizations and death. Morbidity and morbidity are especially common in children under five and the elderly, although any age group can be affected. This study aimed to estimate pneumococcal carriage and determine antibiotic susceptibility patterns of the pneumococci isolated from mother-baby pairs in Ngora district after the rollout of the pneumococcal vaccine. We hypothesized that high carriage of Streptococcus pneumoniae in mothers leads to carriage in their babies and hence a greater chance of contracting pneumonia. Methods: Consecutive sampling was used to select 152 mother-baby pairs from community visits and those seeking care at the health facility. We collected nasal swabs from both baby and mother for culture and sensitivity testing using Kirby-Bauer’s agar disc diffusion method. Data was also collected from the mothers who consented to take part in the study, using an interviewer-administered questionnaire. Results: This study found that there was a low prevalence of pneumococcal carriage in the mother-baby pair in the Ngora district. Only one mother-baby pair (1/152) was found to be colonized with Open Peer Review


Introduction
Pneumonia poses a significant threat to the lives of children below five years worldwide, contributing to a high number of hospitalizations and death [1][2][3] . Morbidity and morbidity is especially common in children under five and the elderly, although any age group can be affected 4,5 . One-third of the overall global childhood mortality due to pneumonia has been attributed to Streptococcus pneumonia 4 . Developing countries are now faced with the double burden of community-acquired and nosocomial pneumonia, with both the under-funded health care systems and individual caretakers of the affected patients bearing negative economic consequences 6 . Streptococcus pneumoniae has been reported to be the most commonly isolated bacteria in both community-acquired and nosocomial pneumonia 5 . A high proportion of mortality due to community-acquired pneumonia has been documented, with up to 81% of these deaths occurring outside of hospital due to challenges related to access to health care in low-and middle-income countries 7 . The evolution of antimicrobial resistance among organisms causing pneumonia has made the consequences of childhood pneumonia even worse.
Over two million children die annually from pneumonia each year, accounting for almost 1 in 5 under-five deaths in low-and middle-income countries on average, as compared to 1 in 66 children in high-income countries 8 .
Pneumonia, together with diarrhea and malaria, was responsible for the deaths of 2.2 million children under five in 2012 in Sub-Saharan Africa, accounting for a third of all deaths in under-fives in this region 9 . In Uganda, pneumonia kills around 24,000 children below the age of five every year 2 . The overall pneumococcal carriage among children under five was estimated to be 56% when the PCV-10 and PCV-13 coverage in Uganda was 42% and 54%, respectively 4 . With improved PCV-10 and PCV-13 vaccine coverage, most of the invasive serotypes should be covered 4 . Pneumococcal carriage among mothers may lead to pneumococcal infections in children by droplet infection. Pneumococcal epidemiology in Africa varies from country to country. This study aimed to estimate pneumococcal carriage and determine antibiotic susceptibility patterns of the pneumococci isolated from mother-baby pairs in Ngora district, Eastern Uganda.

Ethical statement
We obtained ethical approval for this study from Mbale Regional Referral Hospital Research and Ethics Committee (MRRHREC060418). A letter of introduction to the study was obtained following approval from the Busitema University Faculty of Health Sciences community education program office and used by the researchers to seek permission from the Medical Superintendent of Ngora District Health Centre IV. Written informed consent for participation was obtained from participants preceding the actual data collection exercise. Access to collected data was restricted to persons directly involved in the study only. Participants were free to withdraw their consent to participate in the study at any stage of the study. Such a decision did not affect the medical care they received or their possible participation in future research studies in any way. The procedures were verbally explained to the research participants by the researchers and those who agreed to continue with the study could consent and participate.

Research design and setting
A cross-sectional study was carried out in Ngora district Health Centre IV and Ngora Health Center IV catchment area from 7 th April to 5 th May 2018. It serves a population of approximately 142,487. Ngora district is one of the districts in the Teso sub-region and was part of Kumi district until 2010, when it was established as a separate district by a Uganda parliamentary act. Ngora district covers an area of approximately 715.9 square kilometers and is predominantly inhabited by the Iteso and Kumam ethnicities. According to the Ugandan healthcare hierarchy of organizations, a Health Center IV is expected to serve a population of up to 100, 000 people, meaning at its current capacity, this health facility operates above the level of a Health Center IV 10 .

