Volume 174 Issue 5

Healthcare Adverse events associated with rush Hymenoptera venom immunotherapy Glen P Westall, Frank C K Thien, Dan Czarny, Robyn E O'Hehir and Jo A Douglass MJA 2001; 174: 227-230 Abstract - Methods - Results - Discussion - References - Authors' details - - More articles on Immunology and allergy Abstract Objectives: To determine the incidence and nature of adverse events associated with the induction of rush Hymenoptera venom immunotherapy. Design: Retrospective descriptive case study. Setting: The asthma and allergy unit at a major metropolitan teaching hospital, between 1 January 1989 and 30 June 1999. Patients: All patients with anaphylaxis to stings of Hymenoptera insects who received rush venom immunotherapy as inpatients. Outcome measures: Hypersensitivity reactions to venom administration, including angioedema, skin rashes, hypotension and asthma, as well as any other adverse events related to the inpatient stay. Results: 68 venom-allergic patients received 73 courses of rush immunotherapy; 89% were desensitised to honey bee venom, 10% to yellow jacket wasp venom, and one to paper wasp venom. Hypersensitivity reactions occurred after 36 subcutaneous injections (3.8% of all injections given) in 26 patients (38%). Conclusion: In our cohort, immunotherapy was accompanied by a high incidence of adverse systemic events during the induction phase. Immunotherapy should only be given by experienced staff in centres where there are facilities for resuscitation. Stinging-insect anaphylaxis is most often caused by bee and wasp stings, but may also occur as a reaction to ants, march flies, ticks and other insects.1 The insect order Hymenoptera includes the vespids (yellow jacket or European wasp, and paper wasps) and the apids (honey bees). Clinical manifestations after wasp and bee stings include hypersensitivity reactions that can be fatal. There is a minimal degree of immunogenic cross-reactivity between bee and wasp venoms.2 Australian Bureau of Statistics mortality figures after bee stings give a rate of 0.086 per million population per year, or approximately one death per year.3An anaphylactic reaction to insect venom is an absolute indication for venom immunotherapy (see Box 1), as this is protective in more than 80%-90% of individuals in preventing future severe reactions.4 The first commercial venom extracts became available in 1979 and since then indications and treatment schedules have been refined. A number of different dosing schedules for immunotherapy are used, including conventional (induction immunotherapy over several weeks as an outpatient), rush (induction over several days) and ultrarush (induction over several hours) immunotherapy.5-7 In rush protocols, patients are given higher venom doses in a shorter period of time compared with conventional protocols, reaching maintenance dose (100 µg) more quickly and thus offering the patient earlier protection.8 At our allergy clinic, patients usually receive rush immunotherapy. We performed a retrospective analysis of case records of insect-allergic patients to describe and ascertain the incidence of adverse events associated with inpatient rush Hymenoptera immunotherapy. Methods We chose the period 1 January 1989 to 30 June 1999 for our retrospective, descriptive study. Patients were identified by searching hospital records of discharge diagnosis, and we reviewed the records of all patients admitted to the Alfred Asthma and Allergy Unit for rush venom immunotherapy. The cohort therefore represented a highly selected group of venom-allergic individuals. The field sting that caused anaphylaxis was classified according to the system proposed by Mueller (Box 2).9 The causative insect was identified from the history and subsequently confirmed by detection of serum-specific IgE by skin-prick tests or blood-specific IgE serological testing.10 If blood-specific IgE tests were negative or unavailable, venom-specific IgE was sought by skin-prick tests with 100 µg/mL of honey bee, yellow jacket wasp and paper wasp venom; if these tests were negative patients underwent intradermal testing to serial 10-fold dilutions of venom. For diagnosis and immunotherapy, a freeze-dried venom of honey bees (Apis mellifera), paper wasps (Polistes sp.), or yellow jacket wasps (Vespula sp.) was reconstituted in albumin-saline (Bayer Australia Ltd, Pymble, NSW). Patients then had immunotherapy with the venoms corresponding to their sensitivity. Patients were treated over five days as inpatients at the Alfred Hospital, where full emergency resuscitation facilities were available, and with intravenous access being maintained at all times. Escalating doses of venom (0.1 µg to 100 µg) were injected subcutaneously. A total of 13 injections were given 60 minutes apart, and a nurse closely observed the patient after each injection. The patient's general practitioner continued maintenance immunotherapy according to a protocol provided by the Asthma and Allergy Unit. For patients who were unable to attain the recommended maintenance dose of 100 µg of venom, the patient's GP was provided with a treatment schedule on how to increase the dose, up to the maintenance dose. We compared the group of patients who developed hypersensitivity reactions to venom during the course of their induction immunotherapy with a group of patients who had no such reaction to look for any predictive variables. For each group, mean age, sex, stinging insect, severity of initial reaction, atopic status, presence of asthma, and blood-specific IgE to stinging insect was compared. Statistical analysis Statistical analysis was performed with the SAS software package.11 Means and proportions were compared by standard tests (χ2, t tests and Wilcoxon 2-sample test). The 0.05 level of significance was used throughout the analysis and all P values reported are two-sided. Ethical approval The study was approved by the Alfred Hospital Ethics Committee. Results Clinical features Patient demographics and clinical features are shown in Box 2. During the study period, 68 venom-allergic patients received a total of 73 courses of rush immunotherapy (949 injections). When asked about the causative stinging insect, 58 patients identified honey bee, six identified yellow jacket wasp and one identified the paper wasp. All of these responses were confirmed with either positive skinprick tests or blood-specific IgE testing to the same insect. Only three patients could not identify the stinging insect. Of the five patients who received two courses of immunotherapy, two initially complied poorly with the initial course and developed severe anaphylaxis when they were stung again, and three patients ran out of maintenance bee venom during a national shortage. One patient who was taking a β-blocker (a contraindication to immunotherapy) changed to an alternative antihypertensive medication before beginning immunotherapy. Immunotherapy Complications during treatment were recorded as being either IgE mediated (hypersensitivity reaction) or non-IgE mediated (no hypersensitivity reaction); details are shown in Box 3. Allergic complications occurred at all stages of the rush protocol. All patients developed local reactions at the site of injection. Systemic hypersensitivity reactions occurred after 36 of the 949 injections (3.8%) in 26 of the 68 patients (38%). Recurrent adverse reactions prevented 14% of individuals attaining the recommended maintenance dose (100 mg) during their hospital admission. Among patients who had non-IgE-mediated complications, two developed gram-negative sepsis, possibly from infected intravenous sites, and one was initially incorrectly given paper wasp rather than yellow jacket wasp venom. Discussion We found that allergic adverse events during rush immunotherapy were common, occurring in 38% of patients treated, with a risk of allergic reaction per immunotherapy injection of 3.8%. There are very few Australian reports looking at the management of insect venom allergy;1,3,12,13 most such reports come from the United States and Europe.6,14-17 The most striking difference between our findings and those of overseas studies was in the causative insect. In the US and Europe, most courses of insect immunotherapy are prescribed for wasp allergy, while we found that almost 90% of people assessed for desensitisation in our unit had honey bee allergy. This marked difference most likely reflects Australia's geographical location and thus differing local fauna. In particular, the European (yellow jacket) wasp has only been common in Australia in the past 10 years. The importance of this observation relates to marked differences in the efficacy and the rates of adverse events of bee and wasp immunotherapy.18 In our study, the risk of an allergic adverse event and the rate of administration of adrenaline per injection (1.4%) are higher than those reported in other centres that use rapid immunotherapy regimens.6,14-19 However, we treated a considerably higher proportion of patients with honey bee venom than these centres. This is relevant because previous reports have described a higher incidence of side effects in patients treated with honey bee venom owing to its greater potency,4 and our findings support these conclusions. A second explanation for the increased use of adrenaline may lie in our unit's policy of early identification and aggressive treatment of any allergic reactions to venom preparations. Previous investigators have shown that there is a strong correlation between a severe initial sting (Mueller grade IV) and subsequent propensity to develop hypersensitivity reactions after deliberate sting challenges.20 We found no correlation between an initial Mueller grade IV sting and subsequent likelihood of developing anaphylaxis during immunotherapy (P = 0.63). As in other studies looking at predictors of side-effects during treatment,18 we found no association with atopic status or the serum level of venom-specific IgE. The trend for increased adverse events in females that we observed has been previously reported,4,18 despite a predominance of males receiving desensitisation therapy. The finding that two patients developed gram-negative septicaemia led us to introduce a policy of moving the intravenous cannula every 48 hours. One problem with treating insect stings described in the literature is the importance of correctly identifying the stinging insect. This was not a significant problem with our cohort, with only three patients being unsure of the insect at their initial assessment. This may reflect the presence of a "stinger" at the sting site being indicative of honey bee stings and thus aiding identification.21 Over the period examined, 68 patients were treated with a rush immunotherapy protocol, with 86% of patients reaching maintenance dose by the time they were discharged. The maintenance dose is reached within a few days with rush therapy, compared with three months with conventional regimens. The safety and cost-effectiveness of rush therapy have been previously documented,22 and tolerance to rush immunotherapy has been reported to be equal to or better than that for conventional protocols.7 Because allergic reactions occurred within one hour of each injection, we advise general practitioners who administer maintenance injections to observe patients for an hour after each injection. In conclusion, our retrospective review of rush immunotherapy shows a large proportion of reactions to honey bees in Australia compared with Europe and the USA. However, there are only minimal differences in diagnosis, investigation and management compared with overseas practice. The incidence of adverse events during induction was more common with bee venom immunotherapy. Our rush protocol is convenient for the patients who often live in rural areas and would find weekly trips to our unit for outpatient immunotherapy disruptive. Our results support the recommendations that Hymenoptera venom immunotherapy should only be given by experienced staff, in centres where there are facilities for resuscitation.23 References Solley