Physical activity coaching in patients with interstitial lung diseases: A randomized controlled trial.

OBJECTIVES: Physical activity is reduced in patients with interstitial lung disease (ILD) and physical inactivity is related to poor health outcomes. We investigated the effect of a telecoaching intervention to improve physical activity in patients with ILD. METHODS: Eighty patients with ILD were randomized into the intervention or control group. Patients in the intervention group received a 12-week telecoaching program including a step counter, a patient-tailored smartphone application, and coaching calls. Patients in the control group received usual care. Physical activity (primary outcome), physical fitness and quality of life were measured at baseline and 12 weeks later with an accelerometer, 6-min walking test and quadriceps muscle force and the King's Brief Interstitial Lung Disease questionnaire (K-BILD). RESULTS: Participation in telecoaching did not improve physical activity: between-group differences for step count: 386 ± 590 steps/day, p = .52; sedentary time: 4 ± 18 min/day, p = .81; movement intensity: 0.04 ± 0.05 m/s2, p = .45). Between-group differences for the 6-min walking test, quadriceps muscle force and K-BILD were 14 ± 10 m, p = .16; 2 ± 3% predicted, p = .61; 0.8 ± 1.7 points, p = .62 respectively. CONCLUSIONS: Twelve weeks of telecoaching did not improve physical activity, physical fitness or quality of life in patients with ILD. Future physical or behavioural interventions are needed for these patients to improve physical activity.

The association between objectively measured physical activity and the prevalence of comorbidities in lung transplant recipients.

Introduction Lung transplant recipients are often physically inactive and are at risk of developing comorbidities. We investigated whether objectively measured physical activity was associated with the prevalence of comorbidities. Methods Physical activity (accelerometry) and the presence of cardiovascular disease, symptoms of depression and anxiety, diabetes, dyslipidaemia, hypertension, lower extremity artery disease, muscle weakness, obesity and osteoporosis were assessed in 108 lung transplant recipients. Patients were divided into four groups based on daily step count. Results A cohort of 108 patients (60±7 years, 51% male, 20±14 months since transplantation) was included. Active patients (>7500 steps/day) had significantly fewer comorbidities (4 comorbidities) compared to severely inactive patients (<2500 steps/day, 6 comorbidities) and muscle weakness and high symptoms of depression were less prevalent. Severely inactive patients had significantly more cardiovascular comorbidities compared to all other groups. No other significant differences were observed. Conclusion Physically active lung transplant recipients have fewer comorbidities, lower prevalence of muscle weakness and fewer symptoms of depression compared to very inactive patients.

Implementation of an enhanced recovery protocol for lung volume reduction surgery: an observational cohort study.

OBJECTIVES: Lung volume reduction surgery (LVRS) is an established therapeutic option for advanced emphysema. To improve patients' safety and reduce complications, an enhanced recovery protocol (ERP) was implemented. This study aims to describe and evaluate the short-term outcome of this ERP. METHODS: This retrospective single-centre study included all consecutive LVRS patients (1 January 2017 until 15 September 2020). An ERP for LVRS was implemented and stepwise optimised from 1 August 2019, it consisted of changes in pre-, peri- and postoperative care pathways. Patients were compared before and after implementation of ERP. Primary outcome was incidence of postoperative complications (Clavien-Dindo), and secondary outcomes included chest tube duration, incidence of prolonged air leak (PAL), length of stay (LOS) and 90-day mortality. Lung function and exercise capacity were evaluated at 3 and 6 months post-LVRS. RESULTS: Seventy-six LVRS patients were included (pre-ERP: n=41, ERP: n=35). The ERP cohort presented with lower incidence of postoperative complications (42% vs 83%, P=0.0002), shorter chest tube duration (4 vs 12 days, P<0.0001) with a lower incidence of PAL (21% vs 61%, P=0.0005) and shorter LOS (6 vs 14 days, P<0.0001). No in-hospital mortality occurred in the ERP cohort versus 4 pre-ERP. Postoperative forced expiratory volume in 1 s was higher in the ERP cohort compared to pre-ERP at 3 months (1.35 vs 1.02 l) and at 6 months (1.31 vs 1.01 l). CONCLUSIONS: Implementation of ERP as part of a comprehensive reconceptualisation towards LVRS, demonstrated fewer postoperative complications, including PAL, resulting in reduced LOS. Improved short-term functional outcomes were observed at 3 and 6 months.

