ABSTRACT
Objective
To describe the demographics, clinical presentation, treatment, and complications of piercing‐induced auricular perichondritis (P‐IAP) and non‐PIAP.
Methods
We reviewed the medical records of consecutive patients diagnosed and treated for P‐IAP and P‐FAP between 2006 and 2021. The retrieved data included demographics, comorbidities, etiology, location, treatment, complications, and outcome, and they were compared between the two groups.
Results
In total, 217 patients fulfilled the inclusion criteria. Of them, 120 (55.3%) comprised the P‐IAP group, which was characterized by significantly younger age and fewer comorbidities. The 97 non‐PIAP patients tended to present with more systemic and regional signs, while the PIAP group presented with localized signs and significantly more abscess formation and significantly more colonization with Pseudomonas aeruginosa . Time from causative event to symptom onset was significantly shorter in the PIAP group (18.9 ± 38.7 days) compared to the non‐PIAP group (29.8 ± 62.9 days, p < 0.021), which also had longer hospitalization. Complications included drainage or debridement, recurrent disease, > 21 days of antibiotic treatment, or requirement of a central line. Abscess formation was the most significant variable associated with complications (OR 35.75 [95% CI: 13.12–97.35, p < 0.001] for the entire cohort, OR 35.59 [95% CI: 10.84–116.84, p < 0.001] for the PIAP group, and OR 47.30 [95% CI: 5.33–419.67, p = 0.001] for the non‐PIAP group). No other variables reached a level of significance.
Conclusion
PIAP has different demographics, comorbidities, disease course, and microbiology compared to non‐PIAP. Abscess formation is an independent factor for complications.
Level of Evidence
3.
Keywords: auricular perichondritis, cauliflower ear, piercing induced perichondritis
The demographics, clinical presentation, treatment, and complications of piercing‐induced auricular perichondritis (P‐IAP) and piercing‐free auricular perichondritis (P‐FAP) were studied. P‐IAP has different demographics, comorbidities, disease course, and microbiology compared to P‐FAP. Abscess formation is an independent factor for complications.
1. Introduction
Auricular perichondritis (AP) is an inflammation of one of the perichondria of the pinna. The most common inflammatory type is infectious, usually due to Pseudomonas aeruginosa [1, 2]. The infective microorganism directly invades the auricular tissues and perichondrium through either minor skin integrity loss (i.e., local trauma, insect bites, and self‐injury) or major insult, such as in ear piercing, burn, and more [3, 4, 5, 6, 7, 8]. If left untreated, AP can cause auricular deformity (“cauliflower ear”), as well as extension of the infective process to deeper tissues.
Ear piercing is common worldwide, and multiple piercings of the pinna have become increasingly fashionable among both males and females [9]. “High piercing” refers to piercing of one of the auricular cartilages, and it, too, has gained in popularity in recent years [9]. Infection subsequent to high piercing is not uncommon because most piercings are performed by nonmedical practitioners, such as jewelers, hairdressers, or tattooists, who lack basic knowledge of antiseptic principles or do not follow them.
The auricles are exposed body parts in most cultures, and prompt diagnosis and treatment are necessary to prevent potentially devastating cosmetic and health sequelae [10]. Empiric treatment usually consists of ciprofloxacin with or without incision and drainage of the sub‐perichondral abscess. Not uncommonly, meticulous debridement of necrotic cartilage will also be needed.
The objectives of the current study are to compare the demographics, clinical presentation, treatment, and complications of piercing‐induced AP (PIAP) with non‐PIAP.
2. Materials and Methods
2.1. Ethical Considerations
This retrospective anonymized medical chart analysis was approved by the local research institute ethics committee (0046‐23‐TLV), which waived informed consent.
2.2. Patients
The patients in this retrospective cohort study had been diagnosed and treated for AP at a single tertiary referral center between 2006 and 2021. The medical charts of all adults (older than 18 years of age) who were referred, diagnosed, and treated for AP were identified by means of ICD‐9 codes (Supporting Informatio 1) were reviewed. Exclusion criteria were pregnant women, lack of essential data, and drug allergy that limited the choice of antibiotic treatment.
PIAP was defined by AP, which was clinically evident during the first 21 days following piercing and involved the pierced cartilage.
Non‐PIAP was defined by AP in one of the following conditions:
I. Never‐pierced auricle.
II. Disease epicenter that does not involve a pierced cartilage.
III. Disease epicenter that involves a pierced cartilage, which was pierced more than 21 days before the pathology became clinically evident.
The retrieved data included demographics, comorbidities, etiology, location, treatment, and outcome. A complication was defined as a revision of drainage or debridement under local anesthesia (other than the presenting incision and drainage), a recurrence of disease necessitating rehospitalization, the need for long‐term (> 21 days) antibiotic treatment, or for treatment necessitating a central line. The etiology was “piercing” only when the epicenter of the perichondritis was evident on the piercing site regardless of the interval of time between piercing and the appearance of AP.
