Intended for healthcare professionals

Practice Change

Do not diagnose or routinely treat asthma or chronic obstructive pulmonary disease without pulmonary function testing

BMJ 2023; 380 doi: (Published 20 March 2023) Cite this as: BMJ 2023;380:e072834
  1. Samir Gupta, associate professor1 2,
  2. Guylène Thériault, physicianship director3
  1. 1Li Ka Shing Knowledge Institute of St. Michael’s Hospital, Unity Health, Toronto, Canada
  2. 2Division of Respirology, Department of Medicine, University of Toronto, Toronto, Canada
  3. 3Campus Outaouais McGill University, Quebec, Canada
  1. Correspondence to S Gupta samir.gupta{at}

What you need to know

  • More than a third of diagnoses of asthma or COPD made on clinical grounds alone are incorrect

  • A clinical diagnosis of a chronic airways disease without objective testing may harm patients by exposing them to unnecessary medications, while missing their actual diagnosis

  • Given the barriers to objective testing of lung function, including timely test access, it may be reasonable to treat symptoms before a formal diagnosis is made, while arranging for formal tests

Asthma and chronic obstructive pulmonary disease (COPD) are among the most common chronic diseases seen in primary care, affecting 262 million and 212 million individuals worldwide, respectively.1 Respiratory complaints are also the most common reason for a visit to primary care,2 and providers often ascribe lower respiratory symptoms (particularly when recurring or chronic) to asthma or COPD, and initiate treatment based on a clinical diagnosis. However, prospective studies have shown that symptoms elicited during clinical history when asthma or COPD are suspected do not accurately predict disease in real world settings.3 Similarly, features of physical examinations, such as wheeze, are non-specific, and neither individual nor combined physical examination findings have been found to reliably predict obstructive airways disease.4

Evidence from existing studies suggests that more than half of patients who receive a diagnosis of asthma or COPD do not have objective lung function tests performed. In Canadian population based studies using administrative databases and applying validated case definitions, 43% and 36% of patients with a new diagnostic label of asthma5 and COPD,6 respectively, had received pulmonary function testing within one year of their diagnosis. A US administrative database study applied the same methods and found a comparable number (48%) for asthma.7 A smaller Italian study found that 55% and 56% of patients seen in a spirometry laboratory who self-reported a prior physician-diagnosis of asthma and COPD, respectively, had previous spirometry.8 Finally, in the population based Burden of Obstructive Lung Disease (BOLD) sample across 20 countries, 63% of patients who self-reported physician-diagnosed COPD had previously performed spirometry.9

International guidelines, including Choosing Wisely campaigns, Canadian Thoracic Society, the National Institute of Health and Care Excellence (UK), National Heart, Lung, and Blood Institute (US), Global Initiative for Asthma, and Global Initiative for Chronic Obstructive Lung Disease recommend objective pulmonary function testing, starting with pre-/post-bronchodilator spirometry when asthma is suspected in individuals aged ≥6 years and when COPD is suspected in any individual (table 1).10111213141516 In this article, we explore barriers to this practice and possible solutions.

Table 1

International guidelines and recommendations related to avoiding treatment for asthma and/or COPD without objective diagnosis

View this table:

The evidence for change

No randomised controlled trials have been reported that compare outcomes in patients with clinically diagnosed versus objectively diagnosed airway disease; however, several studies show the harms of inaccurate clinical diagnosis and the opportunity cost of absent pulmonary function test results (box 1). A Canadian multicentre prospective cohort study of a representative community based sample from 2012 to 2016 (with most patients diagnosed in primary care) found that 33% of patients who had been diagnosed with asthma by a physician in the previous five years did not have evidence of asthma on pulmonary function testing at the time of the study.17 Of those patients (who were shown not to have asthma on pulmonary function testing) 79% were using asthma medications, exposing them to unnecessary medication costs and possible side effects.17 Around 12% of these patients with current negative testing for asthma did have evidence of a prior positive pulmonary function test for asthma (and were thus likely in clinical remission by the time of the study); however, 67% of patients who did not have asthma on current testing went on to receive an alternative diagnosis from a specialist.