GO. Allergy to sting and biting insects in Queensland. Med J Aust 1990; 153: 650-654. Reisman RE, Mueller UR, Wypych JI, Lazell MI. Studies of co-existing honeybee and vespid venom sensitivity. J Allergy Clin Immunol 1984; 73: 246-252. Harvey P, Sperber S, Kettle F, et al. Bee sting mortality in Australia. Med J Aust 1984; 140: 209-211. Müller U, Helbling A, Berchtold E. Immunotherapy with honeybee venom and yellow jacket venom is different regarding efficacy and safety. J Allergy Clin Immunol 1992; 89: 529-535. Golden DBK, Valentine MD, Kagay-Sobolka A, Lichtenstein LM. Regimens of hymenoptera venom immunotherapy. Ann Intern Med 1980; 92: 620-624. Nataf P, Guinnepain MT, Herman D. Rush-venom immunotherapy: a 3-day programme for hymenoptera sting allergy. Clin Allergy 1984; 14: 269-275. Van der Zwan JC, Flinterrman J, Jankowski IJ, Kerckhaert JA. Hyposensitisation to wasp venom in six hours. Br Med J 1983; 287: 1329-1331. Gillman SA, Cummins LH, Kozak PP, Hoffman DR. Venom immunotherapy: comparison of "rush" vs "conventional" schedules. Annals of Allergy 1980; 45: 351-354. Mueller HL. Diagnosis and treatment of insect sensitivity. J Asthma Res 1966; 3: 331-333. American Academy of Allergy Committee on Insects. In: Levine MI, Lockey RF, editors. Monograph on insect allergy. Library of Congress Catalog Publication Data. Hartland, Wis: Parker Printing of Hartland, 1981. SAS statistical package [computer program]. Version 6.12. Cary, NC: SAS Institute, 1996. Roberts-Thompson PJ, Harvey P, Sperber S, et al. Bee sting anaphylaxis in an urban population of South Australia. Asia Pac J Allergy Immunol 1985; 3: 161-164. Lui CL, Heddle RJ, Kupa A, et al. Bee venom hypersensitivity and its management: patients perception of venom desensitisation. Asia Pac J Allergy Immunol 1995; 13: 95-100. Birnbaum J, Charpin D, Verloet D. Rapid Hymenoptera venom immunotherapy. Comparative safety of three protocols. Clin Exp Allergy 1993; 23: 226-230. Thurnheer U, Müller U, Stoller R, et al. Venom immunotherapy in Hymenoptera sting allergy. Comparison of rush and conventional hyposensitization and observations during long-term treatment. Allergy 1983; 38: 465-475. Bousquet J, Muller UR, Dreborg S, et al. Immunotherapy with Hymenoptera venoms. Allergy 1987; 42: 401-413. Bernstein DI, Mittman RJ, Kagan SL, et al. Clinical and immunologic studies of rapid venom immunotherapy in Hymenoptera-sensitive patients. J Allergy Clin Immunol 1989; 84: 951-959. Youlten LJ, Atkinson BA, Lee TH. The incidence and nature of adverse reactions to injection immunotherapy in bee and wasp venom allergy. Clin Exp Allergy 1995; 25: 159-165. Yunginger JW, Paull BR, Jones RT, Santrach PJ. Rush immunotherapy programs for honeybee sting sensitivity. J Allergy Clin Immunol 1979; 63: 340-347. van der Linden PW, Struyvenberg A, Kraaijenhagen RJ, et al. Anaphylactic shock after insect-sting challenge in 138 persons with a previous insect-sting reaction. Ann Intern Med 1993, 118: 161-168. Visscher PK, Vetter RS, Camazine S. Removing bee stings. Lancet 1996; 348: 301-302. Bernstein JA, Kagen SL, Bernstein DI, Bernstein IL. Rapid venom immunotherapy is safe for routine use in the treatment of patients with hymenoptera anaphylaxis. Ann Allergy 1994; 73: 423-428. Committee on the Safety of Medicines. CSM update: desensitising vaccines. BMJ 1986; 293: 948. (Received 11 Sep 2000, accepted 18 Jan 2001) Authors' details Department of Allergy, Asthma and Clinical Immunology, The Alfred and Monash University, Melbourne, VIC. Glen P Westall, MRCP, MB BS, Registrar; Frank C K Thien, FRACP, MD, Physician; Dan Czarny, FRACP, FRCP, Physician and Associate Professor; Robyn E O'Hehir, FRACP, PhD, Professor and Director; Jo A Douglass, FRACP, MD, Head, Asthma and Allergy Unit. Reprints will not be available from the authors. Correspondence: Dr J A Douglass, Department of Allergy, Asthma and Clinical Immunology, The Alfred and Monash University, Commercial Road, Prahran, VIC 3181. j.douglassATalfred.org.au 1: Indications for venom immunotherapy (adults) Type of reaction Venom- specific IgE Venom immunotherapy Severe systemic life-threatening Positive Yes Moderate systemic (angioedema, asthma, etc) Positive Allergist review Mild systemic (urticaria, pruritus) Positive No Large local Positive No Any type of reaction Negative* No *Blood-specific IgE tests to venom have a false negative rate of approximately 10%, so patients whose blood test results are negative but who have histories suggestive of venom allergy should have appropriate skin tests. Back to text 2: Clinical data of patients who received stings Bee Wasp Number of patients 60 (88%) 8 (12%) Sex Males 47 (78%) 4 (50%) Females 13 (22%) 4 (50%) Median age (years) 37 45.5 Range (years) (13-66) (22-71) Atopic 23 (38%) 1 (12.5%) Asthmatic 10 (17%) 0 Blood-specific IgE test score (mean of 4)* 2.8 2.2 Grade of sting† I 2 (3%) 0 II 11 (18%) 0 III 15 (25%) 4 (50%) IV 32 (53%) 4 (50%) No. of immunotherapy injections 845 104 Hypersensitivity reactions Number of injections 35 (4.1%) 1 (1%) Number of patients 25 (42%) 1 (12.5%) Grade of hypersensitivity reaction* I 21 0 II 4 0 III 7 0 IV 3 1 Adrenaline required Number of injections 12 (1%) 1 (1%) Number of patients 11 (18%) 1 (12.5%) *Blood-specific IgE scored as class 0, 1, 2, 3 or 4. †Mueller et al:9 Grade I - urticaria, pruritus, malaise; Grade II - angioedema, chest tightness, nausea, vomiting, abdominal pain, dizziness; Grade III - dyspnoea, wheeze, stridor, dysphagia, hoarseness; Grade IV - hypotension, collapse, loss of consciousness, incontinence, cyanosis. Back to text 3: Clinical data on patients who developed hypersensitivity reactions (HR) compared with those who had no adverse reactions Cohort with HR Cohort with no HR P Number 26 47 Mean age (years) 33.5 40.3 < 0.05 Sex Males 18 (69%) 36 (77%) NS Females 8 (31%) 11 (23%) Grade of sting (mean of 4) 3.3 3.3 NS Atopic 10 (38%) 10 (21%) NS Blood-specific IgE test (mean of 4)* 2.7 2.9 NS Bee immunotherapy 25 (42%) 35 (58%) NS Wasp immunotherapy 1 (12.5%) 7 (87.5%) NS NS=not significant. *Blood-specific IgE scored as class 0, 1, 2, 3 or 4. Back to text

Glen P Westall · Dan Czarny · Robyn E O'Hehir · Jo A Douglass

Position Statement Non-valvular atrial fibrillation and stroke prevention Graeme J Hankey, on behalf of the National Blood Pressure Advisory Committee of the National Heart Foundation* MJA 2001; 174: 234-239 Abstract - Warfarin versus control - Aspirin versus control - Warfarin versus aspirin - Warfarin combined with aspirin - Warfarin versus other antiplatelet agents - Who to treat and with what? - Who is at high risk of stroke and thromboembolism without treatment? - Who is at high risk of haemorrhage with anticoagulant treatment? - Recommendations for antithrombotic therapy for AF - References - Authors' details - - More articles on Cardiology and cardiac surgery Abstract Atrial fibrillation (AF) affects 5% of people older than 65 years. Among patients with AF, the risk of stroke averages about 5% per year. The risk of stroke increases cumulatively with increasing age, previous transient ischaemic attack or stroke, hypertension, diabetes, impaired left ventricular function and a large left atrium. Management aims to identify and treat the underlying cause, control the ventricular rate, restore and maintain sinus rhythm, and minimise the risk of stroke. Warfarin reduces the risk of stroke by about two-thirds, and aspirin by about one-fifth. The risk of anticoagulant-associated haemorrhage increases with serious concomitant disease, and with poorly controlled hypertension and poorly controlled anticoagulation. All patients with chronic AF should be considered for oral anticoagulant therapy, and the decision based on the balance between the risks of thromboembolism and bleeding. The recommended INR (international normalised ratio) is 2.0-3.0. Treating 1000 "average" AF patients (ie, those with a 5% per year risk of stroke) with warfarin prevents about 30 strokes and causes at least two episodes of major haemorrhage each year. Treating 1000 AF patients with aspirin prevents about 15 strokes each year. Atrial fibrillation (AF) is a common arrhythmia. Its prevalence increases with age, from about 2% in the general population, to 5% in people older than 65 years, and 10% in people older than 75 years1,2 (E4; level-of-evidence codes are described in Box 13). It may occur as a single episode, a series of recurrent episodes ("paroxysmal" AF), or continuously ("permanent" or "chronic" AF). Atrial fibrillation is an important arrhythmia because it may signify underlying heart disease, it may cause symptoms of decreased cardiac output (eg, malaise, effort intolerance) or palpitations, and it is associated with an increased risk of systemic thromboembolism and stroke. This risk of stroke averages about 5% per year among all individuals in AF, which is about 5-6 times greater than for people of the same age who are in sinus rhythm (E32).1,2 The management of AF has four principal objectives: To confirm and document the arrhythmia; To identify and treat the underlying cause; To relieve symptoms of decreased cardiac output by controlling the ventricular rate and restoring and maintaining sinus rhythm; and To reduce the risk of systemic thromboembolism, particularly stroke. All patients, except perhaps the very elderly and infirm, should undergo investigation for underlying causes of AF, including thyroid function tests and echocardiography (E4).4 In haemodynamically stable patients, β-blockade, verapamil or diltiazem can be used to control the heart rate (E4). Recent-onset AF reverts spontaneously within 24 hours in at least half of patients (irrespective of whether or not they are taking digoxin) (E33).5 Patients who have been in AF for more than 48 hours should be considered for anticoagulation therapy and strategies to restore and maintain sinus rhythm. Warfarin should be administered for three weeks before cardioversion is attempted.6 If cardioversion can not be postponed for three weeks, the patient should undergo anticoagulation therapy with intravenous heparin and warfarin,7 and be considered for transoesophageal echocardiography (TOE) (E32). Cardioversion can probably be undertaken safely (with limited risk of stroke) if TOE excludes left atrial and appendage thrombus (and the patient is treated with heparin and warfarin) (E2).8,9 However, if transoesophageal echocardiography identifies left atrial and appendage thrombus, then cardioversion is contraindicated until the patient has been anticoagulated for at least three weeks (E4). The relative merits of cardioversion by electrical shock and medical therapy have been discussed recently.4,10 Direct current cardioversion has never been subjected to a randomised trial, but appears to be the most effective method of restoring sinus rhythm. Its main disadvantage is the need for general anaesthesia. Digoxin and verapamil are ineffective for converting AF to sinus rhythm. Flecainide or sotalol are the preferred medical therapies in younger patients without structural heart disease, and amiodarone in older patients (E2).11 The chances of successful cardioversion are greater if the AF is of recent onset and the left atrial size is normal (E33).12 After successful cardioversion, warfarin therapy should be continued for at least four weeks to prevent clot formation in the "stunned" left atrium (E33).13,14 Antiarrhythmic drug therapy should also be continued to prevent recurrent AF, but this still occurs in 40%-50% of patients after 12 months' follow-up despite drug therapy. If the patient has a low risk of recurrence of AF (eg, "lone" AF) and remains in sinus rhythm for one month after cardioversion, anticoagulation therapy with warfarin can be ceased (E4). In patients at higher risk of recurrence (Box 2), it may be more appropriate to continue warfarin therapy for longer or indefinitely (E4). For patients who are elderly (in whom AF is usually chronic and antiarrhythmic drug therapy may be risky) or have asymptomatic chronic AF, it is often reasonable to avoid attempted cardioversion, accept the