2.3. Statistical Analyses
Categorical variables were described as frequencies and percentages. Continuous variables were assessed for normal distribution using histograms and Q–Q plots and were reported as mean ± standard deviation (STD) or median and interquartile range, as appropriate. Continuous variables were compared using independent t tests or Mann–Whitney U tests, depending upon distribution. Categorical variables were compared using chi‐square tests or Fisher's exact tests. Multivariate analysis was performed using logistic regression. Variables found to be significant in univariate analyses were included in the multivariate model. Statistical significance was defined as a two‐tailed p < 0.05. Data were compiled in Excel format and analyzed using IBM SPSS Statistics for Windows, version 25, IBM Corp., Armonk, NY, USA, 2017.
3. Results
3.1. Cohort
The medical charts of a total of 467 patients were reviewed, of which 217 patients (153 [70.5%] females and 64 [29.5%] males) were diagnosed with AP and met the inclusion criteria (Supporting Information 2). The mean (STD) age of the cohort was 36.5 (19.6) years, and (73.7%) had underlying comorbidities, with diabetes mellitus being the most frequent (n = 57, 26.3%) followed by smoking habits (n = 24, 11.1%) and immunosuppression (n = 5, 2.3%).
One‐hundred and twenty (55.3%) participants had auricular piercings on one pinna. The helix was the most common epicenter of infection (n = 107, 49.3%), followed by the tragus (n = 8, 3.7%), and other auricular subsites (n = 6, 2.8%). The pathology was so extensive, with involvement of more than one auricular subsite, that it was not possible to identify the disease epicenter in 96 (44.2%) cases. Forty‐seven (21.7%) patients developed a local abscess, 10 (4.6%) patients presented with fever (< 38.1°C), and 23 (10.6%) patients were diagnosed as having regional palpated lymphadenopathy (Table 1).
TABLE 1.
Demographics, etiology, presentation, treatment, cultures, laboratory findings, and complications.
| Variables | n (%) or mean ± STD | |
|---|---|---|
| Demographics | Males | 64 (29.5) |
| Age; years, mean ± STD | 36.5 ± 19.6 | |
| Comorbidities | Any comorbidity | 160 (73.7) |
| Diabetes mellitus | 57 (26.3) | |
| Smoker | 24 (11.1) | |
| Immunosuppression | 5 (2.3) | |
| Etiology | Piercing | 120 (55.3) |
| Post traumatic | 13 (6.0) | |
| Herpers zoster | 12 (5.5) | |
| Skin integrity loss | 10 (4.6) | |
| Recent operation | 7 (3.2) | |
| Water exposure | 6 (2.8) | |
| Allergic skin reaction | 4 (1.8) | |
| Unknown | 45 (20.7) | |
| Epicenter | Helix | 107 (49.3) |
| Tragus | 8 (3.7) | |
| Other | 6 (2.8) | |
| Unknown | 96 (44.2) | |
| Presentation | Abcess | 47 (21.7) |
| Lymphadenopathy | 23 (10.6) | |
| Fever > 38.1°C | 10 (4.6) | |
| Treatment (prehospital) | Any treatment | 217 (100) |
| Local treatment | 30 (13.8) | |
| Systemic treatment | 187 (86.2) | |
| Treatment (hospital) | Systemic ATB | 215 (99.1) |
| Local treatment | 106 (48.8) | |
| Systemic steroids | 4 (1.8) | |
| Ciprofloxacin + second ATB | 145 (66.8) | |
| Ciprofloxacin + cefazolin | 109 (50.2) | |
| Ciprofloxacin as a mono ATB | 36 (6.1) | |
| Cultures | Obtained | 97 (44.7) |
| No growth | 14 (6.5) | |
| Pseudomonas aeruginosa | 40 (18.4) | |
| Staphylococcus aureus | 15 (6.9) | |
| CoNS | 7 (3.2) | |
| Other pathogens a | 12 (5.5) | |
| Blood work | WBC count (K/μL) | 9.0 ± 2.4 |
| Absolute neutrophil count (K/μL) | 5.9 ± 2.2 | |
| Neutrophil % | 64.1 ± 10.0 | |
| CRP (mg/dL) | 21.2 ± 33.5 | |
| Complications | Any complication | 64 (29.5) |
| Treatment change | 30 (18.3) | |
| Drainage under local anesthesia | 64 (29.5) | |
| Drainage under general anesthesia | 53 (24.4) | |
| Recurrent or chronic infection | 6 (2.8) | |
| Central venous catheter | 16 (7.4) | |
Abbreviations: ATB, antibiotics; CoNS, coagulase‐negative staphylococci; CRP, C‐reactive protein; STD, standard deviation; WBC, white blood cells.
Enterobacter cloacae complex, Staphylococcus epidermidis, Morganella morganii, Aspergillus flavus, Trueperella bernardiae, Actinomyces spp.
Topical treatment was initiated in 30 (13.8%) patients, and 79 (36.4%) patients were given a systemic (oral) treatment, the most common of which were amoxicillin–clavulanate and cefuroxime (n = 31, 14.3% and n = 17, 7.8%, respectively). Hospitalization stay ranged from 2 to 30 days. The microbial cultures that were obtained for 97 (44.7%) participants isolated the pathogens P. aeruginosa and Staphylococcus aureus (n = 40, 18.4% and n = 15, 6.9%, respectively). The median (STD) white blood count was 9.0 (2.4) K/μL, and the median (STD) absolute and relative neutrophil counts were 5.9 (2.2) K/μL and 64.1% (10.0%), respectively.