Box 1

Possible harms when making a clinical diagnosis of asthma or COPD

In individuals who do not have asthma or COPD

  • Unnecessary treatment17

    • Medication side effects

    • Medication costs (individual and societal)

    • Environmental harm from unnecessary use of metered-dose inhalers

  • Missing or delaying recognition and treatment of the actual diagnosis17

    • The symptoms might have been readily addressed by an alternative treatment (eg, antacid treatment for gastroesophageal reflux disease)

    • The symptoms may have represented a more serious disease (eg, ischaemic heart disease)

  • Wrongfully labelling an individual with a chronic disease

    • May affect insurability or insurance costs

    • May affect quality of life (impact of self-perception of having a chronic disease)18

In individuals who have COPD

  • Underestimating the severity of the disease

    • Undertreatment19

    • Increased admissions to hospital and deaths20


Initial misdiagnosis resulted in delays in diagnosing clinically relevant conditions, including uncommon causes such as ischaemic heart disease and bronchiectasis, and more common asthma mimickers such as rhinitis with post-nasal drip, gastro-oesophageal reflux disease, and anxiety.17 Smaller studies have shown similar results. A similar Canadian prospective cohort study assessed 496 community based patients who had ever received a physician diagnosis of asthma (as opposed to within the past five years), and also found that 30% of such patients did not have evidence of asthma on pulmonary function testing at the time of the study.21 A cross-sectional UK study of 262 patients labelled as having asthma in primary care chart notes found normal current spirometry and methacholine challenge test results in 31% of patients.22 Swedish investigators similarly performed spirometry testing and a methacholine challenge test in 86 consecutive patients who had an asthma diagnosis in primary care charts, and reported that 34% had no evidence of current asthma.23 Other effects of this type of misdiagnosis (sometimes called “overdiagnosis”) include increased health insurance costs, the mental health impacts related to a chronic disease label,18 and environmental harm from unnecessary use of metered-dose inhalers, which use potent greenhouse gases (hydrofluorocarbons) as propellants (box 2).27

Box 2

Environmental harms of misdiagnosis with unnecessary treatment

Pressurised metered-dose inhalers (pMDIs) are used commonly in patients with suspected asthma or COPD.

To generate aerosol clouds, pMDIs use various hydrofluorocarbons, which are potent greenhouse gases (greenhouse gases are also released in pMDI manufacture and disposal). Other environmental effects of pMDIs include water eutrophication and ecotoxicity.24

pMDIs are estimated to comprise 0.03% of yearly global greenhouse gas emissions and 3.1% of emissions related to healthcare.25

The carbon footprint of a canister of salbutamol—one of the most commonly used pMDIs in both asthma and COPD—is approximately equivalent to a 290 km automobile journey.26

Use of these inhalers in patients who do not have either asthma or COPD thus constitutes avoidable environment harm.


In COPD, prospective studies evaluating pre-/post-bronchodilator spirometry in patients with a previous diagnosis of COPD identified in primary care practices in both Australia (31 practices, 341 patients) and Canada (three practices, 1003 patients) showed that 31% to 44% of them did not have the disease.2829 In the BOLD population based sample across 20 countries, 62% of patients were found to have received a misdiagnosis, among which 46% reported use of a respiratory medication.9 In addition to unnecessary treatment in such patients, clinical diagnosis without spirometry has negative effects on patients who do have COPD. Without spirometry, providers tend to underestimate COPD severity, resulting in undertreatment.19 Administrative data from Canada showed that patients given a clinical diagnosis were 10% more likely to be admitted to hospital for COPD or die of any cause than those whose diagnosis included pulmonary function testing.20 In a prospective interventional study across US primary care sites, spirometry results changed management in almost half of patients with asthma or COPD, with 86% of changes made to medications being in line with evidence based guidelines.30

Barriers to change

Barriers to use of pulmonary function testing for asthma and COPD diagnosis are highly dependent on context, given that access to spirometry varies both across and within countries.31 In the primary care setting, where most patients receive diagnosis and management,32 some providers offer in-office spirometry, whereas others refer patients to a pulmonary function laboratory. A recent systematic review of studies conducted in primary care settings applied the Theoretical Domains Framework methodology33 to identify several barriers and enablers to spirometry testing for diagnosis of asthma and COPD, spanning multiple domains that explain provider behaviour (table 2).34 Barriers included the provider lacking knowledge about the role of spirometry, buy-in regarding the need for spirometry (including its impact on treatment decisions and patient outcomes), and skills and reimbursement for accurate test performance and interpretation, as well as lack of test access, workflow impacts, and provider perceptions that a clinical diagnosis is accurate and that patients are reluctant to undergo testing.

Table 2

Barriers and enablers to use of spirometry for diagnosis of asthma or COPD in primary care settings34

View this table:

Most studies in this review were from high income countries, whereas the main barriers in low and middle income countries surround availability of training, personnel, equipment, consumables, and financial resources.31 In the context of the covid-19 pandemic, logistical and cost implications of new protocols to limit aerosol transmission of pathogens, along with provider concerns about the risk of the test, have only amplified access and availability concerns.36

How should we change our practice?