AF and aim for adequate ventricular rate control (digoxin combined with β-blockade, verapamil or diltiazem) and long term anticoagulation therapy (E4). The results of clinical trials in patients with asymptomatic AF (of rate control and antithrombotic therapy versus attempted cardioversion and maintenance of sinus rhythm to avoid warfarin) are awaited. Strategies for reducing the risk of stroke and systemic thromboembolism in patients with AF have been studied in several randomised controlled trials over the past decade.15-25 Warfarin versus control Primary prevention Five large randomised controlled primary prevention trials have shown that, in people with chronic non-valvular AF, warfarin reduced the risk of stroke by about two-thirds (68%; 95% CI, 50%-79%; P < 0.001), from about 4.5% to 1.4% per year overall, with little increase in frequency of major bleeding (warfarin, 1.2%; control, 1.0%), or intracranial haemorrhage (warfarin, 0.3% per year; control, 0.1% per year) (E1).15-19,26 This means that warfarin will prevent about 30 strokes per 1000 patient-years of treatment at a cost of at least two serious bleeding episodes per 1000 patients treated for one year. It must be stressed, however, that this acceptable rate of bleeding was achieved in patients who were carefully selected, screened and closely followed; 53%-93% of eligible patients with AF were not included in the trials because of an increased risk of bleeding. Exclusion criteria included old age (> 75 years), serious illness (liver, kidney, brain or malignant disease), alcoholism, fall risk (eg, syncope), forgetfulness, non-steroidal anti-inflammatory drug therapy, and uncontrolled hypertension. Secondary prevention One secondary prevention trial (the European Atrial Fibrillation Trial [EAFT]) showed that, in people with chronic non-valvular AF and symptoms of previous transient ischaemic attack (TIA) or stroke, who have a risk of stroke of 12% per year, warfarin therapy (target INR, 2.5-4.0) reduced the risk of stroke by about two-thirds (66%; 95% CI, 53%-80%), to 4% per year (E2).20 The annual incidence of major bleeding complications was 2.8% in the anticoagulant group and 0.7% in the placebo group. No intracranial bleeds were identified in patients assigned to warfarin. Thus, warfarin prevents about 80 strokes per 1000 patient-years in patients who have had a TIA or stroke and who are in AF, at a cost of at least 20 serious bleeding episodes per 1000 patients treated for one year. The timing of anticoagulation therapy after recent ischaemic stroke depends on the risk of recurrent thromboembolism (Box 2) and the risk of haemorrhagic transformation of the brain infarct (which is higher within the first two weeks and in patients with large brain infarcts and uncontrolled hypertension [E32]27). Common empirical practice is to treat patients with fibrillating acute ischaemic stroke immediately with aspirin (300 mg daily) and then, depending on the above factors, begin warfarin (5 mg daily) between days three and 14 after stroke onset, aiming to achieve an INR of 2.0.28 However, randomised trials comparing aspirin with heparin during the first two weeks of acute ischaemic stroke among patients in AF show no benefit from early anticoagulation, because any net gains from reduction in recurrent ischaemic stroke are offset by the excess hazards of haemorrhagic stroke (E1).29,30 Aspirin versus control Three primary prevention and three secondary prevention trials have shown that, in people with AF, aspirin reduced the incidence of stroke by 22% (95% CI, 2%-38%), from 5.2% (placebo) to 3.7% (aspirin) per year for primary prevention (absolute risk reduction: 1.5% per year), and from 12.9% (placebo) to 10.4% (aspirin) per year for secondary prevention (absolute risk reduction, 2.5% per year) (E1).31Aspirin was not associated with any significant excess of intracranial haemorrhage (aspirin, 0.16%; control, 0.13%) or major extracranial bleeding (aspirin, 0.5%; control, 0.6%) (E1).31 This means that aspirin might prevent about 10 to 20 strokes per 1000 patient-years of treatment, depending on the type of patient treated and their baseline risk of stroke, with little risk of major bleeding. A speculative interpretation of these data is that, in patients with AF, aspirin prevents strokes due to atherothromboembolism, but not cardiogenic embolism. This interpretation is based on the magnitude of the effect (a 20% relative risk reduction), which is very similar to the effect of aspirin in patients with symptomatic atherothromboembolism of the brain, heart and limbs.32 Whether aspirin combined with adjusted-dose warfarin would be safe and more effective (in preventing both atherothrombotic and cardiogenic strokes) than warfarin alone in patients with AF remains unknown.33 Warfarin versus aspirin The relative benefits and risks of warfarin and aspirin have been studied in three trials,15,20,21 all of which showed that warfarin was associated with half the risk of stroke compared with aspirin (47% relative risk reduction; 95% CI, 28%-61%; P < 0.01) (E1).26 Warfarin combined with aspirin For patients with AF who are at high risk of stroke, adding aspirin (325 mg daily) to low-intensity, fixed-dose warfarin, adjusted to an INR of 1.2-1.5, was not as effective in preventing stroke or systemic thromboembolism as standard adjusted-dose warfarin therapy, maintaining an INR of 2.0-3.0 (event rates, 7.9% per year v. 1.9% per year, respectively; P < 0.0001), and there is no difference in the rates of major bleeding (E2).22 Three subsequent trials also suggested that adjusted-dose warfarin (INR, 2.0-3.0) was superior to low-intensity anticoagulant therapy or an aspirin- anticoagulation regimen (E1).23-25 Warfarin versus other antiplatelet agents An Italian study reported that a new antiplatelet agent, indobufen (100-200 mg twice daily), was as effective as adjusted-dose warfarin (INR, 2.0-3.5) in preventing stroke, systemic embolism, myocardial infarction or vascular death in 916 patients with non-valvular AF and recent (within 15 days) TIA or non-disabling ischaemic stroke (E2).34 The 12-month event rates were 10% in the warfarin group and 12% in the indobufen group (P = 0.47). However, the number of patients and outcome events were quite small, follow-up was short, and it is possible that a true difference was not detected. Future studies are planned to evaluate the safety and effectiveness of other, newer antiplatelet agents (such as clopidogrel, oral glycoprotein IIb/IIIa receptor inhibitors, and oral thrombin inhibitors) and combination antiplatelet therapies (such as aspirin-ticlopidine, aspirin-clopidogrel, and aspirin-dipyridamole) as strategies of thromboprophylaxis in AF. Who to treat and with what? Not all patients with AF benefit from thromboprophylactic treatment. The decision to treat depends on the balance between the risk of thromboemboli without treatment and the risks of thromboemboli and haemorrhage with treatment in each patient, as well as the patient's willingness to accept the potential risks, costs, and inconvenience of treatment in order to possibly benefit. The current profile of individual risk of thromboembolism and bleeding complications (see below) remains imprecise and continues to be refined as new data emerge.7 Who is at high risk of stroke and thromboembolism without treatment? The important independent prognostic factors for an increased risk of stroke among individuals with AF are increasing age, a history of previous TIA or stroke, hypertension, diabetes mellitus, and transthoracic echocardiographic evidence of moderate to severe left ventricular systolic dysfunction (E1).7,26,35-37 Echocardiographic evidence of left atrial enlargement (E2) and left atrial spontaneous echo densities ("smoke"), possibly indicative of stasis of blood, are also significant risk factors for stroke36-39 (E33). These risk factors are cumulative: for people younger than 65 years with no risk factors the untreated annual risk of stroke is about 1%, whereas with one or more risk factors it is about 5%; for people aged 65-75 years with no risk factors the annual risk of stroke is about 4%, and with one or more risk factors it is about 6% per year; and for people older than 75 years with no risk factors the risk of stroke is about 3%-4%, whereas with one or more risk factors it is about 8% (see Box 2) (E1).7,26 Who is at high risk of haemorrhage with anticoagulant treatment? The major risk factors for anticoagulant-associated intracranial haemorrhage include fragile intracranial blood vessels (previous symptomatic cerebrovascular disease, computed tomography brain scan evidence of small vessel disease ["leukoaraioisis"]), high blood pressure (poorly controlled hypertension), and excessive anticoagulation (INR, > 3.5) or factors predisposing to it, such as confusion, dementia, inadequate anticoagulant monitoring, alcoholic liver disease, and a tendency to falls (E2).40,41 Increasing age is a risk factor for all of these risk factors, and is thus a potent risk factor for anticoagulant-associated haemorrhage. Among a subgroup of patients in the Stroke Prevention in Atrial Fibrillation (SPAF) II trial (mean age, 80 years), the rate of intracranial haemorrhage was as high as 1.8% per year in those allocated to warfarin therapy (target INR, 2.0-4.5) and 0.8% among those who were assigned to aspirin (E2).21 Although the target INR in this study was higher than currently recommended (INR, 2.0-3.0), these data suggest that the low rate of intracranial haemorrhage documented in the five primary prevention AF trials15-19 may not apply to very elderly individuals (who were not well represented in many of these trials -- the mean age of the patients studied in the AF trials was 69 years, and only about a quarter were older than 75 years). Recommendations for antithrombotic therapy for AF Current practice necessitates individualisation of therapy after an integrated clinical assessment that evaluates thromboembolic risk due to AF, other potential indications for anticoagulant therapy, risk of haemorrhage, and non-medical factors relating to compliance, capacity to have the INR monitored at least monthly, gait instability, risk of other trauma, and patient values and preferences.42,43 Decision analysis can also be useful.44The role of transthoracic echocardiography (TTE), in addition to excluding structural heart disease in all patients who first present with AF, is to further refine stroke risk in the small group of patients with a low risk of stroke according to clinical factors. Although TOE is more sensitive in detecting left atrial thrombus and spontaneous echo contrast, which are markers for increased risk of thromboembolism,36-39 it is more invasive and is usually only required to improve risk stratification among individuals with a relative contraindication to warfarin or in whom TTE is inadequate. The choices of thromboprophylactic agents for atrial fibrillation include warfarin, which is the most effective but also the most risky treatment, and aspirin, which is less effective than warfarin but safer (E1). The combination of aspirin and low dose warfarin is no more effective than aspirin alone (E1).22,23 The most appropriate treatment regimen is one in which patients at high risk of stroke and low risk of haemorrhage are treated with warfarin, and patients at low risk of stroke or high risk of haemorrhage are treated with aspirin. Who not to treat Individuals with AF who are aged less than 60 years and have no evidence of any concurrent heart disease have a very low risk of a thromboembolic event (about 0.6% per year).45 The potential benefits of aspirin in these patients (which may reduce the risk of stroke by 0.12% per year [20% of 0.6%]) may be offset by an equal potential risk of aspirin-associated haemorrhagic stroke of 0.12%.46 Who to treat with aspirin Aspirin is indicated for individuals in AF who are at fairly low absolute risk of stroke, such as those without any of the independent thromboembolic risk factors listed above, or those at risk of an anticoagulant-related