Artificial neural network identification of exercise expiratory flow-limitation in adults.

Identification of ventilatory constraint is a key objective of clinical exercise testing. Expiratory flow-limitation (EFL) is a well-known type of ventilatory constraint. However, EFL is difficult to measure, and commercial metabolic carts do not readily identify or quantify EFL. Deep machine learning might provide a new approach for identifying EFL. The objective of this study was to determine if a convolutional neural network (CNN) could accurately identify EFL during exercise in adults in whom baseline airway function varied from normal to mildly obstructed. 2931 spontaneous exercise flow-volume loops (eFVL) were placed within the baseline maximal expiratory flow-volume curves (MEFV) from 22 adults (15 M, 7 F; age, 32 yrs) in whom lung function varied from normal to mildly obstructed. Each eFVL was coded as EFL or non-EFL, where EFL was defined by eFVLs with expired airflow meeting or exceeding the MEFV curve. A CNN with seven hidden layers and a 2-neuron softmax output layer was used to analyze the eFVLs. Three separate analyses were conducted: (1) all subjects (n = 2931 eFVLs, [GRALL]), (2) subjects with normal spirometry (n = 1921 eFVLs [GRNORM]), (3) subjects with mild airway obstruction (n = 1010 eFVLs, [GRLOW]). The final output of the CNN was the probability of EFL or non-EFL in each eFVL, which is considered EFL if the probability exceeds 0.5 or 50%. Baseline forced expiratory volume in 1 s/forced vital capacity was 0.77 (94% predicted) in GRALL, 0.83 (100% predicted) in GRNORM, and 0.69 (83% predicted) in GRLOW. CNN model accuracy was 90.6, 90.5, and 88.0% in GRALL, GRNORM and GRLOW, respectively. Negative predictive value (NPV) was higher than positive predictive value (PPV) in GRNORM (93.5 vs. 78.2% for NPV vs. PPV). In GRLOW, PPV was slightly higher than NPV (89.5 vs. 84.5% for PPV vs. NPV). A CNN performed very well at identifying eFVLs with EFL during exercise. These findings suggest that deep machine learning could become a viable tool for identifying ventilatory constraint during clinical exercise testing.

Effect of barometric pressure on single-breath carbon monoxide diffusing capacity.

BACKGROUND: Single-breath diffusing capacity for carbon monoxide (DLCO) quantifies gas transfer in the lungs. DLCO measurement is affected by barometric pressure (Pb) and alveolar partial pressure of oxygen (PAO2). The current equations for adjusting DLCO for Pb and PAO2 may not be accurate given advances in test performance and technology. We quantify changes in DLCO with alterations in Pb in normal and COPD subjects, determine the accuracy of the current Pb and PAO2 adjustment equations and develop updated adjustment equations. METHODS: We measured DLCO in 13 normal and 10 COPD subjects at 1330 m altitude and in a hypobaric/hyperbaric chamber at altitudes of sea-level and 2500 m; six normal subjects were tested at 3600 m. We determined if there were significant differences in DLCO between altitudes. We developed an equation for adjusting DLCO for changes in Pb from sea-level. We compared this equation with the existing Pb adjustment equation in normal and COPD subjects. We determined the accuracy of the current PAO2 adjustment equation and developed a new PAO2 adjustment equation. RESULTS: DLCO significantly increased with decreasing Pb. We developed a Pb adjustment equation that adjusts DLCO measured at altitudes between 1330 m and 3600 m to sea-level values. This Pb adjustment equation yields DLCO results that are not significantly different than the currently recommended equation. We developed a more accurate PAO2 adjustment equation. CONCLUSION: DLCO measurement is significantly affected by altitude. We developed equations that accurately adjust DLCO for changes in Pb and PAO2 in normal and COPD subjects.

Variability in Cardiopulmonary Exercise Testing Biologic Controls.

