WHIST 2019b.

Study characteristics
Methods	Study design: cost‐effectiveness analysis based on the WHIST 2019a RCT) Analytical approach: trial‐based decision model Effectiveness data: SSI (deep) and QoL (EQ‐5D) both derived from WHIST 2019 (UK multicentre RCT, N = 1548) Perspective: NHS and personal social services (PSS) perspectives Utility valuations: EQ‐5D and NHS/PSS resource use values derived from 623 trial participants with complete profiles Measure of benefit: QALY calculated using EQ‐5D‐3L utility scores using UK scoring algorithm Cost data: unit direct medical costs associated with the intervention obtained from the NHS Supply Chain Catalogue 2018/2019. These included cost of standard dressing, the costs of orthotic cast, the cost associated with dressing change, the cost per working hour of the nurse (obtained from the Personal Social Service Research Unit (PSSRU) 2018). The cost of inpatient care derived using the NHS HRG4+ 2017/18 Reference Cost Grouper and the NHS Reference Costs 2017/18. Unit costs of medical items other than those directly attributable to the intervention sourced from the NHS Reference Costs. Medication costs sourced from the BNF. Unit costs for direct non‐medical cost items obtained from PSSRU. The costs of aids and adaptations obtained from the NHS Supply Chain Catalogue. The total cost per patient for additional (private) cost items incurred by patients and their next‐of‐kin obtained from the patients directly. The daily median wage obtained from the Office for National Statistics. Cost data were derived from the key resource inputs of the WHIST 2019 trial and expressed in 2017/2018 UK pounds sterling (£) (completed case analyses); a societal perspective was considered in a sensitivity analysis. Unit costs adjusted to 2017/2018 prices using the NHS Hospital & Community Health Services (HCHS) index for health service resources. No discounting of costs applied due to a short‐time horizon. Analysis of uncertainty: results of ICERs and cost‐effectiveness acceptability curves (CEACs) generated via nonparametric bootstrapping with 1000 replicas for accommodating sampling (or stochastic) uncertainty and varying levels of willingness‐to‐pay. Sensitivity analysis incorporated societal perspective; 3 different willingness‐to‐pay thresholds considered.
Participants	Location: UK hospitals Intervention group: n = 785, control group: n = 763 Mean age: </= 40: 283 (36.1%); > 40: 501 (63.9%), control group </= 40: 278 (36.4%); > 40: 485 (63.6%) Inclusion criteria: adult patients (16 years minimum) presenting to hospital within 72 hours of sustaining major trauma and who required a surgical incision to treat a fractured lower limb. Exclusion criteria: open fracture of the lower limb that could not be closed primarily; evidence that the patient would be unable to adhere to trial procedures or complete questionnaires.
Interventions	Aim/s: to investigate, using appropriate statistical and economic analysis methods, the resource use, and thereby the cost‐effectiveness, of NPWT versus standard dressing for wounds associated with major trauma to the lower limbs.   Group 1 (NPWT) intervention: NPWT using a non‐adherent absorbent dressing covered with a semi‐permeable dressing. A sealed tube connects the dressing to a built‐in mini‐pump that creates a partial vacuum over the wound. NPWT applied as per treating surgeon's normal practice and according to manufacturer's instructions (n = 785 in the trial). Group 2 (control): standard dressing (non‐adhesive layer covered by sealed dressing or bandage) (n = 763 in the trial). Study date/s: October 2016 to March 2016
Outcomes	Outcomes (for data see additional table 1 for WHIST 2019b, and for clinical data WHIST 2019a;) Costs (GBP) QALY (measure of benefit) ICER Probability of being cost‐effective at 3 different thresholds
Notes	Funding: NIHR Authors' conclusions: contrary to the existing literature, incisional NPWT did not provide a clinical or economic benefit for patients having surgical incisions associated with major trauma to the lower limb. Notes: not currently a separate publication for cost‐effectiveness; data taken from monograph which focused on RCT. Quality rating using the CHEERS checklist was 89.1%.