The incidence, aetiology and management of anaphylaxis presenting to an Accident and Emergency department

Transcription

1 Q J ed 1996; 89: The incidence, aetiology and management of anaphylaxis presenting to an Accident and Emergency department A.G. STEWART* and P.W. EWAN From the Allergy linic, Addenbrookes Hospital NHS Trust, University of ambridge linical School, ambridge, UK Received 21 arch 1996 and in revised form 10 September 1996 Summary We retrospectively studied anaphylaxis in an A&E department from computerized records. In 1993 (Study A), of patients seen in casualty, nine had severe anaphylaxis (ANA) with loss of consciousness (LO) or ing (about 1:00). Fifteen had generalized allergic reactions (GR) without LO or ing, but including dyspnoea due to laryngeal oedema or asthma, angioedema and/or urticaria. Thus there were 24 (about 1:2300) generalized reactions involving hypotension and/or respiratory difficulty. A further case diagnosed as hyperventilation syndrome was probably a wasp sting GR. Six cases of urticaria and/or angioedema were also identified. Of the nine with ANA, a possible cause was identified in eight (3 stings; 2 drugs; 3 foods). There was delay in arrival in A&E: hypotension was noted in Introduction three and had resolved spontaneously in six. Only 3/9 were treated with adrenaline: i.v. hydrocortisone and chlorpheniramine was the mainstay of treatment. investigation was recommended nor advice given on future management. Four patients were later referred to our allergy clinic by their GPs. In study B (Aug-Oct 1994), nine cases of ANA were identified (1:1500), eight due to bee or wasp stings. The increased incidence was probably related to more detailed history-taking. Only three were treated with adrenaline. The use of adrenaline for future anaphylaxis was discussed with six patients, and five were referred to our allergy clinic. A reaction to the same allergen had occurred previously in 24%. Improved awareness of anaphylaxis and its management is necessary. Anaphylaxis means the opposite of protection (phylaxis), where prior contact with an antigen leads to an enhanced damaging immunological reaction on re-exposure. It is recognised by a constellation of symptoms and signs which result from the offending antigen cross-linking antigen-specific IgE on mast cells and basophils. This leads to cellular activation and release of mediators which cause pruritus, erythema, urticaria, angioedema, laryngeal oedema, asthma, hypotension and cardiovascular collapse. Unfortunately there is no universally accepted definition of an anaphylactic reaction. Anaphylaxis maybe defined clinically in a variety of ways ranging from any systemic allergic reaction to systemic allergic reactions with profound hypotension and shock. There is the additional difficulty of differentiating true anaphylaxis from 'anaphylactoid reactions', where there is direct degranulation of mast cells and/or basophils without involvement of IgE antibodies. As a result the overall incidence and prevalence of these potentially life-threatening reactions is unknown. ost studies have been on defined groups of patients, usually in hospital, undergoing a therapy or procedure with a known risk of anaphylaxis, e.g. allergen immunotherapy, systemic penicillin or induction of anaesthesia. 1 " 4 There is an apparent increased incidence of anaphylaxis, and a perception amongst the general public that medical awareness of anaphylaxis and its management is inadequate. From our experience Downloaded from by guest on 06 vember 2018 Address correspondence to Dr P. W. Ewan, R entre, Hills Road, ambridge B2 2QH * Present address: edway Hospital NHS Trust, Gillingham, Kent Oxford University Press 1996

