Factors Associated With the Quality of the Patient-Doctor Relationship: A Cross-Sectional Study of Ambulatory Mexican Patients With Rheumatic Diseases.

BACKGROUND: The patient-doctor relationship (PDR) is a complex phenomenon with strong cultural determinants, which impacts health-related outcomes and, accordingly, does have ethical implications. The study objective was to describe the PDR from medical encounters between 600 Mexican outpatients with rheumatic diseases and their attending rheumatologists, and to identify factors associated with a good PDR.
METHODS: A cross-sectional study was performed. Patients completed the PDRQ-9 (Patient-Doctor Relationship Questionnaire, 9 items), the HAQ-DI (Health Assessment Questionnaire Disability Index), the Short-Form 36 items (SF-36), a pain-visual analog scale, and the Ideal Patient Autonomy Scale. Relevant sociodemographic, disease-related, and treatment-related variables were obtained. Patients assigned a PDRQ-9 score to each patient-doctor encounter. Regression analysis was used to identify factors associated with a good PDR, which was defined based on a cutoff point established using the borderline performance method.
RESULTS: Patients were primarily middle-aged female subjects (86%), with substantial disease duration (median, 11.1 years), without disability (HAQ-DI within reference range, 55.3%), and with deteriorated quality of life (SF-36 out of reference range, 73.7%-78.6%). Among them, 36.5% had systemic lupus erythematosus and 31.8% had rheumatoid arthritis. There were 422 patients (70.3%) with a good PDR and 523 medical encounters (87.2%) involved certified rheumatologists.Patient paternalistic ideal of autonomy (odds ratio [OR], 3.029; 95% confidence interval [CI], 1.793-5.113), SF-36 score (OR, 1.014; 95% CI, 1.003-1.025), female sex (OR, 0.460; 95% CI, 0.233-0.010), and being certified rheumatologist (OR, 1.526; 95% CI, 1.059-2.200) were associated with a good PDR.
CONCLUSIONS: Patient-related factors and the degree of experience of the attending physician impact the quality of the PDR, in Mexican outpatients with rheumatic diseases.

Rheumatic diseases comprise a complex group of chronic musculoskeletal conditions, primarily characterized by increased mortality of the affected patients and musculoskeletal pain, which substantially contributes to deteriorating a patient's quality of life.[1] Rheumatologists are considered essential physicians for patients with rheumatic diseases, and early access to these specialists is considered imperative to achieve appropriate outcomes. Differences in access to subspecialty care contribute to the known disparities in morbidity and mortality from some rheumatic diseases and have ethical implications.[2] Patient-centered care, defined as the clinical treatment provided by medical professionals, which focuses on respecting patients' preferences, desires, and values,[3] has been proposed as the optimal conceptual model of care for patients affected by rheumatic diseases.[4] It is founded on the basis of the patient-doctor relationship (PDR), traditionally characterized by patient's trust in the physician with relevant skills and knowledge and concerned with the patient's best interest.[5]

The PDR is a complex dynamic and multidisciplinary phenomenon with strong cultural determinants, which limits the generalization of the results obtained from studies performed in populations with a different anthropologic framework.[6] In the clinical context of bedside and ambulatory clinical care, a good PDR is highly valuable in itself, primarily due to its positive impact on health outcomes,[7] which has been confirmed in patients with rheumatic diseases.[8-10] A recent systematic review of the topic in the field of rheumatology examined factors influencing patient-physician communication and the association between the PDR and health outcomes.[11] The review included 6 quantitative,[9,12] 1 mixed method,[13] and 3 qualitative articles[10,14,15] and concluded that better interaction between the patient-doctor dyad was linked to improved outcomes, such as lower disease activity and organ damage, treatment satisfaction, and fewer adverse effects. In addition, active patient participation in medical consultation and trust in the physician were crucial for perceiving the PDR as good or positive.

