A gender-specific evaluation of a care-oriented curricular change in a Dutch medical school

Abstract

Background: Patients with chronic diseases need doctors who have the appropriate skills to maintain a long-term doctor–patient relationship with an orientation towards care rather than cure and focused on the patient's role in managing their condition. As the number of chronically ill patients is rising, medical education has to develop and evaluate instructional formats to prepare future doctors to provide care tailored to these patients.

Aim: To examine the possibly gender-specific effects on students’ orientations of a patient-oriented programme.

Methods: Three consecutive cohorts of third-year medical students at least twice completed the Ideal Physician questionnaire, which measures care versus cure orientation. Two cohorts participated in a care-oriented curriculum intervention, the third cohort did not. Analysis of variance was used to examine trends in the students’ orientation.

Results: Starting from a rather neutral care-cure orientation, a small but consistent trend towards increased cure orientation was found, which was unaffected by the educational intervention. Female students were more patient centred (p = 0.00) but became increasingly cure oriented.

Conclusion: Our programme was unable to curb the increased cure orientation in students. In addition to appropriate health-care innovations, other or more intensive curricular interventions are probably needed to serve the current and the future influx of chronically ill patients.

Introduction

Although the increasing life expectancy in many Western countries is generally deemed a fortunate circumstance, it is accompanied by the not-so-fortunate downside of increased prevalence of chronic medical conditions, which can be managed but not cured (Ebrahim 1999; Holman 2004). Chronically ill patients make different demands on their doctors’ skills than do patients with acute and curable diseases (Grol et al. 1990; Ebrahim 1999; Nair & Finucane 2003; Holman 2004). The effect on the doctor–patient relationship is that this relationship is not only likely to be a long-term one, but its objective and thus its nature will be different to that of primarily cure-oriented relationships. Considering that one of the main objectives is to maintain the patient's independence for as long as possible, it is clear that patient centredness is the key to this relationship and efforts must be focused on improving patients’ quality of life by helping them develop coping strategies and self-management skills (Barry et al. 2001; Holman 2004). The challenge to medical education is to prepare students adequately for this demographic shift in the patient population and its impact on medical practice. Thus, a key question to be answered by educators is whether and how they can modify the curriculum to equip students with the skills and attitudes to provide the best possible care for chronically ill patients (De Monchy 1990, 1992; Ribeiro et al. 2007). Curriculum interventions designed with this objective in mind have to be tested and evaluated to determine their effectiveness. In this article, we examined the effects of such a curricular intervention in the third-year of a 6-year medical curriculum by measuring changes in students’ views of the attributes of the ideal doctor. We also examined whether gender affected care orientation, because female students have been shown to be generally more care and patient oriented, particularly in the early years of their studies (Tsimtsiou et al. 2007).

Table

Practice points

The increasing number of chronically ill patients needs doctors with a stronger care and patient centredness. Medical education has to develop and evaluate instructional formats to prepare future doctors to provide care tailored to these patients. We evaluated a patient-oriented educational programme, consisting of longitudinal meetings with a chronic patient, assignments and debriefing sessions. The study shows a consistent trend towards increased cure orientations (rather than care), which was unaffected by the educational intervention. We are in need of further studies examining how educational interventions can foster a care-oriented attitude in students.

Methods

Study population

Starting in the academic year 2003/2004, we invited three consecutive cohorts of students in the third year of the undergraduate medical curriculum of Maastricht University, Maastricht, The Netherlands, to complete the Ideal Physician questionnaire (Batenburg et al. 1999; De Valck et al. 2001). Maastricht University offers a 6-year undergraduate medical curriculum, which is problem based, with small group tutorials as the dominant educational format.

In The Netherlands, ethical approval is not required for this type of study. Participation was voluntary and the students were given the assurance that the data would be confidential and their decision whether to participate or not would have no repercussions. The published data are not traceable to individual students.

Patient-oriented curriculum intervention: Year Patient Programme

The Maastricht undergraduate curriculum underwent a comprehensive curriculum reform between 2001 and 2007. The overarching theme of the revised third year is chronic disorders, and the year is divided into four 10-week ‘clusters’ dedicated to chronic disorders in the following domains: abdomen, circulation/lungs, psycho-medical problems and musculoskeletal system. The focus is on clinical aspects of chronic disorders, the impact on patients’ lives and the role of the doctor. Knowledge, skills and personal aspects are addressed. During weekly tutorials, students discuss diseases seen during outpatient encounters in which they interview and examine a patient supervised by a specialist. Students also have community-based experiences, mostly in general practice.