haemorrhage which exceeds the risk of stroke (more than 1% per year) (E1). For some people, such as the elderly and those with hypertension, whose risks of stroke and haemorrhage are both high, the treatment decision can be difficult, and may be determined ultimately by the patient's preferences.42,43Patients taking aspirin should be monitored over time and their treatment changed to warfarin if risk factors emerge; this occurs in 10%-15% of patients being treated with aspirin per year.21 Who to treat with warfarin Warfarin is indicated for individuals with chronic AF who are at high absolute risk of stroke (> 4% per year), such as those with any of the independent thromboembolic risk factors listed above, and a lower risk of haemorrhage (E1) (see Box 2). Similarly, anticoagulant therapy should also be considered in patients with paroxysmal AF, again depending on the thromboembolic risk factors (Box 2) as well as the frequency and duration of the paroxysms. Although clinical trial evidence suggests the stroke rate of patients with paroxysmal AF is similar to that of patients with chronic AF,26 the trials did not specifically examine the benefits of antithrombotic therapy in patients with paroxysmal AF. Furthermore, the range of thromboembolic risk in such patients is likely to be extremely wide, from very low for patients who have one short paroxysm once a year to considerably higher for patients who have daily lengthy paroxysms. What is the optimal target INR? The intensity of oral anticoagulant therapy that provides the best balance between the prevention of thromboembolism and the occurrence of bleeding complications appears to be an INR of between 2.0 and 3.0, but may be lower (INR, 1.8 to 2.0) in patients at greater risk of bleeding (eg, the elderly), and may be higher in patients at greater risk of thromboembolism, such as those with prosthetic heart valves [INR, 3.0-4.0]) (E32).47,48It is important to emphasise that, in people in whom anticoagulant therapy is indicated, the risk of stroke increases substantially when the INR falls below 2.0. Patients with an INR of 1.7 have twice the odds of stroke (95% CI, 1.6-2.4 times), and those with an INR of 1.5 have 3.3 times the odds of stroke (95% CI, 2.4-4.6 times) as those with an INR of 2.047 (E32). What if warfarin therapy needs to be ceased? When cessation of warfarin therapy is required because of other (usually surgical) procedures, it is necessary to stratify the invasiveness of the procedure (minimal versus major) and the short-term risk of thromboembolism. Warfarin can be discontinued for five days before a major procedure and continued at a decreased dose for a minor procedure. Therapy should be reinstituted as soon as possible after invasive procedures. Patients at high risk of thromboembolism (eg, severe mitral stenosis, mechanical mitral prosthesis, left ventricular dysfunction) should be admitted to hospital early for intravenous administration of heparin during warfarin cessation. References Lake FR, Cullen KJ, de Klerk NH, et al. Atrial fibrillation and mortality in an elderly population. Aust N Z J Med 1989; 19: 321-326. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke 1991; 22: 983-988. National Health and Medical Research Council. A guide to the development, implementation, and evaluation of clinical practice guidelines. Canberra: NHMRC, 1999. Kilborn MJ. Atrial fibrillation. Med J Aust 1999; 170: 498-504. Falk RH, Knowlton AA, Bernard SA, et al. Digoxin for converting recent-onset atrial fibrillation to sinus rhythm. Ann Intern Med 1987; 106: 503-506. Stoddard MF. Risk of thromboembolism in new onset or transient atrial fibrillation. Prog Cardiovasc Dis 1996; 39: 69-80. Laupacis A, Albers G, Dalen J, et al. Antithrombotic therapy in atrial fibrillation. Chest 1998; 114 (5 Suppl): 579S-589S. Klein AL, Grimm RA, Black IW, et al. Cardioversion guided by transesophageal echocardiography: The ACUTE Pilot Study. A randomised, controlled trial. Ann Intern Med 1997; 126: 200-209. Bashir M, Grimm RA, Jaber WA, et al. Elderly patients do not have an excessive risk for complications or recurrence following transoesophageal echocardiography-guided cardioversion of atrial arrhythmias: results from the ACUTE registry. J Am Coll Cardiol 2000; 35 (Suppl A): 119A (abstract no. 1097-73). Catherwood E, Fitzpatrick WD, Greenberg ML, et al. Cost-effectiveness of cardioversion and anti-arrhythmic therapy in nonvalvular atrial fibrillation. Ann Intern Med 1999; 130: 625-636. Roy D, Talajic M, Dorian P, et al. Amiodarone to prevent recurrence of atrial fibrillation. N Engl J Med 2000; 342: 913-920. Resnekov L. Present status of electroversion in the management of cardiac dysrhythmias. Circulation 1973; 47: 1356-1363. Lown B, Perlroth MG, Kaidbey S, et al. "Cardioversion" of atrial fibrillation: a report on the treatment of 65 episodes in 50 patients. N Engl J Med 1963; 269: 325-331. Manning WJ, Silverman DI, Gordon SPF, et al. Cardioversion from atrial fibrillation without prolonged anticoagulation with use of transesophageal echocardiography to exclude the presence of atrial thrombi. N Engl J Med 1993; 328: 750-755. Petersen P, Boysen G, Godfredsen J, et al. Placebo-controlled randomised trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation. Lancet 1988; i: 175-179. The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. The Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators. N Engl J Med 1990; 323: 1505-1511. Stroke prevention in atrial fibrillation study: final results. Circulation 1991; 84: 527-539. Connolly SJ, Laupacis A, Gent M, et al, for the CAFA Study Coinvestigators. Canadian atrial fibrillation anticoagulation (CAFA) study. J Am Coll Cardiol 1991; 18: 349-355. Ezekowitz MD, Bridgers SL, James KE, et al, for the Veterans Affairs Stroke Prevention in Nonrheumatic Atrial Fibrillation (SPINAF) Investigators. Warfarin in the prevention of stroke associated with atrial fibrillation. N Engl J Med 1992; 327: 1406-1412. Secondary prevention in nonrheumatic atrial fibrillation and transient ischaemic attack or minor stroke. EAFT (European Atrial Fibrillation Trial) Study Group. Lancet 1993; 342: 1255-1262. Warfarin versus aspirin for the prevention of thrombo-embolism in atrial fibrillation. Stroke prevention in atrial fibrillation II study. Lancet 1994; 343: 687-691. Adjusted-dose warfarin versus low-intensity, fixed-dose warfarin plus aspirin for high risk patients with atrial fibrillation: the Stroke Prevention in Atrial Fibrillation III randomised clinical trial. Lancet 1996; 348: 633-638. Gullov AL, Koefoed BG, Petersen P, et al. Mini-dose warfarin and aspirin in atrial fibrillation. Second Copenhagen Atrial Fibrillation Aspirin and Anticoagulation Study (AFASAK 2). Arch Intern Med 1998; 158: 1513-1521. Vermeer F, Langenberg M, Hellemons BS, et al. Primary prevention of arterial thrombo-embolism in non-rheumatic atrial fibrillation: results of the PATAF study. Eur Heart J 1998; 19 (Abstract Suppl): 154. Pengo V, Zasso A, Barberi F, et al. Effectiveness of fixed minidose warfarin in the prevention of thromboembolism and vascular death in nonrheumatic atrial fibrillation. Am J Cardiol 1998; 82: 433-437. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomised controlled trials. Arch Intern Med 1994; 154: 1449-1457. Hart RG, Boop BS, Anderson DC. Oral anticoagulants and intracranial haemorrhage. Facts and hypotheses. Stroke 1995; 26: 1471-1477. Gallus AS, Baker RI, Chong BH, et al, on behalf of the Australasian Society of Thrombosis and Haemostasis. Consensus guidelines for warfarin therapy. Recommendations from the Australasian Society of Thrombosis and Haemostasis. Med J Aust 2000; 172: 600-605. Berge E, Abdelnoor M, Nakstad PH, Sandset PM, on behalf of the HAEST Study Group. Low molecular-weight heparin versus aspirin in patients with acute ischaemic stroke and atrial fibrillation: a double-blind randomised study. Lancet 2000; 355: 1205-1210. The International Stroke Trial (IST): a randomised trial of aspirin, subcutaneous heparin, both, or neither among 19 435 patients with acute ischaemic stroke. International Stroke Trial Collaborative Group. Lancet 1997; 349: 1569-1581. Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: A meta-analysis. Ann Intern Med 1999; 131: 492-501. Collaborative overview of randomised trials of anti platelet therapy. I: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Antiplatelet Trialists' Collaboration. BMJ 1994; 308: 81-106. Peverill RE. Warfarin or aspirin: both or others? Med J Aust 1999; 171: 321-326. Morocutti C, Amabile G, Fattapposta F, et al, for the SIFA (Studio Italiano Fibrillazione Atriale) Investigators. Indobufen versus warfarin in the secondary prevention of major vascular events in nonrheumatic atrial fibrillation. Stroke 1997; 28: 1015-1021. Predictors of thromboembolism in atrial fibrillation: clinical features of patients at risk. The Stroke Prevention in Atrial Fibrillation Investigators. Ann Intern Med 1992; 116: 1-5. Predictors of thromboembolism in atrial fibrillation: echocardiographic features of patients at risk. The Stroke Prevention in Atrial Fibrillation Investigators. Ann Intern Med 1992; 116: 6-12. Atrial Fibrillation Investigators. Echocardiographic predictors of stroke in patients with atrial fibrillation. A prospective study of 1066 patients from 3 clinical trials. Arch Intern Med 1998; 158: 1316-1320. Fatkin D, Feneley M. Stratification of thromboembolic risk of atrial fibrillation by transthoracic echocardiography: the relative role of left atrial appendage function, mitral valve disease, and spontaneous echo contrast. Prog Cardiovasc Dis 1996; 39: 57-68. Jones EF, Calafiore P, McNeil J, et al. Atrial fibrillation with left atrial spontaneous contrast detected by transoesophageal echocardiography is a potent risk factor for stroke. Am J Cardiol 1996; 78: 425-429. Bleeding during antithrombotic therapy in patients with atrial fibrillation. The Stroke Prevention in Atrial Fibrillation Investigators. Arch Intern Med 1996; 156: 409-416. A randomised trial of anticoagulants versus aspirin after cerebral ischaemia of presumed arterial origin. The Stroke Prevention in Reversible Ischaemia Trial (SPIRIT) Study Group. Ann Neurol 1997, 42: 857-865. Man-Son-Hing M, Laupacis A, O'Connor A, Wells G. Warfarin for atrial fibrillation: the patient perspective. Arch Intern Med 1996; 156: 1841-1848. Gage BF, Cardinalli AB, Owens DK. Cost-effectiveness of preference-based antithrombotic therapy for patients with nonvalvular atrial fibrillation. Stroke 1998; 29: 1083-1091. Thomson R, Parkin D, Eccles M, et al. Decision analysis and guidelines for anticoagulant therapy to prevent stroke in patients with atrial fibrillation. Lancet 2000; 355: 956-962. Kopecky SL, Gersh BJ, McGoon MD. The natural history of lone atrial fibrillation: a population-based study over three decades. N Engl J Med 1987; 317: 669-674. He J, Whelton PK, Vu B, Klag MJ. Aspirin and risk of hemorrhagic stroke. A meta-analysis of randomised controlled trials. JAMA 1998; 280: 1930-1935. Hylek EM, Skates SJ, Sheehan MA, Singer DE. An analysis of the lowest effective intensity of prophylactic anticoagulation for patients with non-rheumatic atrial fibrillation. N Engl J Med 1996; 335: 540-546. Cannegieter SC, Rosendal FR, Wintzen AR, et al. Optimal oral anticoagulant therapy in patients with mechanical heart valves. N Engl J Med 1995; 333: 11-17. Background and evidence basis of recommendations The National Heart Foundation (NHF) Consensus Guidelines for Non-valvular Atrial Fibrillation and Stroke Prevention were written by Clinical Associate Professor Graeme J Hankey on behalf of the National Blood Pressure Advisory Committee of the NHF, which comprises Professor L Wing (chair), Dr A Boyden, Dr A Dart, Dr K Duggan, Clinical Associate Professor G Hankey, Dr M Nelson, Professor I Puddey, Dr M Stowasser, and Dr J Vial. The draft guidelines were circulated for comment to the above members of the committee, who have clinical and research expertise or interests in hypertension, atrial fibrillation, and stroke prevention. Comment was