Pneumococcal colonization prevalence and density among Thai children with severe pneumonia and community controls.

BACKGROUND: Pneumococcal colonization prevalence and colonization density, which has been associated with invasive disease, can offer insight into local pneumococcal ecology and help inform vaccine policy discussions.
METHODS: The Pneumonia Etiology Research for Child Health Project (PERCH), a multi-country case-control study, evaluated the etiology of hospitalized cases of severe and very severe pneumonia among children aged 1-59 months. The PERCH Thailand site enrolled children during January 2012-February 2014. We determined pneumococcal colonization prevalence and density, and serotype distribution of colonizing isolates.
RESULTS: We enrolled 224 severe/very severe pneumonia cases and 659 community controls in Thailand. Compared to controls, cases had lower colonization prevalence (54.5% vs. 62.5%, p = 0.12) and lower median colonization density (42.1 vs. 210.2 x 103 copies/mL, p <0.0001); 42% of cases had documented antibiotic pretreatment vs. 0.8% of controls. In no sub-group of assessed cases did pneumococcal colonization density exceed the median for controls, including cases with no prior antibiotics (63.9x103 copies/mL), with consolidation on chest x-ray (76.5x103 copies/mL) or with pneumococcus detected in whole blood by PCR (9.3x103 copies/mL). Serotype distribution was similar among cases and controls, and a high percentage of colonizing isolates from cases and controls were serotypes included in PCV10 (70.0% and 61.8%, respectively) and PCV13 (76.7% and 67.9%, respectively).
CONCLUSIONS: Pneumococcal colonization is common among children aged <5 years in Thailand. However, colonization density was not higher among children with severe pneumonia compared to controls. These results can inform discussions about PCV introduction and provide baseline data to monitor PCV impact after introduction in Thailand.

Introduction

Streptococcus pneumoniae (pneumococcus) is a leading cause of bacterial pneumonia, meningitis, and sepsis in children worldwide [1, 2] and caused an estimated 294,000 (192,000–366,000) deaths globally in children less than 5 years in 2015 [2]. Pneumococcal colonization at the individual level, although usually asymptomatic, is a necessary precursor to invasive pneumococcal disease [3, 4] and an important source of spread in the community [4]. Young children generally have high colonization prevalence and serve as the primary reservoir for pneumococcus [5].

Prevalence of pneumococcal colonization is usually much higher in low- and middle-income countries [6], and the first acquisition happens at a younger age compared with high-income settings [3]. Although detecting pneumococcus in the nasopharynx or oropharynx of a child with pneumonia does not indicate an etiologic role, the density of colonization has been associated with pneumococcal disease [7-11], and while of uncertain value for individual diagnosis, may help to improve estimates of pneumococcal pneumonia prevalence in surveillance or observational studies. However, colonization prevalence and density vary by geographic location [7-11], so local data are needed to understand local epidemiology. Further, because pneumococcal conjugate vaccine (PCV) is serotype-specific, understanding the serotype distribution among colonized individuals informs estimates of potential vaccine effectiveness and is useful to monitor the impact of vaccine when introduced [3].

The Pneumonia Etiology for Child Health (PERCH) study was a multi-country case-control evaluation of the etiologic agents causing severe and very severe pneumonia among children in low- and middle-income countries in Africa and Asia [12]. A previous analysis of PERCH data from all study sites found an association between pneumococcal colonization density and microbiologically-confirmed pneumococcal pneumonia [7]. Country and regional differences in colonization density [7] warrant more detailed analysis of the Thailand data. While there are a few studies of pneumococcal colonization in Thailand [13-15] and other Southeast Asian countries [16-22], more data are needed in well-characterized populations to describe trends in prevalence and serotype distribution [6]. Using data from the PERCH Study in Thailand, we assessed pneumococcal colonization density among young children hospitalized with pneumonia and its potential association with characteristics of pneumococcal pneumonia, while also describing colonization prevalence and serotype distribution among both ill and healthy children.

Materials and methods

PERCH project

The Pneumonia Etiology Research for Child Health (PERCH) was a multi-center case-control study of the etiologic agents causing severe and very severe pneumonia among children in seven low- and middle-income countries (nine study sites) [12].

Study period, case and control definitions

PERCH enrollment in Thailand occurred in two study sites during January 2012 –December 2013 (Nakhon Phanom) and March 2012—February 2014 (Sa Kaeo). Identification of cases and controls has been described previously [23]. Cases were hospitalized children aged 1–59 months with World Health Organization (WHO)-defined severe or very severe pneumonia (using pre-2013 definitions) [24]. Cases were excluded if they had been hospitalized within the previous 14 days, had been discharged as a PERCH case within the previous 30 days, did not reside in the study catchment area, or had resolution of lower chest wall indrawing following bronchodilator therapy (for severe cases with wheeze).

Controls were randomly selected from the same community as cases and were frequency matched by the following age groups: 1 to <6 months, 6 to <12 months, 12 to <24 months, and 24–59 months [23]. Controls with respiratory tract illness (RTI) were included as long as they did not have severe or very severe pneumonia. RTI was defined as a presence of cough, runny nose, or if a child had (1) at least 1 of ear discharge, wheezing, or difficulty breathing, and (2) either a measured temperature of > 38.0°C within the previous 48 hours or a history of sore throat. PERCH included controls with RTI to minimize bias in the comparison group for estimating pneumonia etiology [25].

Specimen collection and laboratory testing

The standardization of laboratory methods for the PERCH study was described previously [26]. In summary, a flocked nasopharyngeal (NP) swab (flexible minitip, Copan®) and a rayon oropharyngeal (OP) swab specimen were collected from each case and control; the NP and OP swabs from each child were placed into the same vial with 3 ml of universal transport media (UTM) (Copan®). NP/OP specimens were kept at 4–8°C for <24 hours and then stored at -70°C until testing for pneumococcus (lytA gene target) as part of a multiplex quantitative real-time polymerase chain reaction (PCR) assay (FTD Respiratory Pathogens 33, Fast-track Diagnostics, Sliema, Malta) performed using an Applied Biosystems 7500® (ABI-7500) platform (Applied Biosystems, Foster City, CA). Quantification of pneumococcal DNA as a measure of colonization density was performed using plasmid standards provided by Fast-track Diagnostics as described previously [7]. A second NP specimen for S. pneumoniae culture was collected simultaneously with the first swab, placed into a vial with 1 mL skim milk, tryptone, glucose, and glycerin (STGG) broth, kept at 4–8°C for <24 hours and then frozen at -70°C until cultured according to the WHO reference method using a broth enrichment step [27]. Serotyping of pneumococcal isolates was performed by multiplex PCR followed by Quellung when necessary to resolve serotypes using previously published protocols [28]. Specimens that were positive for pneumococcus by PCR but not by culture were not serotyped. Blood from cases and controls was tested for S. pneumoniae by PCR and cultured for cases only. Serum was collected from cases and controls and tested for antimicrobial activity by serum bioassay: placing serum on a filter paper disc and plating a susceptible strain of Staphylococcus aureus and measuring growth inhibition around the inoculated disc [29]. Antibiotic pretreatment was defined as having either a positive serum bioassay or documentation of antibiotics administered at the study hospital prior to specimen collection. C-reactive protein (CRP) was assessed in all cases and in a subset of controls who were positive for pneumococcus by whole-blood PCR, had RTI, or were in the top 25% for total NP/OP PCR pathogen load (across all pathogens tested for).

Clinical characteristics and outcomes

Chest radiographs (CXRs) were performed at admission for cases, and each digital image was assessed by two members of a panel of 14 radiologists and pediatricians who were trained in the standardized interpretation of pediatric CXRs [30, 31]. Clinical characteristics, including oxygen saturation, were assessed on the day of enrollment. Case mortality was assessed at hospital discharge and by contact at 30-day post discharge.

