Emergency medicine

To the Editor: We read with interest the article by Geelhoed and de Klerk1 regarding the mortality benefits of a policy that ensures high numbers of patients are admitted to hospital within 4 hours of arrival at an emergency department (ED). The proof of improved performance following service intervention within a complex institution requires careful consideration. ...

Dylan J Toh · Campbell H Thompson · Josephine S Thomas · Jeffrey Faunt

In reply: We thank Mitra and Cameron, and Toh and colleagues for their remarks. Comments from some physicians that the 4-hour rule might increase mortality provided the initial stimulus for us to look at data related to its introduction in Western Australia. Given our results, their hypothesis seems unlikely to be true. Mitra and Cameron point ...

Gary C Geelhoed · Nicholas H de Klerk

CorrectionTypographical error: In “Emergency department overcrowding, mortality and the 4-hour rule in Western Australia” in the 6 February 2012 issue of the Journal (Med J Aust 2012; 196: 122-126), there was a typographical error in Box 1 (page 123). The total number of admissions for 2008–09 was erroneously preceded by the letter “l”. The number of admissions in 2008–09 was ...

Gary C Geelhoed FRACP, FACEM, MD · Nicholas H de Klerk BSc, MSc, PhD

A new drug with great promise, but be very aware of the risksDabigatran is the first new oral anticoagulant for the prevention of embolism in patients with non-valvular atrial fibrillation (AF) that has been recently ...

Ross I Baker BMedSc, FRACP, FRCPA · Paul Harper MD, FRCP, FRACP · Claire McLintock MB ChB, FRACP, FRCPA

To the Editor: Recent media reports related to Ian “Molly” Meldrum1 are a timely reminder, once again, of the risk of physical injury posed by a fall from a ladder. We carried out a survey of injuries sustained in falls from ladders among patients presenting to the Royal Prince Alfred Hospital (RPAH), an inner-Sydney major trauma centre, between January 2009 and December 2011. We linked data from the hospital trauma registry with data from the New South Wales Department of Health Centre for Epidemiology and Research. The trauma registry routinely collects data on all injury-related admissions and trauma team activations. NSW Health provided data from a free-text search for the term “ladder” within the presenting problems, triage text and diagnoses in the electronic medical records of all patients who presented to the RPAH emergency department (ED) during the period studied. For comparison, we also constructed a survey of injuries sustained by pedestrians struck by a motor vehicle. We used trauma registry data, from a previous study of patients with such injuries, for 409 consecutive patients admitted during 2008–2010. The study was approved by the Sydney Local Health District Ethics Review Committee (RPAH zone). We identified 331 ED presentations related to any ladder incident. Of these, 136 were available in the trauma registry for analysis. The mean age of patients was 54 years (SD, 17 years; 95% CI, 51–57), and a significant proportion (37; 27%) were 65 years or older. No inpatient deaths were reported. The most common body regions injured were upper limb (54; 40%) and head (41; 30%). Severe head injuries were sustained by 14 patients (10%) (Box). Although the proportion of patients with major trauma (injury severity score, > 15) was consistent with previous trauma registry reports,2,3 the proportion with head injury was much higher and may be related to the older age in this cohort. One recent study suggested that the incidence of major injury from ladder falls is increasing, particularly among older people.4 Many presentations for minor injury were not captured by the trauma registry; however, unlike previous reports, we included all patients requiring initial assessment by the trauma team. The proportions with severe head injury (head Abbreviated Injury Scale score, ≥ 3), intensive care admission and major trauma were similar between patients who fell from ladders and pedestrians who were struck by motor vehicles (10% v 12%, P = 0.50; 10% v 9%, P = 0.79; 12% v 15%, P = 0.30, respectively). Based on the significant proportion of people sustaining severe head injury, new ladder-safety initiatives are required, particularly targeting those aged 65 years and over. Demographic and injury profile of patients with injuries sustained in a fall from a ladder who were admitted to an inner-Sydney trauma centre, January 2009 – December 2011 (n = 136) Variable No. (%)* 95% CI Age, years (mean, SD) 54 (17) – Male 111 (82%) 76–89 Time of incident After hours† 30 (22%) 15–29 Weekend 37 (27%) 20–35 Height of fall < 1 m 29 (21%) 14–28 1–5 m 105 (77%) 69–83 > 5 m 2 (1%) 0–5 Hospital admission 95 (70%) 62–77 Intensive care admission 14 (10%) 5–15 Length of stay in hospital, days (mean, SD) 5.4 (8.9) – Injury severity score (mean, SD) 7 (5) – Major trauma‡ 16 (12%) 6–17 Body region injured Head 41 (30%) 22–38 Severe head injury§ 14 (10%) 5–15 Upper limb 54 (40%) 31–48 Lower limb 37 (27%) 20–35 Chest 14 (10%) 5–15 Severe chest injury¶ 7 (5%) 1–9 Abdomen 3 (2%) 0–5 Vertebral column 26 (19%) 12–26 Spinal cord injury 9 (6%) 2–11 AIS = Abbreviated Injury Scale. * Unless otherwise specified. † Any time between 18:00 and 08:00. ‡ Injury severity score > 15. § Any significant intracranial bleed, oedema and/or major skull fracture; head AIS score ≥ 3. ¶ Includes flail chest, major haemothorax or pneumothorax and mediastinal injury; chest AIS score ≥ 3.

