We have demonstrated that the implementation of a clinical pathway involving the use of clinical guided consultation when performing primary care COPD reviews results in a significant reduction in the misdiagnosis of COPD, an increase in the detection of possible serious comorbidity complicating COPD as well as an increase in the uptake of both pharmacological and non-pharmacological management that is aligned to recognised guidelines.

An increasing number of complex clinical guidelines are disseminated to healthcare professionals based in primary care16. Our study has demonstrated that guideline-level practice is often not followed in COPD care and furthermore, the use of clinical decision support software technology such as the CGC leads to greater implementation of guideline level care which reduces variation in the standard of care thus addressing health inequality. The scalability of the clinical decision support system and on a wider level, the clinical pathway is supported by the fact that the identification of appropriate patients using the MIQUEST and SNOMED tools and the subsequent CGC reviews may be performed remotely offering read and write back to the GP clinical record and with the pathway being utilised in 254 GP practices nationally.

The consequences of an incorrect diagnosis of COPD may include failure to recognise the true aetiology of the patient’s symptoms such as cardiac disease as well as the potential adverse effects of pharmacological overtreatment and altered survival9,17,18,19,20,21,22. One reason cited as a cause of COPD misdiagnosis has been the lack of access to high-quality spirometry, and this issue has recently been worsened by the Covid-19 pandemic23,24. However, consistent with the literature, our analysis shows that even when such spirometry has been performed in a patient to guideline standard, the appropriate interpretation of this investigation and correlation with key clinical findings, principles of which are integral to correctly diagnosing any condition including COPD, are often absent25,26. The CGC intelligently interpreted spirometry tests through it’s programmed algorithms to established guidelines and alerted the operator that 13 and 10% of patients on these primary care COPD registers had normal and restrictive spirometry respectively. These patients were relieved of an incorrect diagnosis and further medical review was prompted to seek alternative causes for their symptoms such as previously undiagnosed cardiac disease. The incidence of overdiagnosis reported in our analysis demonstrates that despite the dissemination of clinical guidelines year after year, the messages focusing on the need for robust diagnosis in COPD contained in such guidelines remains hard to achieve27. A study of 1044 patients with a label of COPD referred for pulmonary rehabilitation between 2007 and 2010 revealed a misdiagnosis rate of 20% based on spirometry criteria and also similar to our findings, men were more likely than women to be accurately diagnosed28, In another analysis of over 14,000 COPD patients in primary care in 2013, consistent with our report, 13.1% had no recorded spirometry and that where spirometry was performed, 11.5% had no evidence of airflow obstruction21. We would thus conclude that in clinical practice, the key to achieving a robust diagnosis of COPD however lies beyond simply the interpretation of valid spirometry but in the integration of lung function with a structured clinical assessment, a process aided by the CGC in this analysis. The use of the CGC may therefore support the re-integration of Spirometry back within primary care. An additional utility of such software lies in the setting of newly created “Diagnostic centres” where it may enhance the accuracy and quality of the diagnostic process for those patients with respiratory symptoms who are suspected of having COPD. Those patients in whom the spirometry does not support this diagnosis could then be directly referred for further assessment including cardiac investigations. Further studies evaluating the validity of such a pathway are needed.

Clinical guidelines in COPD emphasise the importance of recognising and treating comorbidity3. The presence of coexisting heart failure is common in COPD yet often remains undiagnosed despite it being an independent predictor of all-cause mortality, whilst uncontrolled hypertension is associated with an increased rate of hospitalisation with the presence of coexisting cardiovascular disease in COPD also being a driver of increased healthcare costs29,30,31,32. The use of the CGC in COPD reviews highlighted new findings suggestive of significant cardiac disease in 7% of patients and prompted the healthcare professional to consider referral for further investigation. The intelligent software was also able to highlight cases of uncontrolled hypertension. Outside the cardiovascular system, co-existing bronchiectasis in COPD is associated with frequent exacerbations, worsening lung function and increased mortality3,33. The GOLD guidelines reference the production of large volumes of purulent sputum as suggestive features of bronchiectasis and the CGC review determined that of all patients with airflow obstruction reporting this symptom, only 10% had previously been diagnosed with bronchiectasis and prompting consideration of this possibility in the remaining 90%3. Further longitudinal studies are needed to determine the outcome of those patients referred for further investigations of suspected comorbidity following CGC review and particularly when compared to “usual care” and whether the application of machine learning to such software systems may further enhance the ability of healthcare professionals to diagnose comorbidity promptly in such scenarios.

The appropriate and timely referral to specialist services is another key component of guideline-based practice. The incorporation of pulse oximetry when performing COPD reviews in primary care allows primary care practitioners to refer patients for consideration of Long-Term Oxygen Therapy34. In addition to CGC identifying those suitable for domiciliary oxygen, in 46% of patients with hypoxia, the CGC highlighted that this could be disproportionate to the degree of lung function impairment further illustrating the role of such technology in aiding diagnostic certainty, improving patient safety whilst simultaneously upskilling healthcare professionals. COPD guidelines also recognise the importance of healthcare professionals delivering smoking cessation advice to their patients during each review yet this is frequently not performed3,35,36,37,38,39,40. The use of the CGC resulted in 99% of patients who smoked receiving smoking cessation counselling and brief intervention during the consultation with the recommendations to refer on for further support if the patient was willing. Furthermore, the database underpinning the CGC captures the “smoking status” of a patient in consultations longitudinally allowing stakeholders accurately to measure the effectiveness of key interventions and also accordingly tailor the nature of local smoking cessation services.