Statistical analysis

Data were single-data entered into a centralised electronic data capture system (Emmes Corporation, Rockville, MD, USA) [32]. Pneumococcal colonization prevalence; demographic, clinical, and laboratory characteristics; and pneumococcal serotype distribution were compared between cases and controls using binary logistic regression adjusted for age. The Kruskal-Wallis test was used to compare the median pneumococcal colonization density across case and control subgroups. Additionally, we determined the percentage of colonizing isolates that were of serotypes included in PCV10 and PCV13 [33]. Serotypes 6A/6B by PCR could not be differentiated for the samples collected in 2013, so we applied the 6B:6A ratio from 2012 to the 2013 data. This calculation was done separately by site and case-control status. Unresolved serotyping data for other serogroups were left unresolved in analyses. Wald confidence intervals (95%) were calculated for the pneumococcal serotype distribution among cases and controls. Statistical significance was defined as p < 0.05. Statistical analyses were performed using SAS 9.4 (SAS Institute, Cary, NC, USA).

Ethical considerations

The study was approved by the Institutional Review Boards (IRB) of the Johns Hopkins Bloomberg School of Public Health (protocol #3075), an IRB of the U.S. Centers for Disease Control and Prevention (protocol #6067), and the Thailand Ministry of Public Health Ethical Review Committee (protocol #17/2554). Parents or legal guardians of participants provided written consent for enrollment.

Results

During the two-year study period, we enrolled 224 cases presenting to hospital with severe (n = 172, 76.8%) or very severe pneumonia (n = 52, 23.2%); two were excluded from this analysis because of missing NP/OP PCR results. Among 659 controls, 254 (38.5%) had RTI (Table 1); 650 controls had PCR results and were included in this analysis. Antibiotic use before specimen collection was documented in 42.0% of cases and 0.8% of controls; an additional 22.8% of cases (7.9% of controls) had received antibiotics by parental report only. Cases in Sa Kaeo were more likely to receive antibiotics before NP/OP specimen collection than those in Nakhon Phanom (51.9% vs. 36.7%, p = 0.03). Two cases but no controls had received PCV.

Table 1

Demographic and clinical characteristics by case and control group in rural Thailand.

Demographic and clinical characteristics	[1] All cases (n = 222)	[2] All controls (n = 650)	[3] RTI controls (n = 250)	[4] Non-RTI controls (n = 400)	All cases [1] vs. all controls [2] p-valuea	All cases [1] vs. RTI controls [3] p-valuea	All cases [1] vs.non-RTI controls [4] p-valuea	RTI controls [3] vs. non-RTI controls 4]] p-valuea
Provincial site
Nakhon Phanom	145 (65.3)	429 (66.0)	160 (64.0)	269 (67.3)	0.91	0.74	0.66	0.38
Sa Kaeo	77 (34.7)	221 (34.0)	90 (36.0)	131 (32.8)
Median age, months (IQR)	14.5 (7–26)	16 (8–33)	16 (9–34)	16 (8–30)	0.030	0.013	0.12	0.29
Age, n (%)
1–5 months	38 (17.1)	90 (13.8)	25 (10.0)	65 (16.2)	0.69	0.16	0.96	0.14
6–11 months	50 (22.5)	150 (23.1)	61 (24.4)	89 (22.2)
12–23 months	68 (30.6)	212 (32.6)	81 (32.4)	131 (32.8)
24–59 months	66 (29.7)	198 (30.5)	83 (33.2)	115 (28.8)
Male Gender	134 (60.4)	330 (50.8)	126 (50.4)	204 (51.0)	0.014	0.037	0.026	0.87
Prior antibiotic use, n (%)
Any documented antibiotic pretreatment prior to specimen collectionb	94 (42.3) c	5 (0.8)	5 (2.0)	0 (0.0)	< .0001	< .0001	< .0001	< .0001
Parental report only	51 (23.0)	51 (7.8)	38 (15.2)	13 (3.2)
No evidence of antibiotic use	77 (34.7)	594 (91.4)	207 (82.8)	387 (96.8)
Feverd	123 (55.4)	3 (0.5)	3 (1.3)	0 (0.0)	< .0001	< .0001	--	--
Observed cough	189 (85.1)	110 (16.9)	110 (44.0)	0 (0.0)	< .0001	< .0001	--	--
S. pneumoniae in the NP/OP
PCR Positive	121 (54.5)	406 (62.5)	172 (68.8)	234 (58.5)	0.28	0.04	0.85	0.009
PCR above thresholde	3 (1.4)	8 (1.2)	7 (2.8)	1 (0.3)	0.61	0.61	0.054	0.023
Culture positive	89 (40.1)	340 (52.4)	155 (62.0)	185 (46.4)	0.07	0.002	0.77	0.0002
PCR or culture	127 (57.2)	417 (64.2)	177 (70.8)	240 (60.0)	0.31	0.043	0.87	0.008
S. pneumoniae in the whole blood
PCR positive	3 (1.4)	5 (0.8)	4 (1.7)	1 (0.3)	0.69	0.83	0.22	0.12
PCR above thresholdf	0 (0.0)	3 (0.5)	2 (0.8)	1 (0.3)	--	--	--	0.50

RTI, Respiratory tract illness; IQR, Inter quartile range. Table restricted to children with available nasopharyngeal/oropharyngeal (NP/OP) PCR results.

aP-values from logistic regression for categorical variables (adjusted for age in months for non-age variables) and Kruskal-Wallis tests for continuous variables. Comparisons of pneumococcus in the NP/OP and whole blood are also adjusted for antibiotic use. Bolded p-values < 0.05.

b Presence of antibiotics by serum, antibiotics at the referral hospital, clinician report of antibiotics prior to specimen collection or antibiotics prior to NP specimen collection based on time of specimen collection and time of antibiotic administration. Only criterion applicable to controls is serum. P-values calculated excluding those with parental report only.

cPrior antibiotic use (documented) stratified among cases stratified by site: 51.9% in Sa Kaeo and 37.2% in Nakhon Phanom, p = 0.03.

dFever defined as measured temperature ≥ 38.0° C. Controls without signs or symptoms of illness in the past 48 hours were assumed to be without fever.

eNP/OP PCR density above 6.9 log10 copies/mL.

fWB PCR density above 2.2 log10 copies/mL.

S. pneumoniae was detected by PCR in the NP/OP of 121 (54.5%) cases, with no difference by pneumonia severity (p = 0.27), site (p = 0.39), CXR positivity (p = 0.23), or antibiotic use before sampling (p = 0.92; Table 2). Colonization prevalence was not significantly different between cases and controls (p = 0.12). Controls with RTI had higher pneumococcal colonization prevalence (68.8%) than those without (58.5%, p = 0.01; S1 Table). Of 9 fatal cases, 5 (55.6%) received antibiotics before NP/OP collection; only one fatal case was colonized with pneumococcus, and this child received antibiotics prior to sampling.

Table 2

Pneumococcal nasopharyngeal PCR positivity and density among positives by case-control group.

