. 2020 Feb 3;15(2):e0226480. doi: 10.1371/journal.pone.0226480

Table 1. Summary of included studies in systematic review of the performance of cardiopulmonary exercise testing (CPET) for identification of post-operative outcomes of interest n = 33.

Study (Year) [Ref]	N	Specialty	Outcomes	Blinded	Prospective	Altered care	Selection for CPET	Variables	Statistics
Older (1993) [5]	187	Major Surgery	Mortality (Total and CVS)	N	Y	Y	Consecutive age 60+	AT^* Ischaemia^*	Chi-Sq
Nugent (1998) [44]	36	Vascular:AAA	Cardiorespiratory Mortality	Y	Y	N	Patients a/w repair Excl. MSK or IC	AT Peak VO₂ O₂ Pulse Ischaemia	M-W, Chi-Sq
Older(1999) [6]	548	Major Abdominal	Mortality (CVS and Other)	N	Y	Y	>60 <60 + previous Myocardial ischaemia. Excl. Thoracic (incl. UGI)	AT	Descriptive
Nagamatsu (2001) [37]	91	Oesophagectomy	Cardiorespiratory	N	N	Unclear	Surgery for SCC Excl. Neo-adjuvant chemotherapy, CCF, COPD	VO₂ Max^* AT	T-T M-W
McCullough (2006) [45]	109	Bariatric	Mortality Cardiorespiratory (MI, Angina, PE)	Y	Y	N	BMI ≥35 (+DM) or ≥40 for lap. roux-en-Y bypass. Excl. Limiting angina, lung disease	Peak VO₂	Descriptive (excluding analysis for composite outcome of death and complications)
Bowles(2008) [28]	124	Colorectal	Mortality	N	N	Y	Over 60 with ‘recognised heart/lung disease’, post anaesthetic assessment	AT	Log. Regression
Forshaw(2008) [46]	78	Oesophagectomy	Mortality Unplanned ICU Cardiopulmonary	N	Y	Unclear	Consecutive	AT VO₂ peak^*	T-T Fisher’s
Struthers(2008) [47]	50	Major abdominal	Mortality(30D)	N	Y	Possibly	Over 65 having ‘major intra-abdominal surgery’ Under 65 with significant myocardial ischaemia, respiratory disease, or CCF. Excl. unstable CVS disease	AT	Sens/Spec (calculated)
Snowden (2010) [48]	123	Major abdominal	Mortality Cardiopulmonary	N	Y	Y	For ‘major surgery’ with METS score ≤ Excl. Colorectal, Urological, and Orthopaedic surgery	AT^*	Chi-sq
Wilson(2010) [32]	847	Colorectal Urology	Mortality	N	N	Y	>55 <55 + cardiorespiratory comorbidity or DM	AT^* VE/VCO₂ ^*	RR
Thompson (2011) [49]	66	Vascular	Mortality(30D) Cardiac (+ Stroke), Respiratory	N	N	Y	All patients	AT^* VE/VCO₂ VE/VO₂ ^*	OR (incl. non operative) AT sig. mortality VE/VO₂ sig. inotropic requirement
Ausania (2012) [26]	124	Hepatobiliary	Mortality Cardiorespiratory	N	Y	N	Patients scoring <7 on a MET score	AT	Chi-sq
Hartley (2012) [29]	415	Vascular	Mortality(30D)	N	Y	Unclear	?All undergoing open/EVAR	AT^* VE/VCO2^* ≥2 abnormal^* Peak VO2^*	Chi-sq
Junejo [33] (2012)	94	Hepatic	Mortality (30D+in-hospital) Cardiorespiratory	N	Y	Y	>65 <65 ‘with comorbidity’, or ‘complex resection’	AT^* VE/VCO2^*	OR AT sig. in-hosp mortality, VE/VCO₂ sig. cardiorespiratory only
Chandrabalan [50] (2013)	100	Pancreatic	Mortality (incl. operative), Cardiac, Respiratory	N	N	Y	Patients for pancreatic surgery	AT	Chi-sq
Goodyear(2013) [30]	85	Vascular:AAA Infrarenal	Mortality(30D)	N	N	Y	Consecutive (not all) patients	AT^*	Fisher’s
Lai (2013) [51]	269	Colorectal	Mortality(30D) Unplanned ICU	N	Y	Y	All major colorectal	AT^* “Unable to achieve AT”	Chi-sq
Moyes (2013) [52]	180	Upper GI	Mortality Cardiorespiratory Unplanned ICU	N	Y	Y	Consecutive after MDT discussion	AT^* VO₂ peak	T-T (AT sig. for Cardiorespiratory only)
Prentis (2013) [53]	69	Cystectomy	Mortality	Y	Y	N	Consecutive	AT	Descriptive
Snowden[35] (2013)	389	Hepatobiliary	Mortality	Y	Y	N	All for open resection	AT^* Peak VO2^* Peak Work^* VE/VCO2 VO2/HR	Chi-sq
Ting(2013) [34]	70	Renal Transplant	Unplanned ICU	Y	Y	N	All >18 Excl.’condition precluding exercise’	AT^* PeakVO₂ ^* O₂ Pulse^* Max Work^* Endurance time VE/VCO₂ slope	Log. regression
Dunne (2014) [24]	197	Hepatectomy	Mortality In hospital Cardiorespiratory	N	N	Y	Initially >65 + ‘significant comorbidity’, or extended op. Changed to all patients	AT Peak VO₂ VE/VCO₂ VE/VO2	Calculated^%
James(2014) [39]	83	Major surgery	Mortality Cardiac	Outcome	Y	Possibly	Over 40s	AT^* VO₂ Peak^*	?T-T
Junejo (2014) [38]	64	Pancreatic	Mortality Cardiorespiratory	N	Y	Unclear	>65 Younger with comorbidity	AT VE/VCO₂ ^* VO₂Max	Log. regression Mortality only
West(BJA 2014) [54]	136	Colorectal	Cardiorespiratory (D5)	Y	Y	N	All over 18 excl: neoadjuvant, IBD, inability to perform	AT^*,	Chi-Sq Fisher’s
West (EJSO 2014) [55]	25	Colorectal (post NACRT)	Cardiorespiratory	Y	Y	N	Surgery post NACRT—main aim to assess impact of NACRT on fitness	AT	Sens/Spec (calculated)
Barakat(2015) [27]	130	Vascular	Mortality Cardiac Respiratory	N	Y	Y	Most patients with AAA>55, able to use treadmill	AT^* VO₂ Max VE/VCO₂ ^*	Log. regression AT sig. cardiac only VE/VCO₂ sig. resp. only
Chan (2015) [56]	94	Colorectal	Mortality(30D), Unplanned ICU	N	N	Y	Subset of patients >80 at surgeons discretion	AT^* VO₂ Max^*	MW (sig. ICU only)
Nikolopoulous (2015) [57]	89	Colorectal	Cardirespiratory Mortality	N	N	Unclear	Consecutive (Open procedures)	AT^*	T-T M-W (Cardioresp. only)
West [58] (2016)	703	Colorectal	Mortality (30D+In hosp.) Cardiorespiratory (D5)	N	Y	Y	Patients for major colorectal, excl: lower limb dysfunction, IBD, neoadjuvant treatment, metastatic	AT^*	Fisher’s
Kanakaraj (2017) [59]	70	Vascular Peripheral	Mortality(30D), Cardiac	Y	Y	N	Those for elective/expedited infra-inguinal bypass surgery	AT Peak VO₂ VE/VCO₂	T-T M-W
Whibley (2018) [31]	73	Upper GI	Respiratory	N	N	Unclear	Subset of those pre-assessed, part of an enhanced recovery protocol	AT VO₂ Max	Chi-sq
Abbott (2019) [41]	1324	Major non-cardiac	Myocardial injury D1-3	Y	Y	Safety only	Over 70 OR Over 40 with higher risk surgery or comorbidity	Chronotropic Incompetence	Log regression
Drummond [25] (2019)	42	Oesophagectomy	Cardiorespiratory, 30D Mortality, Unplanned ICU	N	Y	Unclear	Selected patients with pre/post chemotherapy prior to oesophagectomy for adenocarcinoma	AT (Pre/Post Chemotherapy)	Chi-sq
Lam (2019) [36]	206	Oesophagectomy	Cardiorespiratory	Y	N	Possibly	Consecutive undergoing oesophagectomy for cancer. Excluded 40 who didn’t undergo CPET for unclear reasons	AT Peak VO₂	T-T
Wilson (2019) [60]	1375	Colorectal	Unplanned ICU	N	N	Y	Over 55 or younger with cardiorespiratory risk factors	VE/VCO₂	Log regression

