Influence of different data-averaging methods on mean values of selected variables derived from preoperative cardiopulmonary exercise testing in patients scheduled for colorectal surgery

Ruud F W Franssen; Bart H E Sanders; Tim Takken; F Jeroen Vogelaar; Maryska L G Janssen-Heijnen; Bart C Bongers

doi:10.1371/journal.pone.0283129

. 2023 Mar 16;18(3):e0283129. doi: 10.1371/journal.pone.0283129

Influence of different data-averaging methods on mean values of selected variables derived from preoperative cardiopulmonary exercise testing in patients scheduled for colorectal surgery

Ruud F W Franssen ^1,^2,^*, Bart H E Sanders ³, Tim Takken ⁴, F Jeroen Vogelaar ², Maryska L G Janssen-Heijnen ^2,⁵, Bart C Bongers ^6,⁷

Editor: Lindsay Bottoms⁸

¹Department of Clinical Physical Therapy, VieCuri Medical Center, Venlo, The Netherlands

²Department of Epidemiology, GROW School for Oncology and Reproduction, Faculty of Health, Medicine and Life Sciences, Maastricht University, Maastricht, The Netherlands

³Department of Sports Medicine, VieCuri Medical Center, Venlo, The Netherlands

⁴Child Development and Exercise Center, Wilhelmina Children’s Hospital, University Medical Center Utrecht, Utrecht, The Netherlands

⁵Department of Epidemiology, VieCuri Medical Center, Venlo, The Netherlands

⁶Department of Nutrition and Movement Sciences, Nutrition and Translational Research in Metabolism (NUTRIM), Faculty of Health, Medicine and Life Sciences, Maastricht University, Maastricht, The Netherlands

⁷Department of Surgery, Nutrition and Translational Research in Metabolism (NUTRIM), Faculty of Health, Medicine and Life Sciences, Maastricht University, Maastricht, The Netherlands

⁸University of Hertfordshire, UNITED KINGDOM

Competing Interests: The authors have declared that no competing interests exist.

^✉

* E-mail: rfranssen@viecuri.nl

Roles

Ruud F W Franssen: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Visualization, Writing – original draft, Writing – review & editing

Bart H E Sanders: Data curation, Formal analysis, Methodology, Project administration, Writing – original draft, Writing – review & editing

Tim Takken: Conceptualization, Methodology, Visualization, Writing – review & editing

F Jeroen Vogelaar: Conceptualization, Funding acquisition, Methodology, Supervision, Writing – review & editing

Maryska L G Janssen-Heijnen: Conceptualization, Formal analysis, Funding acquisition, Methodology, Supervision, Writing – review & editing

Bart C Bongers: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Visualization, Writing – original draft, Writing – review & editing

Lindsay Bottoms: Editor

PMCID: PMC10019694 PMID: 36928094

Abstract

Introduction

Patients with a low cardiorespiratory fitness (CRF) undergoing colorectal cancer surgery have a high risk for postoperative complications. Cardiopulmonary exercise testing (CPET) to assess CRF is the gold standard for preoperative risk assessment. To aid interpretation of raw breath-by-breath data, different methods of data-averaging can be applied. This study aimed to investigate the influence of different data-averaging intervals on CPET variables used for preoperative risk assessment, as well as to evaluate whether different data-averaging intervals influence preoperative risk assessment.

Methods

A total of 21 preoperative CPETs were interpreted by two exercise physiologists using stationary time-based data-averaging intervals of 10, 20, and 30 seconds and rolling average intervals of 3 and 7 breaths. Mean values of CPET variables between different data averaging intervals were compared using repeated measures ANOVA. The variables of interest were oxygen uptake at peak exercise (VO_2peak), oxygen uptake at the ventilatory anaerobic threshold (VO_2VAT), oxygen uptake efficiency slope (OUES), the ventilatory equivalent for carbon dioxide at the ventilatory anaerobic threshold (VE/VCO_2VAT), and the slope of the relationship between the minute ventilation and carbon dioxide production (VE/VCO₂-slope).

Results

Between data-averaging intervals, no statistically significant differences were found in the mean values of CPET variables except for the ventilatory equivalent for carbon dioxide at the ventilatory anaerobic threshold (P = 0.001). No statistically significant differences were found in the proportion of patients classified as high or low risk regardless of the used data-averaging interval.

Conclusion

There appears to be no significant or clinically relevant influence of the evaluated data-averaging intervals on the mean values of CPET outcomes used for preoperative risk assessment. Clinicians may choose a data-averaging interval that is appropriate for optimal interpretation and data visualization of the preoperative CPET. Nevertheless, caution should be taken as the chosen data-averaging interval might lead to substantial within-patient variation for individual patients.

Clinical trial registration

Prospectively registered at ClinicalTrials.gov (NCT05353127).

Introduction

Preoperative aerobic fitness is independently associated with postoperative outcomes following major abdominal surgery [1]. Consequently, cardiopulmonary exercise testing (CPET) is increasingly used within multimodal preoperative risk assessment [2], as it provides an objective, non-invasive, and accurate evaluation of a patient’s aerobic fitness that represents the capacity to meet the increased oxygen demand following major abdominal surgery [3, 4]. The advantage of CPET over other risk assessment tools is that CPET encompasses an integrative evaluation of the cardiovascular, pulmonary, and muscular system [5]. In addition, CPET can be used to inform collaborative decision-making, to optimize comorbidities, to triage perioperative care (e.g., ward, intensive care), to advice on preoperative physical exercise training (e.g., risk assessment, contraindications), and to guide and personalize subsequent physical exercise training prescription [6].

During CPET, a patient exercises against a progressively increasing work rate until volitional exhaustion, while breath-by-breath respiratory gasses are analyzed. The large number of data-points that are collected by the breath-by-breath sampling rate can be a challenge for data visualization, as the signal can have high variability. Therefore, data-averaging is performed to optimize graphical data display and to aid CPET interpretation (see Fig 1). Although it is generally accepted that data-averaging methods influence the numerical value of CPET-derived variables, there is no consensus among existing guidelines on the best averaging method [7].

Fig 1 — Visualization of the plot with oxygen uptake (VO₂) and carbon dioxide production (VCO₂) over time without data-averaging (graph A) and using the five different data-averaging intervals: a stationary time-based average of 10 seconds (graph B), 20 seconds (graph C), and 30 seconds (graph D), a rolling average interval of 3 breaths (graph E) and 7 breaths (graph F) in patient 21. See S1 File for a graphical display of the Wasserman plots of patient 21 with the different data-averaging intervals. Note that the number of data points is lower when stationary time-based averaging is used (and decreasing with longer data-averaging intervals) compared to when a rolling average is used. In addition, a lower number of data points leads to smoothing of the VO₂ and VCO₂ curves. Abbreviations: VAT = ventilatory anearobic threshold; VCO₂ = carbon dioxide production; VO₂ = oxygen uptake; VO_2peak = oxygen uptake at peak exercise. Vertical grey dotted lines represent start of the warm-up phase (W), test phase (T), and recovery phase (R).

The most frequently used CPET-derived variables that are associated with postoperative complications in the current literature are the oxygen uptake at peak exercise (VO_2peak), the oxygen uptake at the ventilatory anaerobic threshold (VO_2VAT) [2, 6, 8], and the ventilatory equivalent for carbon dioxide at the ventilatory anaerobic threshold (VE/VCO_2VAT) [9]. Measures that are less frequently used are the slope of the relationship between the minute ventilation and carbon dioxide production (VE/VCO₂-slope), that can be used as an alternative for the VE/VCO_2VAT if the VAT is undeterminable [8], and the oxygen uptake efficiency slope (OUES) [10].