Opioids in geriatric units in 14 Belgian hospitals: prevalence, dosage and associated factors.

Introduction: Adverse effects of opioids are common among older individuals, and undertreatment as well as overuse can be an issue. Epidemiological data on opioid use in older individuals are available, but scarce in hospitalized patients.Aims: The aim of this study is to examine the one-day prevalence of opioid use among older inpatients and identify the factors associated with both opioid use and dosage.Materials and methods: One-day cross-sectional study with data collected from geriatric units across 14 Belgian hospitals. The primary focus of the study is to assess the prevalence of opioid use and dosage, along with identifying associated factors. To achieve this, a multiple binary logistic regression model was fitted for opioid use, and a multiple linear regression model for opioid dose.Results: Opioids were used in 24.4% of 784 patients, of which 57.9% was treated with tramadol, 13.2% with oxycodone or morphine and 28.9% with transdermal buprenorphine or fentanyl. The odds for opioid use were 4.2 times higher in patients in orthogeriatric units compared to other patients (OR=4.2, 95% CI=2.50-7.05). The prevalence of opioid use was 34% higher in patients without dementia compared to patients with dementia (OR=0.66, 95% CI=0.46-0.95). The overall mean daily dosage was 14.07mg subcutaneous morphine equivalent. After adjustment for age, gender and dementia, dosage was only associated with type of opioid: the estimated mean opioid dose was 70% lower with tramadol (mean ratio=0,30,95% CI=0,23-0,39) and 67% lower with oxycodone and morphine (mean ratio=0,33, 95% CI=0,22-0,48) compared to transdermal buprenorphine and transdermal fentanyl.Conclusions: One in four patients received opioid treatment. It is not clear whether this reflects under- or overtreatment, but these results can serve as a benchmark for geriatric units to guide future pain management practices. The utilization of transdermal fentanyl and buprenorphine, resulting in higher doses of morphine equivalent, poses significant risks for side effects.

The memory of airway epithelium damage in smokers and COPD patients.

Chronic obstructive pulmonary disease (COPD), a devastating and irreversible lung disease, causes structural and functional defects in the bronchial epithelium, the (ir)reversibility of which remains unexplored in vitro. This study aimed to investigate the persistence of COPD-related epithelial defects in long-term airway epithelial cultures derived from non-smokers, smokers, and COPD patients. Barrier function, polarity, cell commitment, epithelial-to-mesenchymal transition, and inflammation were evaluated and compared with native epithelium characteristics. The role of inflammation was explored using cytokines. We show that barrier dysfunction, compromised polarity, and lineage abnormalities observed in smokers and COPD persisted for up to 10 wk. Goblet cell hyperplasia was associated with recent cigarette smoke exposure. Conversely, increased IL-8/CXCL-8 release and abnormal epithelial-to-mesenchymal transition diminished over time. These ex vivo observations matched surgical samples' abnormalities. Cytokine treatment induced COPD-like changes in control cultures and reactivated epithelial-to-mesenchymal transition in COPD cells. In conclusion, these findings suggest that the airway epithelium of smokers and COPD patients retains a multidimensional memory of its original state and previous cigarette smoke-induced injuries, maintaining these abnormalities for extended periods.

Mucosal immune alterations at the early onset of tissue destruction in chronic obstructive pulmonary disease.

