Acupuncture has become increasingly accepted as a form of alternative medicine in Western countries, though it has been practiced for hundreds of years by Oriental medical doctors in South Korea. As of 2002, about 13,000 Oriental medical doctors were licensed through the national examination for medical practice. These practitioners are recognized as a type of medical doctor by the South Korean public, however, Oriental medical doctors and other medical doctors are educated, and subsequently practice, in a separate system. Therefore, Oriental medical doctors are relatively unfamiliar with infection control, and usually see patients in general office rooms.

Acupuncture can be complicated by infections caused by environmental microorganisms or patient skin flora [1–3]. Although there have been several acupuncture-associated hepatitis B outbreaks [4, 5], no acupuncture related outbreak of bacterial infections has been reported.

Mycobacterium abscessus is a rapidly growing mycobacterium, which is widespread in natural and drinking water [6, 7]. These species are relatively resistant to chlorine and glutaraldehyde [7–10], thus, are frequently detected in hospital tap water [6–8, 11, 12].

These organisms most commonly cause soft-tissue infection (community acquired or nosocomial) following minor trauma or intramuscular injection. In addition, these organisms have been implicated in surgical wound infections, bacteremia associated with dialysis, catheter-related sepsis and chronic ear infections [8, 12, 13].

There have been sporadic cases reported over the past few years in which non-tuberculous mycobacterial infections complicate acupuncture [9, 14, 15]. Herein, we report on an outbreak of post-acupuncture cutaneous infection, due to M. abscessus, in an Oriental clinic.

We reported an outbreak of acupuncture-associated abscesses due to rapidly growing mycobacteria. Outbreaks of injection-related mycobacteria have been described following injections of histamine, lidocaine, saline solution, vaccines, disinfectant solution and adrenal cortex extract [20–22]. The main distinguishing feature of this study is that no material was injected with the acupuncture needle.

The exact source of M. abscessus in this outbreak was not determined, however, there are several possible explanations. First, the acupuncture needle itself could be the source of infection, likely due to needle length. Acupuncture consists of solid needles, ranging from 15 to 50 mm long, piercing the skin [23]. The acupuncture-associated mycobacteriosis occured most often on the back and the knee-joint areas, where the needles are deeply inserted. Although many patients received acupuncture at the ankle and wrist areas, none of these patients developed a skin lesion. A second possibility is skin contamination before or after acupuncture. In general, the Oriental medical doctor rubbed alcohol on the patient's skin to sterilize it before acupuncture. Patients were also exposed to hot-pack therapy and gel (mixture of Antiphramin^® and ultrasound jelly) massage after acupuncture. Therefore, cultures from environmental objects, including: disposable needles, hot packs, towels, Antiphramin^® gel, alcohol-retained bottles and boiling tank water, were obtained. Although these samples gave negative cultures, towels and hot-pack covers are still a possible main source of infection. Mycobacteria may have been introduced to soft tissue by contaminated towels or hot-pack covers via the acupuncture site. No further cases have been seen since mid-November, after sterilization and regular towel change (figure 1). Finally, though unlikely, the needle could have been carelessly exposed to tap water.

Since the symptoms are relatively mild and indolent, the clinical diagnosis of mycobacteriosis is often delayed. In contrast to other pyogenic bacterial infections with a shorter incubation period, infections due to rapidly growing mycobacteria have longer incubation periods (several days to several months) [12, 14, 21, 22]. It took more than 2 months from initial manifestation to diagnosis, which was also observed in previous study [21]. This suggests that careful monitoring is required for early diagnosis and appropriate treatment. Thirty-eight of 40 patients were treated successfully with administration of oral clarithromycin (500 mg, twice daily) for 3–6 months and some of these patients also received either amikacin or rifampin. None of the patients required hospitalization or surgical management. The mean duration of antibiotic treatment varied from 9 to 12 months in several previous studies, however, other studies saw complete healing after 2–3 months of antibiotic treatment with incision and drainage [15, 21]. This indicates that several factors affect the prognosis, such as the disease type (e.g., cellulite vs. abscess) and the underlying disease. Therefore standard guidelines including antibiotic treatment and surgical management should be established.

With the increasing popularity of acupuncture, a number of surveys concerning the safety of acupuncture have been conducted. Various adverse events were reported, such as: needle pain, tiredness, bleeding, headache, faintness and rare pneumothorax [24, 25]. The importance of infection control has not been emphasized in Oriental medicine despite several sporadic reports of secondary wound infection at acupuncture sites. Cho et al. reported a case of a serious retroperitoneal abscess following acupuncture [26]. Rapidly growing mycobacteria are well known to be resistant to chlorhexidine and relatively resistant to alcohol [27]. The outbreak of infection has been attributed to improper sterilization rather than to iodine and alcohol resistance [20, 27].

There are several limitations that prevented us from identifying the exact contamination source. Because the hot-pack boiling tank was sterilized and the towels and hot-pack covers were changed a few weeks prior to this study, the environmental source of the rapidly growing mycobacteria and their mechanism of distribution were not elucidated by this study. We could not compare patients with controls. Because the Oriental medicine clinic was an outpatient setting and the medical records were poor there was no efficient way to assess unaffected persons.

In conclusion, it is important to educate Oriental medical doctors to wash their hands properly with alcoholic chlorhexadin or povidone iodine, to clean the proposed skin site with povidone iodine or non-contaminated alcohol and allow sufficient drying time, and to sterilize both the environment and equipment. We recommend that Oriental medicine courses provide infection control education. Moreover, future epidemiologic studies are needed to elucidate the exact mechanism of infection.

To our knowledge, this is the first outbreak of acupuncture-associated wound infection caused by rapidly growing mycobacteria. Physicians should suspect mycobacterial infections in patients with persistent cutaneous infections following acupuncture. Furthermore, education on infection control, including hygienic practice, should be emphasized to Oriental medical doctors practicing acupuncture.