Although preoperative risk assessment should be multimodal, CPET-derived thresholds are often used to recognize patients with a low aerobic fitness who have a high risk for adverse surgical outcomes. In major abdominal surgery, often used thresholds to identify patients at high-risk for postoperative complications are a VO2peak <18.2 mL/kg/min and/or a VO2VAT <11.1 mL/kg/min [9]. Studies in healthy individuals have shown that the numerical value of the VO_2peak can differ as much as ~10% depending on the data-averaging method [11–13], indicating that data-averaging might significantly influence threshold determination and subsequently might affect preoperative risk assessment. To date however, there are no studies quantifying the extent to which differences in data-averaging influence the numerical value of preoperative CPET-derived variables such as VO_2peak, VO_2VAT, OUES, VE/VCO_2VAT, and VE/VCO₂-slope. Therefore, the primary aim of this study was to investigate the influence of different CPET data-averaging intervals on the numerical values of CPET-derived variables used for preoperative risk assessment in patients scheduled for elective colorectal cancer surgery. The secondary aim was to elucidate the impact of data-averaging intervals on the classification of patients into low or high risk for postoperative complications based on known risk assessment thresholds.

Methods

This observational cross-sectional study was performed at the VieCuri Medical Center, a large teaching hospital in Venlo, the Netherlands. The current study was executed as a secondary analysis of data collected in a study [14] that was approved by the Medical Ethics Review Committee–Zuyderland/Zuyd (Heerlen, the Netherlands) under reference number METCZ20190150. Reporting was done using the STROBE guidelines for reporting of cross-sectional studies [15]. The study protocol was prospectively registered at ClinicalTrials.gov (NCT05353127).

Participants

Data from consecutive patients considered for colorectal cancer surgery who were ≥18 years of age, had a score ≤7 metabolic equivalents of task on the veterans-specific activity questionnaire, and therefore performed preoperative CPET as a part of a tele-prehabilitation study [14], were collected between July 2020 and September 2021. All patients signed informed consent. Preoperative CPET was conducted after diagnosis and before any intervention or treatment was initiated.

Preoperative cardiopulmonary exercise testing

Patients preoperatively performed incremental CPET up to volitional exertion in upright position on an electronically-braked cycle ergometer (Lode Corival, Lode BV, Groningen, the Netherlands). Prior to the test, patients were asked to refrain from vigorous physical activity, caffeine, and tobacco for 24 hours and meals for 2 hours, but to continue medication as usual. Seat height was adjusted to the participant’s leg length. Before commencing CPET, forced vital capacity and forced expiratory volume in one second was obtained from maximal flow-volume curves (Ergostik, Geratherm Respiratory, Bad Kissingen, Germany) according to ATS/ERS standards [5]. Subsequently, baseline cardiopulmonary values were assessed during a three-minute rest period while seated at the cycle ergometer, thereafter a three-minute warm-up phase took place that consisted of unloaded cycling. After the warm-up, work rate was increased by constant increments of 5, 10, 15, 20, or 25 W/min in a ramp-like manner, depending on the subject’s estimated physical fitness level and aimed at reaching a maximal effort within eight to twelve minutes. Throughout CPET, subjects maintained a pedaling frequency between 60 and 80 revolutions/min. The protocol continued until the patient’s pedaling frequency fell definitely below 60 revolutions/min, despite strong verbal encouragement, or when the patient met the criteria for exercise termination before symptom limitation as proposed in the ATS/ACCP statement on cardiopulmonary exercise testing [5].

During CPET, subjects breathed through a facemask (Hans Rudolph, Kansas City, MO, USA) connected to an ergospirometry system (Ergostik, Geratherm Respiratory, Bad Kissingen, Germany). Before every test, calibration for respiratory gas analysis measurements (ambient air and a gas mixture of 16% oxygen and 5% carbon dioxide) and volume measurements (three-liter syringe) took place. Expired gas passed through a flow meter (triple V volume transducer), an oxygen analyzer, and a carbon dioxide analyzer. The flow meter and gas analyzers were connected to a computer that calculated breath-by-breath minute ventilation, oxygen uptake, carbon dioxide production, and the respiratory exchange ratio. Raw unfiltered breath-by-breath data was retrogradely averaged over five different data display intervals.

Procedures

Preoperative CPET patient data was anonymized and patient characteristics other than anthropometric measures were concealed. A medical and clinical exercise physiologist (BB) and a clinical exercise physiologist (RF) determined VO_2peak, VO_2VAT, OUES, VE/VCO_2VAT, and the VE/VCO₂-slope in all CPETs by means of a predefined set of guidelines (see S2 File). A VO_2peak was conceived “valid” when objective criteria for maximal volitional exertion were reached defined as an RER ≥1.10 or reaching ≥95% of the predicted maximal heart rate at peak exercise. CPET interpretation was performed using Blue Cherry software version 1.3.3.3 (Geratherm Respiratory GmbH, Bad Kissingen, Germany), in which observers interpreted the CPET data together using TeamViewer software (TeamViewer GmbH, Göppingen, Germany). Final determination was based on consensus between the two observers. If the two observers were unable to reach consensus, a third observer (TT) was consulted. Data-averaging-intervals used were stationary time-based averages, calculated by averaging the breath-by-breath data over 10, 20, or 30 seconds and rolling averages calculated by averaging a fixed number of single breath measurements (i.e., 3 and 7), then discarding the first breath and adding a new breath to obtain a new breath averaging block. Determination of the aforementioned CPET variables was repeated for all five different data-averaging intervals.

Apart from the CPET data, the preoperative patient characteristics age, sex, body mass index, smoking status (never, former, current), age-adjusted Charlson comorbidity index, American Society of Anesthesiologists classification, veterans-specific activity questionnaire score, hemoglobin levels (mmol/L), and tumor location were recorded to characterize the study population.

Sample size

A sample size calculation was performed with G*Power [16] for F-test repeated measures within factors. Based on a mean ± standard deviation (SD) value for VO_2VAT of 9.7 ± 2.3 mL/kg/min (based on preliminary analysis of the used data) for a mean difference between data-averaging methods of minimally 0.7 mL/kg/min, the estimated effect size is estimated at ~0.30. With an α of 0.05 and a β of 0.80, a minimum of 15 CPETs are needed to detect the estimated effect size.

Statistical analysis

Continuous data were checked for normality using the Shapiro-Wilks test. To assess the difference between different CPET data-averaging intervals, differences in mean numerical values of VO_2peak, VO_2VAT, OUES, VE/VCO_2VAT, and the VE/VCO₂-slope, between different data-averaging intervals were calculated and analyzed by means of within-factors repeated-measures analysis of variance (ANOVA). In case of a statistically significant difference between methods (P<0.05), post-hoc testing was performed using the Bonferroni correction to identify exact differences. Effect sizes were estimated by calculating the eta squared (i.e., sum of squares of the effect divided by the total sum of squares). To evaluate the influence of data-averaging intervals on preoperative risk assessment, individual numerical values for VO_2peak, VO_2VAT, OUES, and VE/VCO_2VAT were compared with known preoperative risk assessment thresholds. Patients were classified as high-risk when having a VO_2peak <18.2 mL/kg/min [9], VO_2VAT <11.1 mL/kg/min [9], OUES/kg <20.6 [10], and/or VE/VCO_2VAT >30.9 [9]. Cochrane’s Q-test was used to determine whether differences in preoperative risk assessment exist between data-averaging methods. Differences between data-averaging methods were assumed statistically significant when P<0.05.

