Unit costs
Cost of intervention: The intervention is a single dose of IV co-amoxiclav, given post-delivery. Since co-amoxiclav is already widely used within the NHS, its cost could be reliably estimated from the British National Formulary (BNF). The ANODE trial reports an average NHS indicative cost for the 3 main providers of co-amoxiclav powder calculated using the BNF 20178, totalling £2.30. For this economic evaluation, the same methodology was employed but with updated costs from the BNF 202013, yielding an average price of £2.26. A placebo of 0.9% saline was the comparator and just as in the ANODE trial this was attributed no cost.
Cost of complications: In order for the cost of complications to be accurate, the typical patient journey and the likelihood of using certain NHS resources over that journey needed to be considered. This was informed by an NHS flowchart obtained from the Royal Berkshire Hospital NHS Foundation Trust guidelines on postpartum infection (Appendix S3). This highlighted the role of home visits by midwives as the first healthcare point of contact for mothers with suspected postpartum complications and this was consequently included in all cost estimates. Literature searches on standard treatments for each specific complication informed the cost at each terminal node. In certain cases, assumptions were made to estimate the costs of specific complications. A cost for systemic infection but without subsequent development of sepsis was not obtainable; therefore, the treatment cost associated with sepsis was utilised for all outcomes of systemic infection. This is justified as irrespective of the development of sepsis, systemic infections will necessitate similar treatment costs, as reflected in the National Tariff17. Other assumptions included that patients were only re-admitted into hospital for systemic infection, endometritis and wound breakdown and that all patients with “other or confirmed suspected infections” were treated in the community.
Discounting : To ensure all costs were accurate, sources of cost data used were UK-based and from the year of the trial (March 2016-) or newer. Moreover, all but one of the costs were taken directly from NHS reference costs, the British National Formulary (BNF) or government reports. Historical costs were discounted at a rate of 3.5% in line with NICE guidelines18, converting all costs into the year 2020.