Moral competence and medical education in contemporary times: a Brazilian study

Abstract

To evaluate the moral competence of medical students, this article compares different moments of medical training, identifying sociodemographic and academic aspects related to this competence, then discussing the evaluation tool. This is a quantitative, cross-sectional and observational study, with application of the extended version of Lind’s Moral Competence Test and a socio-demographic and academic questionnaire. Low mean scores were identified in the periods evaluated—the first period’s mean was higher than the others—and behavior that deviates from the “physician’s dilemma” in relation to the others, regardless of the period. Low moral competency scores were detected in all periods evaluated, with decline or stagnation over the course and “segmentation phenomenon” of the test. No relevant correlation of sociodemographic and academic variables was identified. Finally, this study found scores for the initial periods lower than those described in the literature, suggesting a generational trend.

Keywords:
Education, medical; Ethics, medical; Moral development; Morals; Ethical analysis

Resumo

A fim de avaliar a competência moral de estudantes de medicina, este artigo compara diferentes momentos do curso, identificando aspectos sociodemográficos e acadêmicos relacionados a essa competência e discutindo a ferramenta de avaliação. Trata-se de estudo quantitativo observacional transversal, com aplicação da versão estendida do Teste de Competência Moral, de Lind, e questionário sociodemográfico-acadêmico. Identificaram-se escores médios baixos nos períodos avaliados – com média do primeiro período superior aos demais – e comportamento destoante do “dilema do médico” em relação aos demais, independentemente do período. Detectaram-se escores de competência moral baixos em todos os períodos avaliados, com declínio ou estagnação no decorrer do curso e “fenômeno de segmentação” do teste, e não se identificou correlação relevante das variáveis sociodemográficas e acadêmicas. Por fim constatou-se que escores dos períodos iniciais inferiores aos descritos na literatura prévia podem sugerir tendência geracional.

Palavras-chave:
Educação médica; Ética médica; Desenvolvimento moral; Princípios morais; Análise ética

Resumen

Este artículo evalúa la competencia moral de estudiantes de medicina desde comparaciones de diferentes momentos del curso, identificando los aspectos sociodemográficos y académicos y discutiendo el instrumento de evaluación. Este estudio es cuantitativo observacional, transversal, que aplicó la versión extendida de la Prueba de Competencia Moral, de Lind, y el cuestionario sociodemográfico y académico. Se identificaron las puntuaciones medias bajas en los períodos evaluados –un promedio del primer período superior a los demás– y los comportamientos que diferían del “dilema del médico” con relación a los demás, independientemente del período. Resultaron bajas las puntuaciones de la competencia moral en todos los períodos evaluados, con descenso o estancamiento durante el curso y un “fenómeno de segmentación” de la prueba, además, no hubo correlación significativa entre las variables sociodemográficas y académicas. Las puntuaciones de los períodos iniciales fueron más bajas a las descritas en la literatura, lo que supone una tendencia generacional.

Palabras clave:
Educación médica; Ética médica; Desarrollo moral; Principios morales; Análisis ético

Despite the historical importance of the study of ethics in medicine, the field does not seem to have followed the technical-scientific development in medical undergraduate courses in the last century. Since the publication of the 1910 Flexner report¹, medical schools have emphasized the scientific and care aspects, often to the detriment of ethical education, keeping it isolated and in the background, part of the “hidden curriculum” ².

Recently, medical ethics has been discussed in the context of restructuring undergraduate curricula, contemplating new proposals for academic approaches and medical training aligned with ethical, reflective, and humanistic principles ³. From a contemporary and humanist perspective ⁴,⁵, the ethics of care can represent the synthesis of moral values applied to medical practice, contemplating the moral virtues, (…) [and] for this, skills such as empathy and the development of a high level of moral reasoning may be essential ⁶.

Ethics has been addressed in medical education under different aspects, such as professionalism ⁵ or as medical humanities or virtues ⁴—altruism, integrity, empathy, advocate, respect for privacy, and dignity, for example. Or, still, according to Eckles and collaborators, honesty, accountability, compassion, service, industry, interprofessional respect, public health, and self-policing (…) ². Castro, Pereira and Bataglia, respect, compassion, and honesty, characteristics fundamental to any morality and that all professional medical caregivers are expected to embrace², bring them closer to the so-called ethics of care⁶. The authors also emphasize the broad diversity and heterogeneity of curricula and concepts or values evaluated which prevented conclusions regarding the quality of the programs ².