Results

A total of 21 CPETs of patients with colorectal cancer (see Table 1 for patient characteristics) were re-assessed using five different data-averaging intervals. Thus, a total of 105 CPETs (five data-averaging intervals × 21 CPETs) were evaluated. Mean ± SD duration of the CPET ramp phase was 586 ± 174 seconds (9:46 ± 2:54 min). A valid VO_2peak was reached in 70 (67.7%) of the evaluated CPETs. VO_2VAT and VE/VCO_2VAT were determinable in 104 out of 105 CPETs (99%). The OUES and VE-VCO₂-slope were determinable in all 105 CPETs.

Table 1. Baseline characteristics of subjects.

Characteristics	n = 21
Age (years)	70.5 ± 12.5
Sex ratio (male; female)	12 (57%); 9 (43%)
Body mass index (kg/m²)	28.6 ± 4.9
Age-adjusted Charlson comorbidity index
≤3	10 (47.6%)
4–5	10 (47.6%)
6+	1 (4.8%)
ASA-classification
I	4 (19.0%)
II	7 (33.3%)
III	9 (42.9%)
IV	1 (4.8%)
Hemoglobin level (mmol/L)	7.4 ± 1.2
Tumor location
Colon	15 (71.4%)
Rectum	6 (28.6%)

Open in a new tab

Data are presented as mean ± standard deviation (SD) or as number (%).

Mean values of the CPET-derived variables ranged from 14.5 mL/kg/min to 14.6 mL/kg/min for VO_2peak, from 9.3 mL/kg/min to 9.7 mL/kg/min for VO_2VAT, from 19.1 to 19.4 for OUES/kg, from 31.2 to 31.9 for VE/VCO_2VAT, and from 33.6 to 35.3 for VE/VCO₂-slope, dependent on the different data-averaging intervals. There was a significant difference in mean values of VO_2peak between groups with different data averaging intervals, but this difference did not remain significant after post-hoc testing. For the variable VE/VCO_2VAT, the 3 breaths rolling average interval was statistically significant different from the time-based 20 seconds (P = 0.004) and 30 seconds (P = 0.005) data-averaging interval, as well as from the rolling average of 7 breaths (P = 0.021; see Table 2). The effect sizes for all variables were ≤0.009.

Table 2. Numerical values of CPET variables using different data-averaging intervals.

	Data-averaging interval
	Stationary time-based average			Rolling average		P-value^a
	10 seconds	20 seconds	30 seconds	3 breaths	7 breaths
VO_2peak (mL/min)	1202 (1008–1396)	1194 (999–1389)	1193 (997–1390)	1201 (1008–1394)	1200 (1005–1396)	0.040 ^c
VO_2peak (mL/kg/min)	14.6 (12.5–16.7)	14.5 (12.4–16.6)	14.5 (12.4–16.6)	14.6 (12.5–16.7)	14.6 (12.5–16.7)	0.012 ^c
Valid VO_2peak (mL/kg/min)^b	16.2 (13.7–18.7)	16.2 (13.6–18.7)	16.1 (13.5–18.8)	16.2 (13.6–18.8)	16.3 (13.7–18.7)	0.104
VO_2VAT (mL/min)	800 (684–916)	775 (656–895)	764 (669–859)	776 (668–884)	761 (669–851)	0.345
VO_2VAT (mL/kg/min)	9.7 (8.5–10.9)	9.4 (8.1–10.7)	9.3 (8.2–10.4)	9.5 (8.3–10.6)	9.3 (8.2–10.4)	0.435
OUES	1559 (1322–1795)	1559 (1338–1779)	1565 (1338–1792)	1582 (1354–1809)	1574 (1351–1798)	0.463
OUES/kg	19.1 (16.6–21.7)	19.2 (16.7–21.7)	19.3 (16.7–21.8)	19.5 (17.0–21.9)	19.4 (16.9–21.8)	0.479
VE/VCO_2VAT	34.2 (31.9–36.5)	34.6 (32.3–37.0)^d	35.1 (32.6–37.5)^d	33.6 (31.4–35.8)^d	34.4 (32.0–36.8)^d	0.001
VE/VCO₂-slope	31.4 (28.6–34.1)	31.8 (28.8–35.0)	31.2 (28.4–34.0)	31.8 (28.9–34.7)	31.8 (29.0–34.7)	0.608

Open in a new tab

Data are presented as mean and 95% confidence interval (CI), unless stated otherwise.

Abbreviations: OUES = oxygen uptake efficiency slope; VE/VCO₂-slope = the slope of the relationship between the minute ventilation and carbon dioxide production; VE/VCO_2VAT = ventilatory equivalent for carbon dioxide at the ventilatory anaerobic threshold; VO_2peak = oxygen uptake at peak exercise; VO_2VAT = oxygen uptake at the ventilatory anaerobic threshold.

^a: as determined by repeated-measures ANOVA (within factors).

^b: as determined by a respiratory exchange ratio at peak exercise ≥1.10 and/or a heart rate at peak exercise ≥95% of the predicted maximal heart rate based on the formula 208 –(0.8 × age in years).

^c: did not remain significant after post hoc testing with Bonferroni correction.

^d: the 3 breaths rolling average interval was statistically significant different from the stationary time-based interval of 20 seconds (P = 0.004) and 30 seconds (P = 0.005), as well as from the 7 breaths rolling average interval (P = 0.021).

Fig 2 depicts within-patient variation in the numerical value of several CPET-derived variables using the five different data-averaging intervals. Although the numerical values for VO_2peak were consistent (maximal within patient difference, 0.4 mL/kg/min, or 5.6%), within patient variation could be as much as 4.0 mL/kg/min for VO_2VAT (40.8%), 5.7 for the OUES/kg (40.3%), 4.7 for VE/VCO_2VAT (13.4%), and 10.4 (37.3%) for VE/VCO₂-slope when using different data-averaging intervals (see Fig 2).

When dichotomizing patients into the high or low risk category for postoperative complications based on the numerical values of each CPET variable, the proportion of patients with a high risk based on their VO_2peak ranged from 76% to 81%, depending on the used data-averaging interval. Based on VO_2VAT the proportion of high-risk patients ranged from 67% to 76%, whereas this ranged from 57% to 67% for OUES/kg and from 76% to 86% for VE/VCO_2VAT. As depicted in Table 3, there were no statistically significant differences in the proportion of patients who were classified as at high risk between different data-averaging-intervals. As depicted in Fig 2, individual values of some patients crossed the risk threshold depending on the data-averaging interval that was used. Based on within-patient variation, the estimated risk could differ for 1 patient when based on VO_2peak (patient 15), for 5 patients based on VO_2VAT (patients 1, 6, 7, 12, and 19), for 2 patients based on OUES (patients 2 and 18), and for 4 patients based on VE/VCO_2VAT (patients 3, 7, 13, and 16), depending on the used data-averaging interval.

Table 3. Effect of different data-averaging intervals on classifying patients as having a high-risk for postoperative complications.

	Data-averaging interval
	Stationary time-based average			Rolling average
	10 seconds, n (%)	20 seconds n (%)	30 seconds (%) n (%)	3 breaths n (%)	7 breaths n (%)	P-value^a
VO_2peak	17 (81%)	17 (81%)	17 (81%)	17 (81%)	16 (76%)	0.406
VO_2VAT	16 (76%)	16 (76%)	14 (67%)	14 (67%)	16 (76%)	0.615
OUES/kg	13 (62%)	13 (62%)	12 (57%)	14 (67%)	14 (67%)	0.231
VE/VCO_2VAT	16 (76%)	18 (86%)	17 (81%)	15 (71%)	16 (76%)	0.334

Open in a new tab

Data are presented as number (%).

Abbreviations: OUES = oxygen uptake efficiency slope; VE/VCO_2VAT = ventilatory equivalent for carbon dioxide at the ventilatory anaerobic threshold; VO_2peak = oxygen uptake at peak exercise; VO_2VAT = oxygen uptake at the ventilatory anaerobic threshold.

