Abstract
Dear Editors,
I read with interest the recent article by Martel et al, 1 which compared high-definition oscillometry (HDO) with direct arterial blood pressure measurement in conscious cats. The veterinary community eagerly awaits the validation of a device for non-invasive blood pressure measurement to the standards set by the 2007 American College of Veterinary Internal Medicine (ACVIM) consensus statement. 2 Though the authors claim they have validated S + B MedVET’s high definition oscillometry device by these ACVIM standards, they, unfortunately, have not.
The authors state in the final sentence of the article’s abstract that ‘The data support that the HDO is the first and only validated non-invasive blood pressure device and, as such, it is the only non-invasive reference technique that should be used in future validation studies’. 1 However, one of the standards set by the ACVIM consensus statement for device validation is that ‘the subject database contains no fewer than eight animals for comparison with an intra-arterial method’. 2 The current study included six cats. The HDO has thus not been validated in this study as a blood pressure monitoring device by ACVIM standards. Also, the ACVIM guidelines state that a device is validated ‘for only the species and conditions in which the validation test is conducted’, so even if the authors reproduced these findings with at least eight awake cats, the device could only be considered validated for use in conscious cats — a far more limited scope than their concluding statement suggests.
Additionally, there is no statement in the ‘Materials and methods’ section that the investigators were blinded to the direct arterial blood pressure measurements while taking HDO readings. Occasionally, readings from non-invasive blood pressure devices, including HDO, must be ignored (‘thrown out’) as a result of patient movement and other errors. While this HDO unit reports some errors automatically, it is unclear by what criteria readings were discarded in the current study. The investigators’ apparent access to the real-time direct pressure data makes this omission particularly troublesome. Possibly, the authors could explain the criteria by which they discarded the readings they considered inaccurate.
Also, the ACVIM consensus statement requires that ‘the correlation between paired measures for systolic and diastolic pressures treated separately is ⩾0.9 across the range of measured values of BP’. 2 The current study reported an overall mean correlation coefficient of >0.9 for systolic blood pressure when all systolic readings (hypotensive, normotensive and hypertensive) were considered together. However, when the blood pressure measurements were divided into hypotensive, normotensive and hypertensive measurement groups, the HDO failed to meet the ⩾0.9 standard for any of the three groups. While the consensus statement is somewhat vague (what exactly does ‘across the range’ mean?), for the practitioner it means that for any measurement in any individual cat (which can only be one: hypotensive, normotensive or hypertensive), the HDO fails to meet the correlation coefficient standard of ⩾0.9. Therefore, it is my opinion that this study’s stated conclusions are misleading.
Finally, and most difficult to address, is the ‘Conflicts of interest’ statement, which states that ‘The authors do not have any potential conflicts of interest to declare’. This statement is false. At least one of the investigators in the current study is intimately involved with S + B MedVET, the manufacturer of the HDO unit being studied, which should be openly declared.