Assessing a full patient-year of in-centre treatment, 156 sessions, the study put standard haemodialysis at 4,427 kg CO2-equivalent and online haemodiafiltration at 4,548 kg, a difference of 2.7%. Strip out patient and staff travel and the gap widens only to 9.3%. Water use differed by 27 litres per session and electricity by under 6%. Varying the case mix from none to all haemodiafiltration moved the total by less than 3%. The authors’ conclusion is unambiguous and worth repeating in full: clinical outcomes, not carbon, should drive the choice of modality, a point that matters given the survival benefit demonstrated for high-dose haemodiafiltration in the CONVINCE trial.

What the same assessment does surface is where the footprint actually sits. Patient travel accounted for 2,548 kg CO2-equivalent per patient-year and staff travel a further 615 kg, together around 71% of the total, on an average round trip of 29 km per session. Procurement came next at 919 kg, or 21%. Electricity, water production and waste were comparatively minor. Two operational levers stood out: raising reverse osmosis recovery from 65% to 70% would save in the order of 40,000 litres of water per patient-year, and a reduced-flow haemodiafiltration prescription cut dialysate use from 125 to 100 litres per session while achieving similar Kt/V. The single most influential variable of all was the carbon intensity of the electricity grid.

Waste, in that accounting, is a small slice, 130 to 143 kg CO2-equivalent of a roughly 4,470 kg patient-year. Which is precisely what makes a second study interesting, because it changes the sign of that number.

Working with the Modena clinical site and its local waste-to-energy plant, the team compared the current practice of incinerating dialysis plastics with a scenario sending 90% to pyrolysis, a chemical recycling process that converts mixed plastic waste into usable outputs. Per tonne of healthcare plastic waste, full incineration carried a burden of 1,996 kg CO2-equivalent. The pyrolysis scenario came out at net −230 kg CO2-equivalent, a negative figure, because the recovered ethylene displaces virgin production, a credit worth −854 kg on its own. The scenario was lower across all 16 impact categories assessed, with net credits in seven of them, and particulate matter formation 48-fold lower. Powering the process with wind energy improved the result further, to −424 kg per tonne.

The proportional significance of that shift depends entirely on the modality. In-centre haemodialysis is dominated by travel, so better waste treatment moves a small share of a large total. Home therapies have no travel at all, and there plastics dominate: the study puts annual plastic waste at around 670 kg for a home haemodialysis patient, 248 kg for automated peritoneal dialysis and 76 kg for continuous ambulatory PD, and gives illustrative savings of roughly 1.49, 0.55 and 0.17 tonnes of CO2-equivalent per patient per year respectively. For home and peritoneal dialysis, in other words, what happens to the plastic is a first-order question.

The obstacles are real and mostly not technical. Pyrolysis plants need capital investment the authors put at €2–10 million, the uncertainty range around the central estimate is wide, and the regulatory position is unsettled: pyrolysis is not uniformly covered by EU rules on recycled plastics in food contact, end-of-waste criteria vary between member states, and residual char generally remains classified as hazardous. The paper is also currently available as an accepted manuscript ahead of final editing.

Taken together the two studies suggest a reordering of priorities. Argue about modality on clinical grounds. Then look hard at travel, at the grid, at how much water the plant recovers, and at whether dialysis plastic is treated as waste or as feedstock.

Related readings

• Comparing environmental footprints of haemodialysis and online haemodiafiltration in Italy
• Healthcare waste: from incineration to pyrolysis: a comparative life cycle assessment of two different healthcare plastic waste management processes

*Photo by Edward Jenner