^a: determined by Cochrane’s Q-test.

Discussion

To our knowledge, the current study was the first study that aimed to investigate whether the selection of different CPET data-averaging intervals would translate into differences in mean values of CPET-derived variables in patients with colorectal cancer who performed CPET for preoperative risk assessment. As CPET-derived variables are used to preoperatively classify patients into having a low or high risk for postoperative complications based on their CRF, the secondary aim of the current study was to investigate whether potential differences in the numerical values of CPET-derived variables would lead to differences in preoperative risk classification. Based on the mean values of the CPET-derived variables there were only statistically significant differences for the variables VO_2peak and VE/VCO_2VAT between different data-averaging intervals. For VO_2peak, the between-group difference did not remain significant after post-hoc analysis, whereas data-averaging group differences VE/VCO_2VAT were statistically significant between the 3 breaths moving average and the 20- and 30-second time-based interval, as well as the 7 breaths moving average.

For VO_2peak, the greatest observed difference between data-averaging groups was 0.1 mL/kg/min. Given that the coefficient of variation (a measure of reproducibility) for VO_2peak is estimated to be between ~5% and ~9% [5] (i.e., between ~0.7 mL/kg/min and ~1.3 mL/kg/min based on mean values of VO_2peak in the current study), the observed maximal difference of 0.1 mL/kg/min is not clinically relevant. The observation that this small difference in VO_2peak is not clinically relevant is further emphasized by the fact that no differences were found between the proportion of patients who were classified as low or high risk based on VO_2peak when using different data-averaging intervals in the current study. Provided that the critical difference of VE/VCO_2VAT in patients with colorectal cancer is assumed to be ~10% [17], the maximal mean difference of 1.5 (5%) measured in the current study is not deemed clinically relevant. The observation that differences in the mean values of the VE/VCO₂-slope are not clinically relevant is also supported by the very small effect size (0.009).

The main purpose of using data-averaging of CPET data is to reduce noise of breath-by-breath fluctuations and to aid CPET interpretation [5]. In the current study there seem to be no clinically relevant differences in CPET-derived variables between different data-averaging intervals. This is a reassuring observation that opens possibilities to be flexible in the use of data-averaging intervals as long as the interval is within certain boundaries. That is, the type and duration of CPET can be taken into consideration when determining the optimal data-averaging interval [7]. For example, using longer averaging intervals in longer tests, or using a rolling average for noisy data. On the other hand, longer intervals might mask dynamic pathophysiological processes such oscillatory breathing. In these circumstances shorter time-based intervals might be optimal [7]. For preoperative exercise testing a stationary time-based average of 10 seconds, or a breath-based rolling average of 3 or 7 seconds might provide a good trade-off between de number of data-points and the duration of the test.

Although the literature is scarce with regard to the influence of data-averaging intervals on the determination of CPET-derived variables (and only available for VO_2peak), results of the current study are in line with a previous publication in which the effect of data-averaging intervals on VO_2peak in 22 healthy athletic subjects was investigated [18]. The authors found that only a stationary time-based data-averaging interval of 60 seconds was significantly different from all other data-averaging intervals (10, 15, 20, and 30 seconds) [18]. In a study evaluating VO_2peak values of 15 patients who were screened for heart transplant surgery (with comparable mean VO_2peak values as observed in the current study), no significant differences were found between stationary time-based data-averaging intervals of 15 and 30 seconds, and a 8 breaths rolling average interval [19]. Moreover, only a 60-second stationary time-based data-averaging interval was statistically significantly different from the aforementioned data-averaging intervals [19]. These long data-averaging intervals (of 60 seconds or more) are probably not used very often in preoperative CPETs and are not recommend by current preoperative CPET guidelines [8].

Based on the results of this study, the recommendation in the preoperative CPET guideline to use a breath-based data-averaging interval of 3–5 breaths or a time-based data-averaging interval of ~20 seconds seems plausible when evaluating the mean (group level) values. Nevertheless, caution should be taken when evaluating individual patients, as different data-averaging intervals caused substantial variation in the numerical values of CPET-derived variables within patients. As depicted in Fig 2, individual values of patients could differ as much as ~40%. In individual patients, the chosen data-averaging interval could induce a shift of that patient from low to high risk or vice versa. This is an important observation, as risk assessment could influence surgical planning for individual patients (e.g., enrollment in prehabilitation program, referring to a higher care unit postoperatively) and the shared decision-making process. It is recognized that preoperative risk assessment is not solely based on risk thresholds determined by CPET, but rather consists of a composite assessment, taking into account the full CPET in combination with other preoperative risk factors such as, but not limited to, malnutrition, comorbidities, and geriatric status. Nevertheless, the influence of the data-averaging interval could be taken into consideration, especially in patients in which the CPET values are close to the risk classification cut-off point. In addition, instead of rigid cut-off points inducing black and white risk assessment, grey zones (intermediate risk) could be introduced to account for individual differences [17].

A limitation of the current study was that VO_2peak was determined over a ~30 second interval [5] regardless of the data-averaging interval that was used. The use of the fixed 30 second interval might have masked some of the variability caused by the data-averaging interval, explaining the very small differences of VO_2peak values between data averaging intervals. A strength of the current study is that variation other than variation coming from the data-averaging interval was minimized. Firstly, by repeating interpretation of the 21 CPETs that were retrospectively formatted using 5 different data-averaging intervals, as opposed to repeated testing of patients with different data averaging intervals. Secondly, to account for inter-observer variability, CPET interpretation was done by two clinical exercise physiologists, based on consensus, and by using a predefined set of guidelines (see S2 File). By doing so, the observed variation between groups of data-averaging intervals was exclusively caused by the used data-averaging interval and not by within-patient biological variation, measurement error, or inter-observer variability.

The current study opens possibilities for clinicians to be flexible in the data-averaging interval that is used for interpretation of the preoperative CPET. Current CPET literature does not provide clear and consistent guidance for clinicians about the choice of a data-averaging interval [7, 8]. As different (patho)physiological patterns might require different data-visualization, future research could focus on investigating optimal data-visualization methods that best fit the aim of the CPET, the properties of the CPET, and the (patho)physiological process the clinician is willing to evaluate.

Conclusion

On a group level there appear to be no clinically relevant differences in the mean values of VO_2peak, VO_2VAT, OUES, VE/VCO_2VAT, and VE/VCO₂-slope between different data-averaging intervals used for interpretation of preoperative CPET in patients with colorectal cancer. In addition, the choice of data-averaging interval does not influence the proportion of patients classified as high or low risk for complications based on their exercise tolerance. Nevertheless, the chosen data-averaging interval might lead to substantial within patient variation for individual patients and should therefore be considered in patients in which the CPET values are close to the risk classification cut-off point.

Supporting information

S1 File. Graphical display of the Wasserman plots of patient 21.

(PDF)

Click here for additional data file.^{(1.1MB, pdf)}

S2 File. Guideline for systematic interpretation of preoperative cardiopulmonary exercise testing.

(PDF)

Click here for additional data file.^{(1.2MB, pdf)}

S1 Dataset

(XLSX)

Click here for additional data file.^{(13.2KB, xlsx)}

Acknowledgments

The authors like to thank Accuramed BVBA (Halen, Belgium) for providing a free unrestricted copy of the Blue Cherry software to support this study.

Data Availability

All relevant data are within the paper and its Supporting Information files.