2 8 A.G. Stewart and P.W. Ewan of patients referred because of anaphylactic reactions in an allergy clinic, this is probably true. A typical story demonstrating this, which we see regularly in our referrals, was reported recently and led to a review article on the management of anaphylaxis. 5 The majority of patients we see who have had anaphylaxis have not been given adrenaline. Once treatment has been given there is often no effort made to investigate the cause of the anaphylaxis. To investigate the incidence and management of anaphylaxis, we carried out a retrospective survey of all attendees in the Accident and Emergency Department at Addenbrooke's Hospital in 1993, who may have suffered anaphylaxis, as defined by syncope or hypotension (study A). The survey was repeated in August to October 1994 (study B). During Study A, the incidence of generalized allergic reactions with respiratory difficulty, but without hypotension was also determined. Addenbrooke's Hospital is a major teaching hospital and regional referral centre, which also acts as a district general hospital for ambridge and the surrounding villages. The A&E department serves a population of about , part of which is city and urban (about ) and part is rural (about ). The nearest A&E departments are at Hinchingbrook Hospital, 15 miles west and West Suffolk Hospital, 32 miles east, while those at Harlow (south) and rfolk and rwich Hospital (north) are further afield. Patients from a wide area would therefore not attend another A&E department. ethods As in most A&E departments, the coding system used was predominantly one for surgical or trauma diagnoses. There was no clear coding for allergy or anaphylaxis. A pilot review (discussion with staff and review of the month of October 1993) suggested that most cases would be detected by a triage diagnosis of allergy or sting and the diagnostic coding of 'skin and subcutaneous'. This identified 487 potential events in The diagnosis of asthma proved unhelpful, as rarely was an allergic cause sought, and no cases had associated systemic allergic symptoms, syncope or shock to suggest an anaphylactic reaction. The search missed one known case of anaphylaxis, identified because of referral to our allergy clinic. On review, the allergic nature of this episode had not been realised and it was coded as 'collapse'. We therefore examined 720 further records, also coded as 'collapse', but found no further cases suggestive of anaphylaxis. All 1207 records were carefully scrutinized (study A). Records which suggested an allergic reaction were noted. Details of the type and timing of symptoms and signs, treatment, attempts to determine or confirm cause, prevent recurrence or refer to the allergy clinic at Addenbrooke's Hospital were recorded. The results were discussed with the A&E staff, although no formal protocols were established. The survey was repeated for 3 months from August to October 1994 (study B). Results The A&E department, Addenbrooke's Hospital, sees patients per year. In 1993, 1207 patients with diagnostic codings likely to include anaphylaxis (see ethods) were identified, and their records reviewed. From these, nine cases of severe anaphylactic reaction, denoted by hypotension or a history of syncope with other systemic allergic features, were identified (Table 1). There was a surprisingly long delay between onset of symptoms and the patients' arrival in A&E ( min). This may explain why only three patients (of six for whom there were data) had a low blood pressure recorded. Only three patients were treated with adrenaline, one of whom was treated by his GP, and another where the patient knew he was allergic to nuts, and self-injected adrenaline prophylactically. The mainstay of therapy was intravenous hydrocortisone and chlorpheniramine (in 7/9 patients). Two patients were admitted. On discharge from casualty, one patient was given prednisolone and only three were given antihistamines for ongoing pruritus and erythema. A probable cause was identified in 8/9 patients. Three cases were due to hymenoptera stings. Two reactions were assumed to be related to drugs (Imigran and cephalexin). The three reactions to foods occurred in patients with known severe food allergies who inadvertently ingested the allergen, (milk in one and nuts in two). The cause of one reaction was unknown. patient had investigations to determine or confirm the cause (e.g. assays for specific IgE antibodies). patient was given advice on how to prevent future reactions, taught simple first aid to deal with future reactions, or given any treatment for future anaphylaxis. -one was given or recommended adrenaline for self-injection and no-one was directly referred to our allergy clinic. However, two of the food-allergic and two of the insect-sting allergy patients were subsequently referred to us by their GPs. A further 15 patients were identified who suffered less severe generalized allergic reactions without loss of consciousness or ing (Table 2). In this group, the reaction consisted of combinations of erythema, Downloaded from by guest on 06 vember 2018