Rationale: COPD is characterized by chronic airway inflammation, small airways changes, with disappearance and obstruction, and also distal/alveolar destruction (emphysema). The chronology by which these three features evolve with altered mucosal immunity remains elusive. This study assessed the mucosal immune defense in human control and end-stage COPD lungs, by detailed microCT and RNA transcriptomic analysis of diversely affected zones. Methods: In 11 control (non-used donors) and 11 COPD (end-stage) explant frozen lungs, 4 cylinders/cores were processed per lung for microCT and tissue transcriptomics. MicroCT was used to quantify tissue percentage and alveolar surface density to classify the COPD cores in mild, moderate and severe alveolar destruction groups, as well as to quantify terminal bronchioles in each group. Transcriptomics of each core assessed fold changes in innate and adaptive cells and pathway enrichment score between control and COPD cores. Immunostainings of immune cells were performed for validation. Results: In mildly affected zones, decreased defensins and increased mucus production were observed, along CD8+ T cell accumulation and activation of the IgA pathway. In more severely affected zones, CD68+ myeloid antigen-presenting cells, CD4+ T cells and B cells, as well as MHCII and IgA pathway genes were upregulated. In contrast, terminal bronchioles were decreased in all COPD cores. Conclusion: Spatial investigation of end-stage COPD lungs show that mucosal defense dysregulation with decreased defensins and increased mucus and IgA responses, start concomitantly with CD8+ T-cell accumulation in mild emphysema zones, where terminal bronchioles are already decreased. In contrast, adaptive Th and B cell activation is observed in areas with more advanced tissue destruction. This study suggests that in COPD innate immune alterations occur early in the tissue destruction process, which affects both the alveoli and the terminal bronchioles, before the onset of an adaptive immune response.

The association of weather conditions with day-to-day variability in physical activity in patients with COPD.

Background: While patients with COPD often cite weather conditions as a reason for inactivity, little is known about the relationship between physical activity (PA) and weather conditions. The present study investigated the association of day-to-day weather changes on PA in patients with COPD and investigated patient characteristics related to being more or less influenced by weather conditions. Methods: In this longitudinal analysis, device-based day-by-day step counts were objectively measured in COPD patients for up to 12 months. Daily meteorological data (temperature, precipitation, wind speed, hours of sunlight and daylight) were linked to the daily step count and individual and multivariable relationships were investigated using mixed-model effects. Individual R2 was calculated for every subject to investigate the estimated influence of weather conditions on a patient level and its relationship with patient characteristics. Results: We included 50 patients with a mean±sd follow-up time of 282±93 days, totalling 14 117 patient-days. Daily temperature showed a positive linear pattern up until an inflexion point, after which a negative association with increasing temperature was observed (p<0.0001). Sunshine and daylight time had a positive association with PA (p<0.0001). Precipitation and wind speed were negatively associated with PA (p<0.0001). The median per-patient R2 for overall weather conditions was 0.08, ranging from 0.00 to 0.42. No strong associations between patient characteristics and per-patient R2 were observed. Conclusion: Weather conditions are partly associated with PA in patients with COPD, yet the overall explained variance of PA due to weather conditions is rather low and varied strongly between individuals.

Baseline characteristics from a 3-year longitudinal study to phenotype subjects with COPD: the FOOTPRINTS study.