Funding Statement

This study was unconditionally financed by the Research and Innovation fund VieCuri Medical Center under reference number E.21.32.004-2. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Moran J, Wilson F, Guinan E, McCormick P, Hussey J, Moriarty J. Role of cardiopulmonary exercise testing as a risk-assessment method in patients undergoing intra-abdominal surgery: a systematic review. Br J Anaesth. 2016;116(2):177–91. doi: 10.1093/bja/aev454 [DOI] [PubMed] [Google Scholar]
2.Steffens D, Ismail H, Denehy L, Beckenkamp PR, Solomon M, Koh C, et al. Preoperative Cardiopulmonary Exercise Test Associated with Postoperative Outcomes in Patients Undergoing Cancer Surgery: A Systematic Review and Meta-Analyses. Ann Surg Oncol. 2021;28(12):7120–46. doi: 10.1245/s10434-021-10251-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Older P, Smith R, Courtney P, Hone R. Preoperative evaluation of cardiac failure and ischemia in elderly patients by cardiopulmonary exercise testing. Chest. 1993;104(3):701–4. doi: 10.1378/chest.104.3.701 [DOI] [PubMed] [Google Scholar]
4.Bongers BC, Dejong CHC, den Dulk M. Enhanced recovery after surgery programmes in older patients undergoing hepatopancreatobiliary surgery: what benefits might prehabilitation have? Eur J Surg Oncol. 2021;47(3 Pt A):551–9. doi: 10.1016/j.ejso.2020.03.211 [DOI] [PubMed] [Google Scholar]
5.American Thoracic S, American College of Chest P. ATS/ACCP Statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med. 2003;167(2):211–77. [DOI] [PubMed] [Google Scholar]
6.Otto JM, Levett DZH, Grocott MPW. Cardiopulmonary Exercise Testing for Preoperative Evaluation: What Does the Future Hold? Current Anesthesiology Reports. 2020;10(1):1–11. [Google Scholar]
7.Dumitrescu D, Rosenkranz S. Graphical Data Display for Clinical Cardiopulmonary Exercise Testing. Ann Am Thorac Soc. 2017;14(Supplement_1):S12–S21. doi: 10.1513/AnnalsATS.201612-955FR [DOI] [PubMed] [Google Scholar]
8.Levett DZH, Jack S, Swart M, Carlisle J, Wilson J, Snowden C, et al. Perioperative cardiopulmonary exercise testing (CPET): consensus clinical guidelines on indications, organization, conduct, and physiological interpretation. Br J Anaesth. 2018;120(3):484–500. doi: 10.1016/j.bja.2017.10.020 [DOI] [PubMed] [Google Scholar]
9.West MA, Asher R, Browning M, Minto G, Swart M, Richardson K, et al. Validation of preoperative cardiopulmonary exercise testing-derived variables to predict in-hospital morbidity after major colorectal surgery. Br J Surg. 2016;103(6):744–52. doi: 10.1002/bjs.10112 [DOI] [PubMed] [Google Scholar]
10.Bongers BC, Berkel AE, Klaase JM, van Meeteren NL. An evaluation of the validity of the pre-operative oxygen uptake efficiency slope as an indicator of cardiorespiratory fitness in elderly patients scheduled for major colorectal surgery. Anaesthesia. 2017;72(10):1206–16. doi: 10.1111/anae.14003 [DOI] [PubMed] [Google Scholar]
11.Dideriksen K, Mikkelsen UR. Reproducibility of incremental maximal cycle ergometer tests in healthy recreationally active subjects. Clin Physiol Funct Imaging. 2017;37(2):173–82. doi: 10.1111/cpf.12283 [DOI] [PubMed] [Google Scholar]
12.Martin-Rincon M, Calbet JAL. Progress Update and Challenges on V. O2max Testing and Interpretation. Front Physiol. 2020;11:1070. doi: 10.3389/fphys.2020.01070 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Martin-Rincon M, Gonzalez-Henriquez JJ, Losa-Reyna J, Perez-Suarez I, Ponce-Gonzalez JG, de La Calle-Herrero J, et al. Impact of data averaging strategies on VO2max assessment: Mathematical modeling and reliability. Scand J Med Sci Sports. 2019;29(10):1473–88. [DOI] [PubMed] [Google Scholar]
14.Franssen RFW, Bongers BC, Vogelaar FJ, Janssen-Heijnen MLG. Feasibility of a tele-prehabilitation program in high-risk patients with colon or rectal cancer undergoing elective surgery: a feasibility study. Perioper Med (Lond). 2022;11(1):28. doi: 10.1186/s13741-022-00260-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335(7624):806–8. doi: 10.1136/bmj.39335.541782.AD [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39(2):175–91. doi: 10.3758/bf03193146 [DOI] [PubMed] [Google Scholar]
17.Rose GA, Davies RG, Davison GW, Adams RA, Williams IM, Lewis MH, et al. The cardiopulmonary exercise test grey zone; optimising fitness stratification by application of critical difference. Br J Anaesth. 2018;120(6):1187–94. doi: 10.1016/j.bja.2018.02.062 [DOI] [PubMed] [Google Scholar]
18.Midgley AW, McNaughton LR, Carroll S. Effect of the VO2 time-averaging interval on the reproducibility of VO2max in healthy athletic subjects. Clin Physiol Funct Imaging. 2007;27(2):122–5. doi: 10.1111/j.1475-097X.2007.00725.x [DOI] [PubMed] [Google Scholar]
19.Johnson JS, Carlson JJ, VanderLaan RL, Langholz DE. Effects of sampling interval on peak oxygen consumption in patients evaluated for heart transplantation. Chest. 1998;113(3):816–9. doi: 10.1378/chest.113.3.816 [DOI] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0283129.r001

Decision Letter 0

Sampath Kumar Amaravadi

8 Nov 2022

PONE-D-22-28377Influence of different data-averaging methods on preoperative risk assessment using cardiopulmonary exercise testing in patients scheduled for colorectal surgeryPLOS ONE

Dear Dr. franssen,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Dec 23 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Sampath Kumar Amaravadi, Ph.D

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following financial disclosure:

“This study was unconditionally financed by the Research and Innovation fund VieCuri Medical Center under reference number E.21.32.004-2.”

Please state what role the funders took in the study. If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

3. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

4. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please delete it from any other section.

5. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

Dear Author,

Kindly revise the manuscript with reviewer comments and submit the same

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Interesting substudy of an approved study. appropriately conducted analysis

A few points to address :

The authors write as if assuming a degree of precision to interpretation, and hard differentation into low or high risk categories which isn`t typical of real life practice .

The authors have referenced the Rose “Grey Zone” paper

Rose GA, Davies RG, Davison GW, Adams RA, Williams IM, Lewis MH, et al. The 434 cardiopulmonary exercise test grey zone; optimising fitness stratification by application of critical 435 difference. Br J Anaesth. 2018;120(6):1187-94

…which makes this point well . The authors of the current study similarly make it clear in their conclusion that any apparent differences in values due to averaging are not clinically relevant . Important that they convey this clearly in the abstract

Patients are generally not classified in a binary fashion as either high or low risk based on whether they are above or below strictly defined single threshold limits / Overall summary of risk tends to be a composite assessment taking into account many CPET measurements , & comorbidities.

