Is lymphangitic streaking associated with different pathogens?

a b s t r a c t

Objectives: Little is known regarding the differences in microbiology associated with cellulitis or abscess with or without lymphangitic streaking. The objective of our study is to assess whether there are differences in the path- ogens identified from wound cultures of patients with paronychia with and without associated lymphangitis. Methods: retrospective cross-sectional study at a tertiary pediatric emergency department over 25 years. We opted to assess patients with paronychia of the finger, assuming that these cases will have a greater variety of causative pathogens compared to other cases of cellulitis and soft tissue abscess that are associated with nail bit- ing. Case identification was conducted using a computerized text-screening search that was refined by manual chart review. We included patients from 1 month to 20 years of age who underwent an incision and drainage (I&D) of a paronychia and had a culture obtained. The presence or absence of lymphangitis was determined from the clinical narrative in the medical record. We excluded patients treated with antibiotics prior to I&D as well as immune-compromised patients. We used descriptive statistics for prevalence and ?2 tests for categorical variables.

Results: Two hundred sixty-six patients met inclusion criteria. The median age was 9.7 years [IQR 4.7, 15.4] and 45.1% were female. Twenty-two patients (8.3%) had lymphangitic streaking associated with their paronychia. Pa- tients with lymphangitis streaking were similar to those without lymphangitis in terms of age and sex (p = 0.52 and p = 0.82, respectively). Overall, the predominant bacteria was MSSA (40%) followed by MRSA (26%). No sig- nificant differences were found between the pathogens in the 22 patients with associated lymphangitis com- pared to the 244 patients without.

Conclusion: Staphylococcus aureus represent the majority of pathogens in paronychia, although streptococcal spe- cies and Gram-negative bacteria were also common. Among patients with paronychia of the finger, there seems to be no association between pathogen type and presence of lymphangitic streaking.

(C) 2021

  1. Introduction

Little is known about the bacteriology of infectious lymphangitis [1]. The medical literature is peppered with case reports discussing zoonotic and atypical bacteria [2], yet Pediatric Emergency Medicine providers more commonly encounter cellulitis, abscess, or insect bite associated with lymphangitic streaking [2,3], suggesting more common skin flora or skin infection pathogens. We sought to identify whether the develop- ment of lymphangitic streaking is associated with a different pathogen than similar infections that were not associated with lymphangitic streaking. Since skin abscess involves primarily Staphylococcus aureus and there is limited ability to identify the pathogen in most cases of cel- lulitis, we opted to focus on paronychia. Finger paronychia was chosen

* Corresponding author at: Harvard Medical School Faculty, Department of Emergency Medicine, Boston Children’s Hospital, 300 Longwood Ave., Boston, MA 02115, USA.

E-mail address: [email protected] (A.A. Kimia).

for two main reasons: 1) its relatively homogenous presentation allowing comparison of similar subgroups and 2) a more heterogeneous pathogen profile that may include oral flora, as some paronychia cases may be associated with nail biting [4,5]. The primary goal of this study was to assess whether there are differences in the pathogens identified from wound cultures of patients with paronychia with and without as- sociated lymphangitis.

  1. Methods
    1. Study design

This is a retrospective cross-sectional study of consecutive patients seen at an urban tertiary care pediatric emergency department (ED). The ED serves approximately 58,000 patients per year. ED electronic medical records (EMR) of all patients presenting over a 25-year period (January 1995 to December 2019) with paronychia of finger were

0735-6757/(C) 2021

reviewed. The study protocol was approved by the hospital Institutional Review Board.

    1. Study population

We included patients from 1 month to 20 years of age who were evaluated in the ED for finger paronychia, underwent an incision and drainage (I&D) of a paronychia, and had a culture obtained. Presence or absence of lymphangitis was determined from the clinical narrative in the medical record. We excluded patients treated with antibiotics prior to I&D as well as immune-compromised patients. repeat visits were excluded if they occurred within a month of the initial visit or if there were no changes in clinical variables. Two authors were assigned to perform the manual EMR review to assure inter-reviewer agreement of case identification using the above criteria. Cases in which there was disagreement among the reviewers were resolved by consensus among the co-authors.

    1. Case identification

Case identification was conducted in two phases: 1) creating and ap- plying a natural language processing (NLP) model and 2) manually reviewing the NLP model output.

    1. Creating and training the NLP model

In order to train an NLP model and a document classifier we needed first to generate a document training set containing ED documents la- beled as those describing finger paronychia and documents that did not. For that purpose, we created a computer-assisted screening tool

[6] similar to key word search tools, using regular expressions [7-9]. The technique of regular expression matching provides a more compre- hensive and inclusive search than key word searching by including pos- sible misspelled and mistyped variations of the key word(s) of interest. The output of the computer assisted tool was then input into a word classifier. Document classification is a standard feature in NLP. Once the document classifier is trained on a certain dataset, it will present a

set of documents similar to the ones flagged as documents of interest. A document classifier, unlike word/text matching, will also provide sta- tistical metrics to the search (for example, accuracy).

