The investigators conducted a prospective randomised controlled trial at a tertiary referral hospital in Dar es Salaam, Tanzania from December 17th, 2015 to March 25th, 2017. All open tibia fractures arriving at the hospital emergency room during the enrollment period were screened for inclusion in the study. Inclusion and exclusion criteria are noted later in this submission.All patients who were eligible after clinical screening were offered enrollment in the study and subsequently completed written informed consent in either English or Swahili depending on the patient's preference. The investigation was approved by ethical review boards at the University of California San Francisco (UCSF) and the Tanzanian National Institute for Medical Research (NIMR).
All enrolled participants were taken to the operating room for initial irrigation and debridement using a standardized protocol aiming for debridement within 24 hours of hospital admission. Preoperative ceftriaxone was given as soon as possible after presentation to the emergency department. Following debridement, the primary surgeon determined if the wound was amenable to immediate or delayed primary closure. If amenable to primary closure, the patient was randomized by a site research coordinator to one of two treatments (Intramedullary nailing (IMN) or external fixation (EF))using a centralized web-based electronic randomization tool, Research Electronic Data Capture (REDCap). The study computer analyst developed the REDCap randomization module with input from senior authors. The REDCap randomization module incorporated block randomization to maintain equal patient distribution between each treatment group and to ensure allocation concealment.
Data from a pilot study at the site of our investigation informed sample size calculations, which showed reoperation rates of external fixation and intramedullary nailing of 38% versus 3.8%. However, with longer follow-up and more stringent criteria for a primary event, The investigators used a more conservative effect size of 20% compared to 5%. Using a power of 80% and alpha of 0.05, The investigators calculated the study would require 88 patients in each treatment group (176 total). Accounting for 20% loss to follow up, The investigators estimated a total sample size requirement of 240 patients (120 per group).
Data monitoring was completed by the adjudication committee at 6 month intervals throughout the study duration and prior to final analysis in December, 2018.