The CPET is not applied as a pass or fail test . CPET derived Individualised Risk estimates are used to add weight to shared decision making about whether to proceed to surgery , and to help determine perioperative care pathway ( eg High Dependency Care versus ward based care pathway ) . It is worth covering this point about clinical application in the discussion

Section by section

Title : “ Influence of different data-averaging methods on preoperative risk assessment using

cardiopulmonary exercise testing in patients scheduled for colorectal surgery”. I think this title is over stating the impact of the study findings. More appropriate to say “ Influence of different data-averaging methods on mean values of selected variables derived from cardiopulmonary exercise testing in patients scheduled for colorectal surgery”

Abstract could define exactly which CPET derived variables were considered in the study – these are defined on p11 , para 2

Methods – clear

Results clear

Table 2 is really good to demonstrate the results

Figure 1 is a graphical representation of averaging – this is a powerful figure to illustrate the point that regardless of averaging method – the thresholds of interest are generally consistent

Figure 2 similarly is powerful to demonstrate the spread of individual patients` values dependent on the method of averaging

P 11 para 2 – several CPET derived variables defined – however the abbrevations (VO2VAT) and(VE/VCO2VAT) are unusual . I`m accustomed to VO2AT and VE/VCO2AT.

Discussion :

Re-states main findings and discusses implications in context . Reasonable and moderate conclusions

Reviewer #2: This study make nonsense. If the aim is to establish whether the averaging interval of VO2 data has an impact in the VO2max value and the medical decisions adopting depending on this value why are the authors analyzing all results using 30-s averages? Moreover, the metabolic cart data are incomplete, please add PETO2 and PETCO2 data.

Are all metabolic carts used disclosed? Or all test were performed with the same metabolic cart?

Please add as as an example the full output data of one of the incremental exercise test with breath by breath data and then the calculated 10, 20 or 30 s averaged values.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Gary Minto

Reviewer #2: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2023 Mar 16;18(3):e0283129. doi: 10.1371/journal.pone.0283129.r002

Author response to Decision Letter 0

22 Nov 2022

Response to reviewers’ comments:

Reviewer #1: Interesting substudy of an approved study. appropriately conducted analysis

A few points to address :

The authors write as if assuming a degree of precision to interpretation, and hard differentation into low or high risk categories which isn`t typical of real life practice .

The authors have referenced the Rose “Grey Zone” paper

…which makes this point well . The authors of the current study similarly make it clear in their conclusion that any apparent differences in values due to averaging are not clinically relevant. Important that they convey this clearly in the abstract

Authors’ response: We fully agree with the reviewer that preoperative risk assessment should be seen in the broader context of CPET interpretation and patient assessment in general.

We added “or clinically relevant” to the conclusion of the abstract

To avoid any unclarities regarding the point put forward by the reviewer about the conclusion in the abstract, we also adjusted the abstract as follows: The section “the proportion of patients classified as high or low risk for complications.” was deleted. And stated more in general as “perioperative risk assessment.”

Authors’ response: In the manuscript, this point was briefly mentioned in the discussion on Page 21, Line 334 - 335. Nevertheless, we agree that we should elaborate more on this. Therefore, we adjusted this part of the discussion on page 21 Line 335-338 as follows:

Although it is recognized that risk assessment is not solely based on the numerical value of any CPET-derived variable

“It is recognized that preoperative risk assessment is not solely based on risk thresholds determined by CPET, but rather consists of a composite assessment, taking into account the full CPET in combination with other preoperative risk factors such as, but not limited to, malnutrition, comorbidities, and geriatric status.”

Section by section

Title : “ Influence of different data-averaging methods on preoperative risk assessment using

Authors’ response: We adjusted the title according to the suggestion of the reviewer. The new title reads as follows: “Influence of different data-averaging methods on mean values of selected variables derived from preoperative cardiopulmonary exercise testing in patients scheduled for colorectal surgery”

Abstract could define exactly which CPET derived variables were considered in the study – these are defined on p11 , para 2

Methods – clear

Results clear

Table 2 is really good to demonstrate the results

Figure 1 is a graphical representation of averaging – this is a powerful figure to illustrate the point that regardless of averaging method – the thresholds of interest are generally consistent

Figure 2 similarly is powerful to demonstrate the spread of individual patients` values dependent on the method of averaging

Authors’ response: We thank the reviewer for his positive evaluation of the manuscript and figures and the suggestion to define the exact variables in the abstract. We added the following to the methods section of the abstract: “The variables of interest were oxygen uptake at peak exercise (VO2peak), oxygen uptake at the ventilatory anaerobic threshold (VO2VAT), oxygen uptake efficiency slope (OUES), the ventilatory equivalent for carbon dioxide at the ventilatory anaerobic threshold (VE/VCO2VAT), and the slope of the relationship between the minute ventilation and carbon dioxide production (VE/VCO2-slope).”

P 11 para 2 – several CPET derived variables defined – however the abbrevations (VO2VAT) and(VE/VCO2VAT) are unusual . I`m accustomed to VO2AT and VE/VCO2AT.

Authors’ response: Indeed, many different terminologies are used within the literature to define CPET-derived thresholds. We prefer to use the term ventilatory anaerobic threshold (VAT) over anaerobic threshold (AT), as is it makes more clear that the thresholds are derived from respiratory gasses (as opposed to i.e., lactate measurement). Nevertheless, if the reviewer or the editor insist on using AT instead, we are willing to adjust the terminology throughout the manuscript.

Discussion :

Re-states main findings and discusses implications in context . Reasonable and moderate conclusions

Reviewer #2:

This study make nonsense. If the aim is to establish whether the averaging interval of VO2 data has an impact in the VO2max value and the medical decisions adopting depending on this value why are the authors analyzing all results using 30-s averages?

Authors’ response: We thank the reviewer for the critical appraisal of our study but we regret that the reviewer qualifies our study as nonsense. Before we can reflect on the reviewers points, we must emphasize that we did not analyze all results using 30-second data-averaging. In addition, we did not aim to establish whether data-averaging intervals impact VO2peak values alone. Instead, we assessed many more variables (as outlined on page 10) using 5 different data averaging intervals. Therefore, the point the reviewer merely refers to the determination of VO2peak values.

Regarding the determination of VO2peak values, let us explain why we chose to analyze VO2peak values over an interval close to 30-seconds regardless of the used data-averaging method. In most (if not all) software packages used for CPET interpretation, VO2peak is calculated over a period that has to be manually set. That is, instead of using one point that is allocated as VO2peak, a lower limit and upper limit have to be set in order to demark the period over which VO2peak is calculated (see also Figure 1, red shaded area in each graph). With this in mind, the method chosen in the current study mimics how VO2peak is estimated in clinical practice. Nevertheless, we do agree with the reviewer that by doing so, the variability introduced by the data-averaging interval is attenuated as was already discussed on page 21, line 344-347. Nevertheless, this also reflects routine practice.

Moreover, the metabolic cart data are incomplete, please add PETO2 and PETCO2 data.

Authors’ response: We are not sure what the reviewer is referring to as we did not disclose any metabolic cart data. We did however include the raw data on which the analyses are based. Is that what the reviewer is referring to?

Are all metabolic carts used disclosed? Or all test were performed with the same metabolic cart?

Authors’ response: All tests are performed on the same metabolic cart as reported in the methods section in the paragraph “Preoperative cardiopulmonary exercise testing” on page 7 and 8. We included a data file with the raw data that was used for the analysis. We did not include the metabolic cart data as we believe that due to the relative small study sample it would be hard to guarantee anonymity of our participants. Nevertheless, on reasonable request the metabolic cart data could be shared after personal communication with the authors.

Please add as as an example the full output data of one of the incremental exercise test with breath by breath data and then the calculated 10, 20 or 30 s averaged values.

Authors’ response: We added an example of a graphical display of the metabolic cart of patient 21 as Supporting information S3.

Attachment

Submitted filename: Response to the reviewers.docx

Click here for additional data file.^{(24.7KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0283129.r003

Decision Letter 1

Lindsay Bottoms

30 Jan 2023

PONE-D-22-28377R1Influence of different data-averaging methods on mean values of selected variables derived from preoperative cardiopulmonary exercise testing in patients scheduled for colorectal surgery.PLOS ONE

Dear Dr. franssen,

Please go through the comments of the reviewers. One reviewer has asked to change the focus and consider removing the second aim. Do consider the comment and provide a response. Otherwise, please address the comments by reviewer 4.