Study population and sampling
Children under five years old and their mothers were selected in a 'mother-baby pair' to determine the relationship between carriage of pneumococci in the baby, the prevalence of pneumococci in mothers, and the child's immunization status. The inclusion criteria were mother-baby pairs from selected villages in the community, and those who visited Ngora Health Center IV for routine immunization with babies under the age of five years. For this study, we defined a baby as any person under the age of five years and a mother was considered as either biological or any other female indirect care of the baby for at least four weeks. This was aimed at comparing samples from the baby, and the person in closest contact with the baby for the longest duration of time in the most recent past to allow presumed cross-transmission/cross-infection. Exclusion criteria were babies above five years, and those presented by male caretakers as those were presumed to spend less time with the babies. Mothers who did not consent or opted out at any stage of the research were excluded as well. The purpose of the research was explained to the mothers and they were asked to voluntarily participate in the research.
Consecutive sampling was used. Every participant meeting the criteria of inclusion was selected at the young child clinic and in randomly selected villages in the community until the required sample size was achieved. This method is relatively easy to employ and reduces the chances for intentional and unintentional manipulation by staff, or errors due to confusion. A total of 152 mother-baby pairs fulfilling the inclusion criteria were sampled. The Sample size was estimated using the Kish and Leslie formula developed in 1965.

Amendments from Version 2
The authors have put more information in the abstract while taking care of the maximum word limit of 300 recommended by the journal.

Data collection and variables
Demographic data was captured using a pre-tested questionnaire. The questionnaire was pre-tested with 10 mother-baby pairs at Mbale regional referral hospital Young Child Clinic (YCC) because the mother-baby pair in Mbale hospital had homogenous characteristics with our target population in Ngora district. After pre-testing the questionnaires, some questions were rephrased for clarity. The Uganda national immunization guidelines require all children to receive their first PCV dose at six weeks (1.5 months), the second dose at 10 weeks (2.5 months), and the third dose at 14 weeks (3.5 months). Therefore, in the questionnaire, full immunization was defined as any child 14 weeks (3.5 months) of age and above, who had received all the three PCV doses as stipulated in the national immunized schedule. Partial immunization included children above six weeks (1.5 months) of age who had received less than three doses of PCV, whether off schedule, or still on schedule as per the national immunization guidelines. All children below five years of age who had not received any single dose of PCV including those below six weeks (still on schedule) were categorized as not immunized.

Specimen collection and transport
Nasopharyngeal specimens were collected from the posterior nasopharynx of the mother and the baby using sterile cotton swab sticks moistened with 0.9% physiological saline. Sample collection was carried out at the young child clinic in Ngora Health Centre IV and the selected villages in the community by qualified hospital laboratory technicians. Separate swabs were used to collect samples from a mother and a baby in a motherbaby pair. To prevent sample contamination, the swab was placed in a casing containing Amies transport medium and immediately placed into a cool box containing ice packs for transportation to Busitema University Microbiology laboratory for culture and susceptibility testing within 12 hours.

Laboratory procedures
Samples were cultured on sheep blood agar and chocolate agar, followed by 24 hours of incubation at 37°C anaerobically. The isolates were identified morphologically by colony appearance and Gram staining. Optochin sensitivity and bile solubility testing were conducted on colonies that were potentially identifiable as S. pneumoniae by an alpha-hemolytic appearance on the culture media and lancet-shaped Gram-positive cocci appearing in pairs.
A 0.5 McFarland standard of S. pneumoniae was made from a 24-hour subculture by suspending colonies in sterile normal saline and inoculated by swabbing onto a plate of Mueller-Hinton agar supplemented with 7% sheep blood for susceptibility testing. Antibiotic susceptibility to penicillin G (1U), chloramphenicol (30µg), tetracycline (10µg), clindamycin (2µg), erythromycin (30µg), and ceftriaxone (30µg) was determined using modified Kirby-Bauer agar disc diffusion methods and the disc zone diameters were interpreted using the Clinical and Laboratory Standards Institute Guidelines 12 .