Complications associated with AP were reported by 64 (29.5%) participants. Forty (18.3%) patients required at least one treatment regimen change, which was unrelated to drug allergy or adverse effect (nonresponders) but rather to insufficient response. Sixty‐four (29.5%) patients required additional incision and drainage during their hospitalization, and six (2.8%) patients developed recurrent or chronic infections. The mean time from symptom onset to hospitalization was 6.9 ± 12.86 days, and the mean duration of hospitalization was 5.3 ± 5.2 days. The mean total treatment duration was 11.1 ± 4.4 days (Table 1, Figure 1). Figure 2 illustrates an example of an AP lesion.
FIGURE 1.
Box and whiskers plot showing the entire cohort as well as stratification according to the existence of piercing and complication as a function of time (days). (A) Latency from insult and disease onset; (B) time from symptom onset to hospitalization; (C) hospitalization time; and (D) total treatment duration.
FIGURE 2.
Examples of three cases of auricular perichondritis. (A) Piercing‐free auricular perichondritis with no abscess formation in a diabetic 68‐year‐old male; (B) antihelix auricular perichondritis in a 32‐year‐old healthy male; (C and D) helical piercing‐induced auricular perichondritis with retroauricular extension in a 22‐year‐old healthy male.
3.2. Comparison Between Patients With and Without Auricular Piercing
Stratification of the cohort according to the presence (n = 120) and absence (n = 97) of auricular piercing revealed major significant differences. The majority of both the piercing and non‐piercing groups was comprised of females (80.0% and 58.8%, respectively, p = 0.001 for both). The patients in the PIAP group were significantly younger, with a mean ± STD of 25 ± 10.8 years compared to 51 ± 18.5 years in the non‐PIAP group (p < 0.001).
The prevalence of a comorbidity was significantly more common in the non‐PIAP group (50.5%) compared to the PIAP group (6.7%), p < 0.001. Additionally, each tested comorbidity, including diabetes (13.4% vs. 0.8%, p < 0.001), smoking (17.5% vs. 5.7%, p = 0.006), and immunosuppression (5.2% vs. 0.0%, p = 0.017), was less prevalent in the PIAP group compared to the non‐PIAP group. The etiology of all of the cases in the PIAP group was defined to be the piercing itself. In contrast, herpes zoster (12.4%), a recent auricular operation (e.g., tympanoplasty [7.2%]), allergic skin reaction (4.1%), post‐trauma (12.4%), and other causes of auricular skin integrity loss (7.2%) were also reported in the non‐PIAP group.
No specific epicenter of the disease was identified in the non‐PIAP group. As a result, 97.9% of the pathologies in the non‐PIAP group were classified as unspecified epicenter, compared to 0.8% in the PIAP group (p < 0.001). A helical epicenter was identified in 87.5% and 2.1% in the PIAP and non‐PIAP groups, respectively (p < 0.001). A tragal epicenter was identified in 6.7% of the PIAP group and in 0.0% of the non‐PIAP group, p < 0.001. The AP in the non‐PIAP group tended to present with more systemic and regional signs, such as fever > 38.1°C (8.2%) and lymphadenopathy (13.4%) compared to the PIAP group (1.7% and 8.3%, p = 0.045 and p = 0.228, respectively), while the presentation in the PIAP group was more local, with abscess formation in 28.3% compared to 13.4% in the non‐PIAP group (p = 0.008) (Table 2).
TABLE 2.
Demographics, etiology, presentation, treatment, cultures, laboratory findings, and complications of piercing and non‐piercing groups.