3 Table 1 Details of patients experiencing severe anaphylactic reactions, with either loss of consciousness or ing in the 1993 study Age and sex Presumed cause Time to arrival at A&E (min) Syncope? BP noted in A&E Symptoms in A&E Treatment TTO Allergy clinic referral 46F 57F 33F F 18F Imigram Bee urry (nut) eflacor ilk Xmas cake (nut) Walking Faint LO Faint LO Faint LO LO LO LO 130/80 80/50 120/75 135/90 70/50 0 as,lo e,p e,p e,p,as c,lo,fa e,p,fa H, H, H, H, A,H, H,,A A,H,** Terfenadine Admit P + Admit TTO, treatment to take home (oral drugs);, not recorded; LO, loss of consciousness; H, intravenous hydrocortisone;, intravenous chlorpheniramine; A, intramuscular adrenaline; P, prednisolone; as, asthma; lo, laryngeal oedema; p, pruritus; u, urticaria; c, cyanosis; e, erythema; fa, facial angioedema. * Patients who have since been seen in the allergy clinic; ** treatment given by GP. x Downloaded from by guest on 06 vember 2018

4 862 A.. Stewart and P.W. Ewan Table 2 Details of patients experiencing systemic allergic reactions without loss of consciousness, or lightheadedness in the 1993 study Age and sex Suspected cause Time to arrival in A&E (min) Symptoms Treatment given in A&E or by GP TTO Allergy clinic referral 47F 26F 5F 40F 6F 23F 29F F 53F 38F 42 F 57F Washing powder actus Unknown Unknown Latex rubber ixed salad Strawberry* hicken* ussels Diazepam Serevent Naproxen 3 24 h 12 h ,as u r,co,a e,p,co p,u,hot,a,a u,a,n e,p,lr u p,u,a,lo H,,S T H, H, H, P Qadmit H = intravenous hydrocortisone;, intravenous chlorpheniramine; S, salbutamol; P, prednisolone; T, terfenadine; co, conjunctivitis; n, nausea; r, rhinitis; as, asthma; lo, laryngeal oedema; p, pruritus; u, urticaria; e, erythema; fa, facial angioedema; LR, local reaction (oedema); TTO, treatment to take home (oral drugs);, not recorded. * Patients who have since been in the allergy clinic. pruritus, urticaria, angioedema, laryngeal oedema and asthma, without symptoms of hypotension after potential allergen exposure. All had dyspnoea due to laryngeal oedema or asthma. ost were treated with an antihistamine. A possible cause was noted in 13 of these 15 patients. Three occurred after wasp stings, three were presumed to be due to drugs (serevent, diazepam, naproxen) and four to foods (chicken, strawberries, mussels, and avocado spinach and bacon salad). The first three of these food-induced reactions occurred in patients who had suffered previous reactions after eating the same food. One reaction was thought to be due to allergy to latex rubber gloves, and this occurred in a person with known rubber-contact hypersensitivity. Biological washing powder or contact with a cactus were possible causes in two, and in two the cause was unknown. ne of these causes was confirmed, and advice or treatment in the event of future reactions was not provided. An additional patient, a young girl, was diagnosed as 'hyperventilation syndrome' after a syncopal episode following her first wasp sting. Her mother was highly allergic to wasps. It is not possible to ascertain whether this was a genuine allergic reaction. Six further cases of urticaria and angioedema were identified, and there was no clear cause. A follow-up survey in August, September and October 1994 revealed nine cases of anaphylaxis, eight of which were due to bee or wasp stings. (Table 3). This apparent increased incidence may have been due in part to increased awareness, but this survey covered the peak bee and wasp sting season (three cases in the previous year were due to bee or wasp stings). During this second survey, all patients stung by bees or wasps were asked more detailed questions, including questions about ing. uch more information was recorded in the notes, even of the mildest sting reaction with local swelling only. Three of these patients were treated with adrenaline by injection, one of whom was given adrenaline prophylactically by a GP because of a history of a previous anaphylactic reaction to a sting. Five patients were ultimately referred to the allergy clinic, two directly from A&E and three following advice from A&E to the GP to make a referral. patient was given injectable adrenaline for future use, however, in six patients (five of whom were seen later in the allergy clinic) this question was discussed in A&E, and the decision on its necessity left to the GP or the allergy clinic. An overview of the data in both surveys is shown in Table 4. There had been an increase in the number of severe anaphylactic reactions (with hypotension) from 1 in 00 to 1 in Discussion In this retrospective study, anaphylactic shock accounted for approximately 1 in 00 casualty attendances in study A and 1 in 1500 in study B. These figures do not include the less severe systemic allergic reactions without collapse. It is likely that this Downloaded from by guest on 06 vember 2018

