Table 2.

Common sources of error in CPET testing

	Source of error	Description	Impact on data
Patient	Failure to follow information disclosed on patient information leaflet	All patients should be given a patient information leaflet and/or advice on what to avoid prior to performing an exercise test.	Various implications, specifically limiting exercise tolerance and impairing gas exchange data
	Poor effort/cooperation/motivation to perform exercise test	Patients need to understand the reason why the test is being performed. Failure to do so may result in suboptimal effort.	Underestimation of all indices, including workload, AT, VO2 and VCO2.
Test operator	Failure to give standardised instruction and encouragement during exercise	Throughout the different phases of exercise, there should be clear and standardised instructions to patients.	Various implications, specifically lack of consistency in data across different test operators
	Failure to select correct load (watts) in view of patient’s activity level/fitness	Incremental workloads that result in exercise duration of <8 or >12 min do not accurately reflect aerobic status.	Various implications, although more commonly underestimation of gas exchange indices
	Lack awareness/guidance on the use of well-defined end of test criteria	Exercise may be stopped too early or too late in what should be a symptom limited exercise test.	If exercise is stopped early (eg, pulse rate), gas exchange indices can be underestimated.
	Incorrect determination/identification of the AT	There should be a clear definition of what AT is and processes in place to promote discussion and review agreement.	Inappropriate estimation of level of fitness or degree of impairment in O2 delivery/use
	Incorrect determination/measurement of slopes (ΔVE/ΔVCO2, OUES and ΔVO2/ΔWR)	The determination of slopes based on linear regression models require correct identification of the start and end points.	Incorrect inferences from data (VE/VCO2 mismatch, cardiovascular impairment, among others)
Equipment	Inaccurate output of power by treadmill/ergometer	Treadmill (speed/grade) and ergometer (resistance) power outputs require yearly servicing (more often if regularly moved).	Various implications, particularly overestimation or underestimation of gas exchange indices
	Non-calibrated weighing scales and stadiometer	Weighing scales and stadiometers require regular servicing and calibration if there is a suspicion of erroneous measurements.	Incorrect estimation of predicted data and consequent inaccurate inferences from recorded data
	Excessive condensation at the point of gas analysis	Gas analysis should meet BTPS conditions, particularly humidity/water vapour pressure.	Various implications, although more commonly underestimation of gas exchange indices
	Volume drift	Thermal or offset volume drift may occur as a result of large fluctuations in temperature or incorrect calibration	Various implications, particularly inaccurate ventilatory and gas exchange indices
	Delayed response time and transit time in gas exchange parameters	Under certain testing conditions, there may be delay from the point of sampling to the point of gas analysis.	Normally, underestimation of gas exchange data due to dispersion of expired gases
	High/low sampling rates/delta time/data averaging of gas exchange data	Data averaging below 30 or above 60 s will affect validity of gas exchange measurements.	Either high fluctuations or excessive attenuation in gas exchange data

AT, anaerobic threshold; BTPS, body temperature and pressure saturated; CPET, cardiopulmonary exercise testing; OUES, oxygen uptake efficiency slope; V_E/VCO₂, ventilatory equivalent for carbon dioxide production; ΔV_E/ΔVCO₂, slope of the ventilatory response; ΔVO₂/ΔWR, slope of the metabolic response.