Launette M Rieb1, Zainab Samaan2, Andrea D Furlan3, Kiran Rabheru4, Sid Feldman5, Lillian Hung6, George Budd7, Douglas Coleman8. 1. Department of Family Practice, University of British Columbia, Vancouver, BC. 2. Department of Psychiatry, McMaster University, Hamilton, ON. 3. Department of Medicine, University of Toronto, Toronto, ON. 4. Department of Psychiatry, University of Ottawa, Ottawa, ON. 5. Department of Family and Community Medicine, University of Toronto, Toronto, ON. 6. Canadian Gerontological Nurses Association, Toronto, ON. 7. Clinical Pharmacist, Vancouver, BC. 8. Family Physician (in private practice), Comox, BC.
Abstract
BACKGROUND: In Canada, rates of hospital admission from opioid overdose are higher for older adults (≥ 65) than younger adults, and opioid use disorder (OUD) is a growing concern. In response, Health Canada commissioned the Canadian Coalition of Seniors' Mental Health to create guidelines for the prevention, screening, assessment, and treatment of OUD in older adults. METHODS: A systematic review of English language literature from 2008-2018 regarding OUD in adults was conducted. Previously published guidelines were evaluated using AGREE II, and key guidelines updated using ADAPTE method, by drawing on current literature. Recommendations were created and assessed using the GRADE method. RESULTS: Thirty-two recommendations were created. Prevention recommendations: it is key to prioritize non-pharmacological and non-opioid strategies to treat acute and chronic noncancer pain. Assessment recommendations: a comprehensive assessment is important to help discern contributions of other medical conditions. Treatment recommendations: buprenorphine is first line for both withdrawal management and maintenance therapy, while methadone, slow-release oral morphine, or naltrexone can be used as alternatives under certain circumstances; non-pharmacological treatments should be offered as an integrated part of care. CONCLUSION: These guidelines provide practical and timely clinical recommendations on the prevention, assessment, and treatment of OUD in older adults within the Canadian context.
BACKGROUND: In Canada, rates of hospital admission from opioid overdose are higher for older adults (≥ 65) than younger adults, and opioid use disorder (OUD) is a growing concern. In response, Health Canada commissioned the Canadian Coalition of Seniors' Mental Health to create guidelines for the prevention, screening, assessment, and treatment of OUD in older adults. METHODS: A systematic review of English language literature from 2008-2018 regarding OUD in adults was conducted. Previously published guidelines were evaluated using AGREE II, and key guidelines updated using ADAPTE method, by drawing on current literature. Recommendations were created and assessed using the GRADE method. RESULTS: Thirty-two recommendations were created. Prevention recommendations: it is key to prioritize non-pharmacological and non-opioid strategies to treat acute and chronic noncancer pain. Assessment recommendations: a comprehensive assessment is important to help discern contributions of other medical conditions. Treatment recommendations: buprenorphine is first line for both withdrawal management and maintenance therapy, while methadone, slow-release oral morphine, or naltrexone can be used as alternatives under certain circumstances; non-pharmacological treatments should be offered as an integrated part of care. CONCLUSION: These guidelines provide practical and timely clinical recommendations on the prevention, assessment, and treatment of OUD in older adults within the Canadian context.