The introduction of structured self-management programmes and written action plans in COPD conveys clinical benefit including a reduction in hospital admissions, improved health related quality of life and a favourable health economic profile and is recommended as guideline standard care3,41,42,43,44,45. The CGC review resulted in a significant increase in the provision of written action plans to include 85% of the cohort. Similarly, guidelines emphasise the importance of correct inhaler technique and the need to check inhaler technique regularly in COPD. Critical errors in inhaler technique are associated with an increased frequency of exacerbations which impacts not only the patients but also health economics46. The CGC reported that 13% of patients had sub-optimal inhaler technique although the fact that this could only be addressed in just under half of cases during the consultation may reflect the fact that remote reviews constituted a significant proportion of the consultations. The Covid-19 pandemic resulted in an increase in the proportion of those patients undergoing remote reviews and more detailed studies are required to fully understand the effectiveness of remote reviews (through video and telephone) in chronic respiratory conditions such as COPD when compared to traditional face to face consultations and whether the use of clinical decision support systems may improve outcomes in specific areas of COPD management when undertaking remote consultations.

Pulmonary rehabilitation (PR) also represents an important guideline standard intervention in COPD which is often underutilised3,47. In an Australian study, referral for PR was described as the least well implemented guideline intervention in COPD and an important barrier for PR referral was low awareness of the program by healthcare professionals, with key recommendations being the identification of suitable patients and streamlining the referral process47. This is supported in our analysis as only 26% of patients who the CGC had identified as being eligible for PR had previously attended a PR program. The use of a clinical decision support system as described here may overcome such barriers by intelligently identifying those patients who are suitable according to guideline standards and subsequently prompting the operator to consider PR referral during the consultation as well as simultaneously upskilling the operator through educational prompts outlining the medical contraindications to PR referral. Thus, an additional utility of technology such as the CGC lies in its incorporation into a clinical pathway that aims to optimise the PR referral process by ensuring all patients referred initially undergo CGC review to ensure correct diagnosis, undergo detailed assessment of significant comorbidity and receive structured holistic self-management.

Whilst the use of inhaled corticosteroid-bronchodilator combination therapy represents evidence-based practice for selected COPD patients with exacerbations, guidelines now also recommend considering de-escalation of such therapy where appropriate. The use of inhaled corticosteroid monotherapy does not constitute guideline-based practice and over-prescription of inhaled corticosteroids and lack of appropriate withdrawal when no benefit exposes patients to adverse effects and negatively impact on health economics3,18,48, In addition to prompting escalation of therapy as per guideline recommendations, CGC review resulted in the discontinuation of inhaled steroid monotherapy in 75% of cases where patients had been prescribed these on entry to the software. The confidence of clinicians to de-escalate medication in any given condition such as COPD is incredibly important yet often overlooked and may further be strengthened by the application of technology such as the CGC enabling accurate diagnosis, staging and management of that condition49. Further qualitative research is needed in this specific area of practice.

A key strength of this evaluation lies in it’s large sample size spread across a wide catchment area and that every patient on the register was reviewed using a structured clinical pathway where on entry, the diagnosis of COPD was either confirmed or refuted based on spirometry using a clinically validated digitally accredited tool. Furthermore, patients with a confirmed diagnosis were appropriately staged in terms of disease severity at this Initial Review consultation. This diagnostic screening process ensured the accuracy and validity of any subsequent guideline standard pharmacological and non-pharmacological recommendations generated from the remainder of the CGC review. One limitation is that the CGC reviews consisted of a mixture of face-to-face and remote consultations (driven in part by the Covid-19 pandemic) which could have affected the delivery of some of the guideline level interventions described. Whilst the consultations described were all performed by trained primary care respiratory specialist nurses, the interventions detailed here represent those driven or primarily prompted by the CGC which a practice nurse conducting a COPD review could have followed. Community spirometry services were also affected by the Covid-19 pandemic which, in this analysis, limited those undertaking the consultations from performing spirometry during the consultation itself. This paper reports on interventions prompted by the CGC in a cross-sectional analysis and whilst the CGC reviews greatly increased the proportion of patients managed within guideline-based recommendations, it is hoped that this will result in an improvement in health-related quality of life for patients and a reduction in exacerbations, healthcare utilisation and hospitalisations in the future. However, the clinical guidelines themselves detail interventions demonstrated to reduce healthcare utilisation and improve health-related quality of life in COPD thus it is not unreasonable to extrapolate that any clinical pathway increasing uptake of these guidelines would achieve the same effect although this does require further prospective study. Finally, whilst preliminary data exists detailing the health economic benefits of the CGC during initial feasibility studies, further analysis is required to define the such benefits realised by utilising this specific clinical pathway as a result of increased clinical guideline implementation.