• Getúlio Rodrigues de Oliveira Filho, Pablo Escovedo Helayel, Diogo Brüggemann da Conceição, Ivo Sebastião Garzel, Patrícia Pavei, and Maurício Sperotto Ceccon.
• Department of Anesthesiology, Hospital Governador Celso Ramos, Nucleus for Teaching and Research in Medical Education, Florianópolis, SC, Brazil. grof@th.com.br
• Anesth. Analg. 2008 Feb 1; 106 (2): 568-73, table of contents.

BackgroundWe aimed to construct learning curves and mathematical learning models for ultrasound basic skills: optimizing needle-ultrasound beam alignment and reaching a target inside a phantom.MethodsThirty subjects participated in the study. Each subject performed 25 trials. Linear ultrasound probes and a bovine muscular phantom were used. In Experiment 1, subjects tried to insert a needle parallel to the ultrasound beam with full imaging of the needle. For Experiment 2, a segment of tendon was inserted longitudinally into the phantom at a depth of 1 to 1.5 cm. Subjects tried to insert the needle until contacting the tendon. Learning curves were constructed using the cumulative sum (cusum) method. Bush and Mosteller's mathematical learning models were constructed for each skill.ResultsOnly 30% and 11% of subjects attained proficiency in Experiments 1 and 2, respectively. The predicted average numbers of trials to achieve 95% success rates as estimated from Bush and Mosteller's models were 37 and 109, respectively.ConclusionsLearning interventional ultrasound basic skills may require a considerable number of trials. Cusum charts revealed that individuals acquire such abilities at variable rates. As skills were assessed in phantoms, our results do not apply to blocks given to real patients.

21 keywords...

hide…