Table 3

Outcome evaluations of contest finalists

First author	Primary outcome measure	Evaluation study design	Primary data collection method	Comparator outcomes	Outcomes of top-ranked submission	Summary findings	Study quality Scale: 0–10
Chansrichavala et al45	Opinion regarding the potential use of videos to raise public awareness of melioidosis	Qualitative focus group discussion	Focus group audio recording	–	–	Top 12 video submissions had strong potential to increase public awareness of melioidosis	9.5
Costello and Stolovitzky6	Predictive accuracy of cancer drug sensitivity	Retrospective observational cohort and cross-validation of predictive algorithms	Clinical biomarkers	0.500	0.583	Finalist algorithm predicted cancer drug sensitivity significantly more accurately than random assignment (p<0.0001)	8.1
Costello and Stolovitzky6	Identification of causal gene mutations and innovative clinical diagnostic protocols	Individual patient genomic sequencing	Clinical biomarkers	NA	NA	Finalist team identified causal genetic mutation of a rare condition and submitted innovative protocols in genome analysis, interpretation and reporting	8.1
Küffner et al50	Deviation of predictive algorithm of ALS disease progression from observed clinical outcomes	Retrospective observational cohort & cross-validation of predictive algorithms	Clinical biomarkers and behavioural indicators	0.559	0.544	Finalist algorithm predicted ALS disease progression more accurately than benchmark standard, but not significantly (p>0.05). Improvement in prediction could potentially reduce size and cost of future clinical trials.	9.4
Lakhani et al52	Speed and accuracy of genomic analysis	Cross-validation of analytical algorithms	Clinical biomarkers	15 616 s; 72% accuracy	16 s; 80% accuracy	Finalist algorithm analysed genomes faster and more accurately than industry standard.	9.4
Zhang et al43	Proportion who had HIV test after 3 months postintervention	Randomised control trial	Survey	0.35	0.37	HIV test promotion finalist video performed as well as professionally created social marketing video (p>0.05) and had greater cost-savings.	9.4
Tang et al44	Proportion who had engaged in condomless sex after 3 months postintervention	Randomised control trial	Survey	0.336	0.323	Crowdsourced sexual health image intervention achieved non-inferiority at 3 months (p>0.05) and had greater cost-savings compared with conventional social marketing alternative.	9.4

ALS, Amyotrophic Lateral Sclerosis.