Michael M Dinh · Helen L Stark · Kendall J Bein

To the Editor: Non-surgical cosmetic procedures have rapidly increased over the past 15 years, and although there are no reliable figures for Australia, there was an eightfold increase in such procedures in the United States between 1997 and 2010.1 Practitioners in the cosmetic field are seeing more frequent complications (eg, burns, scarring) from laser treatments that have been performed by untrained or inexperienced operators, including beauty therapists. These treatments are often performed as walk-in procedures in shopping centres. A cosmetic surgery report to the New South Wales Minister for Health in 1999 outlined a number of issues with laser use in Australia.2 More than 12 years later, very few of these issues have been addressed. There are more regulations on the sale and use of laser pointers in Australia than there are for lasers in health care. The Australian/New Zealand Standard 4173:2004, Guide to the safe use of lasers in health care, limits and classifies lasers used for medical purposes according to the degree of hazard. The Standard recognises that all personnel using and handling lasers should have training appropriate to the task they perform.3 Queensland, Western Australia and Tasmania have some regulations regarding medical grade lasers, but there are currently no regulations enforcing the Standard in other jurisdictions. There is also no industry standard for the necessary skills, qualifications and training for use of lasers, or regulation of these. National guidelines for the use of medical grade (Class 3B and 4) lasers need to be developed urgently and should address several issues to protect consumers: Medical grade lasers should only be sold to medical practices where the operator can be observed by a doctor at least some of the time, and complications can be discussed with the doctor. Medical practices should be licensed under the relevant Radiation Act in each jurisdiction (eg, Radiation Control Act 1990 in NSW) and listed on a website. All users of the laser in each licensed practice should be registered and individually licensed. National guidelines for accredited laser-safety courses should be established, ensuring adequate training and certification for obtaining a licence.

Charles Cope

To the Editor: The past decade has seen significant change in the management of major haemorrhage, including earlier and more aggressive use of fresh frozen plasma (FFP). Australian guidelines now advocate high ratios of FFP to packed red blood cells (PRBC) from the outset of resuscitation.1 Preparation of one unit of FFP for every two units of PRBC is recommended. This approach is supported by a recent meta-analysis demonstrating a reduction in organ failure and mortality with high FFP : PRBC ratios.2 In Australia, 40% of major trauma incidents originate in rural areas,3 some distance from tertiary or even secondary care. Initial resuscitation often falls to general practitioners (with no blood bank) or aeromedical services. In a 2010 international survey of 29 aeromedical services conducted by the Royal Flying Doctor Service (RFDS) (Western Operations), the seven responding Australian services reported difficulty accessing blood products. Only three had immediate access to PRBC, although all could obtain it within 45 minutes. None could obtain FFP without delaying departure, and only four had access within 45 minutes. This difficulty in obtaining blood products, particularly FFP, was evident in an analysis of major haemorrhage management by the RFDS (Western Operations). Of 7585 patients transferred in the 2009–10 financial year, 610 (8%) had a diagnosis that put them at risk of major haemorrhage; of these, 58 (9.5%) demonstrated physiological derangement predictive of requiring massive transfusion and did receive transfusion in flight. However, the FFP : PRBC ratios fell short of accepted standards (Box), particularly in patients originating from non-regional hospital locations. As logistical difficulties with storage and preparation render FFP impractical in remote Australia, alternative sources of coagulation factors must be explored. European groups have proposed replacing FFP with freeze-dried factor preparations (fibrinogen concentrate and prothrombin complex concentrate),4 which are currently available in Australia. These are easier to store, transport and deliver, and may be safer and more efficacious than FFP.4 Evidence that tranexamic acid significantly reduces mortality in trauma-related haemorrhage5 has been met with a guarded response in Australia. The study was conducted in developing countries with limited access to blood products and its applicability here has been questioned, although it may be highly applicable to remote and regional areas where FFP is scarce. Ongoing research and the planned Australian Massive Transfusion Registry will contribute further to the debate. When devising major haemorrhage strategies, however, it should be remembered that managing bleeding in Melbourne and managing bleeding in Meekatharra (in remote Western Australia) are two very different prospects. Ratio of fresh frozen plasma (FFP) to packed red blood cells (PRBC) in patients at risk of requiring massive transfusion, 2009–10 Origin of patient All patients (n = 58) Regional hospital (n = 35) Non-regional hospital (n = 23) Mean number of FFP units delivered: In referring hospital 0.9 1.3 0.3 During transport 0.2 0.3 0.1 Mean number of PRBC units delivered: In referring hospital 2.5 3.3 1.3 During transport 1.8 1.6 2.0 Overall FFP : PRBC ratio 1 : 3.9 1 : 3.1 1 : 8.3