Study groups	All cases N = 222					All controls N = 650
		S. pneumoniae colonization prevalence		S. pneumoniae Colonization density			S. pneumoniae colonization prevalence		S. pneumoniae Colonization density
Characteristics	N	n (%)	p-valuea	Median (IQR) (103 copies/ml)	p-valueb	N	n (%)	p-valuea	Median (IQR) (103 copies/ml)	p-valueb
Overall	222	121 (54.5)c	--	42.1 (2.1,223.3)c	--	650	406 (62.5)	--	210.2 (23.2,991.2)	--
Provincial site
Nakhon Phanom	145	76 (52.4)	0.39	62.0 (2.1,576.5)	0.12	429	274 (63.9)	0.40	324.0 (36.1,1280.9)	< .0001
Sa Kaeo	77	45 (58.4)		19.0 (2.1,86.9)		221	132 (59.7)		82.5 (11.4,482.3)
Age
1–5 months	38	14 (36.8)	0.012	48.7 (0.7,1029.5)	0.31	90	40 (44.4)	< .0001	502.4 (60.6,1536.9)	0.30
6–11 months	50	26 (52.0)		8.3 (1.73,126.0)		150	88 (58.7)		357.7 (51.9,915.2)
12–23 months	68	35 (51.5)		84.9 (9.3,435.0)		212	124 (58.5)		177.0 (14.6,1189.2)
24–59 months	66	46 (69.7)		45.9 (2.1,116.4)		198	154 (77.8)		185.0 (18.4,736.2)
Gender
Male	134	70 (52.2)	0.49	42.2 (9.3,171. 9)	0.43	330	209 (63.3)	0.70	203.4 (27.9,1059.3)	0.78
Female	88	51 (58.0)		18.4 (1.7,570.3)		320	197 (61.6)		222.6 (17.0,953.2)
Prior antibiotic use
Any documented antibiotic pretreatment prior to specimen collectiond	94	50 (53.2)	0.92	23.5 (1.9,124.0)	0.21	5	1 (20.0)	0.08	0.7 (—)	0.12
Parental report only	51	28 (54.9)		38. 6 (2.0, 174.6)		51	35 (68.6)		330.7 (22.2, 1475.6)
No evidence of antibiotic use	77	43 (55.8)		63.9 (3.4, 614.8)		594	370 (62.3)		205.6 (23.7, 965.1)
NP culture positive for pneumococcus
Yes	89	83 (93.3)	< .0001	76.5 (10.2,593.0)	< .0001	340	329 (96.8)	< .0001	321.8 (64.0,1280.9)	< .0001
No	133	38 (28.6)		3.4 (0.5,64.2)		309	76 (24.6)		6.5 (0.9,177.1)
CRP ≥40 mg/L
Yes	37	19 (51.4)	0.40	64.2 (4.8,96.7)	0.88	0	0 (0)	---	---	---
No	165	91 (55.2)		35.0 (1.9,399.7)		88e	61 (69.3)		765.2 (133.6,2163.1)
RSV NPPCR positive
Yes	51	30 (58.8)	0.41	80.8 (10.0,614.8)	0.12	19	16 (84.2)	0.066	380.4 (19.7,1078.0)	0.67
No	171	91 (53.2)		26.2 (1.9,147.6)		631	390 (61.8)		205.6 (23.2,991.2)
Flu A/B NPPCR positive
Yes	7	3 (42.9)	0.32	1.9 (0.4,6.5)	0.08	0	0 (0.0)	---	---	---
No	215	118 (54.9)		48.1 (2.1,399.7)		648	406 (62.7)		210.2 (23.2,991.2)
Any virus NPPCR positive
Yes	190	106 (55.8)	0.44	40.4 (2.1,399.7)	0.69	537	342 (63.7)	0.07	217.9 (19.5,1041.7)	0.85
No	31	15 (48.4)		86.9 (3.2,197.4)		113	64 (56.6)		185.5 (34.7,755.7)
Whole blood lyt A PCR positive
Yes	3	2 (66.7)	0.81	9.3 (0.3,18.4)	0.19	5	4 (80.0)	0.469	3,318.2 (2385.9,3863.2)	0.005
No	219	119 (54.3)		42.3 (2.1,399.7)		609	387 (63.5)		203.4 (20.3,965.1)
Whole blood lytA PCR density ≥ 2.2 log10 copies/ml
Yes	0	0 (0.0)	----	----	----	3	3 (100)	----	3,017.5 (1754.3,3618.9)	0.018
No	222	121 (54.5)		42.1 (2.1,223.4)		611	388(63.5)		205.6 (20.5,967.8)
Observed cough
Yes	189	105 (55.6)	0.74	38.6 (2.1,431.3)	0.65	110	75 (68.2)	0.17	481.9 (159.3,1569.6)	0.0003
No	32	16 (50.0)		61.4 (3.8,83.7)		540	331 (61.3)		160.7 (17.8,919.6)
Hypoxemiaf
Yes	54	29(53.7)	0.79	54.0 (5.6, 635.1)	0.39	-	-		-
No	168	92(54.8)		36.4 (1.9,184.7)		-	-		-
Chest X-ray (CXR)
Positive	99	59 (59.6)	0.23	64.2 (2.9,582.8)	0.18	-	-		-
Negative	96	50 (52.1)		15.2 (1.9,126.0)		-	-		-
CXR positive findings
Any consolidation	43	23 (53.5)	0.32	76.5 (2.1,635.1)	0.90	-	-		-
Other infiltrate	56	36 (64.3)		59.4 (3.8, 362.8)		-	-		-
Pneumonia severity
Severe	170	96 (56.5)	0.27	48.1 (2.6,311.5)	0.29	-	-		-
Very Severe	52	25 (48.1)		6.4 (1.7,171.9)		-	-		-
Died
Yes	9	1 (11.1)	0.043	5.6 (—)	0.56	-	-		-
No	208	118 (56.7)		42.2 (2.1,223.3)		-	-		-
Possible bacterial pneumoniag
Yes	7	4 (57.1)	0.67	7.2 (1.2,49.8)	0.28	-	-		-
No	204	112 (54.9)		40.4 (2.0,210.3)		-	-		-

IQR, Inter quartile range; PCR, Polymerase chain reaction; NP, Nasopharyngeal; CRP, C-reactive protein; RSV, Respiratory syncytial virus; NP PCR, Nasopharyngeal Polymerase Chain Reaction; CXR, Chest radiography; Flu A/B, Influenza A/B.

a. P-values obtained from logistic regression adjusted for age, comparing pneumococcal colonization prevalence by characteristic within study group (cases or controls).

b. P-values obtained from Kruskal-Wallis, comparing pneumococcal colonization density by characteristic within study group (cases or controls).

c. p = 0.12 comparing PCR+ in All cases vs All controls (logistic regression adjusting for age in months); p < 0.01 comparing median density among positives in All cases vs All controls (Kruskal-Wallis)

d. Presence of antibiotics by serum, antibiotics at the referral hospital, clinician report of antibiotics prior to specimen collection or antibiotics prior to NP specimen collection based on time of specimen collection and time of antibiotic administration. Only criterion applicable to controls is serum. P-values calculated excluding those with parental report only.

e. CRP was only assessed in a subset of controls, those who were positive for pneumococcus by whole-blood PCR, had RTI, or were in the top 25% for total NP/OP PCR pathogen load (across all pathogens tested for).

f. Hypoxemia defined as oxygen saturation < 92% on room air at admission or oxygen requirement (if no room air reading available).

g. Possible bacterial pneumonia defined as having CRP> 40 mg/L, alveolar consolidation, and RSV not detected by nasopharyngeal PCR nor induced sputum PCR (excludes children with unknown possible bacterial pneumonia status).

Bolded p-values < 0.05.

Among 222 cases, 89 (40.1%) were positive for S. pneumoniae by NP culture, and an additional 38 cases were positive only by PCR (culture-negative); 83 (68.6%) of PCR+ cases were also culture-positive. No cases were positive for S. pneumoniae by blood culture, and only three cases had S. pneumoniae detected by whole blood PCR. Among 649 controls, 340 (52.4%) were NP culture-positive, and 76 additional controls were PCR+/culture-negative; 329 (81.2%) PCR+ controls were culture-positive. While not statistically different, cases were less likely to be pneumococcal culture-positive in the NP than controls (p = 0.07, adjusted for age and antibiotic use; Table 1).