BACKGROUND: FOOTPRINTS® is a prospective, longitudinal, 3-year study assessing the association between biomarkers of inflammation/lung tissue destruction and chronic obstructive pulmonary disease (COPD) severity and progression in ex-smokers with mild-to-severe COPD. Here, we present baseline characteristics and select biomarkers of study subjects. METHODS: The methodology of FOOTPRINTS® has been published previously. The study population included ex-smokers with a range of COPD severities (Global Initiative for Chronic Obstructive Lung Disease [GOLD] stages 1-3), ex-smokers with COPD and alpha-1-antitrypsin deficiency (A1ATD) and a control group of ex-smokers without airflow limitation (EwAL). At study entry, data were collected for: demographics, disease characteristics, history of comorbidities and COPD exacerbations, symptoms, lung function and volume, exercise capacity, soluble biomarkers, and quantitative and qualitative computed tomography. Baseline data are presented with descriptive statistical comparisons for soluble biomarkers in the individual GOLD and A1ATD groups versus EwAL. RESULTS: In total, 463 subjects were enrolled. The per-protocol set comprised 456 subjects, mostly male (64.5%). The mean (standard deviation) age was 60.7 (6.9) years. At baseline, increasing pulmonary symptoms, worse lung function, increased residual volume, reduced diffusing capacity of the lung for carbon monoxide (DLco) and greater prevalence of centrilobular emphysema were observed with increasing disease severity amongst GOLD 1-3 subjects. Subjects with A1ATD (n = 19) had similar lung function parameters to GOLD 2-3 subjects, a high residual volume comparable to GOLD 3 subjects, and similar air trapping to GOLD 2 subjects. Compared with EwAL (n = 61), subjects with A1ATD had worse lung function, increased residual volume, reduced DLco, and a greater prevalence of confluent or advanced destructive emphysema. The soluble inflammatory biomarkers white blood cell count, fibrinogen, high-sensitivity C-reactive protein and plasma surfactant protein were higher in GOLD 1-3 groups than in the EwAL group. Interleukin-6 was expressed less often in EwAL subjects compared with subjects in the GOLD and A1ATD groups. Soluble receptor for advanced glycation end product was lowest in GOLD 3 subjects, indicative of more severe emphysema. CONCLUSIONS: These findings provide context for upcoming results from FOOTPRINTS®, which aims to establish correlations between biomarkers and disease progression in a representative COPD population. TRIAL REGISTRATION NUMBER: NCT02719184, study start date 13/04/2016.

Sustaining training effects through physical activity coaching (STEP): a randomized controlled trial.

BACKGROUND: Pulmonary rehabilitation (PR) programs improve physical fitness, symptoms and quality of life (QoL) of patients with COPD. However, improved physical activity (PA) is not guaranteed after PR and the clinical benefits fade off after PR discharge. We aimed to investigate whether a 9 months PA-telecoaching program is able to improve PA of patients with COPD, after 3 months of PR and if this leads to maintenance of PR-acquired benefits. METHODS: Patients with COPD enrolled in a 6-month PR program were randomized to a (semi-automated) PA-telecoaching program or usual care, 3 months after PR initiation. The intervention consisted of a smartphone application with individual targets and feedback (for 6 months) and self-monitoring with a step counter (for 9 months). Patients were followed up for 9 months after randomization. Primary outcome was PA (daily step count by accelerometery), secondary outcomes were exercise tolerance, quadriceps force, dyspnea and QoL. RESULTS: Seventy-three patients were included (mean ± SD: 65 ± 7 years, FEV1 49 ± 19%, 6MWD 506 ± 75 m, PA 5225 ± 2646 steps/day). The intervention group presented a significant improvement in steps/day at every visit compared to usual care (between-group differences mean ± SE: 1431 ± 555 steps/day at 9 months after randomization, p = 0.01). Secondary outcomes did not differ between the groups. CONCLUSION: The semi-automated PA-telecoaching program implemented after 3 months of PR was effective to improve the amount of PA (steps/day) during PR and after follow-up. However, this was not accompanied by the maintenance of other PR-acquired benefits. TRIAL REGISTRATION: ClinicalTrials.gov. Identifier: NCT02702791. Retrospectively registered on March 9, 2016. Start study October 2015. https://clinicaltrials.gov/ct2/show/NCT02702791?term=NCT02702791&draw=2&rank=1 .

Factors associated to physical activity in patients with COPD: An ecological approach.

