Elsevier

Social Science & Medicine

Volume 73, Issue 9, November 2011, Pages 1416-1421
Social Science & Medicine

Do general practitioners overestimate the health of their patients with lower education?

https://doi.org/10.1016/j.socscimed.2011.07.031Get rights and content

Abstract

This study sought to ascertain whether disagreement between patients and physicians on the patients’ health status varies according to patients’ education level. INTERMEDE is a cross-sectional multicentre study. Data were collected from both patients and doctors via pre- and post consultation questionnaires at the GP’s office over a two-week period in October 2007 in 3 regions of France. The sample consists of 585 eligible patients (61% women) and 27 GPs. A significant association between agreement/disagreement between GP and patient on the patient’s health status and patient’s education level was observed: 75% of patients with a high education level agreed with their GP compared to 50% of patients with a low level of education. Patients and GPs disagreed where patients with the lowest education level said that their health was worse relative to their doctor’s evaluation 37% of the time, versus 16% and 14% for those with a medium or high education level respectively. A multilevel multivariate analysis revealed that patients with a low educational level and medium educational level respectively were at higher risk of being overestimated by GP’s in respect of self-reported health even if controlling for confounders. These findings suggest that people with a lower education level who consider themselves to have poor health are less reliably identified as such in the primary care system. This could potentially result in lack of advice and treatment for these patients and ultimately the maintenance of health inequalities.

Highlights

► Little is known about the impact of the patient/GP interaction during consultation on health inequalities. ► Evidence suggests that advice, treatment and follow-up in primary care vary depending on patients’ social characteristics. ► Our study highlights disagreement between French patients and their GPs on the patient’s overall health status. ► GPs evaluated patients’ health more positively relative to the patient’s evaluation, when they had a low level of education. ► This could result in lack of advice and treatment for these patients and ultimately the maintenance of health inequalities.

Introduction

The interaction between patients and their general practitioner (GP) is a key element in the efficiency and usage of health services. It varies depending on patient characteristics (Arber et al., 2006). If, as the evidence suggests, the patient–physician relationship affects patients’ health care trajectory, i.e., how they are treated and their compliance with treatments, this could lead to health inequalities that will permeate across the health care system. Indeed, studies have shown that the nature and quality of the relationship between patients and their physicians affects communication, medical advice, satisfaction, and diagnosis (Adams et al., 2008, Arber et al., 2004, Street et al., 2007, Willems et al., 2005). Bensing et al. (2006) highlight that a patient’s inclination to participate in medical decision-making varies by characteristics such as age, gender, education, coping style, and severity of condition. Street et al. (2007) found that the way a physician perceives a patient (intelligent, compliant etc.) affects how he/she treats them during the consultation. Gender and age are also important influences on the doctor-patient relationship, with more smoking and alcohol-related advice being given to men and older patients (Arber et al., 2004, Arber et al., 2006). Doctors also communicate and treat their patients differently according to other social characteristics such as social position and ethnicity. Bao, Fox, and Escarce (2007) found that physicians were less likely to discuss cancer screening tests with patients who had a lower education level, and with patients from low income groups. Black patients, and patients whose ethnicity was other than that of their physician were found to receive significantly less information compared to their white or racially concordant counterparts (Cooper et al., 2003, Gordon et al., 2006).

Further exploration into the black-box of patient-physician interaction may thus highlight how and where some aspects of health inequalities are produced, and indicate how changes can be made in general practice to reduce health care inequalities linked to patients’ social characteristics. During the consultation, the assessment of the patient’s health status is pivotal to the decisions made for the patient’s well being and remains a key element in determining the patient’s healthcare trajectory subsequent to the visit. Assessment of patients’ health by the physician is an essential part of his/her professional activity. This assessment depends on ‘objective’ signs like clinical or biological indicators, and also on questioning the patient, particularly about his/her health status. As such, patients’ self-assessment of their health provides valuable information on their self-perceived health status. Indeed, since the work by Mossey & Shapiro, 1982, many studies have shown that self-rated health is associated with mortality, even after adjusting for different potential confounding factors (DeSalvo et al., 2006, Idler and Benyamini, 1997, Singh-Manoux et al., 2007b). Self-rated health is also strongly associated with both physical and mental health status (Mantyselka et al., 2003, Singh-Manoux et al., 2006).