The median pneumococcal colonization density among PCR-positives was lower for cases (42.1 x 103 copies/mL, interquartile range (IQR): 2.1–223.3) than for controls (210.2 x 103 copies/mL, IQR: 23.2–991.2; p <0.01). Three (1.4%) of 222 cases (2.5% of 121 PCR-positive cases) had colonization densities above a threshold associated with pneumococcal pneumonia identified in the all-site PERCH analysis (6.9 log10 copies/mL) [7], compared to 8 (1.2%) controls (2.0% of 406 PCR-positives) with colonization densities above this threshold (p = 0.84). We found no significant differences in pneumococcal colonization density among cases by site, age, gender, prior antibiotic use, or pneumonia severity (Table 2). Controls with RTI had higher colonization density (505.8 x 103 copies/mL) than those without (109.5x 103 copies/mL, p <0.0001) (S1 Table), and density was higher in controls from Nakhon Phanom (324.0 x 103 copies/mL) than from Sa Kaeo (82.5 x 103 copies/mL; p < 0.0001). Among cases (n = 30) and controls (n = 16) with co-detection of RSV, median colonization density was higher than in RSV-negative children but did not reach statistical significance. Median colonization density did not differ between CXR-positive cases and cases with normal CXRs (p = 0.18), or when restricting to cases with consolidation on CXR vs. cases with other infiltrate (p = 0.90). Of 6 cases with clinical characteristics consistent with possible bacterial pneumonia (CRP ≥40 mg/ L, alveolar consolidation, and no RSV co-infection), 4 were colonized with pneumococcus but the median density was low (7.2 x 103 copies/mL).

Among 121 cases and 406 controls with pneumococcal NP colonization detected by PCR, 83 (68.5%) and 329 (81.0%), respectively, were also culture-positive, allowing for determination of serotype. Colonization isolates from an additional 24 children (N = 7 cases and N = 17 controls) who were NP culture positive for pneumococci but PCR-negative were also serotyped. The pneumococcal colonization serotype distribution was generally similar between cases and controls (Fig 1). The most common colonizing serotypes among cases were vaccine-types: Serotype 6B was detected in 23.1%, followed by 23F (17.8%), 19F (14.4%), 15B/C (10.0%), and 14 (8.9%); 70.0% of colonizing isolates were serotypes included in PCV10 and 76.7% were PCV13-type (Fig 2). Serotypes 6B, 23F and 19F were also the most common serotypes among controls (15.4%, 13.9% and 12.4%, respectively) with little difference between RTI and non-RTI controls; 61.8% were PCV10-type and 67.9% were PCV13-type. After adjusting for prior antibiotic use, age and study sites, there was no significant difference in presence of vaccine-type pneumococci comparing cases to controls for either PCV 10 (OR: 1.7, 95% CI 0.8–3.4, p = 0.12) or PCV 13 (OR: 1.7, 95% CI 0.8–3.4, p = 0.14). Although the pneumococcal colonization serotype distributions were slightly different between the two study sites, 7 of the 8 most frequent serotypes among cases were the same, and the top 6 serotypes among controls were the same (S1 and S2 Figs).

Fig 1

Pneumococcal colonization serotype distribution among pneumonia cases and community controls in rural Thailand.

Bars represent the percent of cases or controls with the serotyped detected, restricted to children with NP culture serotyping data. Error bars are 95% Wald confidence intervals. PCV-10 serotypes: 1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F, and 23F; PCV-13 serotypes: PCV-10 serotypes plus serotypes 3, 6A, and 19A. NT, non-typeable.

Fig 2

Pneumococcal colonization serotype distribution among pneumonia cases and community controls by pneumococcal conjugate vaccine type*.

*Pneumococcal conjugate vaccine (PCV)-10 serotypes: 1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F, and 23F PCV-13 serotypes: PCV-10 serotypes plus serotypes 3, 6A, and 19A (ST 3 displayed separately).Restricted to children with NP culture serotyping data.

Discussion

In children aged <5 years in the PERCH study in Thailand, we found that pneumococcal colonization was common in both children hospitalized with severe pneumonia and community controls. Antibiotic exposure before specimen collection was common among cases and may have contributed to the lower colonization density compared to controls [34]. Among children with severe pneumonia, colonization density was not elevated in children with characteristics of bacterial pneumonia compared to those without such characteristics. Among cases, 70.0%-76.7% of colonizing pneumococcal serotypes were PCV-types, exceeding slightly the percentage PCV-type in controls (61.8%-67.9%). These findings add to the limited knowledge on pneumococcal colonization and density in Southeast Asia and provide important data to advance discussions about potential future introduction of PCV in Thailand.

Previous colonization studies in Thailand have been conducted in selected populations, such as nursery and school attendees [13], infants in refugee camps [14], and hospitalized pneumonia patients and outpatients with influenza-like illness [15]. But such populations are unlikely to represent the true colonization ecology of pneumococcus in the community. Our inclusion of randomly selected community-based controls offers a more representative estimate of community colonization prevalence, which can be important for considering and monitoring the impact of PCV introduction.

Pneumococcal colonization prevalence in our controls (62.5%) was similar to what has been reported among healthy children in low-income countries in Africa and Asia before PCV introduction (64.8% [95 CI 49.8–76.1%]), based on a 2014 meta-analysis [6]. However, that analysis found high heterogeneity by country, with prevalence in several countries (all African) over 80%; the colonization prevalence from our study was somewhat higher than estimates for other middle-income countries (47.8%, 95% CI 44.7–50.8%) [6]. Our controls included 38.5% with RTI, who had a higher colonization prevalence (68.8%) than did the non-RTI controls (58.5%). However, the colonization prevalence among non-RTI controls was still higher than the prevalence reported in other Southeast Asian countries (50.30% in healthy community children aged <5 years in Vietnam [10], 43% - 46% among healthy children in Indonesia [17, 18], 35% from nasal swab specimens in daycare centers in Malaysia [19], and 14.3% among children aged 5–8 weeks and 55.8% among those aged 12–23 months in Laos PDR [21]. These differences may in part be due to variations in specimen collection and testing methods. We tested two specimens (NP and OP), which may have increased detection sensitivity compared to studies that collected specimens from OP or nares only [27]. Some also used culture methods only, but prevalence in our study based only on NP culture (52.4% in controls) was still higher than prevalence reported from other Southeast Asian countries.

While pneumococcal colonization prevalence among our Thai controls was similar to or higher than prevalence reported from many low- or middle-income countries before the PERCH study, colonization prevalence was lower than that at other PERCH sites, which used the same standardized approaches to study enrollment, specimen collection, and testing as at the Thailand site [7]. Therefore, we are confident that differences in colonization prevalence between Thailand and other PERCH study sites likely represent real differences rather than differences attributable to variations in participant selection, specimen collection, or testing procedures. Such differences in colonization could be due to different exposure factors, such as population density, crowded living conditions, regular interaction with many different people, or exposure to other young children such as siblings or via day care [35, 36]. These factors differ by cultural practices and urban vs. rural settings, both between and within countries. Prevalence is also a function of the age of acquisition of carriage, which tends to be earlier in low- compared to high-income countries [3], meaning that middle-income countries like Thailand are likely in between. The lower colonization prevalence among the Thailand cases compared to controls could be attributed to prior antibiotic use, since 42.0% of cases received prior antibiotics compared to 0.8% of controls; however, colonization prevalence in cases did not differ significantly between those with and without prior antibiotic exposure.