One challenge observed in discussions about ethical training in medical education lies in the modalities of evaluation. As highlighted by Bataglia, Morais, and Lepre ⁷, studies on morality applied to moral development emphasize cognitive-evolutionary theories, especially those derived from Piaget’s and Kohlberg’s propositions about judgment and the development of moral reasoning.

Both theories, related to the cognitive development of morality, presuppose the notion of justice and social and environmental interaction and can be considered essential and predictors of moral conduct, whose analyzes lend themselves to evaluating moral and, indirectly, ethical values. In this regard, morality assessment tools stand out, given their potential to combine ethical and moral concepts with teaching-learning theories, allowing the evaluation of moral development, an important gap in ethical education ⁶.

Recent studies carried out in several countries showed stagnation or moral decline during medical education ⁸, suggesting that the traditional teaching model may inhibit students from developing moral competence.

Several authors propose evaluations of moral judgment based on Piaget’s and Kohlberg’s cognitive-evolutionary theories, cited by Bataglia, Morais and Lepre ⁷, especially the Moral Competence Test (MCT), developed by Lind ⁹. This instrument, which measures the degree of consistency with which the individual analyzes arguments and counter-arguments to a character’s decision, has been widely used to evaluate moral competence in medical students. Moral competence emerges therefore as an aspect of ethics that can be evaluated, and as the best predictor of the moral act (or action) ⁷.

Thus, we suggest that undergraduate medical education train professionals poorly qualified to face the ethical challenges of contemporary society ¹⁰ and that ethical education has not followed the scientific-technological development of medicine, leaving gaps in the understanding of these educational processes and the new challenges of contemporaneity.

Given this scenario, this study analyzes the decline in moral competence during undergraduate medical training, correlating this phenomenon with socioeconomic conditions or academic and extracurricular experiences and discusses the MCT test applicability in contemporary times.

Method

Study design and population

This is a quantitative, cross-sectional and observational study was carried out in 2018 at a private higher education institution located in Southeast Brazil, whose curriculum adopts the problem-based learning (PBL) methodology.

By criterion of convenience, the sample included 139 students from the first period (freshmen), fifth period, and sixth (final) year. Freshmen were approached in the inaugural class, before any contact with academic activities, while the others were invited during regular academic activities. Individuals under 18 years of age were excluded.

Students in the 11^th and 12^th periods were approached indiscriminately, as they participate simultaneously in internship activities. After signing the informed consent form (ICF), the participants filled two self-answered questionnaires:

1. Moral Competence Test extended version (MCTxt), validated in Brazil by Bataglia ¹¹. This version comprises three dilemmas: the two original ones (worker and doctor) and a third one (judge);

2. Self-authored questionnaire with sociodemographic, academic, and extracurricular data.

Statistical analysis strategies

• Comparison of students from the three periods regarding each of the MCTxt C scores (global, grouped two by two, and segmented by dilemma) by one-way analysis of variance (ANOVA);

• Evaluation of the influence of the variables “student’s course period” and “dilemmas” on the MCTxt C scores by analysis of variance (ANOVA), based on a model of repeated measures;

• Evaluation of the correlation of the students’ sociodemographic and academic variables with the MCTxt C scores (global or segmented) by multiple linear regressions.

Results

Population studied

A total of 139 students were included, 58 from the first period, 36 from the fifth period and 45 from the sixth year (11^th/12^th periods), representing, respectively, 44%, 63% and 36.5% of the students regularly enrolled in each course period.

The sample studied consisted of mostly female (64.7%), white (77.5%), and single (95.7%) students, from high-income families (41.5% have a family income above 10 minimum wages), high schooling (71.6% of those financially responsible for the family have a college degree or graduation course), and who do not live with family members (63.3%). Age followed the course period, with the median ranging from 19 years in the first period to 25 years in the last year. On average, 15.8% receive some form of scholarship (or equivalent), ranging from 3.4% in the first period to 26.7% in the last year.

