분석적 사고력(Analytic Reasoning)향상이 진단 정확도에 미치는 영향 : 무작위 대조군 연구

(출처 : http://blogs.utas.edu.au/snm-pep/2013/03/26/clinical-reasoning-resources-for-professional-experience-facilitators/)

진단의 실수는 예방가능한 의료과오 중 상당부분을 차지하고 있으며, 다양한 인지적 비뚤림(bias)에 의해서 일어난다. 흔히 생기는 것으로서 Premature closure, Anchoring, Confirmation bias 등이 있다.

Diagnostic errors comprise a large proportion of preventable medical errors and can be a result of numerous types of cognitive bias (Graber et al. 2005). Common cognitive bias include premature closure (‘‘the tendency to stop considering other possibilities after reaching a diagnosis’’), anchoring (‘‘the tendency to perceptually lock onto salient features too early in the diagnostic process’’), and confirmation bias (‘‘the tendency to seek data to confirm, not refute, the hypothesis’’) (Croskerry 2003).

분석적 사고력은 이러한 비뚤림에 대응하는 한 가지 방법으로서, 다른 진단의 가능성을 고려하면서 가설을 지지할 수 있는 근거 뿐만 아니라 반박할 수 있는 근거도 모아야 한다. 그러나 분석적 사고력이 진단의 실수를 줄일 수 있다는 연구는 별로 없다. 표준화 환자(standardized patient, SP)를 활용한 OSCE시험 세팅에서, 우리는 분석적 사고력을 높여주는 것이 진단의 정확도를 높여준다는 것을 무작위 대조군 연구(RCT)로 확인하였다..

Analytic reasoning, one method used to combat bias, is a reflective approach that involves the consideration of alternative diagnoses and of evidence to refute a diagnosis as well as to support it (Norman & Eva 2010). However, there are few studies evaluating the effect of analytic reasoning on reducing diagnostic error (Mamede et al. 2010a, b). We conducted a randomized controlled study to determine whether the enhancement of analytic reasoning increases medical students’ diagnostic accuracy on the objective structured clinical examination (OSCE) using standardized patients.

환자와 대면했을 때 analytic reasoning group의 학생들은 analytic reasoning을 높여줄 수 있도록 디자인 된 표를 받았다. 그리고 이 표를 작성하고 진단을 추측하도록 했다.

After encountering patients, students in the analytic reasoning group received an answer sheet containing a table designed to enhance analytic reasoning. Students in the analytic reasoning group were asked to complete the table with differential diagnoses and symptoms or signs compatible with or differing from each diagnosis, and then provide one most probable diagnosis.

Control group은 analytic reasoning group에게 주어졌던 표는 주지 않고, 진단을 추측하도록 했다.

Students in the control group were simply asked to provide one most probable diagnosis on an answer sheet that lacked the table provided to the test group

Analytic reasoning group에서 진단의 정확도가 더 높았다.

Mean diagnostic accuracy scores were significantly higher in the analytic reasoning group than in the control group (3.400.66 versus 3.050.98; 95% CI of difference, 0.08– 0.62; p¼0.016). In the analytic reasoning group, diagnostic accuracy did not change significantly with respect to case order (p¼0.93 for trend).

---

Med Teach. 2013;35(3):248-50. doi: 10.3109/0142159X.2013.759643. Epub 2013 Jan 18.

Effect of enhanced analytic reasoning on diagnostic accuracy: a randomized controlled study.

Myung SJ, Kang SH, Phyo SR, Shin JS, Park WB.

Source

Seoul National University College of Medicine, Republic of Korea.

Abstract

BACKGROUND:

Diagnostic error can be caused by several types of cognitive bias, which may be reversed by enhancing analytic reasoning.

AIMS:

To evaluate whether enhancing analytic reasoning can increase diagnostic accuracy in objective structured clinical examination (OSCE) in medical students.

METHODS:

All fourth-year medical students, randomly assigned to the analytic reasoning or control groups, undertook the OSCE with four cases using standardized patients. The analytic reasoning group was requested to list differential diagnoses and findings compatible or not compatible with each diagnosis prior to providing a diagnosis, while the control group provided a diagnosis without these processes. Mean diagnostic accuracy scores (perfect score, 4.0) from four cases of OSCE were compared between the two groups.

RESULTS:

One hundred forty-five students were randomly assigned to the analytic reasoning group (n = 65) or the control group (n = 80). The baseline characteristics, including grade point average and the scores from each patient encounter, were comparable between groups. Meandiagnostic accuracy scores were significantly higher in the analytic reasoning group than in the control group (3.40 ± 0.66 versus 3.05 ± 0.98; p = 0.011).

CONCLUSION:

Enhancement of analytic reasoning may improve diagnostic accuracy in novice doctors.