Regret (Bereuen) von diagnostischen Entscheidungen in der Primärversorgung - Fallvignettensurvey unter Allgemeinärzten

Background General physicians have to take diagnostic decisions for their patients almost every day. In medicine, diagnostic errors are natural even if the consequences and longterm (side) effects for patients and doctors may vary considerably. The severity of diagnostic errors ‒ defined as a discrepancy between the assumed diagnosis and the final diag- nosis ‒ between diverse diseases with one cardinal symptom can be judged by weighting the strength of regret as an evaluation tool. Regret is thus defined as an intuitive, stressful reaction that has both, emotional and cognitive aspects, and arises when an action or decision that initially felt correct has turned out to be inferior to a possible way than the one chosen. We suggest that the feeling of subjective regret has a greater impact on diagnostic decision making than many other medical qualifications of a general physician. In our study, we will consider this emotion regarding the symp- tom chest pain and its different connected disorders which excludes the determination of any specific disorders. Our aim is to examine the impact of the discrepancy between the assumed and the final diagnosis ‒ in our study defined as diagnostic error ‒ with regard to the level of regret among general physicians. Factors like gender, professional experience, physician ́s age or the number of inhabitants in the surgery location will be tested with regard to a possible relationship with the strength of regret. Methods The German-speaking general physicians were invited to participate in our study via e- mail distribution lists of several German and Austrian universities in December 2016 and June 2017. We administered a visual analog scale (range 0 to 100) to measure individual differences in regret of decision-making from patients with the symptom 'chest pain' in a primary care setting. Every participant had to value nine of 72 possible, different case reports and every vignette included the comparison between an assumed, initial diagnosis and a final diagnosis. Nine possible medical outcomes have been matched to three disease categories. The categories differ between their progno- sis and therapy because of their different disease severity. We examined the influence of different serious medical outcomes to the strength of regret. Every participant working as a licensed physician in family medicine or as a consulting physician in the ambulant medical sector was included in our study. The web-based survey was comprised of questions on participant demographics (i.e. age, level of train- ing, gender) and workplace information. The questionnaire was presented with the online tool LimeSurveyTM. For data evalua- tion we used the program SPSS by IBM (Version 22). The survey was approved by the ethic committée (AZ 35/14) of the Philipps-University of Marburg. Results 313 general physicians participated in our regret study, but only 254 questionnaires were completely finished. 53.8 % of the participants were male, average age was 51 years and in general, participants worked as general physicians for 17 years. 47.6 % of the general physicians have worked in a group practice, 35 % of them lived in a city with more than 100.000 citizens. The more life-threatening the final diagnosis that was initially not recognized as such, the higher the regret. The two oversight diagnosis myocardial infarction and pulmonary embolism evoked the greatest regret, especially if the general physicians expected a low life-threatening medical outcome. If an assumed medical diagnosis changed through diagnostic researches into a non-life threatening disease, regret was reported to be very low. We couldn ́t prove any close correlation between the strength of regret and the physician ́s age, professional experience or the number of inhabitants in the surgery loca- tion. Female general physicians point to a greater, non-significant regret (p>0.05) than their male colleagues. Discussion Our study among general practitioners shows that regret correlates with the discrepan- cy between the assumed diagnosis and the final diagnosis in chest pain depending on the severity of the diagnosis. The extent of regret should be considered as a relevant factor in the development of (digital) decision support systems and decision rules for physicians. Further research on the impact of Regret in the general medical context is essential.