Furthermore, among Mexican patients and physicians, paternalism has been described as a frequent attitude.[6,16-18] Patients with rheumatoid arthritis (RA), systematic lupus erythematosus (SLE), and other rheumatic diseases were less likely to desire or undertake an active role at the time of their consultation.[18] This was confirmed in our prior study performed in 601 patients, with 10 different rheumatic diagnoses, in which up to 85% had a paternalistic ideal of autonomy.[6] Moreover, paternalism was not objected to by Mexican patients with fibromyalgia, particularly in the context of public health care.[17] Thompson and Whiffen[19] recently suggested that, in some cultural contexts, such as among Latino patients, the sole emphasis on patient autonomy could potentially have negative consequences on the PDR.

With the current work, we aimed to describe the quality of the PDR derived from medical encounters and to identify the factors associated with a good PDR. The objective was intended to answer the following research question: “Which are the patient-related and the physician-related factors associated with a good PDR in adults patients with rheumatic diseases?” We hypothesized that patient-related sociodemographic factors such as age, education level, and sex; the patient-physician concordance in ideal of autonomy; and the physician experience degree might be related to the quality of the PDR.

PATIENTS AND METHODS

Ethics

The study was approved by the Internal Review Board of the INCMyN-SZ (Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán) (reference number: IRE-3005-19-20-1). All the patients from the Outpatient Clinic of the Department of Immunology and Rheumatology (OCDIR) who agreed to participate provided written informed consent.

Before patient enrollment, the study was presented to all the rheumatologists/trainees assigned to the OCDIR during the study period, and all of them agreed to participate.

Study Design, Setting, and Study Population

The study has been previously described.[6] Briefly, it was cross-sectional and performed between July 2019 and February 2020, at the OCDIR of a tertiary-care level and academic center for rheumatic diseases, located in Mexico City. STROBE's guidelines were followed (Appendix, http://links.lww.com/RHU/A394).

The INCMyN-SZ belongs to the National Institutes of Health of Mexico. Patients had Federal government health coverage depending on their socioeconomic level, which was defined by social workers after patient's interview and income-to-needs ratio's assessment. Patients had to pay for their medication, medical assistance, laboratories, and diagnostic imaging studies; however, up to 75% of the patients had at least 70% health coverage.

Eleven certified rheumatologists and 10 trainees in rheumatology were assigned to the OCDIR, and all were self-defined as Mexican. In addition, approximately 5000 patients with at least 1 visit to the outpatient clinic, self-referred as Mexican and with a variety of rheumatic diseases attended the OCDIR. At first visit to the OCDIR, patients were assigned a primary rheumatologist, which was maintained during the entire patient's follow-up, but for patients assigned to trainees in rheumatology, they changed their primary physician every 2 years (training program duration). Patients might be assigned a different primary rheumatologist upon patient's request.

The 10 most frequent diagnoses (n = 4476) based on the attending rheumatologist criteria were SLE in 1652 patients (33%), RA in 1578 (31.6%), systemic sclerosis in 239 (4.8%), systemic vasculitis in 220 (4.4%), primary Sjögren syndrome (PSS) in 190 (3.8%), spondyloarthritis in 174 (3.5%), inflammatory myopathies and primary antiphospholipid syndrome in 150 patients each (3%), mixed connective tissue disease in 94 patients (1.9%), and adult Still disease in 29 patients (0.6%). Finally, 524 patients (10.5%) had other diagnosis.

All the patients who consecutively who consecutively were seen at the OCDIR during the study period, and had a defined rheumatic disease according to the criteria of the attending rheumatologist, were invited to participate. Exclusion criteria included patients on palliative care, with overlap syndrome (but secondary Sjögren syndrome), and with uncontrolled comorbid conditions.

Study Maneuvers

All included patients were invited to evaluate the PDR at the end of their consultation, and to complete the Patient-Doctor Relationship Questionnaire (PDRQ-9)[20] and a PDR Likert Scale. Patients additionally completed the Spanish version of the Health Assessment Questionnaire–Disability Index (HAQ-DI) to assess disability[21] and the Short-Form 36 items (SF-36) to assess health-related quality of life (HRQoL),[22] a visual analog scale (VAS) to assess pain, and the Ideal Patient Autonomy Scale (IPAS).[6] Physicians assigned to OCDIR also completed the IPAS.

