Initial Case Definitions

We initially used broad case definitions to increase the sensitivity of active case finding. We defined a confirmed case-patient as a person who had a P. aeruginosa infection after a recent ear piercing and a probable case-patient as a person who had no cultures taken or no culture growth but had attended the implicated facility or other facilities of the same franchise or had used the same brand of aftercare solution.

Emergency Department Syndromic Surveillance

The NSW Public Health Rapid Emergency, Disease, and Syndromic Surveillance (PHREDSS) system monitors treatment sought at most public hospital emergency departments in the state in near-real time (10). Patients are coded according to their illness and discharge destination (e.g., admitted or discharged home) (10). We identified possible infection cases at hospitals via PHREDSS by using keyword searches of triage text: (infect*|cellulitis*) and (earring*|earing*|pierc*|peirc*).

Initial Investigations

We obtained details of procedures performed on confirmed case-patients, aftercare solutions used, notifications and complaints, and client lists from the initially implicated piercing salon. We identified additional cases of piercing-related infections in residents of greater Sydney, Wollongong, and Newcastle by examining the PHREDSS database.

Statewide Investigations and Case Finding

Beginning on April 30, 2021, weekly PHREDSS line lists of emergency department visits and admissions for piercing-related infections were provided to other public health units (PHUs) across NSW. We asked PHUs to review the medical records of patients on the PHREDSS list who were within their local health district and to contact patients who had positive P. aeruginosa cultures from clinical swab samples; PHUs gathered demographic and hospitalization data, determined when and where the patients had their piercing performed, and what aftercare solution (brand and batch) they had used. We also asked PHUs to obtain any available bottles of aftercare solution(s) used by the patient for laboratory testing. We alerted other states about the outbreak in NSW and asked them to report any confirmed or probable cases.

Microbiological Investigations and Whole-Genome Sequencing

Environmental health officers collected environmental samples and samples of aftercare solutions from the initial piercing salon for culture. We submitted specimens from case-patients and from opened and unopened bottles of aftercare solution to the Institute of Clinical Pathology and Medical Research–NSW Health Pathology for culturing and bacteria identification; positive culture isolates underwent whole-genome sequencing (WGS) by using the Illumina NextSeq platform (Illumina Corp., https://www.illumina.com), and their core genome multilocus sequence types (STs) were determined (11,12).

Ethics

This study was conducted as a public health investigation under the NSW Public Health Act 2010. Therefore, ethics approval was not required.

Descriptive Epidemiology

Figure 1

Figure 1. Weekly emergency department visits across New South Wales, Australia, by patients with piercing-site infections during 2016–2021 compiled for study of community outbreak of Pseudomonas aeruginosainfections associated with contaminated...

PHREDSS time series data showed an increase in emergency department visits for piercing infections in NSW beginning in April 2021 (Figure 1). Using the PHREDSS line lists, we identified 251 persons with a piercing-related infection via active case finding; 62 of those had P. aeruginosa–positive cultures.

We sent samples from 15 previously opened bottles of Protat aftercare solution retrieved from case-patients for microbiological testing. Samples from 10 Protat bottles with the implicated batch number were tested; 9 were positive for P. aeruginosa, and 1 had no bacterial growth. The implicated batch had a use by date of October 1, 2023, and had been supplied to tattooing and piercing establishments during February 16–May 11, 2021 (14). Samples from 5 additional Protat bottles with different batch numbers were also tested and had no bacterial growth (Table 1).

We retrieved 11 unopened bottles of Protat aftercare solution from various chain stores in Sydney and Wollongong on May 14, 2021 (9 with the implicated batch number and 2 with different batch numbers). Samples from 3 unopened bottles of the implicated batch were positive for P. aeruginosa (850–2,000 CFU/mL); the remaining bottles had no growth (Table 1). We received additional samples from 48 unopened Protat aftercare bottles with the implicated batch number from South Australia; 10 of those bottles were cultured and had no bacterial growth.

Figure 2

Figure 2. Phylogenetic analysis of Pseudomonas aeruginosa isolates collected in New South Wales in study of community outbreak of P. aeruginosainfections associated with contaminated piercing aftercare...

WGS was performed for 28 bacteria isolates (16 isolates from clinical samples, 3 from unopened bottles of Protat aftercare solution, and 9 from opened bottles of Protat aftercare solution); 27 of those isolates were P. aeruginosa and belonged to ST988. ST988 is a rare type that has not been identified previously in local isolate collections. Analysis identified 0–9 single-nucleotide polymorphism differences between the 27 isolate genomes (sequences with <25 single-nucleotide differences were regarded as a genomically linked cluster). Cluster analysis showed that all 27 submitted ST988 isolates were genomically linked. The remaining P. aeruginosa isolate was from a clinical sample and belonged to ST247 (Figure 2).

In total, we identified 13 case-patients who had P. aeruginosa–positive clinical isolates belonging to ST988 (2 case-patients had clinical samples taken from 2 different sites). Of the 13 case-patients with ST988 P. aeruginosa infections, 10 had used the implicated batch and 1 had used 2 different batches of Protat aftercare solution; information was not available for 2 cases. The case-patient with ST247 P. aeruginosa infection had used a different aftercare solution batch.

We reinterviewed the case-patient who reported using 2 different batches of Protat aftercare solution to confirm batch numbers. That case-patient had purchased those bottles 6 and 12 months earlier after other piercings and confirmed that additional aftercare solutions had not been purchased at the time of the latest piercing. However, a product was used in-store on their ear during the piercing procedure, and the case-patient was likely exposed to the implicated batch at that time.