In 2003/2004 and 2004/2005, all third-year students participated in the ‘Year Patient Programme’ (YPP; Table 1), designed to enable students to observe the course of chronic disease and its impact on patients. Every student was assigned to one patient whom they visited in their home over a period of 8–9 months. In 2003/2004, the minimum number of visits was four, and in 2004/2005, it was three. During these visits, the students explored with the patients the effects of limitations on their daily activities due to chronic disease and the role of self-management and mood problems. The students also recorded questionnaire data about the patient's general health, social participation and health-care costs. Groups of approximately 10 students met eight times during the year to discuss their YPP experiences. The students were assessed on assignments relating to chronic disease, patients’ self-management and mood problems, and a brief report in which they compared their patient's questionnaire data with those from a comparable subgroup of patients. In 2003/2004, patients were recruited for the programme from patients’ associations, and in 2004/2005, students recruited their own patients. No special selection criteria applied.

Table 1.  Total number of students in the cohorts studied and the number of participating students

			Number of respondents
Year	Care-oriented curriculum intervention	Total number of third-year students	At year start	At year end	At year start and end
2003/2004	+	290	220 (76%)	132 (46%)	112 (39%)
2004/2005	+	293	271 (93%)	259 (88%)	238 (81%)
2005/2006	–	345	232 (67%)	181 (53%)	149 (43%)

Ideal Physician Questionnaire

The Ideal Physician Questionnaire was originally developed by Batenburg and colleagues. It consists of 18 items asking students to rate traits of the ideal doctor on seven-point scale with two contrasting anchors, such as a hierarchical versus a democratic attitude (Table 2).

Table 2.  Attitudes of medical students as measured by the Ideal Physician questionnaire. The results of one-factor (care versus cure orientation) and three-factor solutions (a psychosocial versus biomedical; an empathic versus distant; and a rational versus emotional factor), based upon factor loadings of principal factor analyses with oblique rotation in pooled September/October data in three cohorts (n = 673)

Ideal physician	One-factor solution	Three-factor solution
questionnaire items	Care	Psychosocial	Empathic	Rational
Hierarchical (democratic)	0.35^a	0.22	0.30^a	−0.01
Distant (committed)	0.29	−0.01	0.59^a	−0.01
Disease-centred (person-centred)	0.44^a	0.34^a	0.23	−0.00
Reticent (open)	0.20	−0.10	0.53^a	−0.02
Soloist (team member)	0.11	0.04	0.37^a	0.15
Scientist (care provider)	0.28	0.05	0.23	−0.14
Businesslike (considerate)	0.54^a	0.09	0.54^a	−0.24
Objective (empathic)	0.55^a	0.11	0.41^a	−0.29
Analysing (integrating)	0.38^a	0.19	0.20	−0.14
Specialist (general practitioner)	0.43^a	0.56^a	−0.06	−0.02
Somatic (psychosocial)	0.47^a	0.48^a	0.16	0.01
Intervening (contemplative)	0.33^a	0.47^a	−0.06	0.02
Technician (confessor)	0.52^a	0.39^a	0.07	−0.19
Common sense (compassionate)	0.54^a	−0.10	0.19	−0.66^a
Logical (intuitive)	0.41^a	0.06	−0.06	−0.50^a
Rational (emotional)	0.54^a	0.09	−0.07	−0.66^a
Conducting (following)	0.41^a	0.29	−0.16	−0.33^a
Certain (doubtful)	0.21	0.22	−0.38^a	−0.29

Note: ^aFactor loadings > 0.30 or <−0.30. The scores on the items with such loadings were used to calculate compound scores.

Earlier studies revealed good psychometric properties of the questionnaire with a two-dimensional factor solution (Batenburg et al. 1999; De Valck et al. 2001). A one-dimensional solution was based on a cure versus care orientation and the two-dimensional solution was based on an instrumental dimension (biomedical versus biopsychosocial) and an affective dimension (personal detachment versus empathy). Diederiks and colleagues conducted factor analysis of the same data set collected during 2004/2005 that we used in our study and found two dimensions: a cure versus care dimension and a doctor-centred versus patient-centred dimension (Diederiks et al. 2006).

