Tjitske van Engelen

176 Chapter 8 Exacerbated asthma, thoracic pain of unknown origin 1 Exacerbated asthma, other upper respiratory tract infection Influenza A or B, other thoracic pathology 1 Influenza A or B Other lower respiratory tract infection, influenza A or B 1 Influenza A or B* Other upper respiratory tract infection 1 Other upper respiratory tract infection, extrathoracic pathology* Disagreement 4 Single label cases 3 Dyspnoea of unknown origin 1 Other upper respiratory tract infection Fever of unknown origin 1 Other upper respiratory tract infection Other thoracic pathology 1 Thoracic pain of unknown origin* Multiple label cases 1 Community-acquired pneumonia, extrathoracic pathology 1 Other lower respiratory tract infection, extrathoracic pathology Within categories of agreement rows are sorted first by prevalence and then alphabetically. *Disagreement on the diagnosis was based on discordance on a procedural error or labels from the additional diagnostic categories only. These are considered agreement cases. Overall inter-observer agreement between students Total agreement was observed for 132 of the 240 participants (55%): 107 (45%) were assigned the same diagnostic label(s) and 25 (10%) were directly referred to the expert panel. Cases where the two students disagreed (108, 45%) were sub-classified as partial agreement (33, 14%) and total disagreement (75, 31%) (Figure 1). In 39 of the 75 total disagreement cases one of the two students deemed the case too complex and initially referred this case to the expert panel. Twenty-one of these 39 cases were classified during the consensus meeting with a resident and could be withheld from referral to the expert panel. Reasons for disagreement between students can be found in Supplementary material 3. Inter-observer agreement between students for specific diagnostic labels A total of 523 labels were assigned to the 240 cases: 336 definite diagnostic labels (concordance 65%) and 187 labels from the additional six diagnostic categories (concordance 61%) (Supplementary Table S1). The most prevalent diagnostic labels were LRTI other than CAP (70/336, concordance 49%; κ 0.43, 95% CI 0.27–0.59), CAP (67/336, concordance 81%; κ 0.78, 95% CI 0.66–0.89), exacerbation chronic obstructive pulmonary disease (COPD)(42/336, concordance 76%; κ 0.74, 95% CI 0.59–0.89), exacerbation asthma (30/336, concordance 73%; κ 0.72, 95% CI 0.53–0.90) and influenza A/B (28/336,

RkJQdWJsaXNoZXIy MTk4NDMw