Solutions must target provider, patient, and system-level barriers. Providers can be supported through educational programmes such as that launched recently in Canada (see box, Resources for patients), including mentoring and professional development programmes. Electronic decision support, and audit and feedback strategies might also be helpful but require evaluation.34 Patients with a clinical diagnosis of either asthma or COPD could be directly targeted with educational messages regarding the importance of objective testing, addressing patient level barriers, and with a goal of prompting providers to order testing. System level approaches include improved (private and/or public) funding for spirometry equipment and staff training and remuneration for performing and interpreting tests, and increases in laboratory capacity. Novel models such as travelling spirometry laboratories and new technologies (eg, smartphone spirometers) may one day obviate existing challenges, but will have variable applicability across low and middle income settings.37 System changes will require engagement with decision makers and robust analysis of cost-benefit data, particularly in public-payer systems. Although asthma tests are not perfect and may sometimes yield false positives, and a small percentage of patients may exhibit remission in their asthma over time17 (see above), we believe that efforts to perform diagnostic pulmonary function testing should be limited largely to those who have not previously had a positive objective test. All of these efforts will also be constrained by the considerable diversion of human and health system resources during the covid-19 pandemic.

What can providers do today?

Objective diagnoses should be made for patients who already have a clinical diagnosis (ie, with no prior objective proof of disease) and for those with a new clinical suspicion of disease. We propose a pragmatic approach in settings where spirometry access is limited and/or wait times are long, such as in primary care. In these cases, particularly to prevent potential harm from withholding therapy, we suggest initiating therapy if asthma or COPD is suspected, while simultaneously requesting pulmonary function testing. All patients with negative testing should be re-evaluated if symptoms recur.

Asthma (fig 1)

Fig 1
Fig 1

Algorithm for diagnosing asthma with delayed spirometry access (children ≥ 6 years and adults). This algorithm1138 does not account for asthma/COPD overlap, restrictive lung diseases, and other conditions that may be suggested on spirometry testing.
*Four hours for short-acting beta agonist; 12 hours for short-acting muscarinic antagonist; 24 hours for any twice daily long-acting beta agonist (with or without inhaled corticosteroid); 36 hours for once daily long-acting beta agonist (with or without inhaled corticosteroid); 48 hours for long-acting muscarinic antagonist.39
**Other strategies for asthma diagnosis include: increase in peak flow with a bronchodilator or after a course of controller therapy (≥20% in children ≥ 6years, ≥20% and ≥60 L in adults); diurnal peak flow variability (in adults); increase in FEV1 after a course of controller therapy of ≥12%; and exercise challenge with ≥10-15% decrease in FEV1 post-exercise.11
***If the patient develops acute worsening of asthma symptoms with therapeutic wean/discontinuation, therapy should be restored and a diagnosis of asthma is likely.17
A consensus is emerging that measurement of fractional excretion of nitric oxide may be considered as a diagnostic test for asthma in cases of negative pre-/post-bronchodilator spirometry testing (thus avoiding the need for a methacholine challenge test in some cases)14164041 (NICE recommends that FeNO can be used as an initial diagnostic test in people aged 17 and older). BD=bronchodilator; FEV1 =forced expiratory volume in the first second; FVC=forced vital capacity; LLN=lower limit of normal; ICS=inhaled corticosteroid; LABA=long-acting beta agonist; LTRA=leukotriene receptor antagonist. Adapted from Can Fam Physician.42

Perform pre-/post-bronchodilator spirometry if available.

  • Obstruction with a significant bronchodilator response confirms the asthma diagnosis. However, given its high false negative rate, a negative test requires follow-up methacholine challenge testing for definitive diagnosis (see below), although access is also limited.43

Peak flow testing is an option in settings where spirometry testing availability is limited, as peak flow meters are inexpensive and testing is easy to administer with minimal training (however, results are effort dependent).44

  • Patients who have never received treatment could receive diagnosis by observing an improvement in peak expiratory flow after versus before a bronchodilator or course of controller therapy.11 This allows the initial speculative course of therapy to play a diagnostic role in addition to its potential therapeutic role.

If spirometry and/or peak flow testing is negative, a methacholine challenge test is required.

  • For patients who are taking medications, consider a gradual therapeutic wean with sequential methacholine testing. If the patient develops acute worsening of asthma symptoms with therapeutic wean, restore therapy and a diagnosis of asthma is likely.17

Another diagnostic tool is measurement of fractional excretion of nitric oxide (FeNo). NICE recommends FeNO testing be offered to adults where a diagnosis of asthma is being considered, whereas most guidelines suggest using this test as a next step after negative pre-/post-bronchodilator testing.14164041 However, fractional excretion of nitric oxide measurement is not widely available and is not generally recommended as a replacement for pre-/post-bronchodilator testing. Its role in diagnosing and monitoring asthma continues to evolve.

