The assessment of a specific aspect of the corporate image was through a survey in Spanish (Table 1) that aimed to determine which definition – that of the orthopedist, the rheumatologist, or the rehabilitator – is most meaningful to respondents in the context of four different scenarios. We used the orthopedist definition in Spanish provided by the American Academy Orthopaedic Surgeons.6 The rheumatologist's definition was provided by the CMR,7 and the definition of a rehabilitator was provided in Spanish by the Mayo Clinic.8 The first posed scenario was chronic hand pain, the second was about chronic knee pain, and the third was about arthritis in the hands. In a fourth question, they were asked to choose one of the three definitions that best define the rheumatology specialist (Table 1).

The order of options for each studied question was randomized to reduce order bias, including primacy and recency biases.9 All responses were treated as nominal variables. Additionally, to minimize survey fatigue, the structured survey comprised only nine questions: four studied questions, one exploratory question, and four demographic inquiries. The exploratory question was about the perceived influencers or mavens when about to seek medical consultation by asking respondents: If you had arthritis, who would you ask who to go to?

We surveyed four independent samples to optimize resources and compensate for the absence of reference data in our proof-of-concept study. The first three were non-random samples drawn from the family and friends of three medical students during the same week. Invitations were extended to individuals 18 or older and with at least elementary-level schooling. The invited individuals resided in the metropolitan area of Guadalajara and the surrounding small towns. The invitations were delivered through WhatsApp and were accompanied by a link to Google Forms. A response window of 11 days was provided. Using the Central Limit Theorem for proportions, we determined that gathering 100 cases for each sample would yield information with a normal distribution. The fourth sample was assessed two months later utilizing the services of a polling firm. This sample was gathered through simple, on-street, random interviews conducted in the metropolitan area of Guadalajara from 7 to 11 July 2023. The participants were 18 or older, and we ensured that half of the respondents were women. A minimum of 350 respondents were included in the sample to achieve a reliability level greater than 90% and an error margin of 5%.

We assess the CMR's corporate identity using discourse analysis and a readability assessment from the CMR website.10 Discourse analysis is a qualitative research method used to study written (or spoken) language in the context of its social surroundings.11 By analyzing the CMR website's written content, linguistic patterns, and the framing of information, we aimed to have insights into the CMR's identity projection and its engagement with its target audience. This approach offers a comprehensive perspective on how the CMR shapes its image and message, contributing to a better understanding of its overall corporate communication strategy. This analysis was performed by a trained expert (A V-A) unaware at the time of the aims and results of the surveys. The detailed methods and results are in Supplementary file 1.

For the readability assessment, we used five readability indices for Spanish text: Lecturabilidad de Fernández Huerta, Comprensibilidad de Gutiérrez de Polini, Índice de perspicuidad de Szigriszt-Pazos, Escala INFLESZ, and Legibilidad μ de Muñoz y Muñoz. Each index provides an approximation for the difficulty of the text and has been used to evaluate and improve the clarity of a written communication [reviewed in 11]. The analysis used a devoted website12 with algorithms written in Python for each of the six indices. Each index scores from 0 to 100, where 100 is a very easy-to-read text.

This work did not include any experimentation with humans or animals. The first three samples were made through WhatsApp from each student's contacts. There was no addition of any new contacts for this study, and each student invited only their own contact list. The invitation to answer a questionnaire was generic and included a link to Google Forms. The questionnaire did not record the respondent's telephone number or any other sensitive information. The fourth survey was anonymized to the general population. The information collected did not contain sensitive data and did not allow the identification of any particular individual. Therefore, the study did not require evaluation by an ethics committee.

We hypothesized that at least 60% of respondents would select the rheumatology definition for each scenario. The analysis relied on four underlying assumptions. First, each of the four samples was considered independent, and no effort was made to control for potential confounders or biases, thus limiting the ability to determine the validity, reliability, and generalizability of the findings. Second, each convenience sample complements the others. We further reported frequency distribution tables for the total surveyed population. Third, as all variables were nominal, the differences were assessed using the Pearson Chi-square test (χ2) to evaluate independence. We use Phi or Cramér's V to determine association strength. Confidence intervals for proportions were calculated at a 95% level, and the two-tailed significance level was set at p≤0.05. Fourth, we added an estimate of the distance between specialties by dividing the rheumatology proportion for each scenario by the corresponding proportion for orthopedics and rehabilitation. Therefore, for every rheumatologist chosen, an estimated n number of orthopedists and an n number of rehabilitators were chosen.