Stuart A Gillon · Cem R Kibar

To the Editor: Since the publication of Knight and colleagues’ case report on the topic, five more patients have presented with burns caused by Grevillea pyramidalis ssp. leucadendron, or maangga berry.1 Traditionally, these berries have been used by Aboriginal Australians for tattooing.1 As in the cases described by Knight and colleagues,1 these burns were sustained by Indigenous children aged 8–11 years in northern Western Australia and required specialist burn care. Two patients sustained 1.5% total body surface area, deep partial-thickness burns to their left forearms, which required surgical management. Three patients presented with superficial partial-thickness left forearm burns, which were managed non-operatively via telehealth. Four of the five patients were only referred to the state burns unit 7 days after the initial injury, and one patient was lost to follow-up. Management of these patients’ injuries was limited by compliance issues. Grevillea seeds can be responsible for deeper burns, which require surgical intervention, and superficial burns, which can be treated non-operatively. In this series, the burns occurred in Indigenous children with darker skin, which often has poorer scar outcomes. This demographic, coupled with compliance issues and late referral, make grevillea seed burns a difficult clinical entity to treat. Community education and health staff awareness is imperative in helping us provide optimal treatment in the future.

Joseph Luo · Tania McWilliams · Fiona Wood

Data not correctly labelled: In “Emergency department overcrowding, mortality and the 4-hour rule in Western Australia” in the 6 February 2012 issue of the Journal (Med J Aust 2012; 196: 122-126), there was an error in Box 1 (page 123). Data under “Hospital 1” should have been shown as “Hospital 2”; data under “Hospital 2” should have been shown as “Hospital 3”; and data under “Hospital 3” should have been shown as “Hospital 1”. The sentences in the Results describing this data (page 124) should have read, “The relative mortality rates, and the absolute rate reductions for Hospitals 2 and 3 were similar for the last two periods, while Hospital 1 had less decline”, and “There was some evidence of heterogeneity of monthly decline (P = 0.03) within the three tertiary hospitals, with a smaller monthly decline in Hospital 1”. There was also an error in Box 4 (page 124). The open squares represent data for Hospital 1 (not 2); black triangles represent Hospital 2 (not 3); black diamonds represent Hospital 3 (not 1).

Gary C Geelhoed · Nicholas H de Klerk

Staff safety is improved by clear procedures for managing abuse and assault. Hopper and colleagues describe a scenario familiar to many hospital clinicians and managers: staff reports of verbal abuse and physical assaults from patients.This is often in an organisational context of scarce reliable data about the phenomenon, an ad-hoc management response and no specific training of staff to manage aggression....

Brett McDermott MD, FRANZCP, CertCAPsy

Putting a team-based intervention approach to the test.

Sandy M Hopper FRACP, FACEM · Franz E Babl MD, FRACP, FACEP · Claire E Stewart PGDipENB, RN · Jia Wei Woo BSc, MB BS

Objective: To assess whether emergency department (ED) overcrowding was reduced after the introduction of the 4-hour rule in Western Australia and whether any changes in overcrowding were associated with significant changes in patient mortality rates.