BACKGROUND: Physical activity (PA) is low in patients with Chronic Obstructive Pulmonary Disease (COPD). Identifying modifiable and non-modifiable correlates of PA give understanding of the individual behavior and provide future directions for PA enhancing interventions. As PA is complex and multidimensional, it should be embedded within a thorough framework. OBJECTIVE: To identify correlates of PA in a comprehensive COPD population based on a broad ecological model, including physiological, psychological, socio-demographic and environmental dimensions. METHODS: PA was objectively measured using the Dynaport Movemonitor and a comprehensive data collection of physiological, psychological, socio-demographic and environmental factors were collected. Bivariable and multivariable regression analyses (including principle component analysis) were executed. RESULTS: For this cross-sectional analysis, we included 148 patients with COPD and valid PA data (mean (SD) age 68 (7) years, FEV1 57 (17) % predicted, 5613 (3596) steps per day). Significant bivariable associations were found for physiological (exercise capacity, muscle force, lung function, symptoms, comorbidities), psychological (e.g. fatigue, motivation, perceived difficulty with PA), socio-demographic (dog owning, use of activity tracker) and environmental (season, daylight, temperature) factors. Based on the multivariable regression model, exercise capacity, beliefs on motivation, importance and self-confidence regarding PA and weather conditions were independent correlates of mean steps per day (R2 = 0.35). Movement intensity during walking was only independently associated with exercise capacity and age (R2 = 0.41). CONCLUSION: Although a wide range of potential influence factors were evaluated, variance in PA was only partly explained, supporting that PA is a complex behavior which is difficult to predict.

Machine learning to differentiate pulmonary hypertension due to left heart disease from pulmonary arterial hypertension.

Background and aims: Pulmonary hypertension due to left heart disease (PH-LHD) is the most frequent form of PH. As differential diagnosis with pulmonary arterial hypertension (PAH) has therapeutic implications, it is important to accurately and noninvasively differentiate PH-LHD from PAH before referral to PH centres. The aim was to develop and validate a machine learning (ML) model to improve prediction of PH-LHD in a population of PAH and PH-LHD patients. Methods: Noninvasive PH-LHD predictors from 172 PAH and 172 PH-LHD patients from the PH centre database at the University Hospitals of Leuven (Leuven, Belgium) were used to develop an ML model. The Jacobs score was used as performance benchmark. The dataset was split into a training and test set (70:30) and the best model was selected after 10-fold cross-validation on the training dataset (n=240). The final model was externally validated using 165 patients (91 PAH, 74 PH-LHD) from Erasme Hospital (Brussels, Belgium). Results: In the internal test dataset (n=104), a random forest-based model correctly diagnosed 70% of PH-LHD patients (sensitivity: n=35/50), with 100% positive predicted value, 78% negative predicted value and 100% specificity. The model outperformed the Jacobs score, which identified 18% (n=9/50) of the patients with PH-LHD without false positives. In external validation, the model had 64% sensitivity at 100% specificity, while the Jacobs score had a sensitivity of 3% for no false positives. Conclusions: ML significantly improves the sensitivity of PH-LHD prediction at 100% specificity. Such a model may substantially reduce the number of patients referred for invasive diagnostics without missing PAH diagnoses.

The concomitant use of paracetamol and flucloxacillin. A rare cause of high anion gap metabolic acidosis in the frail oldest old.

OBJECTIVES: This article describes the occurrence of a high anion gap metabolic acidosis (HAGMA) in two older, female patients with a methicillin sensitive staphylococcus aureus (MSSA) infection. Both patients received flucloxacillin and paracetamol. Both of them initially improved but declined rapidly after two to three weeks of treatment. They developed a severe HAGMA resulting in their death. The objective of this article is to determine whether old age is a major risk factor for developing HAGMA when combining paracetamol with flucloxacillin. METHODS: A literature study was conducted using the MEDLINE database, PubMed. The used MeSH terms were 'flucloxacillin, acetaminophen, glutathione synthetase deficiency and acidosis'. Furthermore, we used two cases to illustrate our findings. RESULTS: The origin of the high anion gap metabolic acidosis is the accumulation of 5-oxoproline which is known to occur when combining flucloxacillin with paracetamol due to their interaction with different enzymes of the gamma glutamyl cycle. This leads to the depletion of glutathione and the formation of 5-oxoproline. This phenomenon has a higher risk of occurring in frail older adults as most of them have several predisposing risk factors which result in lower baseline glutathione reserve. These risk factors include old age, malnutrition, assigned female at birth, pre-existing kidney and/or liver dysfunction, uncontrolled diabetes and sepsis. CONCLUSION: The purpose of this article is to raise awareness of this phenomenon and its higher occurrence in frail older adults, which hopefully will lead to an earlier diagnosis with a better outcome for the patient.