Although pneumococcal colonization does not indicate disease, high density of colonization has been associated with confirmed pneumococcal disease in the PERCH study, and a density threshold of >6.9 log10 copies/mL distinguished microbiologically-confirmed pneumococcal cases from community controls (64% sensitivity, 92% specificity) [7]. There were no PERCH cases with laboratory-confirmed invasive pneumococcal disease (IPD) at the Thai site, which may explain the lower colonization density among cases compared to controls, even among cases with alveolar consolidation on CXR. However, higher density in suspected or confirmed pneumococcal cases has been seen elsewhere, including in Southeast Asia. Two studies in Vietnam found pneumococcal NP density among children with radiographically-confirmed pneumonia was significantly higher than density among healthy controls and children with acute lower respiratory tract infection [9, 10]. Studies in South Africa and Mozambique also found higher colonization density among children with IPD compared to controls [8, 11]. The colonization density among PERCH cases in Thailand was lower than the density among cases without microbiologically-confirmed pneumococcal disease at other PERCH sites, suggesting that pneumococcal disease may be less common among those presenting to hospital with severe pneumonia in Thailand compared to the other sites. Receipt of antibiotics prior to specimen collection may also have reduced median colonization density, as antibiotics have been associated with lower colonization density in all PERCH sites [7].

Our study describes the serotype distribution of colonizing pneumococcal isolates in Thailand. We found that a relatively high percentage of colonizing isolates from pneumonia cases and controls were serotypes included in PCV10 (70.0% and 61.8%, respectively), and PCV13 (76.7% and 67.9%, respectively). Serotype distributions among cases and controls were similar, with only small differences in rank order of the top 8 serotypes. Certain serotypes (3, 13, 23A, 6C, 34, 11A/11D, and 18A/18B/18C) were found less frequently in cases than representative community controls, but each of these seroptyes represented <5% of all colonizing serotypes. Enrollment of randomly selected community controls provides an unbiased description of the circulating pneumococcal serotypes. The percentage of colonizing isolates that were PCV-type among cases was lower than previously described for blood isolates from hospitalized children aged <5 years (PCV10 = 74%, PCV13 = 92%) from these same provinces [37], as expected since PCVs include the more invasive serotypes [38]. The proportion of colonizing isolates in our study that were PCV13-type was similar to that from studies in Malaysia (69.5%) [19] and Cambodia (70.2%) [39] but higher than results from Indonesia (56% and 45%) [17, 18]. Previous studies in Thailand have described serotype distribution of colonizing pneumococci among Karen-infants in displaced persons camps [14] and among children with influenza-like illness (ILI) and pneumonia [15]. Although not necessarily representative of colonization in the general community, the top six pneumococcal serotypes from these studies were the same as the top six in PERCH Thailand cases and controls with the exception of 15B, which was not top six in the study of children with ILI and pneumonia [15]. Although Thailand does not include PCV in the National Immunization Program, our data can contribute to future policy discussions about potential introduction. A new conjugate vaccine in Phase II trials includes the same serotypes as PCV13 minus serotype 3 (detected in only 1% of cases); if this vaccine proves to be more affordable than PCV13, it could have important implications for vaccine policy considerations [33]. A recent study concluded that when considering herd-effect, both PCV10 and PCV13 would be cost-effective in Thailand [40]. In the next 3–5 years, Thailand’s Advisory Committee on Immunization Practices (ACIP), a sub-steering committee of Thai National Vaccine Committee, will consider adding PCV to Thailand’s Expanded Program on Immunization [Personal communication].

Our study had some limitations. First, we found no blood culture-confirmed S. pneumoniae among PERCH cases in Thailand, which meant we had no confirmed pneumococcal pneumonia cases as true positives. Second, both study provinces were relatively rural settings, so our findings might not be generalizable to urban settings like Bangkok. Third, all case-control studies are subject to potential selection bias related to control selection. We believe that selection bias was minimized by our approach to control selection; community controls were randomly selected from comprehensive lists of children aged 1–59 months drawn from health services registries, which include virtually all children in each province, and were enrolled year round and frequency matched to the age-group distribution of the cases on a monthly basis. Therefore, our controls represented an unbiased community comparison group in terms of pneumococcal colonization prevalence [25]. Fourth, we did not collect data on the dose and duration of antibiotic exposure prior to enrollment, so we were unable to fully account for the effect of antibiotic use on pneumococcal colonization prevalence and densities. Fifth, serotype data were unavailable for 38 (31.4%) of the 121 pneumococcal PCR-positive cases and 76 (18.7%) of 406 PCR-positive controls who had negative cultures. It is possible that the serotype distribution of PCR-positive/culture-negative samples differed from that of culture-positive samples, which would bias our serotype distribution estimates. Finally, nasopharyngeal isolates from 16.7% of cases and 26.3% of controls were not fully sub-typed for serogroups 6, 15, 9, 11, and 18. For serogroup 6, we estimated the distribution of isolates that were sub-types 6A and 6B by apportioning the unresolved serogroup 6 isolates according to the distribution of the isolates for which we were able to resolve the sub-type; this approach assumes that isolates with unresolved sub-typing had the same sub-type distribution as the fully typed isolates. We had insufficient data to apportion the other unresolved serogroups.

Our two-year, highly standardized study, provides estimates of pneumococcal colonization prevalence, density, and serotype distribution among children aged <5 years hospitalized with severe or very severe pneumonia in Thailand and in community controls. These findings provide important data to inform discussions about PCV introduction in Thailand and offer baseline data on prevalence and serotype distribution that could be monitored after vaccine introduction to evaluate impact. Additional studies could evaluate the potential association of pneumococcal pneumonia and colonization density among young children in Thailand with a sample that includes confirmed cases of pneumococcal disease. Although IPD has been documented to occur commonly in Thailand, including in the PERCH study provinces [41], no IPD cases were detected during the PERCH Study [42]. Additional studies could evaluate the potential association of pneumococcal pneumonia and colonization density among young children in Thailand with a sample that includes confirmed cases of pneumococcal disease.

Disclaimer

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention, the U.S. Department of Health and Human Services, or the U.S. government.

Pneumococcal nasopharyngeal PCR positivity and density among controls by acute respiratory illness symptoms.

RTI, Respiratory Tract Illness; IQR, Inter quartile range; PCR, Polymerase chain reaction; NP, Nasopharyngeal; CRP, C-reactive protein; RSV, Respiratory syncytial virus; NP PCR, Nasopharyngeal Polymerase Chain Reaction; Flu A/B, Influenza A/B. a. Comparison of colonization prevalence: p = 0.01 in RTI controls vs Non-RTI controls. Comparison of median density: p < .0001 in RTI controls vs Non-RTI controls; p < .0001 in All cases vs RTI controls; p < 0.01 in All cases vs Non-RTI controls. b. Presence of antibiotics by serum. P-values calculated excluding those with parental report only. P-values obtained from logistic regression adjusted for age (pneumococcal colonization prevalence) and Kruskal-Wallis (pneumococcal density). Bolded p-values < 0.05.

(PDF)

Click here for additional data file.

Pneumococcal colonization serotype distribution among pneumonia cases and community controls in rural Thailand: Nakhon Phanom.

Bars represent the percent of cases or controls with the serotyped detected, restricted to children with NP culture serotyping data. Error bars are 95% Wald confidence intervals. PCV-10 serotypes: 1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F, and 23F; PCV-13 serotypes: PCV-10 serotypes plus serotypes 3, 6A, and 19A.

(PDF)

Click here for additional data file.

Pneumococcal colonization serotype distribution among pneumonia cases and community controls in rural Thailand: Sa Kaeo.

Bars represent the percent of cases or controls with the serotyped detected, restricted to children with NP culture serotyping data. Error bars are 95% Wald confidence intervals. PCV-10 serotypes: 1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F, and 23F; PCV-13 serotypes: PCV-10 serotypes plus serotypes 3, 6A, and 19A.

(PDF)

Click here for additional data file.

(PDF)

Click here for additional data file.

(PDF)

Click here for additional data file.