30D = 30 Day, AAA = Abdominal aortic aneurysm, AT = Anaerobic Threshold, BMI = Body Mass Index, CCF = Congestive Cardiac Failure, Chi-Sq = Chi-Square test, COPD = Chronic Obstructive Pulmonary Disease. CPET = Cardiopuulmonary Exercise Testing, CVS = Cardiovascular system, DM = Diabetes Mellitus, EVAR = Endovascular Aneurysm Repair, IC = Intermittent claudication, ICU = Intensive care unit, lap. = laparoscopic, METS = Metabolic equivalents, MDT = Multi-disciplinary Team, MI = Myocardial Infarction, N = number of patients who underwent and/or were analysed as having had CPET and may differ from the total number of patients included in some studies, NACRT = Neoadjuvant chemo-radiotherapy, Chi-Sq = Chi- squared test, M-W = Mann-Whitney test, MSK = musculoskeletal pathology, OR = Odd’s Ratio, PE = Pulmonary embolism, RR = Relative Risk, SCC = Squamous cell carcinoma, Sens/Spec = Sensitivity/Specificity, T-T = Student’s T-Test, VE/VCO₂, VE/VO₂ = Ventilatory equivalents of carbon dioxide/oxygen, VO₂ Max = Maximal oxygen uptake. Where stated D refers to postoperative day. Italics indicate supporting information or exclusion criteria

^* next to a CPET variable indicates it was found to be significant (using methods in statistical methods column.

‘Unclear’ was used to indicate if it was not possible to determine whether CPET values could have impacted on any aspect of patient care and thus introduced confounding. Statistical methods/CPET variables shown are those pertaining to our specified outcomes of interest only. ‘Sens/Spec (calculated)’ demonstrates that we extracted data pertaining to these CPET variables and outcomes and no formal testing was presented within the paper. % Indicates that analysis was performed on data requested from the authors after not being presented in the primary paper.