Karlakki 2016.

Methods	Study design: randomised controlled trial Study grouping: parallel Ethics and informed consent: yes Follow‐up period: 6 weeks Sample size estimate: pilot study ITT analysis: yes number randomised: 220 number analysed: 209 Funding: study funded through a grant from Smith & Nephew UK to cover the cost of NPWT dressings and data collection costs. 2 investigators declared they had funding and consultancy fees from Smith & Nephew. Pre‐registration: no
Participants	Location: Oswestry, UK Intervention group: n = 110control group: n = 110 Mean age (SD): intervention group = 69 (9.0)control group = 69.2 (9.0) Inclusion criteria: patients undergoing total hip or knee arthroplasties (for any indication) with any of 3 consultant surgeons Exclusion criteria: patients who had known allergies to dressing, were undergoing revision joint surgery, were unwilling to attend additional clinics, and those on warfarin were excluded.
Interventions	Aim/s: to evaluate the effectiveness of incisional negative pressure wound therapy dressing (iNPWTd) Group 1 (NPWT) intervention: PICO dressing applied over the primarily closed incision by the surgeon in the operating room. Dressing was left on for 4 days, or longer if drainage continued, unless soiled or dislodged. Group 2 (control) intervention: Comfeel dressing applied over the primarily closed incision by the surgeon in the operating room. Dressing was left on for 4 days, or longer if drainage continued, unless soiled or dislodged. Study date/s: July 2012 to April 2014
Outcomes	SSI blisters haematoma hospital readmission Validity of measure/s: not described Time points: 1, 2, and 6 weeks postsurgery
Notes	Investigator contacted for additional details.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "the randomisation was performed using sealed opaque envelopes with a block size of 20 shuffled envelopes" Comment: no sequence generation was required.
Allocation concealment (selection bias)	Low risk	Quote: "the randomisation was performed using sealed opaque envelopes with a block size of 20 shuffled envelopes" Comment: allocation was unknown until envelope opened.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Evidence for participants: not possible Comment: unlikely to affect outcomes Evidence for personnel: not possible Comment: unlikely to affect outcomes
Blinding of outcome assessment (detection bias) All outcomes	High risk	Outcome assessors were aware of group allocation.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	7.3% in intervention group and 2.7% in control group PP analysis Comment: more participants were excluded from the analysis in the intervention group (8 intervention vs 3 control).
Selective reporting (reporting bias)	Low risk	Expected outcomes reported.
Other bias	High risk	Intervention participants were seen in a wound clinic at 1 week, and control participants were not.