2 Dec 2019

PONE-D-19-29663

Pneumococcal colonization prevalence and density among Thai children with severe pneumonia and community controls

PLOS ONE

Dear Mr. Piralam,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The manuscript has been assessed by two reviewers; the comments are available below. The reviewers have raised a number of concerns about the clarity in presentation of the work and the data, they recommend revisions to improve the clarity in presentation and writing and to provide a fuller outline of the methodology and main results. Please carefully revise the manuscript to address all the points raised by the reviewers.

We would appreciate receiving your revised manuscript by Jan 16 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Jose Melo-Cristino, M.D., Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

3. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

4. Thank you for stating the following in the Financial Disclosure section:

'This work was supported by the Bill & Melinda Gates Foundation (https://www.gatesfoundation.org/) grant 48968 to the International Vaccine Access Center, Department of International Health, Johns Hopkins Bloomberg School of Public Health, for the PERCH study.

Representatives from the Bill & Melinda Gates Foundation participated in site selection and in Pneumonia Methods Working Group meetings, which informed the study design. They had no role in the data collection, data analysis, data interpretation, or writing of the report.'

We note that one or more of the authors are employed by a commercial company: The Emmes Company

a. Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form.

Please also include the following statement within your amended Funding Statement.

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”

If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

b. Please also provide an updated Competing Interests Statement declaring this commercial affiliation along with any other relevant declarations relating to employment, consultancy, patents, products in development, or marketed products, etc.

Within your Competing Interests Statement, please confirm that this commercial affiliation does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests) . If this adherence statement is not accurate and  there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

c. Please include both an updated Funding Statement and Competing Interests Statement in your cover letter. We will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This manuscript examines pneumococcal colonisation prevalence and density among Thai children with severe pneumonia and community controls.

This is a well written paper which includes data from the PERCH (pneumonia) aetiology study.

My comments are minor only:

Page 4, line 89: add "usually" to "pneumococcal colonization is usually much higher..."

Page 5, line 107: there have been 2 papers from Laos published in 2019 on pneumococcal community carriage and risk factors for high median density carriage- need to add. Vaccine (Satzke et al) and Plos One (Dunne et al) papers

PCV vaccination status of cases and controls needs to be added as PCV is available in the private market in Thailand with ~30% uptake. Analysis in Table 2 needs to have density by vaccination status

pg 8, stats analysis: add how the data were managed, and how it were summarised before analysed.

Table 2 is difficult to read with regard to the p values and which comparisons they are making eg. the row with Provincial site has 4 p values but it is unclear what comparisons they are referring to- is it the above rows or the rows below? Table needs to be reformatted for clarity.

Table 2

Although there was a lack of statistical significance between many of the study groups with regard to density, the median and IQR differences within the subgroups suggest a there is a difference- just that there is insufficient power to detect it eg median density in RSV positive was 80 (IQR10-614) vs RSV neg 26 (IQR 1.9-147). If two histograms were plotted on one graph for each subgroup, the difference maybe more obvious, although not statistically significant. Ditto for antibiotics vs no antibiotic; CRP; Flu A/B; and al the clinical study groups. Suggest plotting this data as overlapping histograms for each subgroup.

Discussion

pg 19: line 320: please add a reference

Line 339: typo-remove %

Line 343: also compare with Laos community carriage results

Reviewer #2: Piralam and coll. assessed pneumococcal nasopharyngeal (NP) colonization in hospitalized children with pneumonia and in controls in Thailand. They also assessed pneumococcal colonization density among young children hospitalized with pneumonia and the association with characteristics of pneumonia, and distribution of serotypes. They used the data of the large multicentre PERCH study.

224 pneumonia cases and 659 community controls were included in the 2 study sites; both located in Thailand. The results showed a lower proportion of NP colonization and a lower median colonization density in cases compared to controls. No statistical differences were found in sub-groups analyses for colonization density.

Nowadays, PCR on NP samples cannot be used as a diagnostic tool as it has an uncertain value for individual diagnosis. Association between pneumococcal NP colonization density and pneumonia is, to my opinion, a topic of great interest. As a lot of asymptomatic controls have a NP colonization; a high amount of S. pneumoniae might be associated with pneumonia (as the authors mentioned in Introduction section l.93 réf 7-11).

The presented results are unexpected and hypotheses to explain them should be clearly formulated. Overall analyses of the PERCH study showed an association between pneumococcal colonisation density and microbiologically-confirmed pneumonia and the results in the Thai sub-group are opposite (réf 7). Then, potential bias should be carefully raised. I think that the manuscript can easily be improved (especially Results and Discussion sections). Be careful with typographical errors.

Major comments

1. Table 2 has to be improved, particularly to help the reader to understand quickly the Table. It is unclear. For example, in each line of the study characteristics, there are several p-value (for example: age P=0.012, P=0.31 etc.). It is not clear at all what mean these p-values and what is compared in this Table. The authors may add a new column for p-value and precise in header information what is compared. The authors may also perhaps split the Table in 2 Tables as some characteristics were not assessed in Controls. I particularly don’t understand to what refer p-values for characteristics not measured in controls (such as hypoxemia).

2. The authors may add some characteristics in Table 1 (or appendix …) in order to better describe the studied population (ex: for cases % detection of pneumococcus in blood by PCR, symptoms, etc).

3. The text of the results section is quite difficult to read with numerous imprecisions (see minor comments). You may try to clarify and focus on the interesting results.

4. Figure 1 is not understandable. Some serotypes are missing on the x-axis. The authors do not mention if some proportions are different between cases and controls. What mean NT?

5. How was collected the variable “prior antibiotic exposure”? If the quite low proportion of NP carriage of pneumococcus in cases might be explained by a higher proportion of antibiotic intake in cases, then the proportion of NP carriage should be higher in the group without previous antibiotic intake compared to the group without antibiotic intake. What hypotheses can be formulated to explain this?

6. Discussion p. 22 l.374 and after. Did you find other studies showing a higher pneumococcal density in healthy controls compared to pneumonia cases? I think that you have to carefully mention potential alternative explanations to explain this unexpected result: were operators (clinicians who collect the swab on subjects) the same between the hospital study site and community? If controls may have RTI, is it possible that the sensitivity of the swab was higher/increased in controls compared to the sensitivity in pneumonia cases (in the case of a higher amount of mucus, etc)?

7. I think that you have to clearly mention the proportion of cases and controls vaccinated with PCV. If no cases and controls were vaccinated because PCV was not in the national immunization program during the stud, you should clearly mention it.

8. The authors have to clearly present the potential bias. What about a selection bias in the study if the controls are all included in a same population with a high prevalence of NP carriage of S. pneumoniae?

Minor comments

1. P.3 l.67 103 and not 103

2. The authors precise that they performed binary logistic regression adjusted for age; as the study was frequency matched for age. You mean that the study was not paired matched, and then, that it was not possible to perform a conditional logistic regression?

3. P. 17, l.261 I think that it is better to mention IQR with a “-“ and not “,” (for example 2.1-222.3)

4. P.17, l.262 is it possible to add the p-value between proportion of cases and control above the threshold identified in the overall PERCH analysis?

5. P.17, l. 264 mL instead of ml

6. P.17, l. 266 I don’t understand to what the p-value (p=0.31) refers? It may be better to mention “The density is highest […] even if not significant” or mention the p-values after the sentence “none of these differences reached statistical significance”

7. P.17, l. 274 Only 3 cases had pneumococcus detected by PCR in whole blood? You may mention in the population description

8. P. 17, l. 275 Do you mean “were colonized in NP” ?

9. P. 17, l. 278 Do you mean “in whole blood by PCR”?

10. P. 18, l. 286 “with NP colonization”

11. Figures 1 & 2 the caption of y-axis “percent of children” is not clear. You might mention “percent of children with pneumococcal NP colonization”

12. P. 18, l. 299 I think that it is better to mention 95% CI with a “-“ and not “,”

13. P. 19, l. 301 “7 of the 8 most frequent serotypes” instead of “the top 8 serotypes”

14. P. 20, l. 339 remove “%”

15. P. 24, l. 424 “though …. unresolved serotypes”. This sentence is difficult to understand, you may rewrite it. I don’t know the meaning of the word “apporportioned”. Do you mean

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Cédric Dananché

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: review_piralam.doc

Click here for additional data file.