also sought from the Medical Director of the Heart Foundation, Professor Andrew Tonkin. All comments were incorporated into the final document, which was ratified by the Heart Foundation's Cardiovascular Health Advisory Committee. All available evidence from controlled experimental and observational studies was combined with clinical experience to provide recommendations according to the National Health and Medical Research Council Quality of Evidence ratings.3 Authors' details National Heart Foundation of Australia, Melbourne, VIC. Graeme J Hankey, MD, FRACP, Consultant Neurologist and Head of Stroke Unit, Royal Perth Hospital, Perth, WA, and Clinical Associate Professor, Department of Medicine, University of Western Australia. Reprints will not be available from the author. Correspondence: Clinical Associate Professor G J Hankey, Stroke Unit, Royal Perth Hospital, Wellington Street, Perth, WA 6001. gjhankeyATcyllene.uwa.edu.au * L Wing (chair), A Boyden, A Dart, K Duggan, M Nelson, I Puddey, M Stowasser, J Vial 1: Level-of-evidence codes Evidence for the statements made in this article is graded according to the NHMRC system3 for assessing the level of evidence: E1 Level I Evidence obtained from a systematic review of all relevant randomised controlled trials. E2 Level II Evidence obtained from at least one properly designed randomised controlled trial. E31 Level III-1 Evidence obtained from well-designed pseudo-randomised controlled trials (alternate allocation or some other method). E32 Level III-2 Evidence obtained from comparative studies with concurrent controls and allocation not randomised (cohort studies), case-control studies, or interrupted time series without a parallel control group. E33 Level III-3 Evidence obtained from comparative studies with historical control, two or more single-arm studies, or interrupted time series without a parallel control group. E4 Level IV Evidence obtained from case-series, either post-test, or pretest and post-test. Back to text 2: Risk stratification and prophylaxis in atrial fibrillation High risk (6%-12% per year risk of stroke) Age >65 years and hypertension or diabetes Previous transient ischaemic attack (TIA) or stroke Valvular heart disease o Heart failure Recent myocardial infarction Impaired left ventricular function on echocardiography Thyroid disease o Left atrial thrombus or left atrial spontaneous echo contrast (TOE done on basis of clinical suspicion) Treatment: Warfarin (target INR 2.0-3.0) if possible and not contraindicated. Moderate risk (2%-5% per year risk of stroke) Age 65 years and hypertension or diabetes Age >65 years and not in high risk group Treatment: Warfarin (target INR 2.0-3.0) or aspirin 75-300mg daily, depending on individual case and echocardiography findings. Low risk (≤1% per year risk of stroke) Age 65 and no hypertension, diabetes, TIA, stroke, or other clinical risk factors Treatment: None, or aspirin 75-300mg daily. Back to text

Graeme J Hankey

For Debate Disease control in the information era Robert M Douglas MJA 2001; 174: 241-243 Abstract - Information technology - Ownership, privacy and access - Aggregation of individual records - Conclusion - Acknowledgements - References - Authors' details - - More articles on Infectious diseases and parasitology Abstract As a result of advances in information technology, there is now a new capacity to manage, interpret and apply data for the benefit not only of individual patients but of the population as a whole. Population health information systems are currently inadequate to meet the needs of disease control. In a rapidly changing world, effective public health action requires timely and efficient data about what is happening in the whole population. As the national effort to harness information technology to the needs of individual patient care begins, it is desirable that the electronic patient record also becomes the building block for public health research and monitoring. Individual healthcare and population healthcare should be two sides of the one coin. Ownership, privacy and access to the contents of the electronic health record should now be addressed in the context that disease control in the whole population will increasingly depend upon an efficient "real time" information system. In the past 50 years, vaccination and antibiotics have transformed our capacity to manage human infections. Biotechnology is now opening up new possibilities for managing the human genome. Our capacity to change individual patient outcomes through modern clinical treatment rightly commands headlines and attracts public resources. But control of disease in the population as a whole requires more than individual clinical action. Last1 has emphasised that the systematic control of any disease requires a consensus that a particular disease problem exists; an understanding of its cause; the ability to control its cause; a belief that the problem matters; and the political will to control it. Good information systems are essential to support all of these requirements. The relevance of information to disease control is well illustrated by the contrast in management of two communicable disease groups, HIV and respiratory infections. HIV incidence rates are declining in Australia2 at a time when those in other nations, particularly African countries,3are rising. This apparent success is believed to be at least partly attributable to a widely supported national control program that was begun in 1989.4 The program promoted broad community understanding of the nature of the disease and a shared commitment by stakeholders and governments to contain the problem. Unlikely coalitions were formed between scientists, risk groups and community representatives, and the entire community was involved in approaches that were pragmatic as well as innovative. As the epidemic progressed, research focused not only on the behaviour of the virus, but also on the behaviour of the humans who transmitted it. A central ingredient of the HIV public health strategy was the development of an excellent information system at a time when information about other diseases was (and continues to be) in disarray. The data system developed by the National Centre for HIV Epidemiology and Clinical Research helped the nation to view the HIV epidemic as a population problem rather than an individual issue, and enabled the public health community to monitor its progress, and modulate the public health response accordingly. In stark contrast, infections of the respiratory tract, which dominate the clinical experience of primary care practitioners everywhere5 and cause extensive morbidity and absenteeism, are not matched by any systematic national effort to contain them. Perhaps, as they now rarely cause death, we have become complacent about them. Yet, on average, all Australians experience two to three acute respiratory episodes per year and in early childhood the average is five to eight. Apart from the misery they cause, these infections result in widespread misuse of antibiotics and the serious threat of antibiotic ineffectiveness in the longer term.6 It is remarkable that in this field Australia has no public health strategy, no national commitment to the problem, inadequate preventive effort, and no program of either social or biological research. Common respiratory infections are left to patients, clinicians and pharmacists, despite the magnitude of the problem and the negative impact of antibiotic abuse. We tolerate a level of morbidity and misapplication of resources to respiratory infections that does not make public health sense. We collect no systematic information about the problem, and can only guess at its cost to the community. The lack of available information means that the public health perspective is not addressed. Information technology Unlike the banking and tourism sectors, the healthcare sector has not yet harnessed the electronic information revolution to the needs of either individuals or populations. It is time we re-examined the issue of information in healthcare in the light of the new opportunities created by modern technology. Australian health ministers have recently agreed to establish a national health information network built on electronic health records that, through data linkage, can enhance the quality of individual patient care.7 For individual clinical care, which often involves many agencies and professionals, linkage of electronic records is essential to avoid duplication and to ensure that an individual's medical history is accessible and complete wherever the person presents for care (see Box). Because of modern transport and human mobility, fragments of a patient's history may be scattered in many places. Modern technology now permits linkage of data that are "warehoused" in multiple, geographically dispersed electronic sites. At this early developmental stage it is essential that the new system be designed to serve the needs not only of individual patients and their clinicians, but also of organisations concerned with public health monitoring, research and administration. The personal electronic health record, however it is stored and accessed, should also be the building block for "real time" public health surveillance. Improved efficiency of personal clinical care and improved management of public health both require the same data and should become two sides of the one coin. When a patient presents for care by a general practitioner for a respiratory infection or a manifestation of HIV, that information should also automatically become part of national public health monitoring activity. And when the laboratory reports to the GP that the respiratory infection is, or is not, a new strain of influenza that fact should, as well as informing the clinician instantly, feed into a national database that informs public health action. Ownership, privacy and access The development of a national integrated health record and information system poses a range of logistic, ethical, privacy and professional issues. These need to be resolved during the system design phase so that a "rail-gauge" problem (ie, one of incompatibility) does not develop between the States, between the public and private healthcare systems, or between the outputs to clinicians and public health practitioners. A well designed and protected retrieval system would offer major benefits for disease prevention and control in the population as a whole. We need to break out of the mentality that sees medical records as being "owned" and controlled by doctors or hospitals. When a patient contracts for medical care from a doctor, the record that is prepared is paid for jointly by the patient and the community. They, jointly, should be the owners and controllers of the electronic record. Because elements of individual experience are pertinent to the health of the whole community, and because the vast majority of healthcare costs are met by the community, it is important that individual experiences be aggregated to inform public health action. Privacy and confidentiality must be protected and respected, but so also must public good. One can not be allowed to drown out the other. Access to an individual patient record should be controlled by the patient using a unique identifier, such as a thumbprint. An individual's records could be linked by Medicare number with repositories of individual data stored in multiple data warehouses that are themselves linked by the Internet. Through the use of a thumbprint and an access command, patients could authorise different healthcare providers to access different parts of their medical record. While a GP or medical specialist might be given access to the entire record, pharmacists might simply be given access to medication records. Mechanisms to ensure that access to identifiable data is precisely limited and carefully monitored should be part of the system specification. Aggregation of individual records Having created a comprehensive linked record for each person, it is desirable that the individual records become