COVID-19 progression in hospitalized patients using follow-up in vivo CT and ex vivo microCT.

Background: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes coronavirus disease-19 (COVID-19) which can lead to acute respiratory distress syndrome (ARDS) and evolve to pulmonary fibrosis. Computed tomography (CT) is used to study disease progression and describe radiological patterns in COVID-19 patients. This study aimed to assess disease progression regarding lung volume and density over time on follow-up in vivo chest CT and give a unique look at parenchymal and morphological airway changes in "end-stage" COVID-19 lungs using ex vivo microCT. Methods: Volumes and densities of the lung/lobes of three COVID-19 patients were assessed using follow-up in vivo CT and ex vivo whole lung microCT scans. Airways were quantified by airway segmentations on whole lung microCT and small-partition microCT. As controls, three discarded healthy donor lungs were used. Histology was performed in differently affected regions in the COVID-19 lungs. Results: In vivo, COVID-19 lung volumes decreased while density increased over time, mainly in lower lobes as previously shown. Ex vivo COVID-19 lung volumes decreased by 60% and all lobes were smaller compared to controls. Airways were more visible on ex vivo microCT in COVID-19, probably due to fibrosis and increased airway diameter. In addition, small-partition microCT showed more deformation of (small) airway morphology and fibrotic organization in severely affected regions with heterogeneous distributions within the same lung which was confirmed by histology. Conclusions: COVID-19-ARDS and subsequent pulmonary fibrosis alters lung architecture and airway morphology which is described using in vivo CT, ex vivo microCT, and histology.

Patients' acceptance of outcome and experience measurements during hospitalisation for COPD exacerbations: a CICERO Clinical Research Collaboration-European Lung Foundation online patient survey.

Background: The lack of standardised outcome assessments during hospitalisation and follow-up for acute COPD exacerbations has hampered scientific progress and clinical proficiency. The objective of the present study was to evaluate patients' acceptance of selected outcome and experience measurements during hospitalisations for COPD exacerbations and follow-up. Methods: An online survey was held amongst COPD patients in France, Belgium, The Netherlands, Germany and the UK. The European Lung Foundation COPD Patient Advisory Group was involved in the conceptualisation, development and dissemination of the survey. The survey was complementary to a previously obtained expert consensus. We assessed patients' views and acceptance of selected patient-reported outcomes or experiences and corresponding measurement instruments (for dyspnoea, frequent productive cough, health status and hospitalisation experience), and of selected clinical investigations (blood draw, pulmonary function test, 6-min walk test, chest computed tomography, echocardiography). Findings: 200 patients completed the survey. All selected outcomes and experiences were deemed important, and acceptance of their methods of assessment was high. The modified Medical Research Council scale and a numerical rating scale to address dyspnoea, the COPD Assessment Test for quality of life and frequent productive cough, and the Hospital Consumer Assessment of Healthcare Providers and Systems for hospital experiences were the instruments preferred by patients. Consensus on importance of blood draw and spirometry was higher compared with the other investigations. Interpretation: The survey results endorse the use of the selected outcome and experience measurements during hospitalisations for COPD exacerbations. They can be used to optimise standardised and patient-centred care and facilitate multicentric data collection.

Clinical effectiveness of rehabilitation in ambulatory care for patients with persisting symptoms after COVID-19: a systematic review.