According to the World Health Organization (WHO), people over the age of 50 accounted for 39% of deaths from drug use worldwide by 2015, and of those deaths in older adults (age ≥ 65), approximately 75% were linked to the use of opioids. ( Despite these numbers, there is a paucity of data about the many ways opioids specifically affect older adults and about how to care for those who develop an addiction, also referred to as an opioid use disorder (OUD).(There are two primary cohorts of older adults who develop OUD. The first group is made up of those who have been exposed to opioids for many years through drug experimentation, often beginning in adolescence. Some have been identified with and treated for an OUD, and many have had adverse health consequences.( The second group of older adults who may develop OUD is made up of those individuals who were prescribed opioids by a health-care provider for a pain condition.( Some of these individuals may have turned to the illicit market in order to maintain an ongoing supply of opioids following discontinuation of their prescription by a health-care professional. Opioid withdrawal pain can perpetuate the unintended long-term use of opioids for chronic non-cancer pain or due to an OUD, with underlying drivers being opioid-induced hyperalgesia and withdrawal-associated injury-site pain.(In Canada, 43.9% of adults > 55 years of age have used a prescription opioid and 1.1% of that group have done so daily (or almost daily) in the last year.( Though the proportion of people starting opioid therapy in Canada has trended down from 2013 to 2018, those over 65 have consistently received more new opioid prescriptions and have a higher proportion that go on to long-term opioid therapy (24.8%) than any other age group.( From 2007 to 2015, hospitalizations for opioid overdose (referred to as poisonings) in Canada were consistently higher in older adults than in any other age cohort: At over 20 per 100,000, older-adult admissions are almost double that of 15 to 24-year-olds, and represent 30% of all admissions to hospital for opioid poisoning.( Only recently by 2017 have younger adults in Canada started to equal and just surpass older adults in hospitalizations related to opioid poisonings. Most opioid poisonings in adults in Canada are accidental; however, 30% are intentional.( A recent study found that in older adults, opioid misuse was associated with increased odds of suicidal ideation.(According to a US national survey in 2016, 0.8% of adults surveyed met Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV) criteria for OUD in the past year.( There are no comparable Canadian studies that use DSM criteria. However one used WHO criteria and estimated a past year prevalence of OUD in Canadian adults to be 0.9% for drug abuse, and 0.5% for drug dependence. Yet reporting of prevalence data is lacking specific to those over 65 using opioids.(OUD may present more subtly in older adults than their younger counterparts and require a more nuanced approach. ( OUD may overlap with physiologic tolerance alone and may be mistaken for or masked by other medical conditions. Consequences from opioid use may not be recognized in older adults who have stopped work and who have restricted social networks. Social stigma and cognitive impairment may each play a role in under-identification.Adults entering older age are having an effect on the changing demographics of substance use and the need for treatment, its utilization, and its cost. For example, one study from New York City found that, as of 2012, adults > 60 years of age comprised 13.1% of those in opioid treatment programs, up from 1.7% in 2006.( This study also noted a shift from illicit drugs to those obtained by prescription as the primary type of opioid used. Additionally, when compared to their younger counterparts, older US veterans with OUD have higher rates of comorbid mood disorder, post-traumatic stress disorder, hepatitis C, human immunodeficiency virus, and chronic pain, including neuropathy, which has notably increased the cost of care.(The purpose of this article is to outline the issues facing older adults with, or at risk for, an OUD, and to provide a summary of recommendations for the prevention, screening, assessment, and treatment of an OUD in those ≥ 65 years of age. The full guidelines can be accessed electronically (www.ccsmh.ca).
METHODS
An experienced librarian and research assistants conducted a systematic literature search for relevant studies related to opioid use and opioid use disorder in adults and older adults. The databases searched included the Cochrane Library, EMBASE, PsychInfo, International Guideline Library, and PubMed. Included studies for the recommendations were restricted to human and written in English, with publications dates from 2008 through March 2018. The date limitation of the literature was decided a priori based on the changes observed in the last decade related to the opioid crisis in Canada. More recent literature was included in the introduction and discussion.An interdisciplinary guideline development committee was set up through the Canadian Coalition of Seniors’ Mental Health, consisting of two addiction medicine specialists, two psychiatrists, a physical medicine and rehabilitation specialist, a nurse practitioner, a pharmacist, a family physician with expertise in care of the elderly, as well as a person with lived experience. The committee members identified the top quality guidelines on OUD published using the Appraisal of Guidelines for Research & Evaluation Instrument (AGREE II).( We then modified the existing key guidelines using the ADAPTE collaboration process( in order to customize the selected guidelines for older adults and the Canadian context. We formulated questions in the PICOT format (Population, Intervention, Comparator, Outcome and Time) as a working group, and populated the questions with answers obtained from the previously chosen key guidelines, then supplemented this information with evidence from current literature.Working group members drafted recommendations and provided the evidence for each of the recommendations. The person with lived experience provided views and preferences of the target population. The draft recommendations were then reviewed by the whole working group, feedback integrated, so that the final wording of each recommendation was the result of full consensus. Each recommendation was then evaluated using the GRADE system (Grades of Recommendation Assessment, Development, and Evaluation).( GRADE scores the quality of the evidence as high, moderate, low or very low. It was agreed that “low” and “very low” scores would be amalgamated as “low”. The strength of the recommendations was scored as strong or weak.( We continued to discuss each recommendation and evaluation until we reached 100% consensus on each recommendation. The final recommendations were then sent for external peer review. The working group members integrated the majority of the comments from external reviewers into the guidelines after discussion.