We observed an overall low adherence to extracurricular activities, and less than 10% of participants reported attending a course in humanities or social sciences or regular practice of sports, religious or political activities. Participation in artistic and/or philanthropic activities was reported by 45.8% and 29.4%, respectively; extension activities and undergraduate research, as expected, increased during the course, reaching, in the last year, 48.1%, 33.3% and 25.9%, respectively. The preference for clinical specialty ranged from 46.2% in the first period to 63.6% in the last.

Score evaluation: comparison between course periods

Scores were analyzed globally, considering MCTxt with three dilemmas (worker, doctor, and judge), MCT original version with two dilemmas (worker and doctor), grouped two by two and segmented by dilemma (Table 1). The evaluated students showed low scores in all analyzes, with mean MCTxt C total score of 13.1 and mean C score of the original MCT of 17.4. The mean values of the C segmented score with two dilemmas were 20.7 for worker/judge and 13.5 for doctor/judge, and the C segmented scores per dilemma were 38.2 for worker, 21.5 for doctor, and 33.3 for judge.

Table 1
Comparative analysis between course periods regarding the MCTxt C total scores and segmented scores

		Descriptive measures
Dilemmas/period	n	Minimum	Maximum	Mean	Standard deviation
Total	–	–	–	–	–
1 st period	58.	2.3	45.1	15.7	10.5
5 th period	36	1.6	34.7	11.8	7.8
11 th /12 th periods	45	0.1	33.0	11.0	7.2
All	–	–	–	13.1 –	–
Conclusion:	p * =0.019 – 1 st period>(5 th period=6 th year)
Worker/judge	–	–	–	–	–
1 st period	58	1.3	62.2	24.3	15.5
5 th period	36	0.5	44.2	19.8	10.7
11 th /12 th periods	45	0.6	55.1	16.8	11.0
All	–	–	–	20.7	–
Conclusion:	p * =0.021 – 1 st period>6 th year
Doctor/judge	–	–	–	–	–
1 st period	58	1.6	49.4	16.0	11.6
5 th period	36	1.3	44.4	11.6	9.8
11 th /12 th periods	45	0.4	36.0	11.7	8.7
All	–	–	–	13.5	–
Conclusion:	p * =0.068 – 1 st period=5 th period=6 th year
Worker/doctor	–	–	–	–	–
1 st period	58	2.0	52.8	19.4	12.8
5 th period	36	3.7	41.6	16.5	9.8
11 th /12 th periods	45	0.4	37.6	15.6	9.9
All	–	–	–	17.4	–
Conclusion:	p=0.198 – 1 st period=5 th period=6 th year
Worker	–	–	–	–	–
1 st period	58	2.1	81.2	39.4	22.6
5 th period	36	1.3	82.6	40.3	20.8
11 th /12 th periods	45	1.4	83.0	35.0	19.9
All	–	–	–	38.2	–
Conclusion:	p=0.463 – 1stperiod=5thperiod=6thyear
Doctor	–	–	–	–	–
1 st period	58	0.7	88.0	22.9	17.7
5 th period	36	2.5	57.1	20.5	15.5
11 th /12 th periods	45	0.0	76.3	20.6	17.1
All	–	–	–	21.5	–
Conclusion:	p=0.715 – 1stperiod=5thperiod=6thyear
Judge	–	–	–	–	–
1 st period	58	0.0	90.8	38.7	23.3
5 th period	36	0.5	74.5	30.0	21.0
11 th /12 th periods	45	2.2	64.6	28.3	17.3
All	–	–	–	33.3	–
Conclusion:	p=0.028 – 1 st period>6 th year

• Database: 139 students (1^st period: 58 students; 5^th period: 36 students; and 11^th/12^th periods: 45 students).
• p: probability of significance of the one-way analysis of variance with 1 factor
• ^*
  Probability of significance of the one-way analysis of variance, based on Welch’s robust test for equality of means, due to non-homogeneity of variances

In the analyzes by course period, low average MCTxt C total scores (below 20) were detected in all course stages (first, fifth and 11^th/12^th periods), respectively, 15.7, 11.8, and 11.0, with individual scores ranging from 0.1 to 45.1. Using one-way ANOVA, we observed a statistically significant drop between the 1^st period and the others, showing a decline in moral competence between the 1^st and 5^th periods: p=0.019; 1^st period>(5^th period=6^th year).