BACKGROUND: Cardiopulmonary exercise testing is an increasingly common test and is considered the accepted standard for assessing exercise capacity. Quantifying variability is important to assess the instrument for quality control purposes. Though guidelines recommend biologic control testing, there are minimal data on how to do it. We sought to describe variability for oxygen consumption (V̇O2 ), carbon dioxide production (V̇CO2 ), and minute ventilation (V̇E) at various work rates under steady-state conditions in multiple subjects over a 1-y period to provide a practical approach to assess and perform biologic control testing. METHODS: We performed a single-center, prospective study with 4 healthy subjects, 2 men and 2 women. Subjects performed constant work rate exercise tests for 6 min each at 25-100 W intervals on a computer-controlled cycle ergometer. Data were averaged over the last 120 s at each work rate to reflect stepwise steady-state conditions. Descriptive statistics, including the mean, median, range, SD, and coefficient of variation (CoV) are reported for each individual across the 4 work rates and all repetitions. As these data were normative, z-scores were utilized, and a value greater than ± 1.96 z-scores was used to define significant test variability. RESULTS: Subjects performed 16-39 biocontrol studies over 1-y. The mean CoV for all subjects in V̇O2 was 6.59%, V̇CO2 was 6.41%, and V̇E was 6.32%. The ± 1.96 z-scores corresponded to a 9.4-18.1% change in V̇O2 , a 9.6-18.1% change in V̇CO2 , and a 9-21.5% change in V̇E across the 4 workloads. CONCLUSIONS: We report long-term variability for steady-state measurement of V̇O2 , V̇CO2 , and V̇E obtained during biocontrol testing. Utilizing ± 1.96 z-scores allows one to determine if a result exceeds expected variability, which may warrant investigation of the instrument.

Effects of sprint interval training on cardiorespiratory fitness while in a hyperbaric oxygen environment.

Objectives: Hyperbaric oxygen (HBO2) exposure may enhance cardiorespiratory fitness. Exercise training and HBO2 exposure stimulate mitochondrial biogenesis, increase capillary density, and induce adaptive antioxidant mechanisms. We hypothesized that an exercise regimen of sprint interval training (SIT) while breathing HBO2 would lead to a greater improvement in exercise performance compared to the same training breathing ambient air. Methods: Healthy long-term intermediate-altitude residents, ages 20-39 years, with normal spirometry and cardiorespiratory fitness were randomized to two groups: one performing six sessions of a SIT regimen over two weeks in a hyperbaric chamber (1.4 ATA [141.9 kPa], FiO2=1.0); the other performing under ambient pressure conditions (0.85 ATA [86.1 kPa], FiO2=0.21). Training effect was evaluated by comparing incremental cycle ergometry cardiopulmonary exercise testing before and after the training regimen. The primary outcome measure was peak oxygen consumption (V̇O2), while secondary outcomes included additional exercise parameters. The effect of study group on exercise parameters was assessed using two-factor repeated measures ANOVA. Results: Of 58 participants randomized, 49 completed the training program and all cardiopulmonary exercise tests (n=23 HBO2, n=26 ambient). Both groups experienced an increase in peak V̇O2: 8.1% HBO2 and 7.1% ambient; the differences were not significant (p=0.50). Secondary parameters of peak work rate and peak V̇E experienced a significantly higher change in the HBO2 group compared to the ambient group (p=0.05 and p=0.03, respectively). Conclusion: Cardiorespiratory fitness improved after a two-week SIT regimen, but improvement in peak V̇O2 was not significantly different between ambient and HBO2 groups.

A 73-Year-Old Man With Progressive Whole Body Subcutaneous Gas After Pleural Catheter Removal.

CASE PRESENTATION: A 73-year-old man presented to the ED of an outside hospital with asymptomatic chest wall swelling 10 h after discharge from our hospital. Four days earlier, he had presented to our hospital with increased dyspnea, cough, and sputum production. His history was notable for severe COPD with bullous emphysema. Chest imaging demonstrated bilateral opacities and a collection of gas and liquid in the major fissure of the left lung. A catheter was placed into the collection of gas and liquid under imaging guidance. After 4 days, the catheter was removed without event and the patient was discharged from the hospital with an extended course of antibiotics. Imaging performed in the ED revealed gas in the tissues of the chest wall and no evidence of a pneumothorax. He was transported back to our hospital by helicopter.

Airflow Obstruction Categorization Methods and Mortality.