    1. Manual review

The document classifier output was then reviewed by the first au- thor (RK) and a subset of the documents was reviewed by a second re- viewer (AK) to assure adequate inter-rater agreement. A kappa score was calculated for inter-reviewer agreement.

    1. Outcome measures

Our primary binary outcome is a positive wound culture for bacteria and pathogen species.

    1. Statistical analysis

We compared the prevalence and type of bacteria between patients with described lymphangitis versus patients without. As a secondary analysis, given the limitation of clinical documentation, we compared patients with documented lymphangitis to a sub-group of patients in whom lymphangitis was addressed in a negation form (i.e. – ‘no lympahgitis’ or ‘no lymphangitic streaking’).

Data were analyzed using IBM SPSS for Windows V24 (Chicago, 2017). We used Bayesian credible intervals to calculate percentages and confidence intervals for the prevalence of paronychia and lymphangitic streaking in this population. ?2 tests were used to com- pare categorical variables.

  1. Results
    1. Study group

During the study period, there were 1,185,223 ED visits with EMR available for review.

Image of Fig. 1

Fig. 1. Case identification.

Image of Fig. 2

Fig. 2. Wound culture results.

Table 1

Wound culture results

Lymphangitic streaking [+] N = 22

Lymphangitic streaking [-] N = 244



12/22 (55%)

95/244 (39%)



5/22 (23%)

63/244 (26%)


Streptoccocal Species

5/22 (23%)

52/244 (21%)


Gram Negative Bacteria

0/22 (0%)

29/244 (12%)


Coagulase-Negative Staphylococci

0/22 (0%)

2/244 (0.8%)

Negative culture

0/22 (0%)

3/293 (1.2%)

Lymphangitic streaking [+] N = 22

Note describing ‘no Lymphangitic streaking’ N = 66



12/22 (55%)

28/66 (42%)



5/22 (23%)

14/66 (21%)


Streptoccocal Species

5/22 (23%)

12/66 (18%)


Gram Negative Bacteria

0/22 (0%)

11/66 (17%)


Coagulase-Negative Staphylococci

0/22 (0%)

1/66 (1.5%)

Negative culture

0/22 (0%)

0/66 (0%)

    1. Case identification and modeling performance

Our NLP model had a sensitivity of 94.6% (95% C.I. 88.7-100%) with a positive predictive value for finger paronychia of 0.45%. Of the 1099 pa- tients with finger paronychia and no exclusion criteria, 346 patients had no collection to be drained and 411 had an I&D conducted but no pus sent for culture. Two hundred sixty-six patients met study criteria (Fig. 1).

The decision to obtain a wound culture was at the discretion of the treating physician. A comparison between the group of patients who had an I&D and wound culture and the group who had an I&D with no wound culture showed no difference in terms of age (p = 0.49) or gen- der (p = 0.403).

    1. Study cohort

Two hundred sixty-six patients met inclusion criteria. The median age was 9.7 years [IQR 4.7, 15.4] and 45.1% female. Twenty-two patients (8.3%) had lymphangitic streaking associated with their paronychia. Inter-rater agreement on the presence of lymphangitis was excellent with a ? = 0.81. Patients with lymphangitic streaking were no different

from those without lymphangitis in terms of age and sex (p = 0.52 and

p = 0.82, respectively).

    1. Bacteriology

Overall, the predominant bacteria was methicillin-sensitive Staphy- lococcus aureus (MSSA) (40%) followed by MRSA (26%), as presented in Fig. 2. No significant differences were found between the pathogens in the 22 patients with associated lymphangitis compared to the 244 pa- tients without. The prevalence of different bacteria species are pre- sented in the upper portion of Table 1.

    1. Sub-analysis for patients in whom lymphangitic streaking was addressed in the narrative

Documentation addressing lymphangitis varied significantly among patients with no streaking. Some documentation did not address lymphangitis while other documentation (45/244 or 18%) mentioned lymphangitic streaking as part of the criteria to return to the Emergency Department. Of the 244 patients with no lymphangitis, there were 66 patients for whom lymphangitis was addressed in a negation form

(i.e. ‘no lymphangitis,’ ‘no streaking’). The two groups, those in whom lymphangitis was addressed and those where it did not were no differ- ent in terms of age or sex (p = 0.91 and p = 0.1, respectively).

  1. Discussion

We have identified a large cohort of patients presenting with finger paronychia. A minority of the patients had associated lymphangitic streaking. In our pediatric cohort, the pathogens involved were primar- ily skin and oral flora in equal proportions, whether lymphangitic streaking was present or not.

Acute paronychia is defined as an inflammation of fingers or toes in one or more of the three nail folds [4]. The medical literature is pep- pered with controversies regarding the management of paronychia [10]. While most of the literature involves adult populations more than pediatric, published area of controversies involve the role of soaks, the role of added antibiotics [11,12], and the need to send cul- tures from a drainable collection. Fowler et al. presented data on admit- ted adults with Wound infections of the hands and finger [13]. Among a subset of patients with paronychia, only 4% of the wound cultures were positive and most involved polymicrobial results that were not helpful in directing care. With concerns for pseudomonas species infections, some providers recommend soaks of either Burow’s solution or a 1% Acetic acid [14,15] over other measures.