5 Anaphylaxis in A&E 863 Table 3 Details of patients experiencing severe anaphylactic reactions with either loss of consciousness or in the follow up study (August, September, October 1994) Age Sex Suspected cause Time to arrival in A&E (min) LO or BP noted in A&E linical features Treatment given in A&E TTO F F F wasp whelk bee* wasp wasp F = female = male = not recorded as = asthma lo = laryngeal oedema p = pruritus u = urticaria LO LO 140/80 110/70 120/80 105/80 90/ 110/ 90/70 110/70 120/75 c = cyanosis fa = facial angioedema e = erythema ph = pharyngeal oedema n = nausea LO = loss of consciousness D = marked dyspnoea * patients who have since been seen in the allergy clinic. TTO = treatment to take home (oral drugs). Table 4 Nature of reaction e,d,hot D,lo c,p,u,lo c,p,u,ph c,as,n,drowsy,confused,confused D,hot Overall incidence of anaphylactic reactions in an Accident & Emergency department Reactions with loss of consciousness or Generalised allergic reactions with dyspnoea Reactions with hypotension and/or respiratory difficulty ND, not done. Reactions during Survey A (1 year) Number study underestimates the incidence of anaphylaxis, because we may have missed some diagnosed cases (because of the coding system) and there are likely to be cases in whom the diagnosis was not considered. Some patients will have been treated by GPs and not identified in this study. It is probable that some of the episodes recorded as collapse or asthma were anaphylactic reactions, but this cannot be confirmed without more detailed records. A prospective study is required to identify the true incidence. The lack of a universally accepted definition of anaphylaxis is a problem. It seems quite inappropriate to include all systemic allergic reactions, e.g. generalized urticaria and angioedema. We suggest (Incidence) (1 in 00) (1 in 3700) (1 in 2300) H,T,fluids A,H A, (by GP) H H,fluids H, A, (by GP) H, (in A&E) H, H = intravenous hydrocortisone = intravenous chlorpheniramine A = intramuscular adrenaline P = prednisolone Survey B (3 months) Number 9 ND ND nil c c,p nil nil (Incidence) (1 in 1500) that 'anaphylaxis' be reserved for severe systemic allergic reactions with either hypotension or marked respiratory difficulty. It should be noted that in food allergy, laryngeal oedema (which can be severe enough to lead to asphyxia and respiratory arrest) is common whereas hypotension is not. 6 Anaphylaxis could thus occur without hypotension as a presenting feature. There was a long delay between onset of symptoms and arrival in casualty. In some patients this reflects delay whilst the patient's GP was consulted. This probably accounts for the normal blood pressures recorded in patients who undoubtedly had loss of consciousness or ing suggestive of hypo- Downloaded from by guest on 06 vember 2018

6 864 A.G. Stewart and P.W. Ewan tension. In our bee and wasp allergy clinic, we not uncommonly see patients with severe systemic reactions in whom marked hypotension resolves spontaneously. The fact that some of the patients were recovering spontaneously on arrival in casualty accounts in part for the low use of adrenaline by the casualty staff. Although some patients with anaphylactic reactions recover spontaneously (and this might relate mainly to particular causes), this should not be over-emphasized, since many anaphylactic reactions require aggressive therapy. Anaphylaxis can be fatal but is eminently treatable and is easier to reverse if treatment is given early This study will not have identified patients who died from anaphylaxis before medical help arrived. Hypotension and loss of consciousness can lead to serious morbidity. The delay in reaching casualty suggests that if adrenaline