instantly accessible building blocks for defined administrative, monitoring and research databases. Those who use the aggregated datasets to monitor the health of the community should be denied access to personal identifiers, and should not need informed consent to make use of the de-identified data. If analysis of de-identified data revealed new community threats, access to the identification of individuals would be required in the interests of those individuals and of the public. In these circumstances, access protocols would be needed, backed up by audit trails, the right of redress for consumers in the event of misuse, and legal protection for public health practitioners who operate within carefully defined parameters. There are thoroughly reasonable concerns that in the process of aggregation privacy and confidentiality might be compromised. These concerns would need to be addressed in system design. Nevertheless, few systems are absolutely foolproof, and the design would need to include audit trails and monitoring systems to ensure that abuse could be traced and dealt with. Imperfection in this area has not prevented the banking industry from capitalising on the benefits of the new technology, and it should not justify inaction in the sphere of healthcare. Conclusion Currently, Australia's healthcare information systems are inadequate and, partly as a consequence, public health action is seriously deficient. Efficient, real-time information systems are a starting point for effective public health. Public health action can profoundly benefit the whole community by reducing the incidence of disease and the need for clinical care. It makes no more sense to plan public health action without high quality data than it does to prescribe a drug for hypertension without measuring blood pressure. It is essential that as we move towards electronic storage of health records we simultaneously address the public health need for a vastly improved body of data. Acknowledgements This is an edited version of the Sidney Sax Oration, presented to the ACT Branch of the Public Health Association of Australia, 10 August 2000. I am grateful to Dr Chris Mount for comments on the manuscript and for collaboration on many of the ideas discussed in this article, and to Jacquie Steele for preparation of the manuscript. References Last, J. Fouling and cleansing our nest: human-induced ecological determinants of infectious disease. Perspect Hum Biol 1999; 4: 145-147. Law MG, Li Y, McDonald AM, et al. Estimating the population impact in Australia of improved antiretroviral treatment for HIV infection. AIDS 2000; 14: 197-201. Ziegler JB, Ffrench RA. XIII International AIDS Conference, Durban, 9-14 July, 2000 [conference report]. Med J Aust 2000; 173: 572-574. Commonwealth of Australia. AIDS. A time to care: a time to act. Towards a strategy for Australians. Canberra: AGPS, 1988. Douglas RM. Respiratory tract infections as a public health challenge. Clin Infect Dis 1999; 28: 192-194. Schwartz, B. Preventing the spread of antimicrobial resistance among bacterial respiratory pathogens in industrialized countries: the case for judicious antimicrobial use. Clin Infect Dis 1999; 28: 211-218. National Electronic Health Records Taskforce. A health information network for Australia. Canberra: Commonwealth of Australia, 2000. Authors' details National Centre for Epidemiology and Population Health, Australian National University, Canberra, ACT. Robert M Douglas, MD, FRACP, FAFPHM, Visiting Fellow. Reprints will not be available from the author. Correspondence: Professor R M Douglas, National Centre for Epidemiology and Population Health, Australian National University, Canberra, ACT 0200. Bob. DouglasATanu.edu.au Electronic health records as the building blocks for a national health information system Future electronic health records are likely to include the following information on a patient: -- summary of previous medical history; -- current problems; - medications prescribed and dispensed; -- laboratory and radiological results; -- hospital discharge summaries; -- care plans and record of use of community care; -- allergies and adverse reactions; -- elements of social and demographic history. All data for an individual patient should be linked by a common identifier to ensure accuracy and improve safety. Clinical access to the electronic record should be authorised by the patient, and the national system should make it deliverable anywhere in Australia. Patient data should automatically feed into specifically designed national datasets that monitor various elements of the nation's health and can provide information for public health action. Protection of privacy and confidentiality must be designed into the system. Back to text

Robert M Douglas

MJA 2001; 174: 248-253 Abstract - The tools of EBM - Does it improve outcomes? - The future of EBM - Conclusions - Acknowledgements - References - Authors' details Abstract Evidence-based medicine (EBM) integrates clinical experience and patient values with the best available research information. There are four steps in incorporating the best available research evidence in decision making: asking answerable questions; accessing the best information; appraising the information for validity and relevance; and applying the information to patient care. Applying EBM to individual patients requires drawing up a balance sheet of benefits and harms based on research and individual patient data. The most realistic and efficient use of EBM by clinicians at the point of care involves accessing and applying valid and relevant summaries of research evidence (evidence-based guidelines and systematic reviews). The future holds promise for improved primary research, better EBM summaries, greater access to these summaries, and better implementation systems for evidence-based practice. Computer-assisted decision support tools for clinicians facilitate integration of individual patient data with the best available research data. Medical practice is diverse but has some common tasks. One of these is making the best use of available research evidence to diagnose, prevent and treat disease. Imagine yourself in the following situations: As a general practitioner you see a 72-year-old asymptomatic man who wants to know whether he should be screened for colorectal cancer. What do you say? As a nephrologist you see a 50-year-old man with progressive renal failure due to glomerulonephritis. Should you advise treatment with cyclosporin A? As an interventional radiologist or general surgeon should you be using antibiotic- or antiseptic-impregnated central venous lines to prevent line-associated sepsis? Do you know enough of the research evidence to provide sensible answers to these questions? Even if you know today's answers, chances are that they will change as much in the next five years as they have changed in the past five years.1-4 Therein lies the challenge — keeping up with new research information and incorporating it into clinical decision making. This is the task of evidence-based medicine (EBM). In a survey of 625 office-based primary-care physicians and 100 physician opinion leaders in the United States, nearly two-thirds reported that the current volume of scientific information was unmanageable.4 When the researchers asked about the physicians' knowledge of important recent medical advances, they found deficiencies that would adversely affect patient care. Since 1989, when the above study was published, total biomedical knowledge has probably increased by about 50%.5 In addition to identifying a challenge, EBM also provides tools to find, appraise and apply research evidence better.6 These tools are relevant to all users of health information — clinicians, patients and policymakers — but our focus will be on helping clinicians make better use of EBM tools. The tools of evidence-based medicine In a recent survey, Australasian physicians identified insufficient time (74%), limited search skills (41%) and limited access to evidence (43%) as impediments to making better use of research data.7 The survey showed that, to realise the full potential of EBM to improve care, two things are needed: education in EBM, and systems that quickly deliver high-quality evidence at the point of clinical decision making. The EBM process — incorporating the best available research evidence in decision making — has four steps: asking answerable questions; accessing the best information; appraising the information for validity and relevance; and applying the information to patient care. Asking answerable questions Accept that you may not know: While the knowledge explosion continues, making assumptions about the certainty of our knowledge base is risky. Studies of information needs show that one to two questions are generated for each outpatient consultation and five questions for each inpatient consultation.8,9 About a third of these questions are about treatment of a specific condition, and a quarter are about diagnosis.10 EBM tools help answer these "foreground" questions,11 which are specific and relevant to clinical decision making. Other questions concerning basic biological processes, or "background" questions (questions beginning with What is . . . ? and How does . . . ?), are better answered by standard textbooks. For these the EBM framework is not particularly helpful. Framing the question: population, intervention, comparator, outcome: Once the clinical question has been identified, it then needs to be put into a searchable and answerable form. This consists of four parts: a population with a clinical problem; an intervention or exposure; the comparator intervention or exposure; and the outcomes.11 Such questions are specific, and should focus on patient-centred or clinically important outcomes, rather than laboratory-based or surrogate outcomes that do not always correspond with patient benefit.12 For example, the question posed by the 72-year-old man wanting to know about colorectal cancer screening could be rephrased as: "In asymptomatic people at average risk of colorectal cancer (population), does screening by faecal occult blood testing (intervention) reduce mortality from colorectal cancer (outcome) compared with routine care without screening (comparator)?". Aetiology, prognosis, diagnosis or intervention? This four-part question framework can be applied to all types of foreground questions asked by healthcare providers and consumers — Why me? (aetiology), What's wrong with me? (diagnosis), What will to happen to me? (prognosis), and How will intervention change outcome? (intervention). Examples of each type of question are given in Box 1. Because most questions asked by patients and clinicians are about interventions, we will focus on treatment. Accessing the best available information Summarised primary research — evidence-based guidelines and systematic reviews: The ideal information source is valid (contains high quality data), relevant (clinically applicable), comprehensive (has data on all benefits and harms of all possible interventions), and is user-friendly (is quick and easy to access and use). The recent growth of EBM has provided more useful information sources (Box 2), and better access to these information sources (Boxes 2 and 3). Primary research data can now be organised into systematic reviews and evidence-based guidelines. For treatment questions, systematic reviews typically bring together, summarise and synthesise data from randomised controlled trials of a single intervention. Because many interventions are usually possible for the same clinical problem, systematic reviews of these interventions can be further summarised and combined in the form of an evidence-based guideline. To be most useful to clinicians, guidelines should also include diagnostic and prognostic research which provides some guidance for individualising therapy based on disease severity.13 Guidelines and systematic reviews can be stand-alone products, or, more usefully, can be organised into compendia. Primary research — if quality summarised