Gary C Geelhoed FRACP, FACEM, MD · Nicholas H de Klerk BSc, MSc, PhD

The first report on the 4-hour rule needs fleshing out.

Drew B Richardson MB BS(Hons), FACEM, GradCertHE

Do demographic changes alone explain our busy EDs?

Judy A Lowthian MPH, BAppSc(SpPath), LMusA · Andrea J Curtis BSc(Hons), PhD · Damien J Jolley MSc(Epidemiology), MSc, AStat · Johannes U Stoelwinder MD, FRACMA, FACHSE · John J McNeil PhD, FRACP, FAFPHM · Peter A Cameron MBBS, MD, FACEM

The integrated medical curriculum . Raja C Bandaranayake. London: Radcliff Publishing, 2011 (xvii + 128 pp, $43.50). ISBN 9781846195105. THIS MAY BE a timely addition to medical school bookshelves, as so many new medical schools are engaged in curriculum development. Written by a respected Australian medical educator, the book promotes an integrated approach, and then attempts a comprehensive coverage of issues concerning design, implementation, assessment and evaluation, and blending theory with practice. The strength of the book lies in the international experience of the author who has, among other things, worked as a senior academic at the University of New South Wales in Sydney and the Arabian Gulf University in Bahrain. The final two chapters are the most valuable, where this experience is used to describe common pitfalls, and four case studies highlight the lessons learned from well intentioned, but not always successful, attempts to redesign a medical curriculum. However, the case studies are not analysed in great depth and appear almost as a postscript. It may have been better, as in a problem-based curriculum, to present case studies as the lead-in to each chapter, with the analysis combining theory and practice to demonstrate lessons learned about integration. Other strengths of the book are: the commentary on integrated vs integrating curricula, a perceptive issue (Chapter 1); the brief but accessible history of curriculum integration (Chapter 2); the discussion on the advantages and disadvantages of integration (Chapter 5); and the list of evaluation questions to consider (Chapter 7). The major weakness of the book is the relatively scant presentation of a theoretical basis for integration, and methods for achieving, assessing and evaluating integrated learning. Some recent research literature is also not cited, particularly in integrated and workplace-based assessment and evaluation methods. Further, little is said about qualitative evaluation, which may be the stronger approach to exploring how learners perceive and gain from different curriculum approaches. The material is probably of more value to less experienced educators, who should be able to follow the links to the literature and expand their reading. I would suggest that readers work backwards from the final two chapters to the earlier discussions of emerging issues.

Richard B Hays

Objectives: To describe the epidemiology and toxicity of caffeinated energy drink exposures in Australia.Design, setting and subjects: Retrospective observational study analysing data from calls regarding energy drink exposures recorded in the database of an Australian poisons information centre over 7 years to 2010.Main outcome measures: Type of exposure; co-ingestants; symptoms reported; and reported hospitalisations.Results: Callers reported 297 exposures to energy drinks, which showed an increasing annual trend from 12 in 2004 to 65 in 2010. Median age for the 217 subjects with recreational exposure was 17 years (interquartile ratio [IQR], 15–21; range, 11–60) and 57% were male. One hundred recreational users co-ingested other substances, predominantly alcohol (50) or other caffeinated products (44). The number of energy drinks consumed in one session varied greatly (median, 5 units; IQR, 3–8; range, 1–80). Most subjects who reported recreational use reported experiencing symptoms (87%). The most common symptoms were palpitations, agitation, tremor and gastrointestinal upset. Twenty-one subjects had signs of serious cardiac or neurological toxicity, including hallucinations, seizures, arrhythmias or cardiac ischaemia. At least 128 subjects (57 with no co-ingestants) required hospitalisation.Conclusions: Reports of caffeine toxicity from energy drink consumption are increasing, particularly among adolescents, warranting review and regulation of the labelling and sale of these drinks. Educating adolescents and increasing the community’s awareness of the hazards from energy drinks is of paramount importance.

