a) Percent of samples by IS481 signal strength across 8,704 dyad visits. b) Odds ratio of co-detection (OR +95% CI; rows show mothers, columns show infants, N=8,704 unique dyad visits).

Overview of qPCR test results showing A) frequency of IS481 Ct values (by visit), B) frequency of peak IS481 values (by subject and dyad), and C) number of IS481 detections across full study (by dyad).

A) Intermediate and stronger IS481 signals (color) were much more common in infants than mothers (inset numbers show cumulative frequency from left to right; Ct < 35 not shown). B) More infants exhibit stronger peak IS481 values, while dyad members together frequently exhibit weaker peak signals (dashed line). C) Venn diagrams show the frequency of subjects with ≥ 1 (left) to 4 (right) detections (Ct < 45), stratified by infants, mothers, and both dyad members (shaded region).

Study timeline.

A) Timelines of select dyads (mother/infant pairs) showing IS481 qPCR signal strength. Dyad study visits were scheduled at approximately two week intervals (median 16 days). B) Subject visits per week, stratified by IS481 signal strength (color, grey shows N.D.). C) Estimated frequency of IS481 detections by signal strength over time (GAM, mean +95% CI).

Relative risk of confirmation (columns) given initial detection (rows) in an opportunistic sub-set of 1,468 NP samples that were retested.

In a stratified random subsample of 122 NP samples with initial detections, re-test confirmation frequency increased with initial IS481 signal strength.

A) shows confirmation frequency by subsample strata. B) Odds ratio of confirmation by IS481 signal strength (OR relative to null frequency of detection across full sample library; mean and 95% CI estimated from a binomial GLM). Points show initial IS481 Ct stratified by confirmation (shape, color).

For each group of events (rows), the mean hazard ratio (HR, relative to the hazard for N.D. in the alter, Y-axis) is shown as a function of alters’ preceding IS481 signal strength (X-axis).

Overall, HRs are higher in infants than mothers, and are significantly greater than 1 in most cases (asterisk indicates 95% CI non-overlap with HR=1). Here an event signifies the first occurrence of an IS481 signal at least as strong as indicated (row labels). The infrequent occurrence of stronger IS481 signals yields large CI (bottom row, upper limit not shown).

The frequency of reported respiratory symptoms increases with stronger IS481 signals among infant visits with IS481 detections (N=733), both for any symptoms (top) and serious symptoms (bottom: excludes solely cough and/or coryza).

Curves show mean and 95% CI estimated from GAMs (one per row). Dashed lines shows marginal frequency of symptoms in infants (Any=13%; Serious=2.7%; N=8,705). Points show presence of respective symptoms by Ct value.

Relative risk of detection (RR +95%CI) per visit is greater given stronger IS481 Ct values in previous visits (no significant difference between (a) mothers and (b) infants).

Relative risk of contemporaneous detection (RR +95%CI) in ego given IS481 signal strength in alter: a) infant risk given mother’s IS481; b) mother’s risk given infant’s IS481.

Overall, stronger signals predict more frequent detections, with modest differences between the predictive strength of IS481 signals in mothers a) versus infants b).

Relative risk of respiratory symptoms by IS481 signal strength for a) infants and b) mothers.

No symptoms: N.S.

Timelines of study visits showing IS481 signals (color) for the ten infants (A-J) and ten mothers (K-T) with the strongest peak IS481 signals within each group.

Peak IS481 Ct in these infants ranges from 18.1 to 32.6 (mean=27.1) and from 30.0 to 36.4 in these mothers (mean=33.7). Clusters of detections are evident that, however, frequently contain N.D.s. Within each panel, subjects are ordered by detection frequency. Open circles mark visits with any reported respiratory symptoms. Asterisks indicate inclusion in the subgroup analysis below (Fig 7), including 8 of 10 infants and 5 of 10 mothers.

Subgroup analysis: subjects with evidence of persistent colonization, including 50 infants and 54 mothers.

A) Kaplan-Meyer survival analysis of duration of detection (i.e., time interval between the first and last detection) that accounts for right censoring (detection at final study visit, shown by diamonds; includes 20 infants and 20 mothers). Estimated median duration of detection was 55 days. Panels B-D show the number of subjects by (B) number of detections, (C) peak IS481 signal, and (D) frequency of visits with respiratory symptoms. B) Number of detections stratified by censoring (color). For mothers, both left and right censoring are included (i.e., detection at first and last visit). C) Peak IS481 Ct is much higher in infants than mothers and, unlike the full cohort (Figure 1B), is rarely a weaker signal (green: Ct >43). Not shown: 3 infants with peak Ct ranging from 23.4 to 18.1. D) Most subgroup infants experienced respiratory symptoms infrequently, while most subgroup mothers experienced no respiratory symptoms. Color shows serious symptoms (i.e., excluding simple cough and/or coryza), which occurred most frequently in a small number of infants.