Relevant sociodemographic variables (sex, age, formal education, socioeconomic level, religious beliefs, economic dependency, living with a partner, and access to the social security system), disease-related variables (disease duration, years of follow-up at the OCDIR, comorbid conditions and Charlson comorbidity score,[23] participation in clinical trials, previous hospitalizations and number), and treatment-related variables (immunosuppressive treatment and number of immunosuppressive drugs/per patient and corticosteroid use) were obtained from all the patients, in standardized formats after a careful chart review and patient interview to confirm the data.

In all cases, interviews, questionnaires, and scales were applied in an area designated for research purposes by personnel not involved in patient care.

Instruments Description

The PDRQ-9[20] assessed the quality of the PDR experienced by the patient through the quantification of the patient's opinion regarding communication, satisfaction, trust, and accessibility in dealing with the doctor and the treatment that followed. The questionnaire is based on a 5-point Likert scale ranging from 1 (not at all appropriate) to 5 (totally appropriate). PDRQ-9 scores range from 1 to 5, with higher scores translating into a better PDR.

The PDR Likert scale assessed the quality of the PDR experienced by the patient, who is directed to choose among 3 options: inferior, borderline, and superior.

The IPAS[6] is a self-administered questionnaire that assesses patient' ideal of autonomy according to 4 subscales that can be further grouped into 2 subscales: patients with an ideal of physician-centered/paternalistic (with information) autonomy and patients with an ideal of patient-centered autonomy. The IPAS can be applied to the attending physician.

Definitions

A good PDR was defined based on a cutoff point established with the borderline performance method.[24] Briefly, the PDRQ-9 of the patients who rated the PDR Likert scale as borderline were selected (n = 267), and their mean score was calculated as 3.73. Patients who scored the PDRQ-9 with a value >3.73 were considered to have a good PDR, and their counterpart were considered to have a deficient PDR.

Senior rheumatologists were defined as certified rheumatologists with ≥20 years of clinical experience. Certified rheumatologists were defined as rheumatologists who completed their training program and certification process.

Statistical Analysis

Descriptive statistics were performed to estimate the frequencies and percentages for categorical variables and the median, interquartile range (IQR) for continuous variables, of the sociodemographic variables, the disease-related and treatment-related variables, the patient-reported outcomes, and the PDR of the study population.

A PDRQ-9 score was assigned to each patient-rheumatologist encounter. Characteristics of patients with a good PDR were compared with those of patients with deficient PDR, using appropriate tests. Logistic multiple regression analysis was used to establish factors associated with a good PDR, which was considered the dependent variable. The selection of the variables to be included was based on statistical significance in the bivariate analysis (p ≤ 0.10), and a limited number of potential confounder variables was also considered. In addition, the number of variables to be included was previously defined to avoid overfitting the model, and correlations between variables were also analyzed.

Missing data were below 1% and applied to SF-36 questionnaire, 2 missing data; no imputation was performed. In addition, only 496 patients (82.6%) had a predominant ideal of autonomy,[6] and their data were included in the regression analysis.

All statistical analyses were performed using Statistical Package for the Social Sciences version 21.0 (SPSS; Chicago, IL). A value of p < 0.05 was considered statistically significant.