research is not available: When relevant systematic reviews or evidence-based guidelines are unavailable, or if they fail the quick critical appraisal test outlined in the following section, the clinician will need to find primary research studies. Although a randomised controlled trial is the best study type to assess the effects of a healthcare intervention, it is not the best study design to determine the accuracy of a diagnostic test or the prognosis of a condition, and is frequently not feasible for questions of aetiology. Box 1 includes the ideal primary study design for each type of question, along with the appropriate methodological terms that help focus MEDLINE searches on these studies. For users unfamiliar with methodological filters, the free website for MEDLINE, PubMed, has a "clinical queries" option which allows users to select the content area and the type of question (therapy, diagnosis, aetiology or prognosis). The program then automatically incorporates the methodological filters into the search.14 Appraisal of quality and clinical relevance Why quality assessment is needed: Having found the research information, the user then needs to critically appraise the study or studies. Publication does not guarantee quality, and poor-quality studies tend to overestimate the benefits of interventions by about 30%15 — enough to make ineffective interventions appear effective. Likewise, poor-quality studies of diagnostic tests overestimate the accuracy of the test they are evaluating.16 Tools for critical appraisal: Useful tools for critical appraisal developed for the National Health and Medical Research Council (NHMRC)17,18 are summarised in Boxes 3 and 4. They ask three questions: How strong is the evidence? How big is the effect? Does the effect matter to patients? The strength of evidence incorporates the appropriateness of the study design (often called level of evidence), the quality of the study's design and reporting (was bias minimised?) and the statistical precision of the results (could the results be explained by chance?). A few seconds scanning an abstract to see how well it rates on these criteria is often enough to indicate whether the study is worth reading. If the initial scan suggests that the results may be reliable and important, then critical appraisal means focusing on the methods section to see how the study was done (not to see what statistical tests were used, such as whether a χ 2 or t-test was used) and on the results, particularly the figures and tables. Applying the research evidence to decision making Once the best available evidence has been found and appraised, the final step is to apply the research to decision making. To determine whether the results of a trial of a treatment are applicable to a particular patient, it seems reasonable to compare the patient's characteristics with the trial's inclusion criteria. This approach may lead to treating some patients who may experience more harm than benefit.19 An alternative approach helps avoid this problem.20 1. Make a balance sheet of the benefits and harms of the intervention All outcomes (both beneficial and harmful) that are important to the patient and influenced by the intervention need to be considered. For example, for the man with progressive kidney disease in our second scenario, we would need to consider the possible benefits of cyclosporin (reducing the need for dialysis) alongside the possible harms (gum hypertrophy, hypertension and hypertrichosis). 2. From research data, quantify the likelihood of benefits and harms in relative terms How likely is it that the benefits and harms will affect an individual patient? To estimate this we need to know the average effect of the treatment from systematic reviews (or trials, if systematic reviews are not available) and whether the effect varies according to patient and disease factors or whether it is relatively constant and independent of these factors. This type of information comes from subgroup analyses of systematic reviews and large trials. The benefits and harms of interventions are generally best expressed in relative terms (such as relative risks), because the relative effect is often stable across many different patient subgroups. In the trial of cyclosporin for progressive kidney disease, the relative risk of needing dialysis was 0.4 (the risk of dialysis was 0.4 times lower in those treated with cyclosporin than in those not treated with it), but the study was too small to determine whether the effect varied in different subgroups. 3. Convert the relative benefits and harms into absolute terms for your patient using the patient's specific characteristics If the relative beneficial effect of treatment is stable across patients at different levels of risk from their disease, then those at greatest risk will have the most to gain from treatment, and those at least risk from their disease will have the least to gain. The absolute benefit of treatment (how much they have to gain) can be calculated by combining the relative effect of treatment (from randomised trials and systematic reviews) with the risk of the outcome without treatment (from cohort studies of prognosis). This is demonstrated in Box 5 using two groups of patients with kidney disease treated with cyclosporin. While valid data from trials about the average benefit of a treatment are important, we also need valid data (preferably local) about the prognosis of patients without treatment to estimate the absolute benefit for any particular patient. To return to our example, the patient's renal function, blood pressure and degree of proteinuria indicate that he belongs to the low risk group and has a probability of needing dialysis over the next few years of about 10%, which can be reduced to 4% with treatment.20 The same logic can be used to calculate the risk of treatment-related harms. 4. Decide whether the benefits outweigh the harms Having listed all benefits and harms of an intervention and assigned some likelihood for each outcome based on research and individual patient data, the next step is to determine whether, on balance, the treatment is likely to do more good than harm. If the various benefits and harms are roughly equivalent, then this is relatively easy. For example, in weighing up the benefits and harms of thrombolytic therapy for myocardial infarction, it is reasonable to count deaths prevented (from myocardial infarction) as equal to deaths caused (from cerebral haemorrhage) — they are equally undesirable. But not all outcomes are equal. How does a stroke prevented by aspirin compare with a gastrointestinal haemorrhage caused by it? The differential desirability of outcomes can be measured formally, preferably by patients, but more often this integration of probabilities and preferences is informal. Does providing evidence-based care to patients improve outcomes? Observational studies show that treatments proven in randomised trials and systematic reviews seem to work equally well in routine clinical practice.21-23 However, it is still unclear what interventions are most effective in helping clinicians use research data more effectively in decision making.24-26 Guidelines, computer-generated reminders, opinion leaders, and outreach visits (or combinations) have been shown to improve care and patient outcomes, but passive dissemination strategies (eg, conferences and printed educational materials) have not. The future of EBM Better information systems at the point of care Due to time constraints, it is impractical to access and appraise at the bedside all of the primary studies applicable to individual patients. Only access to summarised research information is realistic. This should preferably be in the form of succinct evidence-based guidelines (including benefit-harm balance sheets of all available interventions), formatted to be rapidly and easily integrated with specific patient details. Many examples already exist, such as the evidence-based guidelines for early breast cancer (developed by the National Breast Cancer Centre13), and are widely available in hard copy and on the Internet. Ultimately, given the complexity of the data, widespread use of high-quality evidence requires computer-based information management systems. Such computerised decision support systems for clinicians have already been developed, and are available to help clinicians provide better care.26-28 The clinician's role is to use clinical judgement to integrate the best available research information and the patient's unique circumstances and preferences into a plan of management. Better primary research The evidence base of medicine needs to improve in its scope (both by disease and study type) and quality. Some diseases, like early breast cancer, have a large research base to guide decision making.29 However, for most diseases, many important questions remain unanswered, and for those with available evidence there is often considerable room to improve its quality.16-18 Editors of major medical journals have recently provided guidelines to encourage better design and reporting of randomised controlled trials30 and systematic reviews.31 As well as unequal coverage of diseases, there is also unequal coverage of question types. While there are many randomised controlled trials of treatments, there are too few high quality studies of diagnostic tests, prognoses and interventions to help clinicians use research information more effectively. The research data relevant to the questions in the Introduction are given in Box 7. Conclusions The EBM-oriented clinicians of tomorrow have three tasks: To use evidence summaries in clinical practice; To help develop and update selected systematic reviews or evidence-based guidelines in their area of expertise; and To enrol patients in studies of treatment, diagnosis and prognosis on which medical practice is based. Acknowledgements Thanks to Elisabeth Hodson and John Knight for helpful comments on earlier drafts. References Towler B, Irwig L, Glasziou P, et al. A systematic review of the effects of screening for colorectal cancer using the faecal occult blood test, hemoccult. BMJ 1998; 317: 559-565. Cattran DC, Appel GB, Hebert LA, et al. A randomized trial of cyclosporine in patients with steroid-resistant focal segmental glomerulosclerosis. North America Nephrotic Syndrome Study Group. Kidney Int 1999; 56: 2220-2226. Veenstra DL, Saint S, Saha S, et al. Efficacy of antiseptic-impregnated central venous catheters in preventing catheter-related bloodstream infection: a meta-analysis. JAMA 1999; 281: 261-267. Williamson JW, German PS, Weiss R, et al. Health science information management and continuing education of physicians. A survey of U. S. primary care practitioners and their opinion leaders. Ann Intern Med 1989; 110: 151-160. Wyatt J. Uses and sources of medical knowledge. Lancet 1991; 338: 1368-1372. Sackett DL, Straus SE, Richardson WS, et al. Evidence-based medicine: how to practice and teach EBM. 2nd edition. New York: Churchill Livingston, 2000. Scott I, Heyworth R, Fairweather P. The use of evidence-based medicine in the practice of consultant physicians. Results of a questionnaire survey. Aust N Z J Med 2000; 30: 319-326. Covell DG, Uman GC, Manning PR. Information needs in office practice: are they being met? Ann Intern Med 1985; 103: 596-599. Osheroff JA, Forsythe DE, Buchanan BG, et al. Physicians' information needs: analysis of questions posed during clinical teaching. Ann Intern Med 1991; 114: 576-581. Smith R. What clinical information do doctors need? BMJ 1996; 313: 1062-1068. Richardson WS, Wilson MC, Nishikawa J, Hayward RS. The well-built clinical question: a key to evidence-based decisions. ACP J Club 1995; 123: A12-A13. Bucher HC, Guyatt GH, Cook DJ, et al. Users' guides to the medical literature: XIX. Applying clinical trial results. A. How to use an article measuring the effect of an intervention on surrogate end points. Evidence-Based Medicine Working Group. JAMA 1999; 282: 771-778. National Health and Medical Research Council. Clinical practice guidelines for the management of early breast cancer. 