In 2003/2004 and 2004/2005, we invited all third-year students during tutorials, and in 2005/2006, we invited all third-year students by email (with two reminders at 2-week interval) to complete the ideal physician questionnaire. In 2005/2006, students were asked to go to a specially designed website to complete the questionnaire. Each year, the questionnaire was administered on two occasions: early and late in the academic year, i.e. in September/October and May/June. The students of the third-year cohort of 2004/2005 were also surveyed in their fourth year in 2005/2006.

Statistical analysis

We performed principal factor analysis with oblique rotation of the data collected in September/October in three consecutive years for the three cohorts separately and for the pooled data. We calculated factor scores by averaging the scores on items with loadings >0.30 or <−0.30 on that factor. Mean scores were calculated for six measurements (beginning and end of year in three consecutive cohorts) and compared between measurements and male and female students. Paired t-tests were used for within-year comparisons (beginning and end of year) for each year and for the pooled data of the 3 years. Analysis of variance was performed to compare between the years, i.e. between the means of the September/October measurements and the May/June measurements. In order to examine for differences between the students who did (third-year cohorts of 2003/2004 and 2004/2005) and those who did not participate in the patient-oriented programme (third-year cohort of 2005/2006), we used analysis of variance to compare within year changes between the cohorts. Finally, to examine the effect on the further attitude development in the fourth year, the students from the 2004/2005 cohort who were followed up into their fourth year were separately examined by comparing, with paired t-tests, their orientations in their third and fourth years.

Results

The response is high at the start of 2004/2005 (93%) and low at the end of 2003/2004 (46%) (Table 1). Early and late in 2005/2006, the number of responding fourth-year students was 168 and 150, respectively (not tabulated). Of this cohort, 103 students responded on both occasions.

A three-factor model describes the data best (Table 2). The factors are interpretable and the pattern is consistent across the 3 years. The factors can be interpreted as psychosocial versus biomedical (Cronbach's α = 0.54), empathic versus distant (Cronbach's α = 0.54), and rational versus emotional (Cronbach's α = 0.67). These factors explain 28.3% of the total variance. Correlations of 0.20, −0.53, and −0.10 are found between psychosocial and empathic, psychosocial and rational, and empathic and rational, respectively (Table 2). A one-factor solution also reveals strong similarities across the years, with the same items loading high and low on the care versus cure dimension (percentage of variance explained: 18.1%; Cronbach's α = 0.77). Both solutions are presented in Table 2 for the pooled data collected in September/October of the three years. A forced two-factor solution and an analysis without restrictions (number of factors based upon the number of eigenvalues >1) did not result in similarly interpretable patterns and the analysis did not always converge for a forced two-factor model. The scores on the items with high factor loadings were averaged per student and per factor.

Because the findings are similar for the three factors of the Ideal Physician questionnaire, we report the results for the overall care versus cure factor. It should be noted, however, that the highest and the lowest care versus cure scores are those for the empathic versus distant factor (mean = 4.81; SD = 0.5 for the pooled data) and the rational versus emotional factor (mean = 3.39; SD = 0.7), respectively.

Figure 1 shows the mean care orientation score of third-year medical students across the 3 years of the study. The students’ scores are rather neutral: the observed range never exceeds 3.5–4.5 (mean = 4.01; SD = 0.50 in pooled data) on the scale from 7 (care oriented) to 0 (cure oriented). Within the years, the means showed a tendency, albeit a weak one, towards a stronger cure orientation (p = 0.09 in the pooled data). In 2003, for example, the mean score decreases from 4.11 to 4.04 (p = 0.14). The within-year changes show no signs of an effect of the patient-oriented programme. There are no differences between the years in which this programme was offered and the year in which it was not offered (p = 0.64). This absence of an effect is similar for male and female students (p = 0.14). The means of the early and late measurements in 2003/2004, 2004/2005 and 2005/2006 (4.09, 4.05 and 3.92, respectively; p = 0.00) appear to indicate a somewhat stronger trend towards a cure-oriented attitude.

