Objective To quantify the effects of a thermostatic control system in social (public) housing on the prevalence of dangerous (>60°C) water temperatures and on fuel consumption.

Design Pair-matched double-blind cluster randomised controlled trial.

Setting Social housing in a deprived inner-London borough.

Participants 150 households recruited as clusters from 22 social housing estates. Four small estates were combined into two clusters (resulting in a total of 10 pairs of clusters).

Intervention Social housing estate boiler houses were randomised to a thermostatic control sterilisation programme (heating water to 65°C during 00:00–06:00 h and to 50°C from 06:00 to 00:00 h daily) or to standard control (constant temperature 65°C).

Main outcome measures Water temperature over 60°C (‘dangerous’) after running taps for 1 min and daily fuel consumption (cubic feet of gas).

Results 10 clusters (80 households) were allocated to the sterilisation programme and 10 clusters (70 households) to control, of which 73 and 67 households, respectively, were analysed. Prevalence of dangerous (>60°C) hot water temperatures at 1 min was significantly reduced with the sterilisation programme (mean of cluster prevalence 1% in sterilisation programme group vs 34% in control group; absolute difference 33%, 95% CI 12% to 54%; p=0.006). Prevalence of high (>55°C) hot water temperatures at 1 min was significantly reduced (31% sterilisation vs 59% control; absolute difference 28%, 95% CI 9% to 47%; p=0.009). Gas consumption per day reduced more in the control group than in the sterilisation programme group, although not statistically significantly (p=0.125).

Conclusions The thermostatic control with daily sterilisation was effective in capping hot water temperatures and therefore reduced scald risk. Although expected to save energy, fuel consumption was increased relative to the control group.

Trial registration ClinicalTrials.gov ID: NCT00874692