| Variables | Non‐piercing; n (%) | Piercing; n (%) | p | |
|---|---|---|---|---|
| Demographics | Males | 40 (41.2) | 24 (20.0) | 0.001 |
| Age; years, mean ± STD | 51 ± 18.5 | 25 ± 10.8 | < 0.001 | |
| Comorbidities | Any comorbidity | 49 (50.5) | 8 (6.7) | < 0.001 |
| Diabetes mellitus | 13 (13.4) | 1 (0.8) | < 0.001 | |
| Smoker | 17 (17.5) | 7 (5.8) | 0.006 | |
| Immunosuppression | 5 (5.2) | 0 (0) | 0.017 | |
| Etiology | Piercing | 0 (0) | 120 (100) | < 0.001 |
| Post traumatic | 12 (12.4) | NA | ||
| Herpers zoster | 12 (12.4) | NA | ||
| Skin integrity loss | 7 (7.2) | NA | ||
| Recent operation | 7 (7.2) | NA | ||
| Water exposure | 5 (5.2) | NA | ||
| Allergic skin reaction | 4 (4.1) | NA | ||
| Unknown | 33 (34) | NA | ||
| Epicenter | Helix | 2 (2.1) | 105 (87.5) | < 0.001 |
| Tragus | 0 (0.0) | 8 (6.7) | < 0.001 | |
| Other | 0 (0.0) | 6 (5) | < 0.001 | |
| Unknown | 95 (97.9) | 1 (0.8) | < 0.001 | |
| Presentation | Abcess | 13 (13.4) | 34 (28.3) | 0.008 |
| Lymphadenopathy | 13 (13.4) | 10 (8.3) | 0.228 | |
| Fever > 38.1°C | 8 (8.2) | 2 (1.7) | 0.045 | |
| Treatment (prehospital) | Local treatment | 10 (10.3) | 20 (16.7) | 0.177 |
| Local and systemic treatment | 5 (5.2) | 15 (21.3) | 0.238 | |
| Treatment (hospital) | Systemic ATB | 96 (99) | 119 (99.2) | > 0.999 |
| Local treatment | 45 (46.4) | 61 (50.8) | 0.515 | |
| Systemic steroids | 3 (3.1) | 1 (0.8) | 0.327 | |
| Ciprofloxacin + second ATB | 56 (57.7) | 89 (74.2) | 0.011 | |
| Ciprofloxacin + cefazolin | 42 (43.3) | 67 (55.8) | 0.066 | |
| Ciprofloxacin as a mono ATB | 23 (23.7) | 13 (10.8) | 0.011 | |
| Cultures | Obtained | 34 (35.1) | 63 (52.5) | 0.01 |
| No growth | 5 (14.7) | 9 (14.3) | > 0.999 | |
| Pseudomonas aeruginosa | 7 (20.6) | 33 (52.4) | 0.002 | |
| Staphylococcus aureus | 2 (5.9) | 13 (20.6) | 0.055 | |
| CoNS | 5 (14.7) | 2 (3.2) | 0.092 | |
| Other pathogens a | 10 (29.4) | 2 (3.2) | < 0.001 | |
| Blood work | WBC count (K/μL) | 8.78 ± 2.45 | 9.27 ± 2.36 | 0.157 |
| Absolute neutrophil count (K/μL) | 5.7 ± 2.2 | 6.1 ± 2.2 | 0.13 | |
| Neutrophil % | 64.5 ± 10.1 | 63.8 ± 9.9 | 0.653 | |
| CRP (mg/dL) | 33.4 ± 41.4 | 10.3 ± 18.7 | < 0.001 | |
| Complications | Any complication | 29 (29.9) | 35 (29.2) | 0.907 |
| Treatment change | 16 (16.5) | 14 (11.7) | 0.306 | |
| Drainage under local anesthesia | 29 (29.9) | 35 (29.2) | 0.907 | |
| Drainage under general anesthesia | 21 (21.6) | 32 (26.7) | 0.392 | |
| Recurrent or chronic infection | 3 (3.1) | 3 (2.5) | > 0.999 | |
| Central venous catheter | 13 (13.4) | 3 (2.5) | 0.002 | |
Abbreviations: ATB, antibiotics; CoNS, coagulase‐negative staphylococci; CRP, C‐reactive protein; STD, standard deviation; WBC, white blood cells.
Enterobacter cloacae complex, Staphylococcus epidermidis, Morganella morganii, Aspergillus flavus, Trueperella bernardiae, Actinomyces spp.
P. aeruginosa was significantly more common in the PIAP group (52.4%) compared to the non‐PIAP group (20.6%), p = 0.002, while S. aureus was more common in the PIAP group (20.6%) compared to the non‐PIAP group (5.9%), a difference that did not reach a level of significance (p = 0.055). Other pathogens, namely, Enterobacter cloacae complex, Staphylococcus epidermidis, Morganella morganii, Aspergillus flavus, Trueperella bernardiae, and Actinomyces spp., were isolated more commonly from the non‐PIAP group (29.4%) compared to the PIAP group (3.2%), p < 0.001. White blood and neutrophil counts were similar between the groups, but the C‐reactive protein levels were significantly higher in the non‐PIAP group (33.4 ± 41.4 mg/dL) compared to the PIAP group (10.3 ± 18.7 mg/dL), p < 0.001. Complication rates were similar between the groups, but the necessity for a central venous catheter was more common in the non‐PIAP group (13.4%) than in the PIAP group (2.5%), p = 0.002 (Table 2).
Time from the causative event to symptom onset (latency period) was significantly shorter in the PIAP group (18.9 ± 38 days) compared to the non‐PIAP group (29.8 ± 62.9 days), p = 0.0021. The hospitalization period was longer in the non‐PIAP group (6.13 ± 6.09 days) compared to the PIAP group (4.61 ± 4.2 days), p = 0.030. The latency period (from the causative event to symptom onset) was significantly shorter in the PIAP group (18.9 ± 38.7 days) compared to the non‐PIAP group (29.8 ± 62.9 days), p = 0.021 (Figure 1).