BACKGROUND: Lingering symptoms after acute COVID-19 present a major challenge to ambulatory care services. Since there are reservations regarding their optimal management, we aimed to collate all available evidence on the effects of rehabilitation treatments applicable in ambulatory care for these patients. METHODS: On 9 May 2022, we systematically searched articles in COVID-19 collections, Embase, MEDLINE, Cochrane Library, Web of Science, CINAHL, PsycArticles, PEDro, and EuropePMC. References were eligible if they reported on the clinical effectiveness of a rehabilitation therapy applicable in ambulatory care for adult patients with persisting symptoms continuing 4 weeks after the onset of COVID-19. The quality of the studies was evaluated using the CASP cohort study checklist and the Cochrane Risk of Bias Assessment Tool. Summary of Findings tables were constructed and the certainty of evidence was assessed using the GRADE framework. RESULTS: We included 38 studies comprising 2,790 participants. Physical training and breathing exercises may reduce fatigue, dyspnoea, and chest pain and may improve physical capacity and quality of life, but the evidence is very weak (based on 6 RCTs and 12 cohort studies). The evidence underpinning the effect of nutritional supplements on fatigue, dyspnoea, muscle pain, sensory function, psychological well-being, quality of life, and functional capacity is very poor (based on 4 RCTs). Also, the evidence-base is very weak about the effect of olfactory training on sensory function and quality of life (based on 4 RCTs and 3 cohort studies). Multidisciplinary treatment may have beneficial effects on fatigue, dyspnoea, physical capacity, pulmonary function, quality of life, return to daily life activities, and functional capacity, but the evidence is very weak (based on 5 cohort studies). The certainty of evidence is very low due to study limitations, inconsistency, indirectness, and imprecision. CONCLUSIONS: Physical training, breathing exercises, olfactory training and multidisciplinary treatment can be effective rehabilitation therapies for patients with persisting symptoms after COVID-19, still with high uncertainty regarding these effects. These findings can guide ambulatory care practitioners to treat these patients and should be incorporated in clinical practice guidelines. High-quality studies are needed to confirm our hypotheses and should report on adverse events.

Predicting total lung capacity from spirometry: a machine learning approach.

Background and objective: Spirometry patterns can suggest that a patient has a restrictive ventilatory impairment; however, lung volume measurements such as total lung capacity (TLC) are required to confirm the diagnosis. The aim of the study was to train a supervised machine learning model that can accurately estimate TLC values from spirometry and subsequently identify which patients would most benefit from undergoing a complete pulmonary function test. Methods: We trained three tree-based machine learning models on 51,761 spirometry data points with corresponding TLC measurements. We then compared model performance using an independent test set consisting of 1,402 patients. The best-performing model was used to retrospectively identify restrictive ventilatory impairment in the same test set. The algorithm was compared against different spirometry patterns commonly used to predict restriction. Results: The prevalence of restrictive ventilatory impairment in the test set is 16.7% (234/1402). CatBoost was the best-performing machine learning model. It predicted TLC with a mean squared error (MSE) of 560.1 mL. The sensitivity, specificity, and F1-score of the optimal algorithm for predicting restrictive ventilatory impairment was 83, 92, and 75%, respectively. Conclusion: A machine learning model trained on spirometry data can estimate TLC to a high degree of accuracy. This approach could be used to develop future smart home-based spirometry solutions, which could aid decision making and self-monitoring in patients with restrictive lung diseases.

Pulmonary rehabilitation and physical interventions.

Pulmonary rehabilitation has established a status of evidence-based therapy for patients with symptomatic COPD in the stable phase and after acute exacerbations. Rehabilitation should have the possibility of including different disciplines and be offered in several formats and lines of healthcare. This review focusses on the cornerstone intervention, exercise training, and how training interventions can be adapted to the limitations of patients. These adaptations may lead to altered cardiovascular or muscular training effects and/or may improve movement efficiency. Optimising pharmacotherapy (not the focus of this review) and oxygen supplements, whole-body low- and high-intensity training or interval training, and resistance (or neuromuscular electrical stimulation) training are important training modalities for these patients in order to accommodate cardiovascular and ventilatory impairments. Inspiratory muscle training and whole-body vibration may also be worthwhile interventions in selected patients. Patients with stable but symptomatic COPD, those who have suffered exacerbations and patients waiting for or who have received lung volume reduction or lung transplantation are good candidates. The future surely holds promise to further personalise exercise training interventions and to tailor the format of rehabilitation to the individual patient's needs and preferences.