RESULTS

Population Characteristics

A total of 691 ambulatory patients were invited to participate, 90 of whom declined the invitation, primarily due to time constraints, and 1 patient did not complete the PDRQ-9. The characteristics of the 600 patients are depicted in Table 1 and had been previously described.[6] Briefly, patients were primarily middle-aged women (86%), with a medium-low socioeconomic status (88.8%), long-standing disease (51.8%), comorbid conditions (58.7%), pain under control (68.5%), no disability (55.3%), and HRQoL out of the reference range (73.7%–78.6%), based on published cutoffs for the pain-VAS, HAQ-DI, and SF-36.[25] In addition, they were on immunosuppressive drugs (96.5%). Patients scored high on the PDRQ-9, and 30.5% of the patients rated the PDR with the highest score. The patients' diagnoses were as follows: 219 patients (36.5%) had SLE, 191 (31.8%) had RA, 42 (7%) had systemic vasculitis, 23 each (3.8%) had inflammatory myopathy and primary antiphospholipid syndrome, 25 (4.2%) had systemic sclerosis, 28 (4.7%) had spondyloarthritis, 20 each (3.3%) had PSS and mixed connective tissue disease, and 9 patients (1.5%) had adult Still disease.

TABLE 1

Population's Characteristics (N = 600)

Sociodemographic Characteristics		Disease-Related Characteristics
Female sexa	516 (86)	DD	11.1 (5.3–19.4)
Years of age	48.7 (36.1–59.5)	DD <5 ya	133 (22.2)
Years of formal education	12 (9–16)	DD of 5–10 ya	156 (26)
Medium-low SE levela	533 (88.8)	DD > 10 ya	311 (51.8)
Religious beliefsa	544 (90.7)	Years of follow-up at OCDIR	9.3 (4.1–17.1)
Economic dependencya	356 (59.3)	Comorbid conditionsa	352 (58.7)
Living with a partnera	305 (50.8)	Charlson score	1 (1–2)
Access to SSSa,b	105 (17.5)	Research trials participationa	65 (10.8)
Patient-reported outcomes		Previous hospitalizationsa	81 (13.5)
Pain-VAS score	13 (1–40)	No. previous hospitalizationsc	1 (1–1)
Pain-VAS score ≤30 mma	411 (68.5)	Disease-related treatment
HAQ-DI score	0.38 (0–1.13)	IDsa	579 (96.5)
HAQ-DI ≤0.5a	332 (55.3)	No. ID/patient	1 (1–2)
SF-36 global score	61 (46.5–75.6)	Corticosteroids use a	267 (44.5)
SF-36 physical component ≥79 a (2 MD)	128 (21.4)	Patient-Doctor Relationship
SF-36 emotional component ≥77a (2 MD)	157 (26.3)	PDRQ-9 score	4.6 (3.4–5)
		Highest PDRQ-9 scorea	183 (30.5)

Data presented as median (IQR) as otherwise indicated.

aNumber (%) of patients.

bSSS provides comprehensive health care insurance for public and private employees (including pensioners) and their households, such as outpatient and inpatient health care, hospitalization, paid sick days, disability, and retirement plans. Funded by both the Mexican Federal Government and by contributions from employees and their employers, this system comprises a heterogeneous bundle of autonomous health care institutions, including outpatient clinics, general hospitals, and specialty hospitals (according to Pineda et al 2019, https://doi.org/10.1007/s00296-018-4198-7).

cLimited to patients with previous hospitalizations.

SE, socioeconomic (level); SSS, social security system; IDs, immunosuppressive drugs; MD, missing data; DD, disease duration.

Description of the PDR in the Study Population

The median (IQR) of the PDRQ-9 score in the entire population was 4.6 (3.4–5). Twenty patients (3.3%) rated the PDR Likert scale as inferior, 267 (44.5%) as borderline, and the 313 patients left (52.2%) as superior. Table 2 summarizes the PDRQ-9 global score and individual item scores in the entire population, and the comparison among groups defined according to PDR Likert scale response. As expected, the better the PDR Likert scale, the higher the global and individual items PDRQ-9 scores.