3.3. Analysis of Complications
A total of 153 (70.5%) patients did not experience a complicated disease course, while the remaining 64 (29.5%) did. Male sex was more prevalent in the group that sustained complications (39.1%) than in the group that did not (25.5%), p = 0.046. Age distribution was similar between the groups. None of the tested comorbidities was found to significantly correlate with complicated disease course. A recent surgical intervention involving the auricle was associated with a complicated disease course (p = 0.025), while a herpes zoster infection was associated with noncomplicated disease course (p = 0.014). Abscess formation was the single most predictive variable for a complicated disease course (5.2% vs. 60.9% for a complicated disease course, p < 0.001) (Table 3). Treatment and hospitalization times were significantly shorter in the no complications group (10.6 ± 2.8 and 3.9 ± 2.7 days, respectively) compared to the complications group (13.3 ± 6.3 and 8.1 ± 7.8 days, respectively, both p < 0.001). Patients with complications tended to be hospitalized later (12.2 ± 20.7 days) compared to patients without complications (4.7 ± 5.9 days), p < 0.001 (Figure 1).
TABLE 3.
Demographics, etiology, presentation, treatment, cultures, laboratory findings, and complications of complicated and noncomplicated groups.
| Variables | Noncomplicated; n (%) | Complicated; n (%) | p | |
|---|---|---|---|---|
| Demographics | Males | 39 (25.5) | 25 (39.1) | 0.046 |
| Age; years, mean ± STD | 35.5 ± 18.5 | 38.9 ± 22 | 0.319 | |
| Comorbidities | Any comorbidity | 38 (24.8) | 19 (29.7) | 0.459 |
| Diabetes mellitus | 8 (5.2) | 6 (9.4) | 0.362 | |
| Smoker | 13 (8.5) | 11 (17.2) | 0.063 | |
| Immunosuppression | 4 (2.6) | 1 (1.6) | > 0.999 | |
| Etiology | Piercing | 85 (55.3) | 35 (54.7) | 0.906 |
| Post traumatic | 11 (7.2) | 2 (3.1) | 0.354 | |
| Herpers zoster | 12 (7.8) | 0 (0) | 0.014 | |
| Skin integrity loss | 6 (3.9) | 4 (6.3) | 0.486 | |
| Recent operation | 2 (1.3) | 5 (7.8) | 0.025 | |
| Water exposure | 6 (3.9) | 0 (0) | 0.128 | |
| Allergic skin reaction | 4 (2.6) | 0 (0) | 0.322 | |
| Unknown | 23 (15) | 12 (18.8) | 0.497 | |
| Epicenter | Helix | 78 (51) | 29 (45.3) | 0.446 |
| Tragus | 5 (3.3%) | 3 (4.7) | 0.21 | |
| Other | 2 (1.3) | 4 (6.3) | 0.21 | |
| Unknown | 68 (44.4) | 28 (43.8) | 0.21 | |
| 0.001 | ||||
| Presentation | Abcess | 8 (5.2) | 39 (60.9) | < 0.001 |
| Lymphadenopathy | 19 (12.4) | 4 (6.3) | 0.178 | |
| Fever > 38.1°C | 8 (5.2) | 2 (3.1) | 0.727 | |
| Treatment (prehospital) | Systemic treatment | 37 (38.1) | 42 (35) | 0.623 |
| Local treatment | 20 (13.1) | 10 (15.6) | 0.619 | |
| Treatment (hospital) | Systemic ATB | 151 (98.7) | 64 (100) | 0.582 |
| Local treatment | 78 (51) | 28 (43.8) | 0.331 | |
| Systemic steroids | 3 (2) | 1 (1.6) | > 0.999 | |
| Ciprofloxacin + second ATB | 96 (62.7) | 49 (76.6) | 0.049 | |
| Ciprofloxacin + cefazolin | 65 (42.5) | 44 (68.8) | < 0.001 | |
| Ciprofloxacin as a mono ATB | 31 (20.3) | 5 (7.8) | 0.025 | |
| Cultures | Obtained | 50 (32.7) | 47 (73.4) | < 0.001 |
| No growth | 8 (16) | 6 (12.8) | 0.651 | |
| Pseudomonas aeruginosa | 12 (24) | 28 (59.6) | < 0.001 | |
| Staphylococcus aureus | 14 (28) | 1 (2.1) | < 0.001 | |
| CoNS | 6 (12) | 1 (2.1) | 0.113 | |
| Other pathogens a | 2 (4) | 10 (21.3) | 0.01 | |
| Blood work | WBC count (K/μL) | 9.0 ± 2.4 | 9.1 ± 2.4 | 0.923 |
| Absolute neutrophil count (K/μL) | 5.9 ± 2.2 | 6.1 ± 2.3 | 0.507 | |
| Neutrophil % | 63.8 ± 10.2 | 64.7 ± 9.6 | 0.45 | |
| CRP (mg/dL) | 21.5 ± 36.2 | 20.6 ± 27.5 | 0.162 | |
| Complication | Any complication | |||
| Treatment change | 16 (10.5) | 14 (21.9) | 0.026 | |
Abbreviations: ATB, antibiotics; CoNS, coagulase‐negative staphylococci; CRP, C‐reactive protein; STD, standard deviation; WBC, white blood cells.
Enterobacter cloacae complex, Staphylococcus epidermidis, Morganella morganii, Aspergillus flavus, Trueperella bernardiae, Actinomyces spp.