RESULTS
Selection of the Previous Guidelines for Adaptation
Only one previous guideline published by the Royal College of Psychiatrists (United Kingdom) in the last 10 years for the prevention, assessment, and treatment of OUD specific to older adults was identified in the English language literature searched.( We chose to include this guideline for its direct relevance to older adults, despite a low-quality rating on the AGREE II tool.We closely reviewed the top four rated OUD guidelines focused on adults in general (not just older adults) published in the last 10 years.( We chose the US Veterans’ Affairs and Department of Defense’s guideline (2015) as the best rated guideline to use as a starting point, per the AGREE II tool. We also relied on elements of the Canadian Clinical Practice Guideline for Opioid Use Disorder for its applicability to our local context.(Recommendations from less robust guidelines and clinical practice tips on OUD were considered when the selected guidelines did not address an issue fully or supporting evidence was needed.( We considered costs, benefits, and harms when we were drafting the recommendations.No published randomized, controlled trials were identified on the prevention, assessment, or treatment of OUD in older adults. Therefore, the main evidence used to generate the recommendations was based on the above guidelines, along with systematic and narrative reviews on older adults since 2015.( In addition, numerous clinical and observational studies extrapolated from younger adults, some observational cohorts in older adults, as well as clinical expertise, informed these guidelines. For the section on prevention, key Canadian and US guidelines were identified for the treatment of chronic non-cancer pain (CNCP).(
RECOMMENDATIONS
Prevention of Opioid Use Disorder Among Older Adults
To curb opioid poisonings and the development of OUD, measures need to be implemented for both primary and secondary prevention. These issues are outlined in the questions (A & B) and addressed in the recommendations (1–9) listed below.
Question A: In older adults, what measures can reduce the risk of developing an OUD?
Recommendation 1
In order to avoid the risk of developing an OUD, older adults with acute pain in whom opioids are being considered should receive the lowest effective dose of the least potent immediate release opioid for a duration of ≤ 3 days and rarely > 7 days. ( GRADE Quality: Moderate; Strength: Strong
Recommendation 2
In most circumstances, avoid prescribing opioids for older adults with CNCP. For severe pain that is not responsive to non-opioid therapy in patients without a history of substance use disorder and without active mental illness, a trial of opioid treatment may be considered. Consider obtaining a second opinion before prescribing long-term opioid therapy. After explaining the risks and benefits to the patient, prescribe only in accordance with published guidelines for adults, initiate and maintain opioids at lower doses than for younger adults, and discontinue if function does not improve or if adverse effects arise.( GRADE Quality: Moderate; Strength: Strong
Recommendation 3
Patients and their families should be advised to store opioids safely, never to share their medication, and to return unused medication to the pharmacist for disposal.( GRADE Quality: Low; Strength: Strong
Recommendation 4
Pharmacists and nursing staff are advised to inform the prescriber if there are concerns with co-prescribing, adherence to treatment, or intoxication.( GRADE Quality: Low; Strength: Strong
Question B: In older adults with or at risk for an OUD, what preventive measures can reduce the risk of opioid overdose?
Recommendation 5
In older adults with polypharmacy or comorbidities that increase the risk of opioid overdose (e.g., benzodiazepine use, renal failure, sleep apnea), the lowest effective opioid dose should be used and tapering the opioid and/or other medications should be considered.( GRADE Quality: Moderate; Strength: Strong
Recommendation 6
Once the decision is made to reduce the opioid dose, a slow outpatient tapering schedule (e.g., 5% drop every 2–8 weeks with rest periods) is preferable to more rapid tapering. A faster taper schedule may be attempted under special circumstances of medical need, if the patient is in a treatment setting with medical supervision.( GRADE: Quality: Low; Strength: Weak
Recommendation 7
Dispense naloxone kits to anyone using opioids regularly for any reason (CNCP, OUD, etc.), and train household members and support staff on use.( GRADE Quality: Low; Strength: Weak
Recommendation 8
Include skilled pharmacists and/or nurses on teams to educate patients on appropriate use of opioids and other medications.( GRADE Quality: Low; Strength: Weak
Recommendation 9
Older adults with, or at risk for, an OUD should be given advice on strategies to reduce the risk of opioid overdose, and information on supervised consumption sites, if available in the community.( GRADE Quality: Moderate; Strength: Strong
Screening for and Assessment of Opioid Use Disorder Among Older Adults
Screening and assessment are the starting points for care of a person with an OUD, and many effective approaches to both exist for adults.( Listed below are questions (C–E) along with recommendations (10–12) for screening and assessment of OUD specific to older adults:
Question C: In older adults, when and how should one screen for an OUD?
