10.1371/journal.pone.0149713.g002 Robert Carroll Robert Carroll Chris Metcalfe Chris Metcalfe Sarah Steeg Sarah Steeg Neil M. Davies Neil M. Davies Jayne Cooper Jayne Cooper Nav Kapur Nav Kapur David Gunnell David Gunnell Instrumental variable estimates* of the risk difference in repeat self-harm between those who did, and did not receive a psychosocial assessment.* Public Library of Science 2016 Hospital Presentation BackgroundClinical guidelines RD hospital presentation IV effect estimate indication.ResultsConventional regression analysis UK assessment CI Instrumental Variable Analysis 2016-02-26 08:38:12 Figure https://plos.figshare.com/articles/figure/Instrumental_variable_estimates_of_the_risk_difference_in_repeat_self_harm_between_those_who_did_and_did_not_receive_a_psychosocial_assessment_/3024871 <p>The overall estimate in this figure is the overall instrumental variable estimate, not the pooled estimate from a meta-analysis of the three individual centre estimates.</p>