A nasopharyngeal swab (NPS) is the most commonly collected specimen when laboratory diagnosis of respiratory viral infections, including COVID-19, is required. As an appropriate NPS collection technique is critical for reliable results and for minimizing patient discomfort during the procedure (1), so is the proper use of personal protective equipment to minimize the risk of health care professionals' infection (2). Current recommendations for obtaining NPS are not uniform. The World Health Organization (WHO) recommends leaving the swab in place for a few seconds after reaching the nasopharynx before withdrawing it (3), the Center for Disease Control and Prevention (CDC) recommends “gently rubbing and rolling” the swab after reaching the nasopharynx and leaving it in place for several seconds before withdrawing it (4). To further address these issues and provide additional guidance, Marty et al. published an article with practical instructions and a descriptive video in The New England Journal of Medicine early in the pandemic (5). Nevertheless, we wondered if the procedure could be further changed to reduce the proportion of patients that find it frustrating or even painful, and simplified without compromising the quality of the sample.

The aim of this study was to investigate whether the NPS collection could be simplified to only one rotation with no waiting period in the nasopharynx yet retain the sample quality but decrease the discomfort felt by the participants.

The mean age of the 76 study participants was 36 ± 11 years, and 56 (74%) were female. The observed median discomfort score (DS) for sample A was 3 (first-third quartile; 2–5), whereas for sample B it was 6 (first-third quartile; 4.25–7) and for sample C it was also 6 (first-third quartile; 4–7), as shown in Figure 1A. The DS of sample A was significantly lower when compared to samples B and C (p < 0.001). No statistically significant difference was observed between samples B and C (p = 0.59).

There was no significant difference in the log [UBC copies/sample] between samples A and C (5.2 ± 0.6 vs. 5.3 ± 0.5; p = 0.15). However, sample B had higher log [UBC copies/sample] values (5.5 ± 0.5) compared to samples A and C (p < 0.001 and p = 0.01, respectively), as presented in Figure 1B.

Although younger participants ( ≤ 35 years) reported significantly higher DS in comparison to older participants (sample A, median: 4 vs. 3, p = 0.007; sample B, median: 6 vs. 5, p = 0.02; sample C, median: 7 vs. 5.5, p = 0.056), the two age groups did not differ in mean log [UBC copies/sample]. Moreover, comparable DS were observed between men and women (sample A, median: 3.5 vs. 3, p = 0.843; sample B, median: 5.5 vs. 6, p = 0.815; sample C, median: 6 vs. 6, p = 0.728) as were also log [UBC copies/sample] p > 0.1.

A weak positive correlation between DS and log [UBC copies/sample] was observed (Spearman's rho = 0.202; p = 0.002; Figure 2). Additional analysis on samples A, B and C together showed that NPS collected with minimal subject discomfort (DS < first quartile) had a lower log [UBC copies/sample] than NPS collected with high subject discomfort (DS > third quartile; 4.9 ± 0.4 vs. 5.5 ± 0.3; p < 0.001). When looking separately, only sample A showed that NPS collected with minimal subject discomfort (DS < first quartile) had a lower log [UBC copies/sample] than NPS collected with high subject discomfort (4.9 ± 0.2 vs. 5.7 ± 0.4; p < 0.001) and no link between DS and log [UBC copies/sample] was observed in samples B (5.4 ± 0.6 vs. 5.3 ± 0.4; p = 0.42) and C (5.2 ± 0.5 vs. 5.4 ± 0.3; p = 0.22).

The results of this study show no statistically significant difference in the UBC gene copies/sample between NPS collected with one or five rotations (sample A vs. sample C). These results are also supported by the study from Kinloch et al. (7). They similarly found no difference in DNA/RNA recovery between an “in–out” NPS and a NPS with 10 s of rotation. Our study found a significantly higher DS of NPS with five rotations compared to one rotation NPS. Interestingly, in contrast to our results, Kinloch et al. found no differences in discomfort between the procedures they used (7).

Despite all efforts for uniformity of the collection procedure, differences in log [UBC copies/sample] values were still found between study participants, which may indicate the diversity of the nasal mucosa and anatomy of the individuals (7, 8). In the study by Kinloch et al., RPP30 and RNase p-values also varied markedly regardless of the swab collection technique (7).

Although only a weak correlation between DS and log [UBC copies/sample] was observed between the participants that reported almost no discomfort and those that reported the highest DS, the difference between log [UBC copies/sample] reached a statistically significant value. This observation theoretically means that samples with higher human cell counts would have an impact on specific target (pathogen) detection sensitivity, however due to the study design such a conclusion can't be drawn. Similar results were also observed by Kinloch et al., who showed statistically significant higher DNA/RNA recovery together with higher DS in Asian participants of their study population (7, 8).

Finally, sample B was initially collected because we wanted to gain insight into the DS and quality of the sample if one sample is collected right after another. Contrary to expectations, we found UBC copies/sample in sample B higher than in sample A and even sample C, however, the difference was rather small. We believe that minor trauma to the nasal pseudostratified ciliated epithelium may have occurred with the collection of sample A, which could explain the higher UBC copies/sample in sample B.

A few limitations of the study must be kept in mind when interpreting the results. Firstly, the study population was rather small, due to difficulties of acquiring willing participants after 2 years of frequent NPs collections for SARS-CoV-2 screening. Secondly, only adults were included in the study, therefore the results do not necessarily apply to children, especially young children who do not yet understand why NPs is being collected and are usually already under stress at the collection site which makes them hindered to give an rational DS score. Thirdly, the study was performed on healthy individuals and different results might occur in individuals with acute respiratory infection, who usually secrete increased amounts of mucus and cellular material. Finally, if NPSs were collected by multiple healthcare professionals, although with the same protocol, additional variance might have been observed in both, quality and DS. Future studies should address these questions.

In conclusion, this study shows that a simplified NPS collection procedure has the same recovery of human DNA as a NPS collected with five rotations. However, the collection time is shorter and the experience is less unpleasant for patients. We firmly believe that the NPS collection procedure could be improved globally (3, 4, 7).

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by National Medical Ethics Committee of Slovenia (No. 0120-281/2022/3). The patients/participants provided their written informed consent to participate in this study.

TU: conceptualization, methodology, visualization, and writing—original draft preparation. RK: writing—reviewing and editing, supervision, and reviewing. JŠ: software, data curation, and writing—original draft preparation. MJ: investigation, validation, and reviewing. MP: supervision, reviewing, and editing. All authors contributed to the article and approved the submitted version.