Assessment of research waste part 1: An exemplar from examining study design, surrogate and clinical endpoints in studies of calcium intake and vitamin D supplementation

Abstract

Background: Research waste is estimated to be very common, but assessments of its prevalence and scope are rare. As an example, we assessed research waste in clinical research on calcium intake (assessing study design and endpoint type) and vitamin D supplementation (assessing endpoint type). Methods: We examined 404 randomised controlled trials (RCTs) and observational studies of calcium intake (diet or supplements) and bone mineral density (BMD) or fracture, and 547 RCTs of vitamin D supplements, and assessed the proportion of studies that used surrogate or clinical endpoints. For studies with BMD or fracture as an endpoint, we estimated when the 'tipping' point occurred indicating the need for RCTs with fracture as an endpoint (based on cumulative meta-analyses of BMD RCTs, and chronological review of observational studies), and whether each study published at least 5y after the tipping point was novel, added new clinical knowledge or was research waste. Results: Observational studies of calcium intake and BMD or fracture outnumbered RCTs by 3.3-4.5 times. For both calcium intake and vitamin D supplements, studies using surrogate endpoints outnumbered studies using clinical endpoints by 1.6-3 times. Of 41 RCT publications of calcium intake and BMD or fracture published at least 5y after the tipping point in 1994, we considered that 19 (46%) lacked novelty, another 13 (32%) added no new clinical knowledge, and 30 (73%) were research waste. Of 204 observational study publications of calcium intake and BMD or fracture, 197 (96%) lacked novelty, another 5 (2%) added no new clinical knowledge, and 202 (99%) were research waste. Of 39 RCTs of vitamin D supplementation and BMD or fracture published at least 5y after the tipping point in 1999, 14 (36%) lacked novelty, another 13 (33%) added no new clinical knowledge, and 27 (69%) were research waste. Conclusions: A high proportion of studies of calcium intake since 2000 (95%) and trials of vitamin D supplements since 2005 (69%) on BMD or fracture represent research waste.