P. aeruginosa was more common in the PIAP group (59.6%) than in the non‐PIAP group (24%), p < 0.001, while S. aureus was more common in the P‐FAP group (28.0%) compared to the PIAP group (2.1%), p < 0.001. Other pathogen species (such as E. cloacae complex, Staphylococcus epidermidis, Morganella morganni, A. flavus, T. bernardiae , and Actinomyces spp.) were much more common in the PIAP group (21.3%) compared to the non‐PIAP group (4.0%), p = 0.01. Additionally, treatment change was more than twice as common in the PIAP group (21.9%) compared to the non‐PIAP group (10.5%), p = 0.026. All of the evaluated complications were significantly more common in the PIAP group compared to the non‐PIAP group (Table 3). Moreover, the presence of a complication was an independent predictor for treatment change.
3.4. Complication Predictors
We created a multivariate analysis model in order to identify predictors for complications (Table 4). Abscess formation emerged as the most significant variable associated with complications, with an OR of 35.75 (95% CI: 13.12–97.35, p < 0.001) for the entire cohort, 35.59 (95% CI: 10.84–116.84, p < 0.001) for the PIAP group, and 47.30 (95% CI: 5.33–419.67, p = 0.001) for the non‐PIAP group. Other variables, including sex, age, time from symptom onset to hospitalization, and the presence of piercing were not found to be significant.
TABLE 4.
Multivariate logistic regression model for complication prediction.
| Group analysis | Predicting factor | OR | 95% Cl (low–high) | p |
|---|---|---|---|---|
| Entire cohort | Time from symptoms to hospitalization | 1.036 | 0.999–1.073 | 0.055 |
| Sex | 1.427 | 0.616–3.310 | 0.407 | |
| Age | 1.016 | 0.991–1.043 | 0.214 | |
| Abscess | 35.751 | 13.123–97.350 | < 0.001 | |
| Piercing | 0.598 | 0.598–0.195 | 0.37 | |
| Without piercing | Time from symptoms to hospitalization | 1.02 | 0.961–1.082 | 0.511 |
| Sex | 0.927 | 0.985–1.046 | 0.893 | |
| Age | 1.015 | 0.307–2.799 | 0.322 | |
| Abscess | 47.302 | 5.332–419.671 | 0.001 | |
| With piercing | Time from symptoms to hospitalization | 1.046 | 0.979–1.118 | 0.186 |
| Sex | 2.718 | 0.689–10.577 | 0.307 | |
| Age | 1.027 | 0.976–1.082 | 0.149 | |
| Abscess | 35.589 | 10.841–116.836 | < 0.001 |
Abbreviations: CI, confidence interval; OR, odds ratio.
4. Discussion
AP resulting from a variety of etiologies is a common pathology in clinical practice. One of those etiologies is auricular cartilage piercing, a practice increasing in prevalence and one that is influenced by local culture. Our data show that PIAP is more common among women, and that 80% occurs in the absence and 60% in the presence of high piercing. These findings are in line with those of previous reports, which showed that women's ears are pierced three times more often than men's ears [11, 12].
Even though AP is common in clinical practice, the scientific literature is sparse. In 1981, Bassiouny [5] performed a literature review and added an original case series of patients with AP. They could find only 10 papers with case series that included up to only 191 cases. The most striking finding from that historic paper is the frequency of invasive treatments, usually prior to a trial of noninvasive (antibiotics alone) treatment, even in low‐risk populations (healthy young adults). Those treatments included multiple incisions, partial excision, total chondrectomies, and tubal drainage, and they were described by Stroud (1963) and Stevenson (1964) [13, 14]. A later literature review by Mitchell et al. focused solely on pinna abscess [15]: 76 cases were identified and treated noninvasively, with variable degrees of chondral deformity. The association between piercing, diabetes mellitus, and AP was initially highlighted by Assimakopoulos et al., who reported bilateral PIAP in a diabetic 15‐year‐old female patient 7 days following high piercing [16]. The first medical encounter for such sequelae would usually be with a family physician and not an otolaryngologist, and Meltzer (2005) published a comprehensive paper about complications for body piercing in the official journal of the American Academy of Family Physicians [17].
Age distribution emerged as the main demographic difference between the PIAP and non‐PIAP groups. The former was comprised mainly of patients in their third decade (25 ± 10.8 years), and the latter was comprised mainly of patients in their late fifth and early sixth decades (51 ± 18.5 years). In a survey of more than 10,000 ear‐pierced people, 27% pierced their ears at the age of 16–24 years, and an additional 19% at the age of 25–34 years [12].
Comorbidities played a role in the sequelae of AP, with the older non‐PIAP patients evidencing significantly more comorbidities compared to the younger PIAP patients. This factor more strongly affected the pathogenesis, risk of complications, and disease course of the older group. For example, the prevalence of diabetes among the non‐PIAP patients was over 13% compared to less than 1% in the PIAP, possibly influencing the course of treatment as well as hospitalization time.
The literature regarding the association between DM and complications following body piercing in general and auricular piercing in particular is surprisingly limited. Our search identified only two case reports on diabetic patients with a complicated course following auricular piercing [16, 18]. While DM is not a contraindication to body piercing, cases of endocarditis following piercing in diabetic patients have been reported [19]. Smoking was also hypothesized to be associated with post‐piercing complications due to the fact that smoking increases the risk of contact allergy [20].