The overall aim of the analyses in this paper is to ascertain whether disagreement between patients and physicians on the subject of patients’ health status depends on patients’ social characteristics, which could in turn exacerbate health inequalities. The patient’s self-rated health is compared to his or her GP’s evaluation of the patient’s health to determine: a) whether the pair agrees or disagrees on the patient’s health status, and b) whether agreement varies in relation to the patient’s education level. Neither the patient nor the doctor is ‘correct’ nor ‘incorrect’ with regard to the patient’s health status rating. In fact, we consider that each actor has a different and valid perspective of the patient’s health status, given his or her respective roles during a consultation. The INTERMEDE project (Interactions médecin-patient en médecine générale et inégalités sociales de santé. Analyses interdisciplinaires) aims to identify the impact of the patient-physician interaction in general practice on health inequalities. The quantitative data from the INTERMEDE project will be used for the analyses in this study.

Section snippets

Methods

An in-depth methodological account of both qualitative and quantitative phases of this study is described elsewhere (Kelly-Irving et al., 2009). Briefly, this cross-sectional quantitative phase of the project took place at the GP’s office over a two-week period between September and October 2007. The sample consisted of 27 GPs from three French regions (Ile de France, Midi Pyrenees and Pays de la Loire) who volunteered to participate in the study via GP networks: the Toulouse Department of

Measures

The data of interest in this paper were extracted from a questionnaire where the patients and GPs each responded to questions about the patient’s overall health. There were therefore 585 pairs of responses that allowed for comparison of patients’ self-rated health relative to GPs’ measurements. Health status was determined based on the question: How would you describe your general health? (patient version) How would you describe your patient’s general health (GP version): Very good, good,

Statistical analyses

Disagreement/agreement between GPs and patients was examined using the Kappa statistic (unweighted). Bivariate analyses were carried out using Chi square. Multivariate analyses were carried out using binary logistic regression. We wanted to control for inter-doctor differences in how they interact with patients, and specifically that individual doctors may systematically over- or underestimate their patient’s health relative to the patient’s rating. This doctor-level effect was controlled for

Results

Patient and doctor characteristics are described in Table 1. Perceived health status is the variable of interest in this study, with no significant difference in patient distribution across the categories between men and women (not shown). Patients were more likely to rate their health as average (31%) and poor (7%) compared to their GPs (26% for average and 5% for poor, p < 0.001). The agreement between these two assessments was poor (Kappa = 0.36, p < 0.001) despite a similar distribution

Discussion

Our study evaluates the discordance between the patient’s perceived health status and his or her GP’s evaluation of the patient’s health status. We found that the patient’s education level is significantly related to disagreement between patients and GPs on the patient’s health status. More specifically, our analysis reveals that GPs tended to evaluate patients’ health more positively relative to the individual’s own evaluation, when patients had a low level of education. Even if neither of the

Conclusions

Ultimately, as studies have shown, people who rate their health as poor are more likely to suffer from illness and disease (Barros et al., 2009, Harrington et al., 2010, Mantyselka et al., 2003) and are more likely to die prematurely (Idler and Benyamini, 1997, Phillips et al., 2010, Singh-Manoux et al., 2007a, Singh-Manoux et al., 2007b) than those who rate their health as being good. Our findings suggest that people with a low education level who consider themselves in poor health are not

Acknowledgements

The study was funded by Inserm. Appel à proposition 2005 IVRSP « Inégalités sociales des santé ». Numéro de projet A06024BS. CR received funding from the CRAM. Ethics approval was granted by La Commission nationale de l’informatique et des libertés (CNIL) : La mise en œuvre, à titre expérimental, d’une étude ayant pour objet la relation patient-malade et son incidence sur la construction des inégalités sociales de santé. Volume n°1228223. Paris; 11 Septembre 2007. The INTERMEDE pilot study was

References (35)

  • J.M. Bensing et al.

    Shifts in doctor-patient communication between 1986 and 2002: a study of videotaped general practice consultations with hypertension patients

    BMC Fam Pract

    (2006)
  • CNIL

    La mise en œuvre, à titre expérimental, d’une étude ayant pour objet la relation patient-malade et son incidence sur la construction des inégalités sociales de santé

    (11 September 2007)
  • L.A. Cooper et al.

    Patient-centered communication, ratings of care, and concordance of patient and physician race

    Annals of Internal Medicine

    (2003)
  • C. Delpierre et al.

    Using self-rated health for analysing social inequalities in health: a risk for underestimating the gap between socioeconomic groups?

    Journal of Epidemiology and Community Health

    (2009)
  • K.B. DeSalvo et al.

    Mortality prediction with a single general self-rated health question

    Journal of General Internal Medicine

    (2006)
  • H. Goldstein

    Multilevel statistical models

    (1995)
  • H.S. Gordon et al.

    Racial differences in doctors’ information-giving and patients’ participation

    Cancer

    (2006)
  • Cited by (0)

    View full text