Sample

We included n = 30 physicians in the study (20 female), of which 17 were in specialist training, 13 were medical specialists in endocrinology. Mean age was 35.53 ± 7.35 (range 27–54 years) and the mean duration of work experience in endocrinology was 6.50 years ± 6.98 (range 0.5–26). The approximate number of treated acromegaly patients each year was on average 17.20 ± 16.80 (range 0–60).

Patient-doctor interactions (n = 60)Time-categories

For the patient-doctor interaction, averaged time-categories over all 60 talks yielded that the greatest part of the conversation contained the topic ‘intervention’ (40.3% ± 22.4), followed by ‘symptoms’ (18.0% ± 10.1), ‘QoL’ (13.2% ± 11.5), ‘IGF-I’ (8.9% ± 5.6), ‘prognosis’ (8.5% ± 9.2), ‘medical history’ (7.6% ± 6.7), ‘tumor’ (7.3% ± 6.9) and ‘comorbidities’ (6.6% ± 8.6).

The proportion of each category depended strongly on the focus of the patient profile. For detailed results confer to Figure 3.

Fig. 3

Percentage conversation time divided by the four SPs

Yes/no-categories

Analysis of the yes/no categories for all four SPs for medical details showed that the issue of comorbidities was raised in almost half of the conversations, but interestingly those relevant to the particular patient profile were identified in only one third of the consultations.

Concerning interaction, doctors listened attentively to patients in 81.7% of the consultations. In almost every conversation an opinion on the disease activity and an explanation for this opinion was given.

Shared decision-making in the sense of reaching an agreement with the patient for further procedure was achieved in in 98.3% of the conversations. The patient’s desired involvement in the decision, however, was only asked in 18.3% of cases. For detailed results confer to Table 1.

Table 1 Percentage of yes in yes/no-categories in all 60 conversations

Concerning the specified learning points for the four SPs, stable acromegaly was recognized in the two patients with comorbidities (patient 1 comorbidity depression / patient 2 comorbidity joint pain) with different frequencies (66.7% vs. 100%). In patient 1, recognition of stability (66.7%) and explaining lack of need for therapy change (58.3%) were almost identical in percentage, whereas in patient 2 there was a difference of more than 30.0% in this regard. The respective relevant comorbidity was mentioned explicitly in 66.7% (depression) vs. 94.4% (arthropathy) of the talks. In only 25.0% of conversations it was explained to patient 1 that the current symptoms are an expression of the comorbidity depression. 50.0% of the physicians suggested a therapy for treating the depression, while 83.3% made a suggestion for treating the joint pain.

The fact that acromegaly was florid in patients 3 (therapy escalation) and 4 (change in intervention) was recognized in 83.3% and 100% of the talks. In line with this, a change of therapy was suggested in 91.7% and 100%.

For detailed results confer to Fig. 4.

Fig. 4

Specific categories of the four SPs with superordinate learning points and their mentioning in the talk

SP1: depression

As depression was the least frequently recognized and addressed comorbidity, all case specific yes/no-categories for patient 1 were analyzed in detail, including differences with respect to the (non-)presence of the ACRODAT® profile (see Fig. 5 in supplemental material).

Symptoms of depression were far more recognized in the conversations without ACRODAT® than with the tool, whereby the criterion for addressing the patients’ discomfort about the appearance, shame about nose and feet, lack of desire for sex, mother´s depression and rumination was met in less than half of the talks overall. Accordingly, the depressive episode was mentioned more often without ACRODAT® (41.7%) than with the tool (25.0%). On the other hand, in a larger proportion of the talks with than without ACRODAT® (16.7% vs. 8.3%) it was explained that the current symptoms are predominantly not of a physical nature, but an expression of the depressive episode. A therapy suggestion regarding the depressive episode was made in 25.0% of the conversations with and without the tool each. Only in 25.0% of the interviews with ACRODAT® and in 41.7% of the conversations without ACRODAT® it was recognized that the acromegaly is stable. While 91.7% (50.0% without vs. 41.7% with ACRODAT®) understandingly accepted the patients’ wish for further MRI diagnostic, only 16.7% (all with ACRODAT®) declined this request.

ACRODAT ® tool (n = 30, across subjects)

The Wilcoxon Test yielded no significant differences comparing the time-categories of the conversations (e.g. medical history, QoL) with and without ACRODAT® tool (all Z≤ |1.129|, p ≥ .259).

No significant differences in the yes/no-categories were found using the McNemar test, meaning that the ACRODAT® tool did not influence the quality of the conversations in terms of medical details, interaction and shared decision-making (all p ≥ .065).

Gender differences and professional experience of participants

There were no gender differences of the physicians concerning the conversation parts (all Z≤ |1.408|, p ≥ .159).

Comparing the differences in the conversation parts between two groups of physicians divided by the median number (= 12.5) of acromegaly patients treated per year, there was a significant difference in the percentage of the conversation about medical history (Z=-2.426 p = .015) with physicians who treat more acromegaly patients spending more time on the exploration of the medical history; all other topics did not differ significantly (all Z≤|1.390|, p ≥ .165).

Similarly, there were no significant differences in percentage of the conversation topics between physicians that were medical specialists in endocrinology and those still in education (all Z≤ |1.507|, ≥p.132).

Rating of the ACRODAT® tool by physicians

83.3% found the profile provided by the software tool clear and in logical order and about half of the physicians stated that they found the tool helpful in detecting comorbidities (53.3%) and symptoms that were not addressed (66.7%). Nevertheless, less than half of the physicians would use the tool in their own medical practice (40.0%). For detailed results confer to Fig. 6 in supplemental material.

Order effects of ACRODAT® first or second were checked; the tool was rated worse for all 9 aspects if used in second place, with significant differences for 2 of these aspects (Z≤|2.083|, p ≤ .037).

Comparing physicians in medical education and medical specialists in endocrinology concerning the rating of the software tool, all of the rating categories did not differ significantly (all Z≤|1.560|, p ≥ .119).

Rating of the interactions by the SPs

The simulation persons´ subjective ratings indicated that they were very satisfied with the conversations. With regard to the order of ACRODAT® first or second, the SPs showed a better evaluation of the conversation if ACRODAT® was used first (all Z≤|2.245|, p ≤ .025). However, rating with vs. without ACRODAT® yielded no significant differences in the categories ‘attentive listening’, ‘empathy’, ‘comprehensibility’, ‘responding to questions’, ‘involving the patient in the decisions’, ‘taking time for the patient’, ‘expertise’ (all Z≤|1.453|, p ≥ .146). For detailed results confer to Fig. 7 in supplemental material.