Recommendation 10
Older adults should be screened for an OUD using validated tools, if appropriate (e.g., CAGE-AID, ASSIST, PDUQp, ORT, POMI, COMM). Medication reviews and urine drug screens should be utilized if the patient is taking opioids for CNCP or an OUD.( GRADE Quality: Low; Strength: Strong
Question D: In older adults at risk for an OUD, what are the elements of assessment?
Recommendation 11
Identify a diagnosis of an OUD through completion of a comprehensive assessment, including substance use, medical, pain, psychiatric, cognitive, and psychosocial history within a cultural context, and conduct a brief functional assessment. The use of validated assessment tools may be useful in this process. In addition, a detailed physical examination must be conducted, with an emphasis on signs of intoxication or withdrawal and the sequelae of substance use. Laboratory and other investigations (including urine drug tests) should be performed as appropriate for the medical conditions identified. Reassessment is essential and should be conducted episodically throughout long-term care.( GRADE Quality: Moderate; Strength: Strong
Question E: In older adults with or at risk for an OUD, what considerations are important to patients during the assessment process?
Recommendation 12
A full explanation of findings and diagnosis must be shared with the patient and, if appropriate, caregivers. Therapeutic optimism should be provided (i.e., hope given that addiction is a treatable disorder and that older adults, and especially older women, typically have better treatment outcomes than younger adults).( GRADE: Consensus
Treatment of Opioid Use Disorder Among Older Adults
Pharmacological Treatment
Issues related to pharmacological treatments are asked in questions (F–H) and answered in recommendations (13–26) listed below.
Question F: In older adults with an OUD, what approaches and medications are safe and effective for opioid withdrawal management?
Recommendation 13
Opioid withdrawal management should only be offered in the context of connection to long-term addiction treatment. ( GRADE Quality: Moderate; Strength: Strong
Recommendation 14
Induction onto an opioid agonist is recommended over a non-opioid treatment withdrawal management in older adults with an OUD. If a trial of tapering is attempted, there should be the option to initiate longer-term opioid agonist therapy or opioid antagonist therapy.( GRADE Quality: Moderate; Strength: Weak
Recommendation 15
Buprenorphine-naloxone should be considered first line for opioid withdrawal management in older adults. Methadone is an alternative that may be used, however consider the added risk of adverse events.( GRADE Quality: Moderate; Strength: Weak
Recommendation 16
For symptom control during opioid withdrawal management, adjuvant medications can be used in a time-limited fashion, but with caution due to medical comorbidities, side effect risk, and other concerns related to older age.( GRADE Quality: Moderate; Strength: Weak
Question G: What medications and protocol adjustments are safe and effective in the treatment of an OUD in older adults to improve outcomes?
Recommendation 17
Buprenorphine maintenance should be considered a first-line treatment for an OUD in older adults.( GRADE Quality: Moderate; Strength: Strong
Recommendation 18
Methadone maintenance treatment may be considered for those older adults who cannot tolerate buprenorphine maintenance or in whom it has been ineffective.( GRADE Quality: Moderate; Strength: Strong
Recommendation 19
If renal function is adequate, daily witnessed ingestion of slow-release oral morphine may be considered with caution for those older adults in whom buprenorphine and methadone maintenance have been ineffective or could not be tolerated. Careful supervision of initiation onto short-acting morphine first is recommended, prior to transition to maintenance with the long-acting 24-hour formulation.( GRADE Quality: Low; Strength: Weak
Recommendation 20
For older adults with an OUD for whom opioid agonist treatment is contraindicated, unacceptable, unavailable, or discontinued and who have established abstinence for a sufficient period of time, naltrexone may be offered.( GRADE Quality: Moderate; Strength: Weak
Recommendation 21
Offer medications for an OUD in the context of connection to long-term addiction, mental health, and primary care treatment, where careful monitoring and dose titration can occur. ( GRADE Quality: Moderate; Strength: Strong
Recommendation 22
Advise patients that the use of alcohol, benzodiazepines, and other sedative-hypnotics is hazardous when combined with opioid agonist treatment. If the older adult is living in the community and is already physiologically dependent on one of these substances, then slow tapering of the substance(s) (to elimination, if possible), rather than abrupt cessation, is recommended. If the patient is in hospital, residential treatment, or a long-term care setting and medically managed by an experienced provider, detoxification can progress more rapidly, concurrent with the initiation or stabilization on medications for OUD. ( GRADE Quality: Moderate; Strength: Strong
Recommendation 23
Early take-home dosing for buprenorphine maintenance treatment may be considered, including home induction in patients who are low risk, if they find it difficult to attend the office in withdrawal and if the patient has social supports at home. This approach should not be considered for methadone initiation unless supervised (e.g., reliable caregiver or medical personal administration).( GRADE Quality: Low; Strength: Weak
Question H: In older adults, compared with younger adults, what dose/protocol adjustments are needed when initiating or maintaining medications to treat an OUD?