27 Jan 2020

Please find the response to both of major and minor comments from reviewers as details as below;

Major comments

1. Table 2 has to be improved, particularly to help the reader to understand quickly the Table. It is unclear. For example, in each line of the study characteristics, there are several p-value (for example: age P=0.012, P=0.31 etc.). It is not clear at all what mean these p-values and what is compared in this Table. The authors may add a new column for p-value and precise in header information what is compared. The authors may also perhaps split the Table in 2 Tables as some characteristics were not assessed in Controls. I particularly don’t understand to what refer p-values for characteristics not measured in controls (such as hypoxemia).

Response: The p-values in the table are comparing S. pneumoniae colonization prevalence and density by characteristics within each study group (cases and controls). There are no comparisons between cases and controls displayed in the table. We have revised the structure to shift the p-values from the rows to dedicated columns. We have also added a footnote explaining the p-values. The revised table 2 was placed in the manuscript but the changes to the table are not tracked

2. The authors may add some characteristics in Table 1 (or appendix …) in order to better describe the studied population (ex: for cases % detection of pneumococcus in blood by PCR, symptoms, etc).

Response: We appreciate the suggestion to add more information on characteristics of the cases and controls in table 1. Many of the important characteristics are included in table 2 (number with pneumococcus detected in blood, number with cough, deaths, etc), although we recognize that the information is stratified by pneumococcus detection in the NP/OP. We are willing to add some of these characteristics to table 1 but worry it might look a little redundant. Also, a full description of all cases and controls in Thailand is pending publication right now. As soon as that paper is accepted by the journal, we will add a reference to that paper into our paper under review at PLoS One. If the editor would like us to add further details to table 1, we can do so quickly.

3. The text of the results section is quite difficult to read with numerous imprecisions (see minor comments). You may try to clarify and focus on the interesting results.

Response: Thank you for the comment. We have revised the results section in response to the reviewer’s specific comments below. In the process, we also made some additional adjustments for clarity, including shortening sections that were not critical to the paper’s main message.

4. Figure 1 is not understandable. Some serotypes are missing on the x-axis. The authors do not mention if some proportions are different between cases and controls. What mean NT?

Response: Some serotypes were dropped when exporting the figure. We have revised the figures to correct this issue and include all serotypes (see revised Figure 1).

5. How was collected the variable “prior antibiotic exposure”? If the quite low proportion of NP carriage of pneumococcus in cases might be explained by a higher proportion of antibiotic intake in cases, then the proportion of NP carriage should be higher in the group without previous antibiotic intake compared to the group without antibiotic intake. What hypotheses can be formulated to explain this?

Response: “Any documented antibiotic pretreatment prior to specimen collection” was defined as presence of antibiotics by serum, antibiotics at the referral hospital, clinician report of antibiotics prior to specimen collection or antibiotics prior to NP specimen collection based on time of specimen collection and time of antibiotic administration. “Parental report only” defined as no documented evidence but report by parent or caregiver. In reviewing these definitions, we decided to make the comparisons cleaner by excluding from the analysis those with antibiotic pretreatment by parental report only. The revised p-values in table 2 and supplementary table 1 are only for the comparison between those with documented evidence of antibiotic pretreatment and those with no evidence of antibiotic pretreatment. Although the differences in NP/OP PCR positivity by antibiotic use were not statistically significant, we did observe differences in NP culture positivity by antibiotic use.

6. Discussion p. 22 l.374 (Shifted to Page 24 line 382) and after. Did you find other studies showing a higher pneumococcal density in healthy controls compared to pneumonia cases? I think that you have to carefully mention potential alternative explanations to explain this unexpected result: were operators (clinicians who collect the swab on subjects) the same between the hospital study site and community? If controls may have RTI, is it possible that the sensitivity of the swab was higher/increased in controls compared to the sensitivity in pneumonia cases (in the case of a higher amount of mucus, etc)?

Response: Despite extensive review of the literature, we did not find other studies showing lower colonization density among pneumonia cases compared to controls. We speculated possible explanations in the discussion starting with line 375. Although statistically significant, given the wide range of densities seen, one could argue that the absolute difference in median density was not substantial. The study staff collecting the NP/OP swabs from cases were not the same as the staff collecting specimens from controls. However, as noted in the discussion, all clinical staff underwent rigorous training in standard specimen collection, with required demonstration of proficiency and periodic refresher trainings. Controls with RTI had significantly higher colonization density compared to controls without RTI, which supports the reviewer’s suggestion that increased mucous could increase pneumococcal yield.

7. I think that you have to clearly mention the proportion of cases and controls vaccinated with PCV. If no cases and controls were vaccinated because PCV was not in the national immunization program during the stud, you should clearly mention it.

Response: Only 2 cases were vaccinated, and no controls received one or more dose of PCV during the study period. We added this information in Page 20, line 304-305

8. The authors have to clearly present the potential bias. What about a selection bias in the study if the controls are all included in a same population with a high prevalence of NP carriage of S. pneumoniae?

Response: We believe that potential selection bias was minimized by the approach to control selection. Community controls were randomly selected from comprehensive lists of children aged 1 - 59 months drawn from health services registries, which include virtually all children in each province. Controls were enrolled year round and frequency matched to the age-group distribution of the cases on a monthly basis. The control selection process is detailed in a separate paper that has been submitted for publication and will ultimately be referenced in this paper.

Minor comments

1. Page 4, line 89 (shifted to Page 5 line 89): add "usually" to "pneumococcal colonization is usually much higher..."

Response: Revised as suggested

2. Page 5, line 107: there have been 2 papers from Laos published in 2019 on pneumococcal community carriage and risk factors for high median density carriage- need to add. Vaccine (Satzke et al) and Plos One (Dunne et al) papers

Response: Added 2 papers from Laos PDR.

3. PCV vaccination status of cases and controls needs to be added as PCV is available in the private market in Thailand with ~30% uptake. Analysis in Table 2 needs to have density by vaccination status

Response: The PCV vaccination status was added in Page 10 line 202. Given that there were only 2 cases who received PCV and no controls who received even one dose of PCV, we did not feel it necessary to include density by vaccination status in Table 2.

4. Page 8 (shifted to Page 9 line 173-174), stats analysis: add how the data were managed, and how it were summarized before analyzed.

Response: We added the following sentence to the Statistical Analysis section and reference to Watson et al 2017 (Data Management and Data Quality in PERCH, a Large International Case-Control Study of Severe Childhood Pneumonia) which described the data management processes in PERCH. “Data were single-data entered into a centralized electronic data capture system and data cleaning occurred throughout the study (Emmes Corporation, Rockville, MD, USA).”

5. Table 2 is difficult to read with regard to the p values and which comparisons they are making eg. the row with Provincial site has 4 p values but it is unclear what comparisons they are referring to- is it the above rows or the rows below? Table needs to be reformatted for clarity.

Response: The table was reformatted as the described above (major comment # 1)

6.Table 2

Although there was a lack of statistical significance between many of the study groups with regard to density, the median and IQR differences within the subgroups suggest a there is a difference- just that there is insufficient power to detect it eg median density in RSV positive was 80 (IQR10-614) vs RSV neg 26 (IQR 1.9-147). If two histograms were plotted on one graph for each subgroup, the difference maybe more obvious, although not statistically significant. Ditto for antibiotics vs no antibiotic; CRP; Flu A/B; and al the clinical study groups. Suggest plotting this data as overlapping histograms for each subgroup.