Figure 1. Mean care orientation of third-year medical students who participated in the patient-oriented programme (2003/2004 and 2004/2005) and who did not participate in this programme (2005/2006) (theoretical minimum and maximum, 0 and 7).

The mean age of the student cohorts is 21 years (SD = 2.21 in pooled data) and age is not related to care versus cure orientation (Pearson correlation = 0.04; p = 0.25 in pooled data). The percentage of women increases from 67 in 2003/2004 to 71 in 2004/2005 and 74 in 2005/2006 (p(χ²) = 0.22) and the female students score significantly more highly on care orientation than their male counterparts (4.11 versus 3.78; p = 0.00 in the pooled data). Clearly, the trend towards a stronger cure-oriented attitude in the consecutive cohorts cannot be explained by differences between the years in gender distribution.

Figure 2 shows the care versus cure orientation of 82 students who completed the questionnaire in their third year in 2004/2005 and in their fourth year in 2005/2006. These students attended the patient-oriented programme in third year and then, in fourth year, did not participate in any further dedicated curricular component aimed at enhancing patient orientation. There is a clear trend towards an increased cure orientation with a (significant) acceleration between the end of the third year and the beginning of the fourth year (mean scores 4.04 and 3.73, respectively; p = 0.00 for the paired difference). The women score higher on care orientation than do the men at the four measurements (3.99 versus 3.57; p = 0.00 in pooled data).

Figure 2. Mean care orientation of medical students in their third (2004–2005) and fourth years (2005–2006); all experienced the YPP curricular change in their third year (n = 82).

Categorical analysis of the data from the students with a stronger care orientation (>4.5) or a stronger cure orientation (<3.5) shows similar trends towards a cure orientation within and between the years (although in some cases the numbers are small). There is no difference in care orientation between the students who responded only once and the students responding twice in one year. Less than 5% of the students left items unanswered. There are no age- or gender-related differences in these patterns of response and in missing item scores.

Discussion

Life expectancy is increasing in many countries, including the Netherlands (Mackenbach 2005). The baby boom after World War II and biotechnological advances have resulted in more patients living longer, albeit with disease in many cases. The result is a substantial demographic shift in the patient population towards a larger proportion and a greater number of elderly persons with and without chronic disease. The doctor's role in the relationship with patients with chronic disease is characterized by a stronger orientation towards care rather than cure and towards more interactive and longitudinal modes of consultation (Barry et al. 2001; Holman 2004). In this relationship, care management occupies a more central position than does disease management, with emphasis on patients learning to cope with their disease and self-manage their lives (Lorig & Holman 2003). Medical education should anticipate this change in the patient population which the next generation of doctors will serve by paying more attention to fostering care-oriented attitudes in students (De Monchy 1990, 1992; Ribeiro et al. 2007). We examined the effects of a programme for third-year students aimed at promoting a patient-oriented attitude towards chronically ill patients by exposing students to one patient with a chronic disease during a whole year and focusing their attention on aspects of chronic disease through assignments, debriefing sessions and assessment. We report the results of a study in which we administered the Ideal Physician questionnaire to three cohorts of third-year medical students in order to determine whether the newly introduced patient-oriented programme was successful in modifying students’ attitude in the desired direction.

The results show that students see the ideal doctor as a person with a relatively neutral orientation on a scale from care to cure orientation. We found a small but consistent trend towards a stronger cure orientation in three subsequent cohorts (2003/2004, 2004/2005 and 2005/2006) of third-year medical students (p = 0.00), in each cohort between the beginning and the end of the year (p = 0.09), and most markedly between the end of the third year and the beginning of fourth year in the 2004/2005 cohort (p = 0.00). The first two cohorts we studied participated in a care-oriented, community-based programme. Based on our results, we can only conclude that this programme was not successful in achieving its intended objective. Although the female students were somewhat more care oriented, they exhibited the same trend towards cure orientation as did their male counterparts.