Recommendation 24
Reduce initial doses of medications for treatment of an OUD (e.g., by 25–50%); slow dose escalation frequency (e.g., by 25–50%); use the lowest effective dose to suppress craving, withdrawal symptoms and drug use; and monitor closely (especially for sleep apnea, sedation, cognitive impairment, and falls with opioid agonists).( GRADE Quality: Low; Strength: Strong
Recommendation 25
The threshold to admit an older adult with social, psychological, or physical comorbidities to either residential or hospital care for opioid withdrawal management or induction onto medications for an OUD should be lower than for a younger adult.( GRADE: Quality: Moderate; Strength: Strong
Recommendation 26
In older adults on medication for an OUD requiring management of mild-to-moderate acute pain or CNCP, non-medication and non-opioid strategies are recommended. For those on an opioid agonist for an OUD who have severe acute pain that has been unresponsive to non-opioid strategies, a short-acting opioid may be considered for a short duration (1–7 days) along with a taper, if necessary (1–7 days).( GRADE Quality: Moderate; Strength: Weak
Other Comments on Medications for OUD
In younger adults, for the palliative end of the OUD spectrum in whom all other modalities have been unsuccessful, there is currently available in Canada injectable opioid agonist treatment (iOAT). No data on iOAT in older adults were identified. The risk of adverse events with injectable medications rises with age, as does immunosuppression and mobility issues for accessing the three-times-daily visit to a clinic needed for iOAT. BC is currently the only province providing and paying for iOAT, and the cost is higher than with all other opioid agonist and antagonist treatments. Due to all of the above issues, we cannot endorse iOAT treatment in older adults at this time and thus have not included it in these guidelines.
Psychosocial Treatment
Although robust data on behavioural interventions in older adults with OUD are lacking, an understanding of the importance of incorporating psychosocial supports is clear: Older adults may have an accumulation of losses (job, spouse, family, friends, role, home), and may struggle with lack of social support.( The key question (Question I) regarding psychosocial interventions in older adults with an OUD is outlined below, along with related recommendations (27–31).
Question I: In older adults, what psychosocial interventions are effective in the treatment of an OUD to improve outcomes?
Recommendation 27
Psychosocial interventions should be offered concurrently with medications for an OUD, at a pace appropriate for age and patient needs, but they should not be viewed as a mandatory requirement for accessing pharmacotherapy.( GRADE Quality: Moderate; Strength: Strong
Recommendation 28
Contingency management may be offered as part of opioid treatment programs and used if accepted by the patient.( GRADE Quality: Moderate; Strength: Weak
Question J: For older adults with an OUD, what are the treatment considerations for special populations (specifically Indigenous peoples)?
Recommendation 29
Traditional healing practices used by Indigenous communities can be integrated with buprenorphine treatment to improve outcomes for an OUD.( GRADE Quality: Low; Strength: Weak
Setting
Setting can refer to treatment setting and also the living situation of the older adult.
Question K: For older adults with an OUD, what specific treatment recommendations are suggested for non-community–based facilities?
