We undertook a systematic review of the diagnostic accuracy of artificial intelligence-based software for identification of radiologic abnormalities ( computer-aided detection , or CAD) compatible with pulmonary tuberculosis on chest x-rays (CXRs). We searched four databases for articles published between January 2005-February 2019. We summarized data on CAD type, study design, and diagnostic accuracy. We assessed risk of bias with QUADAS-2. We included 53 of the 4712 articles reviewed: 40 focused on CAD design methods (“Development” studies) and 13 focused on evaluation of CAD (“Clinical” studies). Meta-analyses were not performed due to methodological differences. Development studies were more likely to use CXR databases with greater potential for bias as compared to Clinical studies. Areas under the receiver operating characteristic curve (median AUC [IQR]) were significantly higher: in Development studies AUC: 0.88 [0.82–0.90]) versus Clinical studies (0.75 [0.66–0.87]; p-value 0.004); and with deep-learning (0.91 [0.88–0.99]) versus machine-learning (0.82 [0.75–0.89]; p = 0.001). We conclude that CAD programs are promising, but the majority of work thus far has been on development rather than clinical evaluation. We provide concrete suggestions on what study design elements should be improved.

Objective: To identify the relative importance of extrinsic determinants of doctors’ choice of specialty. Design: A self‐administered postal questionnaire. Setting: Australian vocational training programs. Participants: 4259 Australian medical graduates registered in September 2002 with one of 16 Australian clinical colleges providing vocational training programs. Main outcome measures: Choice of specialist vocational training program; extrinsic factors influencing choice of program, and variation by sex, age, marital status and country of birth. Results: In total, 79% of respondents rated “appraisal of own skills and aptitudes” as influential in their choice of specialty followed by “intellectual content of the specialty” (75%). Extrinsic factors rated as most influential were “work culture” (72%), “flexibility of working arrangements” (56%) and “hours of work” (54%). We observed variation across training programs in the importance ascribed to factors influencing choice of specialty, and by sex, age and marital status. Factors of particular importance to women, compared with men, were “appraisal of domestic circumstances” (odds ratio [OR], 1.9), “hours of work” (OR, 1.8) and “opportunity to work flexible hours” (OR, 2.6). Partnered doctors, compared with single doctors, rated “hours of work” and “opportunity to work flexible hours” as more important (OR, 1.3), while “domestic circumstances” was more important to doctors with children than those without children (OR, 1.7). In total, 80% of doctors had chosen their specialty by the end of the third year after graduation. Conclusions: Experience with discipline‐based work cultures and working conditions occurs throughout medical school and the early postgraduate years, and most doctors choose their specialty during these years. It follows that interventions to influence doctors’ choice of specialty need to target these critical years.