These controversies are well represented in the management of our patients, where decisions to perform I&D or use soaks and whether to obtain cultures from drainable collections varied significantly.

Lymphangitic streaking [1,16] has been associated with a variety of

uncommon pathogens in a variety of case reports, suggesting that type of pathogens may be represented in higher frequency if infectious lymphangitic streaking is present. Pathogens such as arthropods [2], herpetic whitlow [17], and Rickettsial infections [18] were among these, especially if insect bites were suspected as the mode of acquiring these infections [3,19]. Our cohort was limited to paronychia and we have identified common oral and skin flora to associated with lymphangitic streaking.

    1. Limitations

As a retrospective study, our data depend on the quality of the clin- ical documentation. As such, describing lymphangitic streaking is af- fected by individual practice of thorough documentation as well as beliefs that lymphangitic streaking in the setting of paronychia is clini- cally pertinent. We hope that our secondary analysis mitigated this lim- itation to some extent. Secondly, our sample sizes were small, despite screening a very large corpus of data (over 1 million records). Our in- ability to show difference sin prevalence and bacteria type may have been affected by the sample size. We do, however, believe that if such differences exist then they are not striking.

In summary, our current study demonstrated that the presence of lymphangitic streaking associated with acute paronychia may not

have a substantial effect on the choice of antimicrobials used or the de- cision to obtain a wound culture when managing these patients. Future directions should include other Skin infections for which wound culture is feasible, especially if insect bites are associated with the initial infec- tion/exposure.

Financial support


Declaration of competing interest

None of the authors have any conflict of interest either financial or other. That holds true to both authors and family members.

This project was not funded and done at kind as part of the authors academic interest.


  1. Kano Y, Momose T. Acute lymphangitis. Cleve Clin J Med. 2020;87(3):129-30.
  2. Marque M, et al. Superficial lymphangitis after arthropod bite: a distinctive but underrecognized entity? Dermatology. 2008;217(3):262-7.
  3. Meyer Sauteur PM, Orban P, Theiler M. Recurrent superficial lymphangitis after in- sect bites. Arch Dis Child. 2018;103(3):296.
  4. Leggit JC. Acute and chronic paronychia. Am Fam Physician. 2017;96(1):44-51.
  5. Shafritz AB, Coppage JM. Acute and chronic paronychia of the hand. J Am Acad Orthop Surg. 2014;22(3):165-74.
  6. AA Kimia. DocumentReviewTools; 2018 [[cited 2021 01.02.2021]; Available from:].
  7. Kimia AA, et al. An introduction to natural language processing: how you can get more from those electronic notes you are generating. Pediatr Emerg Care. 2015;31 (7):536-41.
  8. Friedl J. Mastering Regular Expressions. 3rd ed.. Safari Books Online (Firm); 2006 [online resource 534 pages].
  9. Nield T. An Introduction to Regular Expressions : Decoding Simple Regex Features to Match Complex Text Patterns. Sebastopol, CA: O’Reilly Media; 2019; 1 [online re- source (1 volume)].
  10. Raff AB, Kroshinsky D. Cellulitis: a review. JAMA. 2016;316(3):325-37.
  11. Schmitz GR, et al. Randomized controlled trial of trimethoprim-sulfamethoxazole for uncomplicated skin abscesses in patients at risk for community-associated methicillin-resistant Staphylococcus aureus infection. Ann Emerg Med. 2010;56 (3):283-7.
  12. Duong M, et al. Randomized, controlled trial of antibiotics in the management of community-acquired skin abscesses in the pediatric patient. Ann Emerg Med. 2010;55(5):401-7.
  13. Fowler JR, Ilyas AM. Epidemiology of adult acute hand infections at an urban medical center. J Hand Surg Am. 2013;38(6):1189-93.
  14. Nagoba BS, et al. Acetic acid treatment of pseudomonal wound infections–a review. J Infect Public Health. 2013;6(6):410-5.
  15. Madhusudhan VL. Efficacy of 1% acetic acid in the treatment of chronic wounds in- fected with Pseudomonas aeruginosa: prospective randomised controlled clinical trial. Int Wound J. 2016;13(6):1129-36.
  16. Simma L, Barrett MJ. Ascending upper limb lymphangitis. BMJ Case Rep. 2017;2017.
  17. Lopez Valle A, Pena M Puerta, Honrubia V Monsalvez. Lymphangitis associated with herpetic whitlow. Med Clin (Barc). 2020;155(11):518.
  18. Miguelez Ferreiro S, Rodriguez D Navalpotro. Tubular acute lymphangitis caused by Rickettsia sibirica mongolitimonae. Enferm Infecc Microbiol Clin. 2020;38(10): 506-7.
  19. Piccolo V, et al. Superficial lymphangitis after insect bite. J Pediatr. 2013;163(1) [299 e1].