Naren Gunja MB BS, FACEM · Jared A Brown BPharm(Hons), MPH

Teaching children how to behave around dogs can reduce the incidence Dogs never bite me. Just humans. Marilyn Monroe (1926–1962) Dogs have been companions of humans for over 12 000 years and have become an inseparable part of rural and urban life. Many breeds continue to do valuable work — managing livestock on farms, guiding the visually impaired, sniffing out contraband, and as guard dogs. Most of the 3.4 million dogs in Australia are family pets, with 36% of households owning a dog.1 Unfortunately, as a result of this close relationship, dog bites are common. Statistics on dog bites often seem alarming, but most bite injuries are relatively minor, not requiring hospital admission.2 In 2008 and 2009, 928 children aged 0–14 years attended accident and emergency departments in Queensland with dog bites, equating to nine bites per week (Access Information Service, Queensland Health). The Queensland Trauma Registry (QTR) contains data on 186 children (aged 0–14 years; mean age, 5.5 years) who were admitted to Queensland hospitals for more than 24 hours in 2003–2009, equating to one such hospitalisation for a dog bite injury every 2 weeks (own unpublished data). No children died from dog bites in this period. Of the dog bite injuries recorded in the QTR, 88% occurred in the home environment, and almost all (98%) required at least one operation under general anaesthetic. In 2003, the Royal Children’s Hospital conducted a telephone survey on 45 consecutive children who were admitted with serious dog bites during 1997–2002. This survey showed that the child had physically interacted with the dog immediately before the attack in 63% of cases. Most attacks (92%) occurred in a setting familiar to the child, usually the family house or garden, or that of a relative or friend. Most bites were to the head and neck region (72%) and most children (93%) were left with permanent cosmetic scarring (own unpublished data). These results are consistent with those of other Australian studies.2,3 While most dog bite reports attempt to identify the breed, this information is commonly not documented in the patient record and, when stated, relies on correct identification by the family. No one breed stands out, but breeds commonly mentioned in Australian articles include Rottweilers, German Shepherds, Cattle Dogs and the Bull Terrier group.2,3 Data on an array of other breeds and cross-breeds provide evidence that any dog is capable of biting. An American study examined 238 dog-bite-related deaths over 20 years to 1998, and identified at least 25 different breeds.4 Breed identification may not be accurate. For example, the general public might have difficulty differentiating between several breeds in the Bull Terrier group. Staffordshire Bull Terriers, American Staffordshire Terriers and American Pit Bull Terriers all appear very similar. To add further confusion, a recent Queensland Supreme Court ruling stated that there was no distinction between the American Staffordshire Terrier and the American Pit Bull Terrier. Dogs are highly regulated in Australia with mandatory registration laws. It is now prohibited to import certain breeds, including American Pit Bull Terriers, Japanese Tosa, Dogo Argentino and Fila Brasiliero. State laws on the ownership of such breeds vary, but generally require desexing (unless registered for breeding), microchipping, the use of a leash and muzzle in public places and a locked enclosure at home with a warning sign. The laws are based not so much on evidence that these breeds are inherently dangerous, but rather that they are breeds which have been traditionally used for dog fighting. Much debate continues whether this is good and fair legislation, as studies have failed to show improvement in the incidence of bites after the legislation was passed.5 Breed-specific legislation fails to take into account that any breed of dog can be dangerous in the hands of an irresponsible owner who fails to provide good and early training. Further, these restrictions may create the risk of higher numbers of unregistered animals or irresponsible owners simply turning to other breeds. It is clear that all breeds bite, and that the severity and commonality is related to the size of the dog and how many of a particular breed exist.6 In the Netherlands, where good data on breeds and bites exist, breed-specific legislation was repealed because they found no one breed more dangerous than any other. They recommended there be a focus on owners of dogs in cases where the people bitten did not interact with the dog that bit them.6 Training for dogs and education for dog owners and children can reduce the incidence of dog bites. A recent United States study reported that knowledge about dog bite prevention among young, school-aged children is poor.7 However, education programs in the primary school setting have been shown to alter a child’s interaction with dogs.8 In general, children should be taught: to ask permission from the owner before slowly approaching an unfamiliar dog; never to run from a dog or scream; to stand still if approached by a strange dog and, if knocked over, roll into a ball and lie still; to avoid eye contact with the dog by looking at their own feet; not to disturb a dog that is sleeping, eating, or caring for puppies; and not to pat a dog without supervision or without allowing it to see and sniff them first.9 Furthermore, dog owners should take their dogs to obedience classes, and dogs should be taught to obey commands from all family members. Dogs used for hunting or as guard dogs should not be allowed to mix with children.9

Roy M Kimble MD, FRCS, FRACS · Natalie Dallow · Richard Franklin PhD · Belinda Wallis BBEnv(Dist)