Response: Thank you for the suggestion. We made histograms of colonization density by various characteristics (See separate attachment with histograms) to see if they revealed distinctions between groups. We see a lot of overlap in density between groups for each characteristic. We do not think the histograms add information that will be helpful to the reader and prefer not to include them in the manuscript.

7. P.3 l.67 103 and not 103

Response: Revised to 103

8. The authors precise that they performed binary logistic regression adjusted for age; as the study was frequency matched for age. You mean that the study was not paired matched, and then, that it was not possible to perform a conditional logistic regression?

Response: Since the study was age-frequency matched, as opposed to individually matched, we didn’t think it is necessary to use conditional logistic regression.

9. P. 17, l.261 (shifted to Page 18 line 265-266) I think that it is better to mention IQR with a “-“ and not “,” (for example 2.1-222.3)

Response: Revised

10. P.17, l.262 (shifted to Page 18 line 269) is it possible to add the p-value between proportion of cases and control above the threshold identified in the overall PERCH analysis?

Response: 3/222 (1.4%) of case and 8/650 (1.2%) of controls in Thailand had pneumococcal NP/OP PCR density above the threshold identified in the overall PERCH analysis (6.9 log10 copies/ml) (p=0.84). The p-value was added to the text.

11. P.17, l. 264 (shifted to Page 18 line 268) mL instead of ml

Response: Revised as suggested

12. P.17, l. 266 (shifted to Page 18 line 271) I don’t understand to what the p-value (p=0.31) refers? It may be better to mention “The density is highest […] even if not significant” or mention the p-values after the sentence “none of these differences reached statistical significance”

Response: This comment refers to the below sentence in the Results:

“Pneumococcal colonization density among cases was highest in those aged 12-23 months (p=0.31), and although slightly higher in males than females (p=0.43), in Nakhon Phanom vs. Sa Kaeo (p = 0.12), and in cases with severe vs. very severe pneumonia (p = 0.29), none of these differences reached statistical significance.” These p-values were obtained from a Kruskal-Wallis test, comparing pneumococcal colonization density by characteristic within study group (cases or controls). Upon further review, we have decided to revise the sentence to note the lack of statistically significant differences and remove the p-values.

13. P.17, l. 274 Only 3 cases had pneumococcus detected by PCR in whole blood? You may mention in the population description

Response: This is correct. We have added this detail to Page 18 line 258-259.

14. P. 17, l. 275 Do you mean “were colonized in NP” ?

Response: Yes, this sentence referred to NP colonization among cases and controls who were positive for pneumococcus by PCR in whole blood. Because there were so few cases and controls positive for pneumococcus in whole blood, we decided to delete this information from the text. The information remains in table 2, and we believe that this change simplifies the text.

15. P. 17, l. 278 Do you mean “in whole blood by PCR”?

Response: Yes, but because there were so few cases and controls positive for pneumococcus in whole blood and the data are in table 2, we deleted the sentence from the text.

16. P. 18, l. 286 (shifted to Page 20 line 292) “with NP colonization”

Response: We added ‘NP’ to make sure this was clear.

17. Figures 1 & 2 the caption of y-axis “percent of children” is not clear. You might mention “percent of children with pneumococcal NP colonization”

Response: It was changed to “Percent of children with pneumococcal NP colonization” as suggested

18. P. 18, l. 299 (shifted to Page 20 line 304) I think that it is better to mention 95% CI with a “-“ and not “,”

Response: Revised

19. P. 19, l. 301 (shifted to Page 20 line 306) “7 of the 8 most frequent serotypes” instead of “the top 8 serotypes”

Response: Revised as suggested

20. Page 19: line 320 (shifted to Page 21 line 327): please add a reference

Response: We added a reference supporting the potential association between antibiotic use and lower colonization density.

21. P. 20,

Line 339 (shifted to Page 22 line 346): remove “%”

Response: Removed

22. P. 22

Line 343 (Shifted to Page 22 line 351-352): also compare with Laos community carriage results

Response: Added a sentence to reference the recently published cross-sectional carriage study from Laos PDR

23. P. 24, l. 424 (Shifted to Page 26-27 line 431-437) “though …. unresolved serotypes”. This sentence is difficult to understand, you may rewrite it. I don’t know the meaning of the word “apporportioned”. Do you mean “apportioned”? What are the “unresolved serotypes”?

Response: The “unresolved serotypes” refers to isolates for which we could determine serogroup but not fully resolve the sub-type (e.g., we could identify that the isolate was serogroup 6 but not determine whether A or B). Serotypes 6A/6B by PCR could not be differentiated for the samples collected in 2013, so we applied the 6B:6A ratio from 2012 to the 2013 data. This calculation was done separately by site and case-control status which was described in the Statistical analysis, page 9 line 178-181. For the other serogroups we did not have sufficient data on the sub-types to perform this same calculation. We corrected the typographical error, with “apporportioned” revised to “apportioned”

Submitted filename: Response to reviewers_20200101.docx

Click here for additional data file.

5 Feb 2020

PONE-D-19-29663R1

Pneumococcal colonization prevalence and density among Thai children with severe pneumonia and community controls

PLOS ONE

Dear Mr. Piralam,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

You have now addressed most of the issues raised by the reviewers and the paper already clarifies most of the points brought up. However, two outstanding points remain. These are questions 2 and 8 of the reviewer. Please add the reference of the accepted paper you allude to to the current submission or, if the paper was not yet accepted, please provide the requested details in table 1 and a more completed discussion of a potential selection bias and the steps you took to minimize it in the text.

We would appreciate receiving your revised manuscript by Mar 21 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Jose Melo-Cristino, M.D., Ph.D.

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

14 Feb 2020

I would like to provide a point-by-point response to the two remaining requests related to our manuscript:

1. Major comment no. 2; the authors may add some characteristics in Table 1 (or appendix …) in order to better describe the studied population (ex: for cases % detection of pneumococcus in blood by PCR, symptoms, etc).

Response: We have added key laboratory and clinical characteristics to table 1, which complements the information in table 2 (stratified by pneumococcal NP/OP positivity). These variables include sub-site, sex, pneumococcal positivity in the NP/OP by PCR and culture, pneumococcal positivity in whole blood, and clinical symptoms (cough, fever). Also, a full description of all cases and controls in Thailand is pending publication. If that paper will be accepted before publication of this paper, we will add a reference to that paper into our paper under review at PLoS One.

2. Major comment no. 8; the authors have to clearly present the potential bias. What about a selection bias in the study if the controls are all included in a same population with a high prevalence of NP carriage of S. pneumoniae?

Response: We recognize that all case-control studies are subject to potential selection bias. To acknowledge this point, we added the following sentences to the limitations paragraph in the Discussion section (lines 412-419): Third, all case-control studies are subject to potential selection bias related to control selection. We believe that selection bias was minimized by our approach to control selection; community controls were randomly selected from comprehensive lists of children aged 1 - 59 months drawn from health services registries, which include virtually all children in each province, and were enrolled year round and frequency matched to the age-group distribution of the cases on a monthly basis. Therefore, our controls represented an unbiased community comparison group in terms of pneumococcal colonization prevalence.

Submitted filename: Response to Reviewers_20200214.docx

Click here for additional data file.

9 Apr 2020

Pneumococcal colonization prevalence and density among Thai children with severe pneumonia and community controls

PONE-D-19-29663R2

Dear Dr. Piralam,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Jose Melo-Cristino, M.D., Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

14 Apr 2020

PONE-D-19-29663R2

Pneumococcal colonization prevalence and density among Thai children with severe pneumonia and community controls

Dear Dr. Piralam:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Jose Melo-Cristino

Academic Editor

PLOS ONE