Why did our care-oriented curricular intervention fail to curb the trend towards increasing cure orientation? We might be able to answer this question if we could identify modifiable determinants of the cure-directed tendency. Researchers who detected similar trends as students progress through the medical curriculum, have pointed to the influence of role modelling of practising doctors and teachers, the hidden curriculum (what is taught but not documented) and the null curriculum (what is documented but not taught), and cultural influences as sources of potentially elitist and paternalistic attitudes and behaviours in students (Tsimtsiou et al. 2007). As students progress through the curriculum, their initial idealism, reflected in empathy and acknowledgement of the psychosocial context of health and disease, appears to be increasingly eroded due to the above influences. Another explanation may be that medical students are too young to get excited about care management in older patients with chronic disease. Most students embark upon their studies inspired by noble motives and the ideal of providing instant help to patients with acute problems. Their interest is fuelled by complicated diseases in need of speedy and decisive clinical interventions, particularly those characterized by a technological tour de force (Holman 2004; Diederiks et al. 2006). It is not unthinkable that longitudinal, patient-centred care management simply does not fit the bill for young students. We will have to develop new curricular experiments to foster a patient-centred attitude (Batenburg & Smal 1997; Nair & Finucane 2003). Continuous, consistent and explicit attention to these issues in all stages of the curriculum is likely to bring the best results.

Perhaps we can learn from the female students, who were reported by, for example, Batenburg et al. (1999) and De Valck et al. (2001) to be more attuned to care-oriented attitudes. This may be related to Western culture being more accepting of women discussing feelings as opposed to men (Tsimtsiou et al. 2007). However, in our study both the female and the male students show the same trend towards cure orientation. They are also equally unaffected by the curriculum change. Others have also reported attenuation of gender differences in the final years of the undergraduate medical curriculum, due to female students incorporating cure- and doctor-centred attitudes more quickly than male students (Tsimtsiou et al. 2007). However, since our data were limited to third and fourth years in a 6-year curriculum, we can neither support nor refute this supposition. In summary, although gender-related differences appear to be relevant to attitude development, it is unclear how we could harness these in developing effective curricular interventions.

Two issues warrant further consideration. First, it is unclear why consecutive third-year cohorts should show an increase in cure orientation with every year, particularly considering that the curriculum of their first 2 years was identical in content and delivery. Perhaps what we see is a cohort effect of younger student generations placing (somewhat) less value on empathy, care and patient-centredness as traits of the ideal doctor. However, this is mere conjecture and requires studies in more cohorts. Second, the three-factor solution that we found and the moderate correlations of the factors suggest that the students may see the ideal doctor as someone who can be characterised as care oriented on a rational versus emotional dimension, but also as cure oriented on a biomedical versus psychosocial dimension or any other combination. The implication of this view would be that a care orientation and a cure orientation are not necessarily mutually exclusive but, instead, the two can go together and such a combined orientation may even be a characteristic of doctors that are well equipped to deal with the current and future needs of patients (Batenburg et al. 1999; De Valck et al. 2001). In addition to educational innovation aimed at fostering care orientations in future doctors, the chronically ill patients are likely to be best served also by appropriate health care innovations, such as the use of physician assistants and nurse practitioners, and trying to find more effective ways of communication between health-care workers.

Conclusion

The increasing numbers of elderly patients with chronic diseases are best served by doctors who are interested in longer-term care and patient-centred interaction. Third-year students showed a growing cure orientation, even during exposure to a curriculum intervention aimed at fostering a care- and patient-oriented attitude. Although female students were initially more care oriented than their male counterparts, they showed the same tendency towards cure orientation and were equally unaffected by the curricular change. Without neglecting the cure-oriented aspects of medical education and acknowledging the importance of appropriate health care innovations, we are left wondering what steps we, as educators, should take to equip our future doctors with the appropriate care-oriented attitude to meet the care needs of a large part of the patient population in the near future.

Acknowledgements

We thank all the students who participated in this study by responding to the questionnaires. We would also like to thank Dr Mereke Gorsira for further editing of the text.

Conflict of interest: None.

Additional information

Notes on contributors

Hans Bosma

HANS BOSMA is an associate professor and lecturer in the Department of Social Medicine, Maastricht University.

Jos Diederiks

JOS DIEDERIKS is an associate professor of Medical Sociology in the Department of Health Sciences, Maastricht University.

Albert Scherpbier

ALBERT SCHERPBIER is a professor of medical education and scientific director of the Institute for Education Health, Medicine and Life Sciences, Maastricht University.

Jacques Van Eijk

JACQUES VAN EIJK is a professor of Medical Sociology and former coordinator of the ‘Year Patient Programme’.