The higher prevalence of comorbidities in the non‐PIAP group could potentially be attributed to impaired wound healing (mainly DM and smoking). Additionally, the non‐PIAP group presented for medical attention (latency interval) significantly later than the PIAP group. We could only speculate that the reason lay in the higher index of suspicion for any auricular abnormality among the PIAP group following the piercing procedure. In the absence of auricular piercing, and especially in the presence of a peripheral sensory neuropathy (as in the case of DM), it is reasonable to consider that the patient will present later for medical advice. DM patients reportedly suffer a comparatively longer latency period (patient delay) in various pathologies, including sudden sensorineural hearing loss [21], retinopathy [22], and more. It is therefore reasonable to consider that the combination of comorbidities and a prolonged latency interval probably results in a higher prevalence of complicated non‐PIAP.
The helix has become the major disease epicenter in the PIAP group as a result of the recent piercing location trends which are widely illustrated in social media platforms. In contrast, the expression of non‐PIAP has no clear epicenter. Such diffused spread of the condition in non‐PIAP is associated with fever and lack of localized abscess. Moreover, while fever is rare in PIAP, abscess formation adjacent to the piercing spot is more common, and it was found to be a pivotal finding in the disease course. Abscess formation was the only independent factor for complications in both the PIAP and non‐PIAP groups. While a small abscess could be undetectable in non‐PIAP, making the diagnosis difficult and causing treatment (incision and drainage) delay, an abscess in PIAP is far easier to identify, especially when it is close to the piercing site. The formation of an auricular abscess is an uncommon complication of AP, which typically does not have an abscess component [4, 23]. A proposed treatment algorithm is suggested in Figure 3. The algorithm is supported by the limited literature on this topic. The initial treatment strategy for complicated AP with abscess formation should include incision and drainage + intravenous antibiotics, regardless of other markers for severe disease, such as lymphadenopathy or systemic fever [4]. The presence of comorbidities, such as DM, heavy smoking, or immunosuppression, should also prompt intravenous antibiotics [23, 24]. Alternatively, we consider that local treatment alone or in combination with oral antibiotics can be safely initiated in the absence of comorbidities, abscess, or systemic symptoms.
FIGURE 3.
Proposed treatment algorithm for auricular perichondritis. ATB, antibiotics; I&D, incision and drainage; IV, intravenous.
There was a significant difference in pathogens between the PIAP and non‐PIAP groups. However, the limited number of cultures that had been obtained and the limited number of culture growths precluded the ability to formulate specific guidelines for antibiotic treatment. However, the authors recommend obtaining cultures prior to the initiation of systemic antibiotic treatment for every newly diagnosed patient with AP, regardless of the etiology (PIAP vs. non‐PIAP). This recommendation is based upon two considerations: (1) non‐PIAP is associated with high numbers of non‐pseudomonal AP, and (2) the increased resistance to fluoroquinolones has become a matter of notable concern, calling for the need to seek alternative treatment strategies [11, 23].
5. Study Limitations
This is a retrospective study based upon medical charts. Additionally, 217 cases comprise a relatively small number given the prevalence of AP associated with increasingly widespread high piercing. Another limitation that should be borne in mind is that the study might be exposed to selection bias. A considerable proportion of patients with auricular cellulitis or AP initially seek medical advice with family physicians or dermatologists. It is possible that full resolution following conservative treatment made further meetings with an otolaryngologist unnecessary. Another limitation is the inclusion criterion of patients > 18 years of age given that a significant number of teenagers undergo auricular piercing earlier and they were not included in the cohort.
6. Conclusion
PIAP displays different demographics, comorbidities, disease course, and microbiology from those of non‐PIAP. Abscess formation is an independent risk factor for complications associated with non‐PIAP.
Ethics Statement
This retrospective anonymized study was approved by the local research institute ethics committee (0046‐23‐TLV), which waived informed consent.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Data S1. Supporting Information.
Ungar O. J., Bursztyn N., Shilo S., et al., “Auricular Perichondritis in the Cartilage Piercing Era,” The Laryngoscope 135, no. 12 (2025): 4693–4702, 10.1002/lary.32409.
Funding: The authors received no specific funding for this work.