The segmented sub-analyzes, with two dilemmas (worker and judge), and segmented by dilemma (using the judge’s dilemma), identified a decline between the 1^st and the remaining periods, since, in the first case, the average scores ranged from 24.3 to 16.8, and in the second, from 38.7 to 28.3. Both showed a significant difference between the first period and the sixth year (worker/judge: p=0.021; judge: p=0.028). In the remaining analyzes with C segmented scores, however, we found no statistically significant differences between the students’ scores in the three course periods.

In an analysis with category classification using the C total scores of the original MCT, as defined by Lind ¹², we observed that 64% of the students were classified in the lower categories (“none,” “very low,” or “low”), ranging from 60.4% in the 1^st period to 66.1% in the last year, whereas 15.1% were in the “high to very high” category, ranging from 22.4% to 11.1% (Table 2). Such a decline showed no statistical significance in the three periods evaluated (p=0.601).

Table 2
Comparative analysis between MCT C total score category classification and course period

Classification	MCT C score	Course period
		1st year		3rd year		6th year		General
		n	%	n	%	n	%	n	%
None	0.0 to 4.9	6	10.4	6	16.7	8	17.8	20	14.4
Very low	5.0 to 9.9	9	15.5	4	11.1	8	17.8	21	15.1
Low	10.0 to 19.9	20	34.5	14	38.9	14	31.1	48	34.5
Sufficient	20.0 to 29.9	10	17.2	9	25.0	10	22.2	29	20.9
High to very high	30.0 to 100	13	22.4	3	8.3	5	11.1	21	15.1
TOTAL		58	100.0	36	100.0	45	100.0	139	100.0

• Database: 139 students (1^st period: 58 students; 5^th period: 36 students; and 11^th/12^th periods: 45 students).
• p=0.601; p: probability of significance referring to Pearson’s chi-square test

Segmentation phenomenon

ANOVA, based on repeated measures planning, was used to evaluate whether the course period and the dilemmas influenced the moral competence score (MCTxt). The analysis did not confirm the influence of course periods on the dilemma-segmented MCTxt C scores, but showed a statistically significant influence, independent of the dilemma factor, on the mean of the MCTxt C scores, that is, regardless of the student’s course period and the interaction between dilemma and course period, demonstrating the “segmentation phenomenon” ¹².

Table 3 shows that, on average, MCTxt C segmented scores for the “worker’s dilemma” and the “judge’s dilemma” do not differ significantly from each other; however, their mean scores are significantly higher than the C segmented scores for the “doctor’s dilemma”— (worker=judge)>doctor—, regardless of course period. Figure 1 demonstrates the discordant behavior of the “doctor’s dilemma,” with scores always lower than the others, regardless of period, characterizing the “segmentation phenomenon” of the moral judgment competence ⁹.

Table 3
Evaluation of the influence of course period and dilemma factors on MCTxt C segmented mean scores

Period descriptive measures	Course
	Dilemma	n	Minimum	Maximum	Mean	Standard deviation
1 st period	Worker	58	2.1	81.2	39.4	22.6
	Doctor	58	0.7	88.0	22.9	17.7
	Judge	58	0.0	90.8	38.7	23.3
5 th period	Worker	36	1.3	82.6	40.3	20.8
	Doctor	36	2.5	57.1	20.5	15.5
	Judge	36	0.5	74.5	30.0	21.0
11 th /12 th periods	Worker	45	1.4	83.0	35.0	19.9
	Doctor	45	0.0	76.3	20.6	17.1
	Judge	45	2.2	64.6	28.3	17.3
ANOVA result based on a model with 2 independent factors:
Variation source:	Period – (F2; 118=1.225; p=0.298)
	Dilemma×period – (F4; 236=1.065; p=0.374)
	Dilemma – (F 2; 236 =21.656; p<0.001)

• Conclusion for dilemma: (worker=judge)>doctor, regardless of course period
• Database: 139 students (1^st period: 58 students; 5^th period: 36 students; and 11^th/12^th periods: 45 students).
• p: probability of significance of the analysis of variance based on a repeated measures model; F: analysis of variance statistics based on a repeated measures model

Figure 1
Boxplot with MCTxt C segmented scores by course period (1) and by dilemma (2)

Sociodemographic and academic aspects

To evaluate whether students’ sociodemographic and academic variables would be related to moral competence scores, we applied a multiple linear regression model to the C total and dilemma-segmented scores of the MCTxt. We selected ten variables (course period, gender, race/ethnicity, marital status, family income, education of the person financially responsible for the family, scholarship holder, living with family, intended specialty, and completion of a humanities/social course), considered as possible independent/predictor variables.