COPD (fig 2)

Fig 2
Fig 2

Algorithm for diagnosing COPD with delayed spirometry access. This algorithm10121645 does not account for asthma/COPD overlap, restrictive lung diseases, and other conditions that may be suggested on spirometry testing.
*Four hours for short-acting beta agonist; 12 hours for short-acting muscarinic antagonist; 24 hours for any twice daily long-acting beta agonist (with or without inhaled corticosteroid); 36 hours for once daily long-acting beta agonist (with or without inhaled corticosteroid); 48 hours for long-acting muscarinic antagonist.39 FEV1=forced expiratory volume in the first second; FVC=forced vital capacity; LLN=lower limit of normal. Reproduced from Can Fam Physician.42

After a brief therapeutic washout, presence or absence of obstruction on post-bronchodilator spirometry is sufficient to rule in or rule out the diagnosis, respectively (note that peak flow testing is not used for COPD diagnosis).

  • While post-bronchodilator spirometry is required for COPD diagnosis, we note that the degree of reversibility does not require interpretation, as it does not reliably augment diagnosis, predict response to treatment, nor differentiate from asthma (this requires additional clinical context).12

Resources for patients

Choosing Wisely Canada, College of Family Physicians of Canada, Canadian Thoracic Society “Let’s Clear The Air” campaign web based educational materials, downloadable (pdf) handouts, and a brief cartoon video for people with an existing diagnosis of asthma (website/handout and video) or COPD (website/handout). Materials describe misdiagnosis and its harms, and prompt patients to discuss objective testing with their provider if they have not previously had it (free online, no registration required):

American Academy of Asthma, Allergy and Immunology and consumer reports resource for people with symptoms of asthma, describing the reason objective tests are needed for asthma diagnosis and management, risks of not having the test, benefits of having the test, and when the test should be performed (free online, no registration required; also available in Spanish):

Education into practice

  • Do you routinely order a pre-/post-bronchodilator spirometry test for your patients with chronic respiratory symptoms before, or at the same time as, initiating inhalers?

  • What proportion of the patients in your practice with a diagnosis of asthma and/or COPD have previously had a pre-/post-bronchodilator spirometry test or other confirmatory test?

How patients were involved in the creation of this article

Sharon Lapkin is a person living with asthma who has experienced the entire journey from symptom onset to formal diagnosis. She reviewed and commented on each version of the manuscript, adding valuable editorial input on communication and patient-sensitive language, as well as a review of the appropriateness of suggested patient-facing material.

How this article was made

We searched Medline to identify published articles pertaining to misdiagnosis of asthma and COPD in primary care settings, and the impacts of inappropriate diagnosis on treatment. We also systematically searched Medline, Allied and Complementary Medicine Database, Cochrane Central Registrar of Controlled Trials, Cumulative Index to Nursing & Allied Health, Embase, PsychInfo, and Scopus for any qualitative or quantitative articles describing barriers and/or enablers to use of spirometry for diagnosis of asthma or COPD in primary care settings (both in-office and in-laboratory). We also consulted the latest international asthma and COPD guidelines and experts in asthma, COPD, and primary care.


  • “Change” articles aim to alert clinicians to the immediate need for a change in practice to make it consistent with current evidence. We welcome any suggestions for future articles (email us at

  • Competing interests: We have read and understood The BMJ policy on declaration of interests and declare the following interests: SG is an unpaid chair of the Canadian Thoracic Society’s Canadian Respiratory Guidelines Committee and former chair of the Canadian Thoracic Society’s Choosing Wisely Working Group; SG is the unpaid former co-chair of the Health Quality Ontario Asthma Quality Standards Group; SG developed and studied an electronic quality improvement tool for asthma management; SG has received peer-reviewed research grant funding from public and not-for-profit granting organisations (Lung Health Foundation, Canadian Institutes of Health Research, Public Health Agency of Canada, Temerty Faculty of Medicine, Ryerson University, Government of Ontario); GT is the unpaid primary care co-lead for Choosing Wisely Canada.

  • Provenance and peer review: commissioned; externally peer reviewed.

  • TheBMJ thanks Wendy Levinson and Karen Born at Choosing Wisely for valuable advice and supporting the selection of topics. Choosing Wisely had no input into the peer review process or editorial decision.


Log in

Log in through your institution


* For online subscription