TABLE 2

Global and Individual Item PDRQ-9 Scores in the Entire Population and Comparison Between Patients Classified According to their PDR Likert Scale

		PDR Likert Scale
	Study Population, N = 600	Inferior, n = 20	Borderline, n = 267	Superior, n = 313	p value
1. “My doctor helps me”	5 (3–5)	3 (2–3)	4 (3–5)	5 (4.5–5)	≤0.001
2. “My doctor has enough time for me”	4 (3–5)	2 (1–3)	3 (3–4)	5 (4–5)	≤0.001
3. “I trust my doctor”	5 (4–5)	3 (2–3)	4 (3–5)	5 (5–5)	≤0.001
4. “My doctor understands me”	4 (3–5)	2 (2–3)	4 (3–5)	5 (4–5)	≤0.001
5. “My doctor is dedicated to help me”	5 (3–5)	2 (2–2.8)	4 (3–5)	5 (5–5)	≤0.001
6. “My doctor and I agree on the nature of my medical symptoms”	5 (3–5)	2.5 (2–3)	4 (3–5)	5 (4–5)	≤0.001
7. “I can talk to my doctor”	5 (3–5)	2 (1–2.8)	3 (3–5)	5 (4–5)	≤0.001
8. “I feel pleased with my doctor's treatment”	5 (4–5)	2 (2–3)	4 (3–5)	5 (5–5)	≤0.001
9. “I find my doctor easily accessible”	5 (3.3–5)	2 (2–2.8)	4 (3–5)	5 (5–5)	≤0.001
PDRQ-9	4.6 (3.4–5)	2.4 (1.8–2.7)	3.8 (3–4.6)	4.9 (4.4–5)	≤0.001

Items 2 (“My doctor has enough time for me”), 4 (“My doctor understand me”), and 7 (“I can talk to my doctor”) obtained lower scores than the remaining items (p ≤ 0.001 for any comparison), and the differences persisted within the patients grouped according to the PDR Likert scale response (inferior, borderline, and superior), as summarized in Table 2 and Figure.

Individual item PDRQ-9 scores in the patients grouped according to their PDR Likert Scale category (inferior, borderline, superior). Color online-figure is available at http://www.jclinrheum.com.

Finally, among the 600 patient-doctor encounters, 523 (87.2%) involved certified rheumatologists, whereas the remaining 77 encounters (22.8%) involved trainees in rheumatology. Patient-doctor encounters from the former group (with certified rheumatologists) were rated with higher global and individual item PDRQ-9 scores, compared with their counterparts, as summarized in Table 3, but for item 6, “My doctor and I agree on the nature of my medical symptoms” and 7, “I can talk to my doctor.”

TABLE 3

Comparison of Global and Individual Item PDRQ-9 Scores Between Patient-Doctor Encounters that Involved Certified Rheumatologists and Those That Involved Trainees in Rheumatology

	Patient-Doctor Encounters With Certified Rheumatologists, n = 523	Patient-Doctor Encounters With Trainees in Rheumatology, n = 77	p value
PDRQ-9	4.7 (3.6–5)	4 (3.2–4.7)	0.001
1. “My doctor helps me”	5 (3–5)	4 (3–5)	0.004
2. “My doctor has enough time for me”	4 (3–5)	3 (3–5)	≤0.001
3. “I trust my doctor”	5 (4–5)	4 (3.5–5)	0.002
4. “My doctor understands me”	5 (3–5)	4 (3–5)	0.010
5. “My doctor is dedicated to help me”	5 (3–5)	4 (3–5)	0.031
6. “My doctor and I agree on the nature of my medical symptoms”	5 (3–5)	4 (3–5)	0.076
7. “I can talk to my doctor”	5 (3–5)	4 (3–5)	0.061
8. “I feel pleased with my doctor's treatment”	5 (4–5)	4 (3–5)	0.019
9. “I find my doctor easily accessible”	5 (4–5)	4 (3–5)	0.021