Outbreak Case Characteristics

After receiving the WGS analyses, we refined case definitions to be more specific. We used the confirmed case definition to determine associations with the suspected product, whereas the other case definitions tracked case incidence over time and assisted with ongoing case finding. We defined a confirmed case as a person in Australia with a body piercing infection caused by P. aeruginosa who had a piercing date after February 1, 2021, and ST988 detected in a clinical isolate. We defined a probable case as a person in Australia with a body piercing infection caused by P. aeruginosa who had a piercing date after February 1, 2021, but sequencing and typing data for a clinical isolate were not available; in addition, the piercing was performed at a franchise A store and the person had either used Protat aftercare solution or P. aeruginosa ST988 had been isolated from their aftercare product (regardless of store attended or product used). We defined a probable case either if a person in Australia had a body piercing infection caused by P. aeruginosa who had a piercing date after February 1, 2021, at a franchise A store but sequencing and typing data for a clinical isolate were not available; or the person had used Protat aftercare solution; or P. aeruginosa ST988 had been isolated from their aftercare product (regardless of store attended or product used). We defined a suspected case as a person in Australia with a body piercing infection caused by P. aeruginosa who had a piercing date after February 1, 2021, but the piercing location or aftercare product used was not known; a suspected case was also defined as a person in Australia with a body piercing infection and piercing date after February 1, 2021, but culture specimens were not collected or had no bacterial growth and the piercing site or use of aftercare solution was either not yet known, the store was not a franchise A store, or Protat aftercare solution was not used.

Of the 251 case-patients with piercing-related infections, 13 were confirmed, 40 were probable, 9 were possible, and 189 were suspected cases. Confirmed and probable cases predominantly comprised female patients (48/53, 91%) with a median age of 19.6 (range 15.5–59.6) years; 80% of those resided in metropolitan Sydney, Wollongong, or Newcastle. The median number of days between the piercing date and emergency department visit was 15 (range 3–35) days.

A higher percentage of confirmed and probable case-patients (30/53, 57%) required hospital admission compared with possible and suspected cases (9/198, 5%). In addition, confirmed and probable case-patients were more likely to have had piercings in cartilaginous areas, such as the tragus or helix, than possible and suspected cases and were much less likely to have been pierced in the earlobe or other noncartilaginous areas. We observed only minor differences in age, gender, and time (from piercing procedure to emergency department visit) between the confirmed/probable and possible/suspected groups. (Table 2)

All 13 confirmed case-patients attended either the initial implicated piercing facility or another franchise A facility and used Protat aftercare solution (Table 3). Of those, 10 case-patients were able to state specifically that they used the implicated batch of Protat aftercare solution; 1 stated that the Protat aftercare solution they had used came from a different batch, and 2 stated they had used Protat aftercare solution postpiercing but were unable to identify the batch number. The case-patient who claimed to have used a different batch of Protat aftercare solution was likely administered the implicated batch immediately after piercing in the store (Table 3).

Outbreak Control Measures

On May 4, after identifying several additional P. aeruginosa piercing-related infections in clients who had attended franchise A stores across multiple NSW local health districts, the implicated franchisee agreed to cease using and selling Protat aftercare solution pending further investigation. NSW Health issued a clinician alert on May 13 to advise of an increase in hospital admissions because of P. aeruginosa infections after piercing procedures and to encourage clinicians to consider P. aeruginosa in patients with infections at piercing sites.

After confirmation of the presence of P. aeruginosa in unopened bottles of Protat aftercare solution, the distributor/manufacturer issued a voluntary recall of the contaminated batch from market shelves on May 14 and issued a product recall media release on May 31. On June 1, the Australian Competition and Consumer Commission published a recall notice for the implicated batch (14), which had been supplied to franchise A stores in NSW, Queensland, Victoria, and South Australia, as well as other tattooing and piercing establishments in all states of Australia and New Zealand. The commission also liaised with overseas regulators about the safety recall in Australia (14).

On May 28, after performing their own independent review, the manufacturer informed NSW Health that production/processing pathway testing of the implicated batch had detected a positive total microbial count, and results were PCR positive for P. aeruginosa. Root cause analysis conducted by the manufacturer identified several potential opportunities for contamination during manufacturing: handling of raw materials used for the manufacture of the product might have been compromised; flushing of the previous product out of the filling system might not have been performed sufficiently; the spray ball in the mixing tank might have been contaminated; the mixing tank’s lid might have been left open and unattended, potentially introducing contamination through water droplets; and the flow plate for tanks might have been contaminated when transfer connectors were switched between tanks. In addition, the prescribed microbial testing for this product was suspected to be insufficient.

Figure 3

Figure 3. Timeline of key events during the community outbreak of Pseudomonas aeruginosa infections associated with contaminated piercing aftercare solution, Australia, 2021. Graph shows weekly numbers of P. ...

In response to the contamination, the manufacturer began preservative challenge testing on the product formula; increased microbial testing for opportunistic pathogens, including P. aeruginosa, S. aureus, and Candida spp.; and performed antimicrobial disinfectant fogging of the manufacturing facilities. They also planned to take additional preventive actions to address the risk for future contamination, such as reassessing cleaning and hygiene procedures and maintenance and operation of equipment and parts; providing refresher training for all manufacturing and production operators, emphasizing vigilance and aseptic techniques; and potentially reformulating the product on the basis of challenge test results. The effect of the outbreak control measures was evident in the decline in emergency department visits beginning in late May 2021 and their return to baseline levels by mid-June 2021 (Figure 3).