Relata-se a experiência da realização de prova prática no concurso para seleção de médicos residentes no Centro Universitário Serra dos Órgãos, Teresópolis (RJ). Houve uma primeira fase com prova teórica (PT) de múltipla escolha e uma segunda fase com prova prática presencial (PP), ambas eliminatórias e classificatórias. As cinco áreas básicas do ensino médico foram avaliadas por meio do exame clínico de dois pacientes - um clínico ou cirúrgico e outro da área materno-infantil, conforme sorteio. A PP objetivou avaliar atitudes, habilidades e conhecimentos. Houve capacitação de dois tipos de bancas: a primeira com docentes das áreas clínica e cirúrgica, dita clínico-cirúrgica (CC), e a segunda com docentes das áreas de pediatria, tocoginecologia e saúde coletiva, chamada materno-infantil e preventiva (MIP). Não houve correlação significativa entre as notas da PT e as da PP. A PP foi determinante para a classificação final em cinco dos seis programas oferecidos. Houve correlação entre as notas das bancas CC e MIP (r = 0,483, p < 0,001). Houve uniformidade das notas intrabanca e interbanca. A PP é exeqüível, segura e importante instrumento para a adequação do ensino de graduação em Medicina às diretrizes curriculares.
This paper relates the experience of a hands-on test model for the selection of medical residents in Serra dos Órgãos Universitary Center, Teresópolis, RJ. There was a first phase with a multiple-choice theoretical test (TT) and a second one with hands-on evaluation performed at the bedside (HO), both with eliminatory and classificatory effect. The five basic areas of medical teaching were evaluated through clinical examination of two patients: one from the clinical or surgical units and another from the maternal-child unit, selected at random. The HO test aimed to evaluate attitudes, skills and knowledge. There were two types of examination boards: the first was composed by teachers from the clinical and surgical areas, the so called clinical-surgical (CS) area and the second by teachers from the pediatric, obstetric and public health areas, called maternal-child and preventive (MCP) area. There was no correlation between TT an HO. The HO was responsible for final classification in five of the six residency programs offered. There was a correlation between CS and MCP scores (r = 0.483, p < 0.001) and uniformity in the evaluation in each board as well as between the different boards. The HO evaluation model showed to be feasible and secure and is an important instrument for adapting medical teaching to the Brazilian curricular guidelines.