Objective: To determine whether emergency department (ED) patients’ self-rated levels of anxiety are affected by exposure to purpose-designed music or sound compositions with and without the audio frequencies of embedded binaural beat.Design, setting and participants: Randomised controlled trial in an ED between 1 February 2010 and 14 April 2010 among a convenience sample of adult patients who were rated as category 3 on the Australasian Triage Scale.Interventions: All interventions involved listening to soundtracks of 20 minutes’ duration that were purpose-designed by composers and sound-recording artists. Participants were allocated at random to one of five groups: headphones and iPod only, no soundtrack (control group); reconstructed ambient noise simulating an ED but free of clear verbalisations; electroacoustic musical composition; composed non-musical soundtracks derived from audio field recordings obtained from natural and constructed settings; sound composition of audio field recordings with embedded binaural beat. All soundtracks were presented on an iPod through headphones. Patients and researchers were blinded to allocation until interventions were administered. State–trait anxiety was self-assessed before the intervention and state anxiety was self-assessed again 20 minutes after the provision of the soundtrack.Main outcome measure: Spielberger State–Trait Anxiety Inventory.Results: Of 291 patients assessed for eligibility, 170 patients completed the pre-intervention anxiety self-assessment and 169 completed the post-intervention assessment. Significant decreases (all P < 0.001) in anxiety level were observed among patients exposed to the electroacoustic musical composition (pre-intervention mean, 39; post-intervention mean, 34), audio field recordings (42; 35) or audio field recordings with embedded bianaural beats (43; 37) when compared with those allocated to receive simulated ED ambient noise (40; 41) or headphones only (44; 44).Conclusion: In moderately anxious ED patients, state anxiety was reduced by 10%–15% following exposure to purpose-designed sound interventions.Trial registration: Australian New Zealand Clinical Trials Registry ACTRN 12608000444381.

Tracey J Weiland BBSc(Hons), PhD/MPsych · George A Jelinek MB BS, MD, FACEM · Keely E Macarow BA, MA, PhD · Philip Samartzis GradDipArt · David M Brown DipArt, MA · Elizabeth M Grierson LicDip, MA, PhD · Craig Winter MB BS, MBA, FACEM