Estimating individual treatment effects on COPD exacerbations by causal machine learning on randomised controlled trials.

RATIONALE: Estimating the causal effect of an intervention at individual level, also called individual treatment effect (ITE), may help in identifying response prior to the intervention. OBJECTIVES: We aimed to develop machine learning (ML) models which estimate ITE of an intervention using data from randomised controlled trials and illustrate this approach with prediction of ITE on annual chronic obstructive pulmonary disease (COPD) exacerbation rates. METHODS: We used data from 8151 patients with COPD of the Study to Understand Mortality and MorbidITy in COPD (SUMMIT) trial (NCT01313676) to address the ITE of fluticasone furoate/vilanterol (FF/VI) versus control (placebo) on exacerbation rate and developed a novel metric, Q-score, for assessing the power of causal inference models. We then validated the methodology on 5990 subjects from the InforMing the PAthway of COPD Treatment (IMPACT) trial (NCT02164513) to estimate the ITE of FF/umeclidinium/VI (FF/UMEC/VI) versus UMEC/VI on exacerbation rate. We used Causal Forest as causal inference model. RESULTS: In SUMMIT, Causal Forest was optimised on the training set (n=5705) and tested on 2446 subjects (Q-score 0.61). In IMPACT, Causal Forest was optimised on 4193 subjects in the training set and tested on 1797 individuals (Q-score 0.21). In both trials, the quantiles of patients with the strongest ITE consistently demonstrated the largest reductions in observed exacerbations rates (0.54 and 0.53, p<0.001). Poor lung function and blood eosinophils, respectively, were the strongest predictors of ITE. CONCLUSIONS: This study shows that ML models for causal inference can be used to identify individual response to different COPD treatments and highlight treatment traits. Such models could become clinically useful tools for individual treatment decisions in COPD.

Collaboration between explainable artificial intelligence and pulmonologists improves the accuracy of pulmonary function test interpretation.

BACKGROUND: Few studies have investigated the collaborative potential between artificial intelligence (AI) and pulmonologists for diagnosing pulmonary disease. We hypothesised that the collaboration between a pulmonologist and AI with explanations (explainable AI (XAI)) is superior in diagnostic interpretation of pulmonary function tests (PFTs) than the pulmonologist without support. METHODS: The study was conducted in two phases, a monocentre study (phase 1) and a multicentre intervention study (phase 2). Each phase utilised two different sets of 24 PFT reports of patients with a clinically validated gold standard diagnosis. Each PFT was interpreted without (control) and with XAI's suggestions (intervention). Pulmonologists provided a differential diagnosis consisting of a preferential diagnosis and optionally up to three additional diagnoses. The primary end-point compared accuracy of preferential and additional diagnoses between control and intervention. Secondary end-points were the number of diagnoses in differential diagnosis, diagnostic confidence and inter-rater agreement. We also analysed how XAI influenced pulmonologists' decisions. RESULTS: In phase 1 (n=16 pulmonologists), mean preferential and differential diagnostic accuracy significantly increased by 10.4% and 9.4%, respectively, between control and intervention (p<0.001). Improvements were somewhat lower but highly significant (p<0.0001) in phase 2 (5.4% and 8.7%, respectively; n=62 pulmonologists). In both phases, the number of diagnoses in the differential diagnosis did not reduce, but diagnostic confidence and inter-rater agreement significantly increased during intervention. Pulmonologists updated their decisions with XAI's feedback and consistently improved their baseline performance if AI provided correct predictions. CONCLUSION: A collaboration between a pulmonologist and XAI is better at interpreting PFTs than individual pulmonologists reading without XAI support or XAI alone.