Recommendation 30
If experienced, clinicians may manage older adults with a mild-to-moderate OUD; however, for patients with more severe or complex disorders, it is recommended that personnel or teams with advanced substance use disorder management skills be accessible to support clinicians and to enhance their capacity to care for patients in all settings. The threshold for an admission to hospital or drug and alcohol treatment facility under the care of an Addiction Medicine Specialist is lower than for younger adults, and closer follow-up is needed on discharge to ensure appropriate community-based support. ( GRADE Quality: Moderate; Strength: Strong
Recommendation 31
Older adults with an OUD who are admitted to a hospital, drug and alcohol treatment facility, or non-medical facility with access to medical care (e.g., prisons and shelters) should be offered opioid agonist treatment at the onset of withdrawal (advisable within 1–3 days), with bridging pharmacological treatment on discharge with confirmed transfer of care. ( GRADE Quality: Moderate/Strength: Strong
Question L: For older adults with an OUD, what systemic factors support recovery?
Recommendation 32
The cost of medically-recommended pharmacological and non-pharmacological treatment for an OUD in older adults should be covered by the public health plan.( GRADE Quality: Moderate; Strength: Strong
DISCUSSION
No previous guidelines or clinical trials for the prevention, screening or management of OUD have been undertaken in adults over 65 despite the increasing prevalence of OUD in this age group. Our guidelines represent a systematic effort to identify evidence from literature, taking into consideration the context, the expertise of individuals working with older adults, the input from individuals with lived experience, and the multidisciplinary guideline team. These guidelines provide 32 recommendations covering aspects of opioid management from prevention, screening, assessment, pharmacological treatment, psychosocial interventions, treatment setting, and steps to recovery. Many of the recommendations are in alignment with the other two recent Canadian guidelines on OUD treatment in adults.(The challenges and limitations in writing these guidelines were many, most important of which was the lack of direct evidence to inform the guidelines of the efficacy or effectiveness of current treatment approaches in this population. Other limitations include lack of a consistent definition for older adults and successful treatment outcome measures for OUD. Finally, there are limited data and empirical evidence for the management of OUD in special populations of older adults related to sex, gender, and ethnic group.Studies with a focus on older adults are needed to provide empirical evidence for the effectiveness of pharmacological and psychosocial treatments. The deliberate inclusion of subgroups in these studies will provide data for future guidelines to improve their external validity. Future studies to define meaningful and patient-centered treatment outcomes targets are needed. Also, the role of other interventions that are less studied, including self-help groups, is yet to be delineated. Canadian-based data are needed on the rates of opioid use and OUD in older adults, as well as rates of overdose and mortality, in order to design targeted prevention strategies and test the effectiveness of OUD treatment and opioid overdose mitigation.
Facilitators and Barriers for the Application of These Recommendations
In Canada, the need for expanded resources for treatment of older adults with OUD will likely rise in the coming years. Evidence-based treatment for OUD in adults in general is cost-effective and decreases both morbidity and mortality.( The lifetime savings to provincial and national health care based on this model could be substantial. All components of the Canadian health-care system must prepare to provide addiction treatment to older adults, including those affected by an OUD.The uptake of these guidelines will be monitored through tracking CCSMH website visits, orders of print version, distribution of office handouts, and journal article reads. There are plans to update the guidelines approximately every five years, as needed, if new literature becomes available that may significantly change management.
CONCLUSION
Older adults are susceptible to adverse health consequences of opioid use. There is a growing need for opioid management guidelines for older adults as the population in Canada ages. The current recommendations are intended to provide healthcare workers and policy-makers with evidence-informed, clinically relevant direction and advice on the prevention, screening, assessment, and treatment of OUD in older adults. We hope practitioners will find them both a practical and useful clinical aide, and that community members will find them a helpful education resource. The expanded version of these guidelines can be accessed electronically (www.ccsmh.ca).
Authors: Bohdan Nosyk; Huiying Sun; Elizabeth Evans; David C Marsh; M Douglas Anglin; Yih-Ing Hser; Aslam H Anis Journal: Addiction Date: 2012-05-08 Impact factor: 6.526
Authors: J L Brożek; E A Akl; E Compalati; J Kreis; L Terracciano; A Fiocchi; E Ueffing; J Andrews; P Alonso-Coello; J J Meerpohl; D M Lang; R Jaeschke; J W Williams; B Phillips; A Lethaby; P Bossuyt; P Glasziou; M Helfand; J Watine; M Afilalo; V Welch; A Montedori; I Abraha; A R Horvath; J Bousquet; G H Guyatt; H J Schünemann Journal: Allergy Date: 2011-01-17 Impact factor: 13.146