RATIONALE: Current guidelines recommend using forced expiratory volume in 1 second (FEV1) % predicted to categorize the severity of airflow obstruction. There are limitations to using FEV1 % predicted for this purpose, including bias associated with demographic factors and the inability to correct for "lung size." Other methods for grading the severity of airflow obstruction have been proposed to address these limitations. OBJECTIVES: Our objectives were to categorize airflow obstruction severity using these methods and then determine which method results in a categorization most closely associated with mortality. METHODS: Study subjects were patients aged 40-80 years tested in our pulmonary function test laboratories in the period 2002 to 2013 with airflow obstruction based on an FEV1/forced vital capacity (FVC) less than the lower limit of normal. Categorization of airflow obstruction severity was determined using four methods: FEV1 % predicted; FEV1 % predicted adjusted by FVC % predicted; FEV1/FVC confidence interval approach; and FEV1 z-scores. Receiver operating characteristic curve analysis was used to determine which categorization method best predicts 5-year survival. RESULTS: We identified 2,000 patients with airflow obstruction. Important differences in the categorization of airflow obstruction severity were observed using the different methods. More patients were categorized as having severe obstruction using FEV1 % predicted and FEV1 z-scores compared with FEV1 % predicted adjusted by FVC % predicted and FEV1/FVC confidence interval approach. FEV1 % predicted was the best predictor of 5-year survival among the four methods studied. CONCLUSIONS: In our study, categorizing airflow obstruction severity using FEV1 % predicted best predicted 5-year survival. This validates the current guideline recommendation that FEV1 % predicted be used to categorize the severity of airflow obstruction.

Repeatability and Meaningful Change of CPET Parameters in Healthy Subjects.

INTRODUCTION/PURPOSE: Cardiopulmonary exercise testing (CPET) plays an important role in clinical medicine and research. Repeatability of CPET parameters has not been well characterized, but is important to assess variability and determine if there have been meaningful changes in a given CPET parameter. METHODS: We recruited 45 healthy subjects and performed two symptom-limited CPET within 30 d using a cycle ergometer. Differences in relevant CPET parameters between CPET-1 and CPET-2 were assessed using a paired t-test. Coefficient of variation (CoV) and Bland-Altman plots are reported. Factors that may be associated with variability were analyzed (sex, age, time of day, fitness level). The coefficient of repeatability was calculated for peak oxygen consumption (V˙O2) and V˙O2 at lactate threshold (LT) to establish a 95% threshold for meaningful change. RESULTS: There were no significant differences between tests in the parameters reported. Specifically, we found overall low CoV in peak V˙O2 (4.9%), V˙O2@LT (10.4%), peak O2 pulse (4.6%), peak minute ventilation (V˙E; 7.4%), V˙E/V˙CO2@LT (4.0%), and V˙E/V˙O2@LT (4.8%). The CoV for peak respiratory exchange ratio@LT was significantly affected by diurnal factors; age, sex, and fitness level did not affect variability. The 95% threshold for meaningful change was 0.540 L·min in peak V˙O2 and 0.520 L·min in V˙O2@LT. CONCLUSIONS: Repeatability of CPET parameters is generally higher than previously reported. There were no significant differences in variability related to sex, age, and fitness level; diurnal factors had a limited effect. The threshold for meaningful change in peak V˙O2 and for V˙O2@LT should be considered when gauging a response to therapies or training.

Hospital variation and temporal trends in palliative and end-of-life care in the ICU.

OBJECTIVES: Although studies have shown regional and interhospital variability in the intensity of end-of-life care, few data are available assessing variability in specific aspects of palliative care in the ICU across hospitals or interhospital variability in family and nurse ratings of this care. Recently, relatively high family satisfaction with ICU end-of-life care has prompted speculation that ICU palliative care has improved over time, but temporal trends have not been documented. DESIGN/SETTING: Retrospective cohort study of consecutive patients dying in the ICU in 13 Seattle-Tacoma-area hospitals between 2003 and 2008. MEASUREMENTS: We examined variability over time and among hospitals in satisfaction and quality of dying assessed by family, quality of dying assessed by nurses, and chart-based indicators of palliative care. We used regression analyses adjusting for patient, family, and nurse characteristics. MAIN RESULTS: Medical charts were abstracted for 3,065 of 3,246 eligible patients over a 55-month period. There were significant differences between hospitals for all chart-based indicators (p < 0.001), family satisfaction (p < 0.001), family-rated quality of dying (p = 0.03), and nurse-rated quality of dying (p = 0.003). There were few significant changes in these measures over time, although we found a significant increase in pain assessments in the last 24 hours of life (p < 0.001) as well as decreased documentation of family conferences (p < 0.001) and discussion of prognosis (p = 0.020) in the first 72 hours in the ICU. CONCLUSIONS: We found significant interhospital variation in ratings and delivery of palliative care, consistent with prior studies showing variation in intensity of care at the end of life. We did not find evidence of temporal changes in most aspects of palliative care, family satisfaction, or nurse/family ratings of the quality of dying. With the possible exception of pain assessment, there is little evidence that the quality of palliative care has improved over the time period studied.