In most situations analyzed, the variables showed little probability of significance, with low joint relationship with the outcome variable of interest, except for “living with family,” which would be associated with higher scores, showing statistical significance in the C total score (p=0.044) and in the judge’s dilemma (p=0.029). Another highlight was the “intended specialty,” with the option for clinical specialty showing non-significant correlation (p=0.062) only in the worker’s dilemma. However, these variables do not seem to sufficiently explain the variability of the score, given the low explained variability (R²) of the model.

Discussion

Considerations on the population and study design

The highly homogeneous socioeconomic profile of the study population, with a predominance of white, single, childless individuals and families with high income and schooling, may have been influenced by the private status of the educational institution, and does not represent the social diversity of the Brazilian university population, as described by the National Institute of Educational Studies and Research Anísio Teixeira (Inep) ¹³.

The lower percentage of scholarship recipients (or equivalent) in the 1^st period may be due to the fact that the freshmen data were collected before the release of the required benefits—which can occur after the start of classes—and/or to changes in the national school funding policy. In comparing course periods, we observed that the participants have similar sociodemographic characteristics, except for age and being a scholarship holder (or equivalent).

The reduced number of male individuals may somehow compromise the scores, given the possible influence of gender on the C score (MCT), as found by Rego ¹⁰ and Feitosa and collaborators ¹⁴,¹⁵. Moreover, the homogeneity of the sample and the low frequency of participation in extracurricular activities may have impacted the analysis of the correlation of these factors with the scores, which hindered drawing profiles or defining clusters among students that could be related to the development of moral competence.

The cross-sectional design led to the comparison of different students at three course stages and limited the evaluation of the development of moral competence over the undergraduate training. A longitudinal cohort study, although more costly, might allow us to analyze the same students and monitor them throughout the course. Thus, it would be possible to evaluate more accurately individual characteristics that influence, positively or negatively, the development of moral competence.

Score evaluation

We found low moral competence scores in the three analyzed course stages, with a mean MCTxt C total score of 13.1 (dropping from 15.7 in the first period to 11 in the last year).

Using the original MCT (two dilemmas), which allows for better comparison with international studies, we identified a mean C score of 17.4 (ranging from 19.4 to 15.6). As for the C segmented score with the worker/judge dilemmas—excluding the doctor’s dilemma with the worst performance—, we observed a mean score of 20.7, dropping from 24.3 (sufficient) to 16.8 (low). Regardless of the C score variation chosen, mean scores are lower than those presented by national studies and, even worse, by international ones.

When analyzing the results classified into Lind ¹² categories, means were situated at the “low” level of moral competence in all periods, with most students in the “low” (or lower) categories, except when excluding doctor’s dilemma from the score, in which case they would be classified as having a “sufficient” level of moral competence.

Medical students

Comparing our results with those of other national and international studies of similar design and carried out at different times, we observed different levels of scores. In Brazil, Feitosa and collaborators ¹⁴ report, at a university in the Northeast region, MCT C score dropping from 26.2 to 20.5, from the first to the eighth period (mean of 23.4). In a public school in São Paulo, Serodio, Kopelman, and Bataglia ¹⁶ observed MCT C score ranging from 22.6 to 20.3, and MCTxt score ranging from 19 to 17.7.

The present scores are about six points below than those found in the Northeast study (MCT) and four and five points below than those of São Paulo (MCT and MCTxt, respectively). Simultaneous to the present study, Otani ¹⁷, in a study carried out in two medical schools in Paraná using MCTxt, observed very similar scores, with means of 16 and 15.2, respectively, in the public and private institutions. When calculated using MCT (with two dilemmas), the means were slightly higher: 19.5 and 21.3, respectively.