Factors Associated With Good PDR

We first compared medical encounters rated by the patients as deficient PDR (defined as PDRQ-9 score ≤3.73, n = 178) with medical encounters rated by the patients with good PDR (n = 422), and the results are summarized in Supplemental Table, http://links.lww.com/RHU/A393. Compared with their counterparts, patients from the former group were more likely to be female (90.4% vs 84.1%, p = 0.053) and scored higher pain-VAS (18 [3–42.8] vs 11 [0–38.5], p = 0.042), were more likely to have disability based on the HAQ-DI score (58.5% vs 47.8%, p = 0.019), and were less likely to have the SF-36 physical component within the reference range (14.7% vs 24.2%, p = 0.009). Accordingly, patients with deficient PDR had lower SF-36 global scores (56 [43.5–70.5] vs 63.1 [47.9–76.5], p = 0.001); also, they were more frequently involved in patient-doctor encounters with trainees in rheumatology. Finally, they were less likely to have a paternalistic ideal of patient autonomy (74% vs 89.4%, p ≤ 0.001), and were less likely to be concordant with their doctor's ideal of autonomy (62.3% vs 77.7%, p = 0.001, and 65.1% vs 80%, p = 0.001 for patient-doctor concordance with a physician-centered/paternalistic ideal of autonomy).

The following variables were included in the multiple logistic regression analysis to identify factors associated with a good PDR, which was considered the dependent variable: whether the patient was female, pain-VAS and SF-36 scores (highly correlated to SF-36 emotional component within the reference range), HAQ-DI score out of reference range, patient-doctor encounters with variable experience of the attending rheumatologist (trainees, certified rheumatologists), patient-doctor concordance in the ideal of autonomy (highly correlated to patient-doctor concordance in paternalistic ideal of autonomy), and patient's paternalistic ideal of autonomy. Results showed that patient paternalistic ideal of autonomy (odds ratio [OR], 3.029; 95% confidence interval [CI], 1.793–5.113; p ≤ 0.001), patient SF-36 global score (OR, 1.014; 95% CI, 1.003–1.025; p = 0.011), patient's female sex (OR, 0.460; 95% CI, 0.233–0.010; p = 0.026), and being a certified/senior rheumatologist (OR, 1.526; 95% CI, 1.059–2.200; p = 0.024) were associated with a good PDR.

DISCUSSION

The study focused on the quality of the PDR, where participants are greatly influenced by the social and cultural factors that define each other.[26] Accordingly, the results complement the current knowledge of the topic, which has been conceived based on studies primarily performed in developed countries (United States, North European countries, United Kingdom, and Japan) and in populations with a different anthropologic background.[9-15,27]

First, the study revealed that the majority of the primarily Mexican female patients with long-standing rheumatic diseases perceived a good PDR, which was more evident among medical encounters that involved certified rheumatologists. Moreover, some components of the PDR were rated lower by the patients, particularly, patients' perception of the time spent with the clinician and being understood by and of accessibility to talk to the doctor.

Similar results have been observed[27] and could be explained by the substantial follow-up of the underlying rheumatic disease of the patients included, which might have biased the PDRQ-9 score to higher values. In addition, 10 certified rheumatologists were involved in the majority of the medical encounters; clinicians' knowledge and clinical expertise shape treatment preferences and have the potential to influence the shared decision-making (SDM), which improves the quality of care.[28] Meanwhile, trust in the physician develops over time, characterizes long-term PDR, and impacts patient satisfaction with care, which might be considered a surrogate of a good PDR.[27] Moreover, medical encounters with trainees lack physician continuity on repeat clinical visits, which has been associated with less positive perception of physician style and physician trust, which ultimately affects the PDR.[29] Finally, time constraints have been recognized as a caveat of the quality of care in busy medical practice and limit the application of SDM.[30,31] In RA patients, a longer consultation time of 10 minutes has been associated with a slightly higher SDM score.[32] In addition, a patient's lower score of being understood by and accessibility to their doctor might reflect the well-known misalignment between patients and physicians' values, preferences, and perception of shared goals.[33,34]

A second relevant finding was that, in our population, the patient paternalistic ideal of autonomy, the patient's quality of life, and the degree of experience of the attending rheumatologist were risk factors associated with a good PDR, whereas female sex was protective.