Please submit your revised manuscript by Mar 16 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Lindsay Bottoms

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #3: (No Response)

Reviewer #4: (No Response)

Reviewer #5: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #3: No

Reviewer #4: Yes

Reviewer #5: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: No

Reviewer #4: Yes

Reviewer #5: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #3: Yes

Reviewer #4: Yes

Reviewer #5: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #3: Yes

Reviewer #4: Yes

Reviewer #5: Yes

**********

6. Review Comments to the Author

Reviewer #1: (No Response)

Reviewer #3: The introduction is not specific to the findings of this study until line 108. Although you mention you have addressed reviewer #1 concerns about the interpretation of CPEX utility, comments like lines 106-107 are misleading. I absolutely agree that threshold values are far from agreed upon across surgical specialties. CPEX is at best a single test that should be interpreted in context. I do not believe that patients would be denied surgery on the result on this single test in many institutions - unless you have data on this. This should be made clearer in your paper, not just in the conclusion, which seems at odds now with the introduction.

I do believe that it is useful to the examine the influence of data-averaging methods for this test, which the authors have done well. This, in my opinion, should be the focus of the paper. I do not believe the secondary aim of this study is worthwhile. I understand the point being made, but clarifying the influence of different data averaging method is sufficient. We can interpret the clinical application ourselves.

I therefore recommend changing the focus of the paper to reflect the above.

Reviewer #4: General comments

The authors of this study addressed a relatively simple question, yet potentially important from a practical perspective, whether the way data from CPET is averaged affects parameters estimated from the CPET in a clinical population.

Overall, the manuscript is well written, and contains all technically relevant information.

I think the authors and previous reviewers have done a thorough job, and thus my comments remain relatively minor. I hope my comments help improve the manuscript.

The main comments (see below) refers to the interpretation of some of the results in the discussion.

Specific comments

Line 28 abstract – please abbreviate cardiorespiratory fitness as CRF.

Line 64-66 – perhaps you can make the point/link (or make this point clearer) that a key goal of prehabilitation is to increase CRF, and increases in CRF have a clinically relevant positive effect on subsequent complications.

Line 100 and throughout the study: I would like see the authors’ view on whether the term ‘anaerobic threshold’ should still be used, even if that’s still common in this field? See for a review https://pubmed.ncbi.nlm.nih.gov/33112439/. The term gas exchange threshold perhaps best captures this first threshold? You also refer in the supplementary material 2 to RCP, but this was not reported in the manuscript. Was this (RCP) only determined for the purpose of determining the VE-VCO2 slope? I fully appreciate the terminology in this field is far from standard and different labs/fields use different terms, I can see the terminology was already addressed in previous reviews, but would like to bring this up again.

Figure 1. This may be an error on my end – but the resolution of the figure is not great. Please check before publication. Same in other figures.

Line 172 – how was disagreement defined? E.g. As a 5%? 10? difference between two initials assessments?

Line 281 – Would it make sense to change from “exercise tolerance” to CRF? I would argue that CRF is then what underpins exercise tolerance – to be ability perform a task without reaching task-failure.

The discussion and conclusion should better reflect some of the results reported, in my view– specifically that “within patient variation could be as much as 4.0 mL/kg/min for VO2VAT (40.8%), 5.7 for the OUES/kg (40.3%), 4.7 for VE/VCO2VAT (13.4%), and 10.4 (37.3%) for VE/VCO2-slope when using different data-averaging intervals (see Figure 2).”

I can see this is addressed in the last part of the discussion, and also mentioned in the conclusion, but I would like to ask if this can be stressed / made more clear? For example, in the abstract you state: "nor does the choice of data-averaging interval influence

perioperative risk assessment", but the results show the opposite (at least at the individual level, which to me is more important than data at the group level in this instance), see lines 263 onwards.

Reviewer #5: My perception is that the study is very well organized in writing and the statistic methods were correctly used. However kindly allow me one point to address, as per below:

- The authors should also include the effect size in the statistics. This will give a more accurate picture of what the findings represents.

Another point to enphasize is that the discussion refers to the main finding of the study, which is very appropriate.

This type of initiative will always be very welcome since post surgical complications can be significantly affected (reduced) by the use of apropriate preoperative procedures, bearing in mind that the aim of this study is to manly investigate the methods which are presently being clinically used.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Gary Minto

Reviewer #3: No

Reviewer #4: Yes: Daniel Muniz-Pumares

Reviewer #5: No

**********

PLoS One. 2023 Mar 16;18(3):e0283129. doi: 10.1371/journal.pone.0283129.r004

Author response to Decision Letter 1

7 Feb 2023

Response to reviewers’ comments:

Reviewer #1: Interesting substudy of an approved study. appropriately conducted analysis

A few points to address :

The authors write as if assuming a degree of precision to interpretation, and hard differentation into low or high risk categories which isn`t typical of real life practice .

The authors have referenced the Rose “Grey Zone” paper

…which makes this point well . The authors of the current study similarly make it clear in their conclusion that any apparent differences in values due to averaging are not clinically relevant. Important that they convey this clearly in the abstract

Authors’ response: We fully agree with the reviewer that preoperative risk assessment should be seen in the broader context of CPET interpretation and patient assessment in general.

We added “or clinically relevant” to the conclusion of the abstract

Although it is recognized that risk assessment is not solely based on the numerical value of any CPET-derived variable

Section by section

Title : “ Influence of different data-averaging methods on preoperative risk assessment using

Abstract could define exactly which CPET derived variables were considered in the study – these are defined on p11 , para 2

Methods – clear

Results clear

Table 2 is really good to demonstrate the results

Figure 1 is a graphical representation of averaging – this is a powerful figure to illustrate the point that regardless of averaging method – the thresholds of interest are generally consistent

Figure 2 similarly is powerful to demonstrate the spread of individual patients` values dependent on the method of averaging

P 11 para 2 – several CPET derived variables defined – however the abbrevations (VO2VAT) and(VE/VCO2VAT) are unusual . I`m accustomed to VO2AT and VE/VCO2AT.

Discussion :

Re-states main findings and discusses implications in context . Reasonable and moderate conclusions

Reviewer #2:

Moreover, the metabolic cart data are incomplete, please add PETO2 and PETCO2 data.

Are all metabolic carts used disclosed? Or all test were performed with the same metabolic cart?

Please add as as an example the full output data of one of the incremental exercise test with breath by breath data and then the calculated 10, 20 or 30 s averaged values.

Authors’ response: We added an example of a graphical display of the metabolic cart of patient 21 as Supporting information S3.

Attachment

Submitted filename: Response to the reviewers.docx

Click here for additional data file.^{(28.7KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0283129.r005

Decision Letter 2

Lindsay Bottoms

3 Mar 2023

Influence of different data-averaging methods on mean values of selected variables derived from preoperative cardiopulmonary exercise testing in patients scheduled for colorectal surgery.

PONE-D-22-28377R2

Dear Dr. franssen,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Lindsay Bottoms

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #3: All comments have been addressed

Reviewer #4: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #3: Yes

Reviewer #4: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: Yes

Reviewer #4: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #3: Yes

Reviewer #4: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #3: Yes

Reviewer #4: Yes

**********

6. Review Comments to the Author

Reviewer #3: Much more focused paper, which addresses a scientific question well. My concerns have been addressed by the authors.

Reviewer #4: I have no further comments.