Table 3 and Fig. 1 show the frequency distribution of responses across scenarios and samples. In the hand scenario, the frequencies were similar among the four samples, with a slight preference for orthopedics. However, orthopedics emerged as the most preferred choice in the knee scenario despite significant differences among samples. The arthritis scenario displayed significant preference differences among samples, with a slightly higher selection frequency for rheumatology. When defining a rheumatologist, the differences among samples were also significant, with orthopedics and rheumatology preferences nearly equal.

Fig. 1.

Visual representation of the distances between specialties.

(0.13MB).

Fig. 2 provides a visual representation of the distances between specialties. On average, respondents opted for 1.13 orthopedists and 0.70 rehabilitators in the hand scenario for each rheumatologist chosen. In the knee scenario, respondents chose 2.36 orthopedists and 0.64 rehabilitation specialists, whereas a preference for 0.85 orthopedists and 0.58 rehabilitators was observed in the arthritis scenario. Finally, 1.0 orthopedists and 0.67 rehabilitators were selected for each rheumatologist when defining a rheumatologist.

Fig. 2.

Frequency distribution of responses across scenarios and samples.

(0.15MB).

Age was associated with the choice of a specialist in the hand scenario. The youngest age group (i.e., ages 18–30 years) preferred orthopedists over rheumatologists (50.0% vs. 31.2%), and the 41–50-year age group more frequently chose the rheumatologist over the orthopedist (45.8% vs. 34.7%, χ2=17, DF=6, p=0.009; Phi=0.156). Age was also associated with the choice of definition of rheumatology. The 31–40-year age group preferred rheumatology (45.1%) or orthopedics (40.7%) rather than rehabilitation (14.3%; χ2=18.6, DF=6, p=0.005). Age was not associated with specialty in the knee (p=0.263) or arthritis (p=0.769) scenarios.

Schooling was significantly associated with a preference for a specialist in the arthritis scenario. As schooling progressed, the choice of rheumatology increased from 27.3% for those in primary school to 29.4%, 38.5%, and 49.4% for those in secondary school, high school, and university, respectively (χ2=27, DF=6, p<0.001, Cremér's V=0.13). Schooling was also significantly associated with the choice of definition of rheumatology. Respondents with middle (28.4%) and high schooling (27.0%) chose the CMR's definition less frequently than individuals with elementary (34.5%) and university (46.4%) (χ2=26.9, DF=9, p<0.001, Cremér's V=13).

Sex was not significantly associated with the specialist selection in the hand (p=0.4), knee (p=0.6), arthritis (p=0.4), and rheumatologist definition (p=0.1) scenarios. We observed no association between education and preference for a specialist in the hand (p=0.2) or knee (p=0.6) scenarios. Self-reported rheumatic issues were not associated with the chosen specialist in any of the scenarios: hand (p=0.6), knee (p=0.2), arthritis (p=0.8), or rheumatologist definition (p=0.5).

Table 4 shows the results of the exploratory question about who would be asked which doctor to go to. Although responses differed significantly among samples, the mother figure was the most frequently mentioned in each sample. Men preferred mother, father, and doctor (37%, 15%, and 15%, respectively), while women preferred mother (41.1%), doctor (15.5%), and friend (10.4%) (χ2=28.7, DF=8, p<0.001). Age was associated, as expected, since 52% of the young group mentioned their mother (52%), but just 14.7% of the older group did it. On the contrary, the older group preferred son/daughter (76.5%) (χ2=158.7, DF=24, p<0.001). The influencer was not associated with the scenarios.

The full results of the discourse analysis are included in Supplementary file 1. The analysis showed that the enunciation of the CMR website was discursively positioned as a collective “we” that referred to rheumatology specialists and particularly to “colleagues,” and the patient appears as a “necessary actor” within the medical social practice. Although the patient is presented as a potential recipient (addressee), they are never treated as interlocutor. Still, the patient has no agentive representation (i.e., no possibility of intervention in their diagnosis and healing process). Specifically, the physical “patient” was also discursive: they talk about the patient, but the patient is not allowed to act.13 The semiotic structure of the CMR's mission and vision is imprecise and lacks a statement of values and purpose. The overall readability scores corresponded to very difficult and arid text. Specifically, the scores for the Lecturabilidad de Fernández Huerta, the Comprensibilidad de Gutiérrez de Polini, the Índice de perspicuidad de Szigriszt-Pazos, the Escala INFLESZ, and the Legibilidad μ de Muñoz y Muñoz were 26, 29, 21, 21, and 42, respectively.