Systematic review of clinical prediction models to support the diagnosis of asthma in primary care
npj Primary Care Med 2019;29
Asthma is commonly misdiagnosed, with overdiagnosis leading to potentially harmful treatment and unnecessary healthcare cost, and underdiagnosis risking avoidable morbidity and mortality.
Asthma is a clinical diagnosis, with no definitive reference standard that can confirm or refute the diagnosis. Conflicting recommendations in national and international guidelines are evidence of the uncertainty about the best combinations of clinical features and tests for asthma diagnosis.
Clinical prediction models help healthcare professionals to assess the probability of a diagnosis and enhance shared decision making. Daines and colleagues set out to identify, compare and synthesise existing clinical prediction modules that could support the diagnosis of asthma in children and adults in the primary care setting. They searched Medline, Embase, CINAHL, TRIP and US National Guidelines Clearinghouse databases from 1 January 1990 to 23 November 2017. They screened titles, abstracts and full texts for eligibility, extracted data and assessed risk of bias. From 13,798 records, they reviewed 53 full-text articles. Seven clinical prediction models to support the diagnosis of asthma in primary care were identified.
This review highlighted the paucity of current criteria to inform diagnostic algorithms. All seven of the selected studies were at high risk of bias and could not be recommended for diagnosing asthma in routine clinical practice. Wheeze, allergy, allergic rhinitis, symptom variability and exercise-induced symptoms were associated with asthma and could be considered as predictors in future prediction models. Cough, respiratory tract infections and nocturnal respiratory symptoms were consistently associated with asthma.
In the future, establishing a data-driven approach to asthma diagnosis could resolve current discrepancies in guidelines and enable the unacceptable level of asthma misdiagnosis to be reduced.