This is the first reported case of eye injury caused by a crayfish antenna. A 20-year-old male crayfish diver sustained a scleral penetrating injury that led to a subconjunctival abscess. The foreign body was histologically similar to a crayfish antenna. Clinical recordA 20-year-old man who worked as a crayfish diver presented with a unilateral red watery right eye, initially developed while at sea. He was unable to continue diving, mainly due to photophobia, and on reaching land he was transferred to our hospital by aeroplane. He had no other medical history, regular medications or known allergies. Seven days before presentation, he had a sore ear on the right side with pain radiating down the right hand side of his neck and discharge from his nose; however, he denied any trauma to the right eye. At first presentation to the ophthalmologist, the patient’s vision was 6/36 in the right eye and 6/6 in the left eye. The vision in his right eye improved to 6/18 with pinhole. He had peripheral corneal superficial punctate keratitis, 3+ cells in the anterior chamber, and an elevation of the conjunctiva, inferior to the cornea, 4.5 mm in diameter. Examination of his retina was unremarkable, and he had a small nasal pterygium. Examination of the left eye was unremarkable. The intraocular pressure was 12 mmHg in both eyes. He was provisionally diagnosed with right eye acute anterior uveitis and treated with topical combined prednisolone acetate plus phenylephrine hydrochloride and 2% homatropine hydrobromide for mydriasis. At a scheduled review 2 days later, the condition of the patient’s right eye had deteriorated. He had developed an anteriorly discharging subconjunctival abscess. A combined duplex B-scan ultrasound showed a hypoechoic, hypervascular soft tissue mass in, or external to, the sclera and inferomedial to the iris. The mass was 7 mm deep and protruded into the globe. The patient was immediately taken to the operating theatre for excision and exploration of the subconjunctival abscess. Intraoperative examination with a three-mirror gonioscope lens before excision showed an unidentifiable membranous foreign body in the anterior chamber. The foreign body was thought to be a parasite due to its longitudinal appearance. The subconjunctival abscess was excised and explored. A tract from the subconjunctival space through the full scleral thickness and into the suprachoroidal space was explored, and the foreign body in the anterior chamber was removed. The scleral defect was covered with a conjunctival graft using cardinal and continuous sutures, and the overlying conjunctival defect was repaired with a conjunctival rotation. After surgery, the patient was treated with intravenous ceftriaxone and topical ofloxacin (3 mg/mL), prednisolone acetate plus phenylephrine hydrochloride and homatropine hydrobromide eye drops, oral analgesia, and three doses of albendazole (400 mg). Two days after surgery, the patient’s condition had improved considerably. His vision had improved to 6/6 with pinhole and the anterior chamber cells had improved to 2+ cells. He was discharged home on topical prednisolone acetate plus phenylephrine hydrochloride, ofloxacin and homatropine hydrobromide eye drops, and oral prednisolone. Histological results of the initial surgical specimen showed a small foreign body several microns in length intimately associated with deeply placed active inflammation (Box 1). There was fibroproliferative activity and an infiltrate of inflammatory cells; these were associated with tissue necrosis and microabscess formation. Culture of a preoperative conjunctival swab was negative for microbial growth. Three days after discharge, the patient returned to hospital with a recurrence of the subconjunctival abscess, for which he underwent emergency excision and drainage. The previous conjunctival graft was found to be necrotic and was removed. A thin fibrovascular membrane covering the scleral defect remained. A swab taken from the abscess was later positive for methicillin-resistant Staphylococcus aureus, which was successfully treated with intravenous vancomycin, oral clindamycin and topical chloramphenicol. At 4 weeks’ follow-up, the patient’s unaided visual acuity in his right eye had stabilised to 6/6 and the intraocular pressure was 7–8 mmHg. The conjunctival flap had dehisced, but there was a steady increase in the thickness of the fibrovascular covering of the scleral perforation. The original foreign body extracted during surgery was compared with a specimen of crayfish antennae and was found to be similar to the very outer layer (1–2 μm) of the specimen (Box 2). It most likely became lodged while the patient was handling crayfish or cleaning the crayfish tank without eye protection, which he was doing extensively during the morning that the red eye developed. The crayfish caught in this area are of several different types, although the Panulirus ornatus (tropical rock lobster) make up 95% of the local population.1 DiscussionPenetrating eye injuries (PEIs) are serious and can result in blindness, either from direct trauma to ocular structures or secondary infection and inflammation. Given the significant consequences of PEIs, early diagnosis and management is imperative to improve outcomes. Despite the vast majority of PEIs being obvious through history of trauma, some are less evident to both patient and clinician. We report a case of delayed diagnosis of PEI due to a unique mechanism. A population-based cross-sectional study of eye injuries in Australia found that they were more common among men, people living in rural areas and tradespeople, and that the workplace was the most common location where the injury occurred.2 A review of 109 penetrated or ruptured globes over 4 years in Adelaide identified that 80% of patients were men; the most common cause of injury was hammering metal; and the final visual acuity was 6/12 or better in 40% of patients, and no perception of light in 26%.3 Another group found that of 6308 patients treated at their hospital over 12 months, 6% were admitted for severe ocular trauma and most of these were young men.4 Workplace-related injuries accounted for 44% of all ocular injuries and 19% of cases of severe eye trauma.4 The annual medical costs for eye injuries have been estimated at $155 million for the projected 116 000 cases nationwide.4 In 2006, a review of ocular trauma over 7 years at Cairns Base Hospital estimated the incidence of open globe injuries to be 3.7 per 100 000 population.5 An international study of PEIs in rural areas found that 16% were work-related.6 Most injuries were related to maintenance or repair work, wood chopping or machine use. A small number were the result of cow butting. At the final review, 64% of these eyes were blind.6 Crayfish antenna penetration is a mechanism that has not previously been reported. Thorns of plants have been previously reported to cause PEIs in children, but the most common aquatic-related PEI is with fish hooks.7,8 Appropriate preventive measures, such as use of protective eyewear and implementation of education programs, could be effective in decreasing the incidence of these aquatic-related ocular traumas. Our case highlights the risk of PEIs for young men while at work. It demonstrates the possibility of a small organic foreign body associated with a large inflammatory reaction and a full-thickness scleral penetration that could have resulted in blindness. For clinicians, our case emphasises the importance of early suspicion of ocular trauma in high-risk patients with red eyes, particularly if they have an unexpected lack of response or deterioration to treatment despite an initially negative history of trauma. 1 Tip of crayfish antenna (arrow) embedded in scleral tissue 2 Tip of crayfish antenna

Khoi A Tran MMed, MB BS, BPharm · Matthew Green MSc, MB BS · Jayne Camuglia BSc, MB · Stephen O'Hagan MB BS, FRANZCO