A research coding method for the basic patient-centered interview.

Area of Science:

• Medical Education
• Communication Skills Assessment
• Qualitative Research Methods

Background:

• Effective medical interviewing requires proficiency in both data gathering and emotion handling.
• Existing coding methods may lack reliability or comprehensive assessment of patient-centeredness.
• Standardized coding is crucial for evaluating physician-patient interactions and training effectiveness.

Purpose of the Study:

• To develop and validate a reliable coding method for medical interviews.
• To specifically assess residents' skills in data gathering and emotion handling during patient interactions.
• To establish a criterion standard for research coding of patient-centered communication.

Main Methods:

• Two trained undergraduates rated 25 videotaped simulated patient (SP) interviews.
• A coding scheme with 33 dichotomized items was used, derived from an earlier Likert scale method.
• Inter-rater reliability was measured using Guetzkow's U, Cohen's Kappa, and percent agreement.

Main Results:

• High inter-rater reliability was achieved, with overall Kappa at 0.90 and percent agreement at 97.5%.
• Individual item agreement ranged from 84% to 100%, indicating consistent coding.
• The coding method effectively captured residents' skills in eliciting various patient story components and overall patient-centeredness.

Conclusions:

• A simple, highly reliable coding method for medical interviews was successfully developed.
• The method is weighted to emphasize personal and emotional elements, enhancing its clinical relevance.
• This validated coding procedure is recommended as a criterion standard for research and can inform patient satisfaction tools.