Houwing 2008.

Methods	Trial design: RCT Trial grouping: cluster Ethics and informed consent: yes Follow‐up period: 4 weeks Sample size estimate: no ITT analysis: yes; number randomised: 79; number analysed: 79 Funding: not stated
Participants	Inclusion criteria: written informed consent was obtained from each participant. If the mental capability of the participant to decide on participation was uncertain, the legal representative of the participant was asked for consent participants had to be able to participate for an evaluation for 4 weeks participants had to rest on an anti‐PU mattress participants had to be at high risk of developing PU according to the Braden scale using a cut‐off point of 20 Exclusion criteria: patients who were treated with another, unrelated ointment or cream people who were to undergo or had undergone surgery < 2 weeks prior people with existing PU people with dark skin because of difficulty in assessment Pretreatment men: 27.7% control; 25% placebo; 37.9% DMSO not incontinent: 16.7% control; 6.3% placebo; 0% DMSO age (median): 81.5 years control; 85 years placebo; 80.5 years DMSO no significant differences in participant characteristics
Interventions	Intervention group: massage using a "DMSO‐cream". This cream consisted of 5% dimethyl sulfoxide in Vaseline‐cetomacrogol cream, combined with a 30° position change. This procedure was repeated every 6 h for 4 weeks Placebo: 3‐min massage of the buttock, heel, and ankle regions with an "indifferent" cream (Vaseline‐cetomacrogol) combined with a 30° position change. This procedure was repeated every 6 h for 4 weeks Control: 30° position change, repeated every 6 h for 4 weeks
Outcomes	PU incidence Outcome type: dichotomous outcome Reporting: fully reported Unit of measure: numbers Direction: lower is better Data value: endpoint
Identification	Sponsorship source: not stated Country: the Netherlands Setting: 8 nursing homes Comments: no comments Author's name: R Houwing Institution: Department Dermatology, Deventer Ziekenhuis Email: houwingr@dz.nl Address: Department Dermatology, Deventer Ziekenhuis, The Netherlands
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Throw of a dice (additional information from the study author)
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Stated as double‐blind
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinded, quote: "presence of a pressure ulcer confirmed by two external observers"
Incomplete outcome data (attrition bias) All outcomes	Low risk	None excluded
Selective reporting (reporting bias)	Low risk	Outcome measure was the presence of a PU, this was reported by the study authors
Other bias	Unclear risk	There is no indication that the cluster design was accounted for in the analysis