Data management
Collected data were entered in Microsoft Excel 2010, cleaned, coded, and imported to SPSS Version 16.0 statistical package for analysis. Data were analyzed using descriptive statistics, frequencies, and bivariate analyses (cross-tabulations). The primary outcome was a pneumococcal carriage and the secondary outcome was antibiotic resistance. Statistical frequency distribution tables and graphs were used for data presentation in terms of proportions, absolute values, percentages, and confidence intervals for point approximations at a 95% level of confidence, with P<0.05 considered as statistically important.
Data quality control Laboratory procedures were performed by laboratory scientists under the close supervision of a clinical microbiologist to ensure quality results were obtained. Data were double entered into Microsoft Excel for accuracy and reliability. ATCC 49619 S. pneumoniae was used as a control strain for quality assurance during isolation and drug susceptibility testing.

Demographic characteristics of the study participants
At the young child clinic, a total of ninety-three pairs were sampled, two of whom were excluded because the babies were presented by male caretakers, and one mother opted out due to fear of discomfort associated with the procedure for sample collection. Ninety pairs were recruited. In the community, a total of sixty-nine pairs were sampled, five were excluded because they presented children above five years of age, while two were presented by male caretakers. Sixty-two were recruited, making a total of 152 mother-baby pairs.
The study participants comprised 152 mothers and 152 babies. Of the 152 babies, 74 were male and 78 were female with the age range of 0-59 months. The youngest mother was 16 years, whereas the oldest was 44 years. None of the mothers who participated in the study reported having formal employment 13 .

Prevalence of pneumococci in mother-baby pairs
During the study, 304 samples were collected; 152 from the mothers and 152 from the children, making 152 mother-baby pairs. All samples were cultured, and antibiotic susceptibility was carried out on the isolated pneumococci. Out of 152 samples from the mothers, only five (5/152) isolates of pneumococci were obtained whereas seven (7/152) were isolated from the babies. Only one mother-baby pair (1/152) was found to be colonized with pneumococci in both mother and baby and the rest of S. pneumoniae colonized either the mother or baby.

Immunization coverage
During data collection, the immunization status of the baby was categorized as fully immunized, not immunized, and partially immunized among different age groups (Figure 1). There was high immunization coverage among the children above 12 months old but lower in the 3.5-<12 age group.

Antibiogram
The antibiotic susceptibility testing was done on positive isolates for both mother and baby.
Generally, a high trend of anti-microbial resistance was observed among the S. pneumoniae isolated ( Table 1). The highest resistance patterns were recorded with chloramphenicol (50%) and tetracycline (50%), whereas the lowest resistance was recorded in clindamycin (17%).
Babies that were fully immunized had a lower likelihood of being colonized by S. pneumoniae than their non-immunized counterparts P<0.05. Other factors examined by this study were not significantly associated with colonization with S. pneumoniae among the babies.