The authors have addressed my previous points satisfactorily - thank you for that.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: No

Reviewer #4: Yes: Daniel Muniz Pumares

**********

PLoS One. doi: 10.1371/journal.pone.0283129.r006

Acceptance letter

Lindsay Bottoms

8 Mar 2023

PONE-D-22-28377R2

Influence of different data-averaging methods on mean values of selected variables derived from preoperative cardiopulmonary exercise testing in patients scheduled for colorectal surgery

Dear Dr. Franssen:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Lindsay Bottoms

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 File. Graphical display of the Wasserman plots of patient 21.

(PDF)

Click here for additional data file.^{(1.1MB, pdf)}

S2 File. Guideline for systematic interpretation of preoperative cardiopulmonary exercise testing.

(PDF)

Click here for additional data file.^{(1.2MB, pdf)}

S1 Dataset

(XLSX)

Click here for additional data file.^{(13.2KB, xlsx)}

Attachment

Submitted filename: Response to the reviewers.docx

Click here for additional data file.^{(24.7KB, docx)}

Attachment

Submitted filename: Response to the reviewers.docx

Click here for additional data file.^{(28.7KB, docx)}

Data Availability Statement

All relevant data are within the paper and its Supporting Information files.

[pone.0283129.ref001] 1.Moran J, Wilson F, Guinan E, McCormick P, Hussey J, Moriarty J. Role of cardiopulmonary exercise testing as a risk-assessment method in patients undergoing intra-abdominal surgery: a systematic review. Br J Anaesth. 2016;116(2):177–91. doi: 10.1093/bja/aev454 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref002] 2.Steffens D, Ismail H, Denehy L, Beckenkamp PR, Solomon M, Koh C, et al. Preoperative Cardiopulmonary Exercise Test Associated with Postoperative Outcomes in Patients Undergoing Cancer Surgery: A Systematic Review and Meta-Analyses. Ann Surg Oncol. 2021;28(12):7120–46. doi: 10.1245/s10434-021-10251-3 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0283129.ref003] 3.Older P, Smith R, Courtney P, Hone R. Preoperative evaluation of cardiac failure and ischemia in elderly patients by cardiopulmonary exercise testing. Chest. 1993;104(3):701–4. doi: 10.1378/chest.104.3.701 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref004] 4.Bongers BC, Dejong CHC, den Dulk M. Enhanced recovery after surgery programmes in older patients undergoing hepatopancreatobiliary surgery: what benefits might prehabilitation have? Eur J Surg Oncol. 2021;47(3 Pt A):551–9. doi: 10.1016/j.ejso.2020.03.211 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref005] 5.American Thoracic S, American College of Chest P. ATS/ACCP Statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med. 2003;167(2):211–77. [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref006] 6.Otto JM, Levett DZH, Grocott MPW. Cardiopulmonary Exercise Testing for Preoperative Evaluation: What Does the Future Hold? Current Anesthesiology Reports. 2020;10(1):1–11. [Google Scholar]

[pone.0283129.ref007] 7.Dumitrescu D, Rosenkranz S. Graphical Data Display for Clinical Cardiopulmonary Exercise Testing. Ann Am Thorac Soc. 2017;14(Supplement_1):S12–S21. doi: 10.1513/AnnalsATS.201612-955FR [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref008] 8.Levett DZH, Jack S, Swart M, Carlisle J, Wilson J, Snowden C, et al. Perioperative cardiopulmonary exercise testing (CPET): consensus clinical guidelines on indications, organization, conduct, and physiological interpretation. Br J Anaesth. 2018;120(3):484–500. doi: 10.1016/j.bja.2017.10.020 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref009] 9.West MA, Asher R, Browning M, Minto G, Swart M, Richardson K, et al. Validation of preoperative cardiopulmonary exercise testing-derived variables to predict in-hospital morbidity after major colorectal surgery. Br J Surg. 2016;103(6):744–52. doi: 10.1002/bjs.10112 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref010] 10.Bongers BC, Berkel AE, Klaase JM, van Meeteren NL. An evaluation of the validity of the pre-operative oxygen uptake efficiency slope as an indicator of cardiorespiratory fitness in elderly patients scheduled for major colorectal surgery. Anaesthesia. 2017;72(10):1206–16. doi: 10.1111/anae.14003 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref011] 11.Dideriksen K, Mikkelsen UR. Reproducibility of incremental maximal cycle ergometer tests in healthy recreationally active subjects. Clin Physiol Funct Imaging. 2017;37(2):173–82. doi: 10.1111/cpf.12283 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref012] 12.Martin-Rincon M, Calbet JAL. Progress Update and Challenges on V. O2max Testing and Interpretation. Front Physiol. 2020;11:1070. doi: 10.3389/fphys.2020.01070 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0283129.ref013] 13.Martin-Rincon M, Gonzalez-Henriquez JJ, Losa-Reyna J, Perez-Suarez I, Ponce-Gonzalez JG, de La Calle-Herrero J, et al. Impact of data averaging strategies on VO2max assessment: Mathematical modeling and reliability. Scand J Med Sci Sports. 2019;29(10):1473–88. [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref014] 14.Franssen RFW, Bongers BC, Vogelaar FJ, Janssen-Heijnen MLG. Feasibility of a tele-prehabilitation program in high-risk patients with colon or rectal cancer undergoing elective surgery: a feasibility study. Perioper Med (Lond). 2022;11(1):28. doi: 10.1186/s13741-022-00260-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0283129.ref015] 15.von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335(7624):806–8. doi: 10.1136/bmj.39335.541782.AD [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0283129.ref016] 16.Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39(2):175–91. doi: 10.3758/bf03193146 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref017] 17.Rose GA, Davies RG, Davison GW, Adams RA, Williams IM, Lewis MH, et al. The cardiopulmonary exercise test grey zone; optimising fitness stratification by application of critical difference. Br J Anaesth. 2018;120(6):1187–94. doi: 10.1016/j.bja.2018.02.062 [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref018] 18.Midgley AW, McNaughton LR, Carroll S. Effect of the VO2 time-averaging interval on the reproducibility of VO2max in healthy athletic subjects. Clin Physiol Funct Imaging. 2007;27(2):122–5. doi: 10.1111/j.1475-097X.2007.00725.x [DOI] [PubMed] [Google Scholar]

[pone.0283129.ref019] 19.Johnson JS, Carlson JJ, VanderLaan RL, Langholz DE. Effects of sampling interval on peak oxygen consumption in patients evaluated for heart transplantation. Chest. 1998;113(3):816–9. doi: 10.1378/chest.113.3.816 [DOI] [PubMed] [Google Scholar]

PERMALINK

Influence of different data-averaging methods on mean values of selected variables derived from preoperative cardiopulmonary exercise testing in patients scheduled for colorectal surgery

Ruud F W Franssen

Bart H E Sanders

Tim Takken

F Jeroen Vogelaar

Maryska L G Janssen-Heijnen

Bart C Bongers

Roles

Abstract

Introduction

Methods

Results

Conclusion

Clinical trial registration

Introduction

Fig 1.

Methods

Participants

Preoperative cardiopulmonary exercise testing

Procedures

Sample size

Statistical analysis

Results

Table 1. Baseline characteristics of subjects.

Table 2. Numerical values of CPET variables using different data-averaging intervals.

Fig 2.

Table 3. Effect of different data-averaging intervals on classifying patients as having a high-risk for postoperative complications.

Discussion

Conclusion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Sampath Kumar Amaravadi

Roles

Author response to Decision Letter 0

Decision Letter 1

Lindsay Bottoms

Roles

Author response to Decision Letter 1

Decision Letter 2

Lindsay Bottoms

Roles

Acceptance letter

Lindsay Bottoms

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases