This study is the first to access the extent to which spirometries are performed by French GPs in private practice. Our results show a gradual increase in the number of spirometries performed in patients aged between 40 and 75 years, between 2010 and 2018 by “non-expert GPs”: the GPs of interest for public health strategies to increase the number of spirometries performed in private practice. Indeed, between 2010 and 2018, both the number of spirometries performed by GPs and the number of GPs performing spirometries significantly increased. Thus, over time, more patients have had access to spirometry in French primary care.

Between 2010 to 2018, the proportion of spirometries performed per GP decreased (to 6.8 spirometries per GP in 2018), while the number of GPs performing spirometry increased (+147%), as did the number of spirometries performed each year (+106%). Thus, overall, more and more patients benefited per GP spirometry: from 5,145 in 2010 to 10,976 in 2018. The decrease in the proportion of spirometries performed per GPs could be explained by a “screening effect”. Indeed, GPs do not yearly screen all their patients, aged 40–75 years with COPD risk factors. However, yearly spirometry could be performed, by GPs, to follow up respiratory disease evolution, e.g., COPD and asthma, particularly when the pathology is not complex and not severe. The fact that the mean number of spirometries performed per patient exceeded 1 implies that GPs do not only perform spirometries for screening and diagnosis, but also for disease follow-up. However, it is noteworthy that certain patients performed excessive numbers of spirometries per year, e.g., the maximum number of spirometries performed by a patient in 2010 was 14. Although marginal, this could be due to coding errors by GPs when completing the SNDS database.

The more GPs perform spirometries the better the quality of the spirometries. Tollånes et al.17 reported that the more Norwegian GPs performed spirometries the better they interpreted the spirometry results. In our study, more and more “non-expert GPs” were practicing spirometry, but the annual volume was low (6.8 spirometries per GP in 2018): thus the quality of spirometry performed may be questionable. However, a recent study evaluated spirometries, performed by French GPs, for COPD detection and reported a 93% positive predictive value for spirometries performed by GPs compared to reference spirometries for confirmation of diagnosis18. The low volume of spirometries performed by GPs, that we report, does imply that GPs need to have a professional activity that is compatible with performing spirometries19. Indeed, to perform a spirometry a dedicated consultation is needed. This can be difficult, particularly considering the increased tension, particularly in rural areas, due to medical desertification. In this context, it is understandable that certain GPs do not want to be trained to perform spirometries and to integrate this procedure in their practices, this despite the public health need.

Even though the number of spirometries performed by GPs, for patients aged 40–75 years, has increased by 147% from 2010 to 2018, only 2.8% of French “non-expert GPs” performed spirometries in 2018. This is insufficient considering the need for both the early detection of COPD and the follow up of non-severe COPD and asthma, relative to the prevalence of these diseases. Moreover, when prescribed, a larger proportion of patients underwent spirometries when the spirometry was performed by GPs compared to pneumologists. Indeed, a French study evaluating professional practices showed that 89% of eligible patients (n = 184) agreed to have spirometry performed by their GPs for the early detection of COPD, and 66% performed this specific consultation. Among them, 82% were satisfied that their GPs prescribed and performed their spirometries20. Overall, a qualitative study, performed in patients following a routine spirometry test to confirm COPD diagnosis by their GPs, showed that patients were pleased with the spirometry in primary care21. In patients prescribed a spirometry, fewer patients perform spirometries when the delays for an appointment were long: delays are often shorter with GPs than with pneumologist. Indeed, GPs are often geographically easier to access and more availability than pneumologists22.

The study screening also highlights the 277,177 GPs, with expertise in spirometry, that perform more than 60 spirometries per year: between 1 and 2 spirometries per week. Currently, with the reorganisation of French primary care to develop multidisciplinary groups, it may be more pertinent to encourage GPs with expertise to perform more spirometries, instead of encouraging more GPs to perform few spirometries. We observed in our study that about 25% of patients with spirometries performed by GPs were referrals from other primary healthcare providers. Thus, in France, GPs without spirometry expertise are referring patients to GPs with this expertise. These referrals allow trained GPs to perform more spirometries and improves interpretation of results, and making spirometry more accessible for patients.

This is the first study to access the extent to which GPs performed spirometries in France, between 2010 and 2018, based on data from the SNDS. The patients included in our study are similar to those enrolled in studies accessing early detection of COPD in primary care14,23,24. The Score Santé database was used to obtain the annual numbers of GPs practicing in France. This database only concerns GPs in private practice. However, in the last few years, primary healthcare centres have developed that employ GPs. Therefore, using the data from the Score Santé database may overestimate the proportion of GPs performing spirometries. During the period under study (2010–2018), and at present, most GPs were in private practice. In contrast, the number of spirometries performed in 2017 and 2018 may be slightly underestimated. During these years, GPs that performed spirometries for referred patients could have coded the spirometry as an “exceptional consultation”, and not GLQP012, without needing to specify that it was spirometry. These spirometries would not have been extracted from the SNDS for our study. In the SNDS, the GLQP012 code does not specify whether the spirometry was performed for detecting or monitoring a respiratory pathology; nor does it indicate the pathology of interest (COPD, asthma, or another respiratory disorder). Notwithstanding, considering the age of our population (40 to 75 years old) and the increased prevalence of COPD, compared to asthma, in this age group, we suppose that most of the spirometries performed were to detect or follow up COPD. Finally, it is difficult to compare our data with those of other countries owing to the substantial differences in healthcare and primary care structures and organisations.

Our study shows that, in France, in patients 40–75 years old, the annual number of spirometries rose between 2010 and 2018, as did the number of non-expert GPs performing them. Furthermore, between 2010 and 2018, the number of patients undergoing spirometries increased. Currently, COPD is largely underdiagnosed. Early diagnosis of COPD, particularly in primary care, is essential. Although our study shows that more and more French GPs are performing spirometries, it is critical that GPs be encouraged to develop expertise to perform spirometries to detect respiratory disorders, particularly COPD, and to follow up respiratory diseases.