The four participating casualty clinics registered a total of 832 patients with chest pain as their main symptom, of which the first 100 unique patient and physician pairs, with completed structured telephone interviews, were included in the study.
The included patients’ (n = 100) age ranged from 18 to 92 years (median age 46 years), 58% males with a median age of 45 years, and 42% females with median age 51 years. The study included 60 male physicians and 40 female physicians. GPs constituted 67%, the rest were interns in general practice (11%) or hospital-based physicians (22%).
Table 2 describes the physicians’ approach to diagnosing patients with chest pain by registering the selected importance of different aspects of the diagnostic process. 99% believed that the patient’s symptoms and history was fairly (19%) or very important (80%) (mean 4.8/5 on Likert scale), while all of the physicians stated that a “positive” ECG-finding was fairly (10%) or very important (90%) (mean 4.9). “Negative” ECG-findings (mean 2.8) and effect of sublingual nitro-glycerine (mean 3.0) were considered to be the least important aspects.
Figure 1a and b show the risk score sums from the Pearson Risk Scale (Figure 1a) and Tolerance of Risk Scale (Figure 1b). Both scales divide the physicians into three groups; “risk-avoiding”, “middle-scorers” and “risk-seeking”.
Table 3 presents “physician risk attitudes” derived from the Pearson Risk Scale. There was no significant difference in the length of work experience between male and female physicians (p = 0.072). The “middle-scoring” group constituted two thirds (66 of 100), while the groups “risk-avoiders” and “risk-seekers” were equally divided with 17 physicians each. When analysing “risk-avoiders” against the rest, we found no significant differences in length of experience (p = 0.155) or gender (p = 0.913). Analysing “risk-avoiders” against the rest using the Tolerance of Risk scale also showed no significant differences (length of experience p = 0.085; gender p = 0.148).
Table 4 describes the physicians’ tolerance of risk and uncertainty (dimension A) and concerned all patients out-of-hours. The strongest agreement in dimension A was found in the statement “I think my risk assessment is reasonably good, and I’m reasonably safe”, in which 94% agreed to the statement (67% a little; 27% strongly; mean 4.2). We found the weakest agreement in the statement “I don’t worry about my decisions after I’ve made them”, 46% disagreed (5% strongly; 41% a little), while 50 % agreed (42% a little; 8% strongly).
The other three dimensions (B-D) concerned chest pain patients only. Dimensions B – D measured attitudes to hospital admission, including patient related and relative related influence on decision making.
In dimension B, we found that half of the physicians (51%, mean 3.0) worry about complaints being made about them, but few let fear of complaints from the Board of Health Supervision influence their practice (16%, mean 2.1).
Dimension C examined attitudes to hospital admission. 69% (mean 3.6) agreed that admitting someone to hospital enables them to get a second opinion, but 75% (mean 3.7) also agreed that admitting someone to hospital put patients in danger of being “over-tested”.
The last dimension (D) concerned patient-related factors. There was a strong agreement that the patient’s clinical status was the most important factor (96% agreed, mean 4.6) in deciding to admit a patient or not. Half of the physicians were more likely to admit the patient if the patient himself wanted to be admitted (51% agreed, mean 3.2), or if a family member wanted the patient to be admitted (46% agreed, mean 3.1).
Overall mean scores from all items in the four dimensions were also compared with mean scores within the three risk groups derived from the Pearson Risk Scale. In dimension A, concerning all patients out-of-hours, there is a clear trend in most items that the “risk avoiders” differ from the rest, and there is a significant difference in the statement “When it comes to OOH-medicine I’m quite cautious” (p = 0.024). In dimension B, we found a significant difference in the statement “I don’t worry about a complaint being made about me” (p = 0.006), where the group “risk avoiders” had a mean score of 2.2 versus the mean score of 3.2 for the rest of the physicians. There were no significant differences when testing the “risk avoiders” against the rest in each of the five items in dimension C. In the last dimension (D), we found significant differences in the statements “I am more likely to admit a person if they want to be admitted” (p = 0.039), “If members of the family say there’s nobody to look after someone, I see that as a problem for the family rather than the doctor” (p = 0.034) and “I am more likely to admit someone if they live alone” (p = 0.008).