2nd edition. Canberra: NHMRC, 2000. <http://www.health.gov.au/nhmrc/advice/pdfcover/eabrscov.htm> Hunt DL, Jaeschke R, McKibbon KA. Users' guides to the medical literature: XXI. Using electronic health information resources in evidence-based practice. JAMA 2000; 283: 1875-1879. Moher D, Pham B, Jones A, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998; 352: 609-613. Lijmer JG, Mol BW, Heisterkamp S, et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 1999; 282: 1061-1066. National Health and Medical Research Council. How to use the evidence: assessment and application of scientific evidence. Canberra: NHMRC, 2000. Liddle J, Williamson M, Irwig L. Method for evaluating research and guideline evidence. Sydney: NSW Health, 1996. Glasziou PP, Irwig LM. An evidence based approach to individualising treatment. BMJ 1995; 311: 1356-1359. Chitalia VC, Wells JE, Robson RA, et al. Predicting renal survival in primary focal glomerulosclerosis from the time of presentation. Kidney Int 1999; 56: 2236-2242. Mitchell JB, Ballard DJ, Whisnant JP, et al. What role do neurologists play in determining the costs and outcomes of stroke patients? Stroke 1996; 27: 1937-1943. Soumerai SB, McLaughlin TJ, Spiegelman D, et al. Adverse outcomes of underuse of beta-blockers in elderly survivors of acute myocardial infarction. JAMA 1997; 277: 115-121. Krumholz HM, Radford MJ, Ellerbeck EF, et al. Aspirin for secondary prevention after acute myocardial infarction in the elderly: prescribed use and outcomes. Ann Intern Med 1996; 124: 292-298. Grimshaw JM, Russell IT. Effect of clinical guidelines on medical practice: a systematic review of rigorous evaluations. Lancet 1993; 342: 1317-1322. Oxman AD, Thomson MA, Davis DA, Haynes RB. No magic bullets: a systematic review of 102 trials of interventions to improve professional practice. CMAJ 1995; 153: 1423-1431. Walton R, Dovey S, Harvey E, Freemantle N. Computer support for determining drug dose: systematic review and meta-analysis. BMJ 1999; 318: 984-990. Chatellier G, Colombet I, Degoulet P. Computer-adjusted dosage of anticoagulant therapy improves the quality of anticoagulation. Medinfo 1998; 9 Pt 2: 819-823. Montgomery AA, Fahey T. A systematic review of the use of computers in the management of hypertension. J Epidemiol Community Health 1998; 52: 520-525. Early Breast Cancer Triallists' Collaborative Group. Tamoxifen for early breast cancer: an overview of the randomised trials. Lancet 1998; 351: 1451-1467. Begg C, Cho M, Eastwood S, et al. Improving the quality of reporting of randomized controlled trials. The CONSORT statement. JAMA 1996; 276: 637-639. Moher D, Cook DJ, Eastwood S, et al. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of Reporting of Meta-analyses. Lancet 1999; 354: 1896-1900. Authors' details Department of Public Health and Community Medicine, University of Sydney, NSW. Jonathan C Craig, MM(ClinEpi), PhD, FRACP, Senior Lecturer; Paediatric Nephrologist, Centre for Kidney Research, Children's Hospital at Westmead; and Coordinating Editor, Cochrane Renal Group, NSW. Les M Irwig, PhD, FFPHM, Professor of Epidemiology. Martin R Stockler, MSc(ClinEpi), FRACP, Senior Lecturer; also at Department of Medicine, and NHMRC Clinical Trials Centre, University of Sydney; and Consultant Medical Oncologist, Sydney Cancer Centre, Royal Prince Alfred Hospital and Concord Repatriation General Hospital, NSW. Reprints will not be available from the authors. Correspondence: Dr J C Craig, Department of Public Health and Community Medicine, Edward Ford Building A27, University of Sydney, NSW 2006. joncAThealth.usyd.edu.au . 1: How to ask answerable clinical questions, where to look, and how to search at a glance Question type Diagnosis Harm/Aetiology Prognosis Intervention Population In people with suspected colorectal cancer Do newborns What proportion of children with febrile seizures In asymptomatic people In patients with central venous lines Intervention/ exposure how accurate is faecal occult blood testing given parenteral vitamin K have a develop recurrent episodes does testing for faecal occult blood lead to does antiseptic impregnation Outcome for diagnosing colorectal cancer higher incidence of leukaemia fewer colorectal cancer deaths cause fewer line- associated infections Comparator compared with colonoscopy than newborns not given vitamin K than routine care without screening than ordinary catheters Best feasible primary study design Cross-sectional analytic study Cohort study (best), population-based case-control study (next best) Cohort study Randomised controlled trial Randomised controlled trial Best MEDLINE search term for study type Sensitivity.tw Risk.tw Exp cohort studies/ Clinical trial.pt if no hits with Randomised controlled trial.pt tw (text word search ["risk.tw" finds the word "risk" in the title or abstract]); pt (publication type ["clinical trial.pt" finds studies which are classified as clinical trials]) Back to text 2: Types of research evidence and usefulness for decision making Type of evidence Advantages Disadvantages Evidence-based guideline Very comprehensive -summarises all relevant research information about all possible interventions for a common clinical problem -improved power to detect small and important differences Very useful applicability information -explores the trade-off of benefit and harm according to the level of risk in different patient subgroups Can be difficult to use if not formatted with the end-user in mind May quickly become out of date Systematic review Moderately comprehensive -summarises all relevant research information about a common intervention Less random error -improved power to detect small and important differences Useful applicability information -analyse variability of effects among different patient subgroups Generally only one of many possible interventions considered Often insufficient data about potential harms Generally provides little information from cohort studies for estimating disease risk to individual patients Primary study Very specific information available Not comprehensive -only one of (usually) many studies available Insufficient for clinical application Back to text 3: Useful sources of evidence-based guidelines, systematic reviews and general EBM resources Name Form and purpose Web access (accessed February 2001) Clinical Evidence Compendium of research evidence of interventions for common medical conditions <http://www.clinicalevidenceonline.org/> Cochrane Library Compendium of systematic reviews and randomised controlled trials <http://www.update-software.com/cochrane/cochrane-frame.html> <http://www.ausdoctors.net/> (follow link to Library) Guideline websites Provided by medical colleges or specialty groups Canadian Medical Association - <http://www.cma.ca/cpgs/index.asp> National Guideline Clearinghouse (US) - <http://www.guideline.gov/index.asp> NHMRC -<http://www.health.gov.au/nhmrc/publicat/cp-home.htm> MEDLINE Compendium of published research PubMed - <http://www.ncbi.nlm.nih.gov/PubMed/> Filter for guidelines (practice guideline as a publication type) Filter for systematic reviews (meta-analysis as a publication type) McMaster University Health Information Research Unit Evidence-based medicine: how to practice and teach EBM6 General "how to do" book <http://hiru.mcmaster.ca/ebm.htm> NHMRC Guidelines toolkit How to review, use, apply, implement and communicate evidence Hardcopy and pdf version available <http://www.health.gov.au/nhmrc/advice/contents.htm> JAMA user's guides to the medical literature JAMA series about all aspects of medical literature (26 articles so far) <http://medicine.ucsf.edu/resources/guidelines/users.html> ScHARR Internet guide for EBM <http://www.shef.ac.uk/~scharr/ir/netting/> Back to text 4: Checklist of critical appraisal items (dimensions of evidence) Item Definition Strength of evidence Level Was the best feasible study design used? (see Box 1) Quality How good was the study design and reporting? (see Box 4) Statistical precision How small was the P-value? How narrow were the confidence limits? (What is the degree of uncertainty about the true effect?) Size of effect How large was the effect? Relevance of effect Does the outcome matter to patients? Adapted from a National Health and Medical Research Council (NHMRC) publication.18 Back to text 5: Checklist of quality items for different types of studies and questions Systematic review (all questions) Was a comprehensive and explicit search strategy used? Were the included studies assessed for quality? Were the characteristics and results of the studies summarised appropriately? Were sources of heterogeneity explained? Evidence-based guideline Was a comprehensive and explicit search strategy used? Have all relevant interventions and outcomes been considered, covering both benefits and harms? Is the level and quality of evidence for the recommendations given? Do the recommendations explore the trade-off of benefit and harm according to the level of risk in different patient subgroups? Randomised controlled trial for intervention questions Was allocation to treatment groups concealed from those responsible for recruiting the subjects? Were all randomised participants included in the analysis? Was there a blinded assessment of outcomes? Cross-sectional analytic study for diagnosis questions Was the test compared with a valid reference (gold) standard? Were the test and reference standard measured independently? Was the choice of patients assessed by the reference standard independent of the test results? Cohort study for prognosis questions Was there a representative sample of patients at a well defined point in the course of the disease? Was follow-up sufficiently long and complete? Were all potentially important prognostic factors assessed? Adapted from a National Health and Medical Research Council (NHMRC) publication,17 and Liddle et al.18 Back to text 6: Comparisons of effects of using cyclosporin in two patient subgroups with different risks of dialysis over two years (assuming the same relative treatment effect, 60% reduction in risk) Risk of dialysis Patient No. of patients who have dialysis subgroups If untreated If treated averted for every 100 treated Low risk Normal kidney function Mild proteinuria 10% 4% 6 Normal blood pressure High risk Very abnormal kidney function Marked proteinuria 100% 40% 60 Hypertensive Back to text 7: Research data relevant to the questions posed in the beginning of the article Question Information Source Quality of the study Study result Does colorectal cancer screening reduce mortality from colorectal cancer compared with routine care? Cochrane Library Search term: "colorectal neoplasms" 7 hits, #6 relevant1 Level: systematic review of RCTs (highest level) Quality: high Statistical precision: narrow confidence limits Size of effect: 23% reduction in mortality Relevance: high For every 10000 screened biennially over 10 years eight deaths prevented 2800 extra colonoscopies Does cyclosporin A prevent dialysis in focal and segmental glomerulonephritis? MEDLINE Search terms: "cyclosporine" and "glomerulonephritis" and "randomised controlled trial" as a publication type (.pt) 3 hits, #1 relevant2 Level: single RCT Quality: inadequate allocation concealment Statistical precision: wide confidence limits Size of effect: 33% reduction in ESRD Relevance: high For every 100 treated for 4 years 25 fewer develop ESRD uncertainty remains (small, potentially biased result) Does antiseptic impregnation of central venous lines reduce line-associated sepsiscompared with standard lines? MEDLINE Search terms: "local anti-infective agents" and "central venous catheterisation" and "meta-analysis" as a publication type (.pt) 2 hits, #1 relevant3 Level: systematic review of RCTs (highest level) Quality: high Statistical precision: narrow confidence limits Size of effect: 44% reduction in line sepsis Relevance: high For every 100 lines two fewer episodes of line sepsis RCT=randomised controlled trial.

Jonathan C Craig · Les M Irwig · Martin R Stockler