EDITORIAL Nephrol Dial Transplant (2023) 0: 1–3 https://doi.org/10.1093/ndt/gfad054 Advance Access publication date 16 March 2023 Cold haemodialysis: the instrumental power of large cohorts Christian Combe 1 ,2 and Sébastien Rubin 1 ,3 1 Department of Nephrology, Transplantation, Dialysis, Centre Hospitalier Universitaire de Bordeaux, Bordeaux, France, 2 Unité INSERM U1026, University of Bordeaux, Bordeaux, France and 3 Division of Cardiovascular Sciences, Faculty of Medicine, Biology and Health, University of Manchester, Manchester, UK Correspondence to: Christian Combe; E-mail: christian.combe@chu-bordeaux.fr; Twitter: @ChristianCombe In this issue of NDT, Zoccali et al.[1] evaluate the impact of cool dialysate temperature on the occurrence of intradialytic hypotension (IDH) in an incident cohort of >8000 patients treated by haemodialysis (HD) from NephroCare Spanish and Portuguese centres in Europe. Over a follow-up of slightly more than 1 year they found that a 0.5°C reduction in dialysate temperature in 27% of patients was associated with a small (-2.4%) and non-signifcant reduction in the risk of IDH. Because of potential bias by indication—sicker patients being treated with cooler dialysate to minimise IDH— and to minimize the efect of unmeasured and measured confounders, they used instrumental analyses assessing the associations between dialysate temperature at the centre level and the occurrence of IDH. In this type of analysis carried out at the facility level with multiple adjustments for case mix, with the same 0.5°C lower dialysate temperature, risk reduction for IDH occurrence was 33%, with a P-value <.001. It is noteworthy that there was no association between death and temperature reduction [1]. These results are quite diferent from those reported recently in the Lancet [2] by the MyTEMP research group in Ontario, Canada. In the largest randomised controlled trial (RCT) ever performed in HD, involving 15 413 patients in 84 centres over 4 years, there was no impact of a dialysate temperature of 35.8°C versus 36.4°C on a composite outcome of cardiovascular-related death or hospital admission with myocardial infarction, ischaemic stroke or congestive heart failure. There was also no efect on IDH. In contrast, patients in the cold dialysate group were more likely to feel uncomfortable during HD sessions. Diferences between the two studies provide an opportunity to discuss several clinical and methodological points that go beyond the question of dialysate temperature: • Is a 0.5°C diference in dialysate temperature clinically relevant? • What are the diferences between the two largest studies performed so far in terms of patients’ characteristics and techniques used? • How can the results of the gold standard RCT be com- pared with a cohort study analysed with the instrumental variable methodology? CLINICAL RELEVANCE OF A 0.5°C DIALYSATE TEMPERATURE DIFFERENCE An individual’s body temperature rises by ≈0.5°C on average during HD treatment when the dialysate temperature is ≈37°C. This happens because the whole-body temperature is normally <37°C: body temperature rising is associated with vasodilatation and vascular instability during HD [3]. As detailed in both studies [1, 2], several small studies have shown that IDH incidence might be reduced by the use of so-called cold dialysate. A drawback of the use of cold dialysate is that patients feel uncomfortable when the dialysate temperature is low, as happened in the Ontario study [2]: this might be an indicator of the clinical signifcance of a 0.5°C diference in dialysate temperature. In an RCT performed in 73 patients new to HD, individualized cooled dialysate had a measurable impact: it slowed the progression of HD-associated cardiomyopathy over 12 months [4]. Rather than dialysate temperature, body temperature or energy balance might be more relevant to the occurrence of IDH. In an important randomised crossover trial performed in 95 hypotension-prone patients, Maggiore et al.[5] evaluated the impact of ‘thermoneutral dialysis’. In this HD system, through a feedback loop connecting the arterial and venous blood temperature sensor with the dialysate thermostat in the machine, the device can modulate dialysate temperature to maintain an unchanged energy fow rate across the extracor- poreal circuit, i.e. perform thermoneutral HD [5], which leads to stability of body temperature during HD. Maggiore et al. [5] found that systolic and diastolic blood pressures and heart rate were more stable during thermoneutral dialysis, with fewer episodes of IDH. The results of this study are important since patients were selected because they were prone to hypotension during HD sessions. © The Author(s) 2023. Published by Oxford University Press on behalf of the ERA. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com Downloaded from https://academic.oup.com/ndt/advance-article/doi/10.1093/ndt/gfad054/7079141 by guest on 19 May 2023