is needed it should be given in the community, either by self-administration in patients who have previously had anaphylaxis or by the GP or the ambulance crew. A high proportion of patients (8/33 or 24%) gave a history of a previous allergic reaction to the suspected cause. Only one of these (a nut-allergic patient) had been given adrenaline for selfadministration in case of inadvertent ingestion of nuts. However, he had not been given appropriate advice on how to avoid nuts hidden in food, and had eaten a curry, a food with a known risk of containing hidden nuts. The patient who reacted to hristmas cake knew she was allergic to nuts but again had not had advice on avoidance. The anaphylactic reaction to milk was very severe. This occurred in a toddler whose parents believed he was milkallergic because of previous reactions, and avoided cows milk. Unfortunately this diagnosis had not been accepted by his physician so that the parents had not kept milk completely from his reach, and he took milk himself from the family refridgerator. He was subsequently seen in the allergy clinic, where history confirmed the diagnosis of milk allergy, and this was backed up by positive skin-prick tests. The family was then given advice on avoidance and an emergency treatment plan which included inhaled adrenaline through a volumatic spacer and a preloaded adrenaline syringe, and training in the use of these. The treatment plan was introduced to the nursery school where staff were also taught. Referral to an allergy clinic is important, first to identify or confirm the allergen, so that advice on avoidance can be given, which should greatly reduce the likelihood of further anaphylaxis. The allergist would also provide appropriate management which can be self- (or parent-) administered in the community as soon as serious symptoms occur. During the course of this study, there was improved knowledge amongst the A&E senior house officers on future management of anaphylaxis, since during the second survey patients were being referred to the allergy clinic for further investigation and management. This occurred despite staff rotations, the increased awareness being in a group of doctors not involved in the initial survey. This suggests that knowledge and experience gained by the original cohort was being passed on. There appeared to be a reluctance to use or recommend adrenaline, both for the acute reaction and as a drug to carry for early treatment of further episodes of anaphylaxis. The latter is best decided in an allergy clinic where accurate diagnosis can be made, risk of recurrence assessed, advice on avoidance given and teaching on self-injection given with trainer syringes. All patients given adrenaline for selfadministration should have a written treatment plan. This study highlights the importance of proper management of anaphylaxis. Intramuscular adrenaline is the drug of choice for severe reactions. Improved education of doctors should reduce morbidity and mortality, through advice on allergen avoidance and through detailed self-management plans, ensuring that patients receive treatment appropriate to the severity of their reactions. Prospective studies are needed to determine the true incidence of anaphylaxis. Acknowledgements We are grateful to r. aimaris and r H.. Sherriff, onsultants in Accident and Emergency, for advice and help with access to records. PWE is supported by the edical Research ouncil. This study was prompted by a workshop on anaphylaxis organized by Dr R.S.H. Pumphrey in ambridge, December References 1. Youlten LJF, Atkinson BA, Lee TH. The incidence and nature of adverse reactions to injection immunotherapy in bee and wasp venom allergy. lin Exp Allergy 1995; 25: Ewan PW, Stewart A A prospective study of systemic allergic reactions to venom immunotherapy. lin Exp Allergy 1993; 23: Idsoe O, uthe T, Wilcox RR, deweck AL. Nature and extent of penicillin side-reactions with particular reference to fatalities from anaphylactic shock. Bull WHO 1968; 38: Fisher cd, Baldo BA. Anaphylactoid reactions during anaesthesia. lin Anaesthesioh 984; 2: Fisher. Treatment of acute anaphylaxis. Br edj 1995, 311: Ewan PW. A clinical study of 62 patients with peanut or nut allergy in 62 consecutive patients: new features and associations. Bred) 1996; 312: Downloaded from by guest on 06 vember 2018