References
- 1. Keene W. E., Markum A. C., and Samadpour M., “Outbreak of Pseudomonas aeruginosa Infections Caused by Commercial Piercing of Upper Ear Cartilage,” JAMA 291, no. 8 (2004): 981–985. [DOI] [PubMed] [Google Scholar]
- 2. Fisher C. G., Kacica M. A., and Bennett N. M., “Risk Factors for Cartilage Infections of the Ear,” American Journal of Preventive Medicine 29, no. 3 (2005): 204–209. [DOI] [PubMed] [Google Scholar]
- 3. Martin R., Yonkers A. J., and C. T. Yarington, Jr. , “Perichondritis of the Ear,” Laryngoscope 86, no. 5 (1976): 664–673. [DOI] [PubMed] [Google Scholar]
- 4. Prasad H. K. C., Sreedharan S., Prasad H. S. C., Meyyappan M. H., and Harsha K. S., “Perichondritis of the Auricle and Its Management,” Journal of Laryngology & Otology 121, no. 6 (2007): 530–534. [DOI] [PubMed] [Google Scholar]
- 5. Bassiouny A., “Perichondritis of the Auricle,” Laryngoscope 91, no. 3 (1981): 422–431. [DOI] [PubMed] [Google Scholar]
- 6. Kaplan A. L. and Cook J. L., “The Incidences of Chondritis and Perichondritis Associated With the Surgical Manipulation of Auricular Cartilage,” Dermatologic Surgery 30, no. 1 (2004): 58–62. [DOI] [PubMed] [Google Scholar]
- 7. Prasad K. C., Karthik S., and Prasad S. C., “A Comprehensive Study on Lesions of the Pinna,” American Journal of Otolaryngology 26, no. 1 (2005): 1–6. [DOI] [PubMed] [Google Scholar]
- 8. Templer J. and Renner G. J., “Injuries of the External Ear,” Otolaryngologic Clinics of North America 23, no. 5 (1990): 1003–1018. [PubMed] [Google Scholar]
- 9. Erez Davidi M. D., Alona Paz M. D., Hava Duchman M. D., et al., ““High” Ear Piercing and the Rising Incidence of Perichondritis of the Pinna,” BMJ 322, no. 7291 (2001): 906–907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ungar O. J., Oron Y., Cavel O., et al., “Case Series and Systematic Review of Treatment Outcomes for Auricular Pseudocysts,” Otology & Neurotology 42, no. 5 (2021): 774–782. [DOI] [PubMed] [Google Scholar]
- 11. Zhang X., Zhang Y., Pu C., Wang L., Ni Y., and Huang T., “Etiology, Microbiological Isolates, and Antibiotic Susceptibilities in Inpatients With Refractory Auricular Perichondritis: A 10‐Year Retrospective Study,” Infection and Drug Resistance 17 (2024): 377–386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Bone A., Ncube F., Nichols T., and Noah N. D., “Body Piercing in England: A Survey of Piercing at Sites Other Than Earlobe,” BMJ 336, no. 7658 (2008): 1426–1428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Stroud M. H., “A Simple Treatment for Suppurative Perichondritis,” Laryngoscope 73, no. 5 (1963): 556–563. [DOI] [PubMed] [Google Scholar]
- 14. Stevenson E. W., “Bacillus Pyocyaneus Perichondritis of the Ear: An Effective Method of Treatment,” Laryngoscope 74, no. 2 (1964): 255–259. [DOI] [PubMed] [Google Scholar]
- 15. Mitchell S., Ditta K., Minhas S., and Dezso A., “Pinna Abscesses: Can We Manage Them Better? A Case Series and Review of the Literature,” European Archives of Oto‐Rhino‐Laryngology 272 (2015): 3163–3167. [DOI] [PubMed] [Google Scholar]
- 16. Assimakopoulos D., Tziouris D., and Assimakopoulos A. D., “Bilateral Auricular Perichondritis and Diabetes Mellitus,” Otolaryngology—Head and Neck Surgery 140, no. 3 (2009): 431–432. [DOI] [PubMed] [Google Scholar]
- 17. Meltzer D. I., “Complications of Body Piercing,” American Family Physician 72, no. 10 (2005): 2029–2034. [PubMed] [Google Scholar]
- 18. Antoszewski B., Jędrzejczak M., and Kruk‐Jeromin J., “Complications After Body Piercing in Patient Suffering From Type 1 Diabetes Mellitus,” International Journal of Dermatology 46, no. 12 (2007): 1250–1252. [DOI] [PubMed] [Google Scholar]
- 19. Patel M. and Glenn Cobbs C., “Infections From Body Piercing and Tattoos,” Infections of Leisure 3, no. 6 (2009): 331–348. [DOI] [PubMed] [Google Scholar]
- 20. Linneberg A., Nielsen N. H., Menné T., Madsen F., and Jørgensen T., “Smoking Might Be a Risk Factor for Contact Allergy,” Journal of Allergy and Clinical Immunology 111, no. 5 (2003): 980–984. [DOI] [PubMed] [Google Scholar]
- 21. Chen I., Eligal S., Menahem O., et al., “Time From Sudden Sensory Neural Hearing Loss to Treatment as a Prognostic Factor,” Frontiers in Neurology 14 (2023): 1158955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Zhao M., Chandra A., Liu L., Zhang L., Xu J., and Li J., “Investigation of the Reasons for Delayed Presentation in Proliferative Diabetic Retinopathy Patients,” PLoS One 19, no. 2 (2024): e0291280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Klug T. E., Holm N., Greve T., and Ovesen T., “Perichondritis of the Auricle: Bacterial Findings and Clinical Evaluation of Different Antibiotic Regimens,” European Archives of Oto‐Rhino‐Laryngology 276 (2019): 2199–2203. [DOI] [PubMed] [Google Scholar]
- 24. Sosin M., Weissler J. M., Pulcrano M., and Rodriguez E. D., “Transcartilaginous Ear Piercing and Infectious Complications: A Systematic Review and Critical Analysis of Outcomes,” Laryngoscope 125, no. 8 (2015): 1827–1834. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1. Supporting Information.