Discussion
We determined the prevalence of pneumococcal carriage and factors associated with colonization of pneumococci in a mother-baby pair in our study. Out of the 304 nasal swabs cultured, only 12 (3.95%) were positive for pneumococci, seven (4.61%) in children under five and five (3.29%) in mothers. We report a low carriage of pneumococci among mothers and babies that were included in this study. In contrast, a previous study in Iganga/Mayuge reported high carriage rates of over 50% in children aged under five years 4 . In the Iganga/Mayuge study, participants were selected based on presentation with pneumonia symptoms as defined by WHO guidelines, as opposed to our study, which included all children that fulfilled our selection criteria and did not take into account signs and symptoms of pneumonia. In a similar study carried out in Kenya, 90.0% of children were colonized with pneumococci. Both the Iganga/Mayuge and Kenyan studies were carried out before the introduction of PCV-10, accounting for the difference in carriage observed in our study. Different studies have shown varied carriage rates of pneumococci among children under five in Uganda and elsewhere 4,14,15 , with most of them reporting a higher carriage rate than reported in our study. A systematic review reported the carriage rate in Africa to range between 21-94% 16 , with more studies done on children than in the adult population. The high immunization coverage for PCV-10 in Ngora district could further explain the low carriage rate of pneumococci in our study as opposed to the Iganga/Mayuge study, which indicated a high carriage rate of 56% at a lower immunization coverage of 42% for PCV-10 and 54% for PCV-13. There was a statistically significant relationship between the pneumococcal carriage and the immunization status of the babies in our study (Table 2). Pneumococcal carriage is a prerequisite for disease; therefore, our findings suggest that full immunization with PCV-10 is protective against the pneumococcal carriage and hence pneumonia caused by S. pneumoniae. Several studies have reported a decrease in the burden of invasive pneumococcal disease and serotype distribution since the introduction of the PCV vaccines 17-20 . The immunization coverage for the first dose of PCV-10 (PCV1) in Ngora district in the current study was 97.78% (133/136), and 2.22% (3/136) of children above six weeks had not received PCV1. Of the 152 participants, 10.53% [16] were children below six weeks and were therefore not eligible for immunization with PCV10. The immunization coverage for PCV3 was 90.99% (101/111). Children below fourteen weeks who had not received PCV3 were excluded from the denominator because they were not eligible.  We also report a low carriage rate of pneumococci among the mothers. A similar study in coastal Kenya indicates that pneumococcal carriage was associated more with children under five than adults 15 . The low carriage rate of pneumococci among adults has been attributed by other studies to the development of natural immunity 21 . The upper respiratory tract appears a disadvantageous niche for S. pneumococci due to the development of mucosal host defenses such as sIgA 22,23 . Also, vaccination with PCV has resulted in the development of herd immunity in the adult population against S. pneumoniae 15 . Cases of pneumococcal colonization are, however, reported to rise in the elderly population due to immune senescence 24 , with many countries not considering the importance of immunization to this group of people.
In our analysis, there was no statistically significant association between the carriage of pneumococci and the sex/gender of the child. This finding is similar to the results of a systematic review in Africa, which noted that there was no association between pneumococcal carriage and gender 16 , although one study associated pneumococcal carriage with males and the other that reported association of carriage with females.
A trend of antimicrobial resistance was observed in chloramphenicol, tetracycline, and erythromycin. Other studies have similarly reported resistance of pneumococci to commonly used antibiotics 25,26 . In a study of erythromycin-resistant S. pneumoniae, 81% of the isolates were resistant to tetracycline and 76% were multi-drug resistant, whereas 12% were resistant to clindamycin, tetracycline, chloramphenicol, and kanamycin combined 25 . As opposed to our study, low resistance rates to tetracycline, erythromycin, chloramphenicol and ceftriaxone were reported in Tanzania 26 . Again, contrary to our findings, an earlier study in Uganda reported no resistance to erythromycin and ceftriaxone 27 , indicating the emergence of antimicrobial resistance against those drugs, which may be attributed to the irrational use of antibiotics in Uganda and also because such drugs are given empirically since there is no laboratory capacity to carry out culture and sensitivity studies.

Limitations
We were not able to serotype the pneumococci isolated to determine the circulating serotypes and to link pneumococcal carriage to the development of the disease.

Conclusions and recommendations
We report low pneumococcal carriage in the mother-baby pair in the Ngora district. There was no significant relationship between pneumococcal carriage in the mother and prevalence in the baby. The relationship between pneumococcal carriage and immunization status suggests that PCV-10 is protective against the pneumococcal carriage. Resistance of S. pneumoniae to commonly used antibiotics was high.

Data availability
Underlying data OSF: Pneumococcal carriage and antibiotic susceptibility patterns in mother-baby pairs in a rural community in Eastern Uganda: a cross-sectional study. https://doi.org/10.17605/OSF.IO/H9X7R 13 Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).

Open Peer Review
Reviewer Expertise: Streptococcus pneumoniae resistance, epidemiology and typing I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.