A quality improvement study on the relationship between intranasal povidone-iodine and anesthesia and the nasal microbiota of surgery patients

Abstract

Introduction: The composition of the nasal microbiota in surgical patients in the context of general anesthesia and nasal povidone-iodine decolonization is unknown. The purpose of this quality improvement study was to determine: (i) if general anesthesia is associated with changes in the nasal microbiota of surgery patients and (ii) if preoperative intranasal povidone-iodine decolonization is associated with changes in the nasal microbiota of surgery patients.

Materials and methods: One hundred and fifty-one ambulatory patients presenting for surgery were enrolled in a quality improvement study by convenience sampling. Pre- and post-surgery nasal samples were collected from patients in the no intranasal decolonization group (control group, n = 54). Pre-decolonization nasal samples were collected from the preoperative intranasal povidone-iodine decolonization group (povidone-iodine group, n = 97). Intranasal povidone-iodine was administered immediately prior to surgery and continued for 20 minutes before patients proceeded for surgery. Post-nasal samples were then collected. General anesthesia was administered to both groups. DNA from the samples was extracted for 16S rRNA sequencing on an Illumina MiSeq.

Results: In the control group, there was no evidence of change in bacterial diversity between pre- and post-surgery samples. In the povidone-iodine group, nasal bacterial diversity was greater in post-surgery, relative to pre-surgery (Shannon's Diversity Index (P = 0.038), Chao's richness estimate (P = 0.02) and Inverse Simpson index (P = 0.027). Among all the genera, only the relative abundance of the genus Staphylococcus trended towards a decrease in patients after application (FDR adjusted P = 0.06). Abundant genera common to both povidone-iodine and control groups included Staphylococcus, Bradyrhizobium, Corynebacterium, Dolosigranulum, Lactobacillus, and Moraxella.

Conclusions: We found general anesthesia was not associated with changes in the nasal microbiota. Povidone-iodine treatment was associated with nasal microbial diversity and decreased abundance of Staphylococcus. Future studies should examine the nasal microbiota structure and function longitudinally in surgical patients receiving intranasal povidone-iodine.

Fig 1. Study design components.

A quality improvement study schematic outlining sample collection, nasal povidone-iodine (PI) application, and sequencing method.

Fig 2. Taxa composition with relative abundance.

Bar plots display relative abundance (%) of top 10 phyla and genera composition between pre- and post-surgery nasal microbiota in the control group (A and C) and PI group (B and D). Rare taxa are classified as “uncultured”. The legend displays the color coding of genera and phyla to which these taxa belong.

Fig 3. Alpha diversity of bacteria community.

Box and whisker plots to compare bacteria community alpha diversity between pre-surgery and post-surgery in control (Prec, and Postc) and povidone-iodine (pi) (Prepi and Postpi) groups. (A) and (B) Chao’s richness, (C) and (D) Shannon’s diversity, and (E) and (F) Inverse Simpson’s diversity. P-value was based on pairwise sample comparison in Wilcoxon signed-rank test. Significant values (p-values) between pre- and post-surgery are displayed on the plots. P-values <0.05 are considered significant.

Fig 4. Beta diversity of bacterial community composition.

Differences of nasal bacterial community compositions across treatments groups. Non-metric multidimensional scaling (nMDS) plot of the Bray-Curtis dissimilarity for bacterial community of nasal samples collected pre- and post-surgery from (A) control group and (B) povidone-iodine group. Ellipses illustrates standard error around centroid of pre- and post- nasal community samples in each treatment group and colored by treatment. P- value <0.05 was considered a significant difference between bacterial community composition between pre- and post-surgery samples. The nMDS stress levels for PI group = 0.20 and control group = 0.25.