It is generally accepted that active rheumatic patients' participation in their interaction with rheumatologists is associated with health care satisfaction, which might be considered a surrogate for a good PDR.[11,27,30] Nonetheless, the PDR is a complex and dynamic construct that is shaped by components highly nuanced by the cultural background.[27] Several studies have confirmed that Mexican patients with rheumatic diseases do not desire or undertake an active role at the time of their consultation.[6,17,18,35] Singh et al[36] found that 40% of the United States and Canadian patients with cancer experienced discordance between the preferred and the experienced decision-making role, and highlighted the need to deliver the type of experience that the patients prefer in terms of their decision-making role. In the current study, the majority of patients (and physicians) referred a paternalistic ideal of autonomy,[6] which explains its association with a good PDR. Also, and in agreement with our results, Ishikawa et al[12] studied 115 Japanese RA patients who were under the continuous care of 8 rheumatologists. They found that, among patients who preferred autonomous decision-making, the likelihood of being understood was positively associated with the extent of reported participation in visit communication, whereas such a relationship was less evident among those with a lower preference for active decision-making.

Studies involving patients with rheumatic diseases suggest that the nature of PDR can have a significant impact on HRQoL, which can be assessed with the SF-36.[11] A possible explanation was proposed by Freburger et al,[27] who argued that sicker patients deal with the health care system more frequently and are more likely to have problems with the care they receive and blame the physician because they are not getting better. The authors evaluated trust in the rheumatologist among 713 patients with RA, osteoarthritis, and fibromyalgia from North Carolina and found that patients with poor health and HRQoL reported lower levels of trust (a component of the PDRQ-9). Similarly, Beusterien et al[9] found that positive physician interaction with patients led to greater satisfaction with treatment and more favorable emotional health among 302 SLE patients from the United States.

The association between the degree of experience of the attending rheumatologist and a good PDR might be explained based on 3 arguments. First, experienced rheumatologists might be perceived by patients as paternal authoritative figures, and there is a respect for such figures among Hispanic patients.[31] Second, in the Hispanic community, there is an imbalance in social status between the patient and the physician, which favors a high-power distance culture, where patients expect the physician to take a more authoritative approach to the medical encounter, which is in line with the preferred ideal of autonomy of our patients.[6,7] Third, as previously stated, experienced rheumatologists might build solid and trustful relationships with their patients, which are particularly relevant for the PDR in Mediterranean and Latin American cultures.[29,37]

Finally, sex disparities in patients' experiences have received little attention, although our results were confirmed in nonrheumatic populations.[38] Men have generally reported better experiences with specific aspects of outpatient care, and the opposite has been true regarding dissatisfaction with nursing care and staff attitude.[38] Regarding inpatient experiences, in general, women also reported fewer positive experiences than men, with the exception of doctor communication, which is a component of the PDR.[38]

There are a few limitations of the study, which need to be considered. First, we used the PDRQ-9 to assess the PDR, but it is limited to patients' perspective; in addition, it has a substantial ceiling effect, which translates into a poor capacity to discriminate among patients who scored high on the PDR. Second, the study had a cross-sectional design, and only associations can be inferred. Third, the study was conducted at a single academic center where patients referred might have particular characteristics; therefore, the results may not be generalizable to other populations. Finally, relevant variables that are known to affect the PDR, such as patient-physician sex disparity, were not considered in the regression analysis.

CONCLUSIONS

The PDR is a complex dynamic and multidisciplinary phenomenon that needs to be approached from a cultural perspective. The PDR might also be conceived as a highly valuable outcome in itself, the quality of which influences disease outcomes, patient's satisfaction with care and adherence to treatment, and clinician satisfaction at work. In Mexican outpatients with rheumatic diseases, we found factors associated with a good PDR that were related to patient characteristics and of the clinician. Insights from this study are of great value for the development of strategies targeted at building solid relationships and improving communication among patients and doctors.