When compared with European studies, the difference is even greater (Figure 2). Portuguese students ¹⁶ had a mean MCT C score of 29.7 (ranging from 28.3 in the first to 31.3 in the eighth period), whereas German students ⁸ had a mean MCT C score ranging from 48, 4 to 44.6 throughout the course. In these comparisons, we found a difference greater than 12 points below the Portuguese average (40% lower) and about 30 points below the German study average (60% lower). Even if we consider the score of the two best performing dilemmas (that is, excluding the doctor’s dilemma), the result is quite inferior, with about nine points below the Portuguese study and more than 25 points below the German students.

Figure 2
Comparison of first-period medical students’ moral evaluation scores in different contexts

Compared to Brazilian studies on public schools, the present article presents lower scores from the first year, which may be related to the private and less competitive nature of the studied institution in relation to public universities. According to Bataglia, Morais and Lepre ⁷, this is due to the fact that students from more competitive universities (that is, more sought-after institutions) have higher moral competence scores.

In comparing national and international studies, we observed a difference between the scores of studies conducted in Portugal and Brazil and those conducted in Germany. Lind ⁸ discusses the influence of religion (Catholic versus Protestant) and obedience to authority and institutions, as well as the experience of societies exposed to wars and revolutions on moral competence. These regional, historical, political, and cultural differences are also reflected in different MCT performances, with the “segmentation phenomenon” possibly influencing negatively the score of Latin countries, especially Americans, reflecting a possible limitation of the instrument in this population.

One must also consider the potential impact of generational issues in the score. Most of the German studies ⁸ were developed in the 1990s, addressing groups of students from the well-known generation X, which includes individuals born between 1965 and 1980. In contrast, the studies by Feitosa and collaborators ¹⁴,¹⁵ and by Serodio and collaborators ¹⁶, developed in the 2010s, represent the generation Y (internet generation, or millennials), born between 1981 and 1994. Finally, the present study and that of Otani ¹⁷ address individuals born between 1995 and 2010, representing generation Z.

According to Calliari ¹⁸, generations can have an impact on the educational process, and the prevalence of technologies is an important differentiating factor between them. Studies have described, for example, reduced reading habits as opposed to an increased exposure to electronic and technological resources. This different interaction with technology seems to impact information processing and thinking models, as well as social and professional interests.

Perception and social interaction also appear to be changing over generations, including in Brazil ¹⁹, where generation Z is seen as more sociable, but society is perceived as more egocentric, with less universal values than the previous ones.

This characterization in generations ¹⁹ is not restricted to the temporal unit, nor is it a deterministic or homogeneous variable, but rather contemplates specific collective identities and reflects the historical-social context of each population group. Some authors ²⁰,²¹ claim that generational issues, especially between generations Y and Z, would strongly impact teaching-learning processes and the development of attitudes and skills among medical students.

Conversely, the frequently used moral evaluation instrument—the MCT—has the premise of valuing the humanitarian approach as a reference for higher moral levels. Thus, one may question whether the apparent decline in scores observed in the studies, which include samples from different generations, is related to generational changes.

Stagnation and decline of moral competence in medical undergraduate training

By correlating the MCTxt scores (total or segmented) to the course periods in the univariate analysis, we identified statistically significant drops in the score between the first period and the others, when considering the C total scores (with three dilemmas), the C segmented scores with two dilemmas (worker/judge), and the judge’s scores. In contrast, the multivariate analysis, which evaluated the interaction between scores, dilemmas, and periods, found no influence of the period on the scores. Moreover, analysis of the relationship between categories and periods showed no statistically significant differences.

Some univariate sub-analyzes showed a drop in moral competence scores, characterized as “moral decline” during medical training, as widely cited in national and international literature ²,¹⁰,¹⁴,²²,²³, described by some authors as “ethical erosion” ²⁴. Importantly, this decline occurred mainly between the first and fifth periods, corroborating the study by Hojat ²⁵, which claims that the “demon is in the third year,” period considered critical for moral regression.

Although the multivariate analysis did not confirm such a decline, it identified moral stagnation due to the absence of the expected moral competence development. Thus, whether as decline or stagnation, the present research corroborates studies by authors such as Self, Baldwin, and Wolinsky, according to which the experience of medical education appears to inhibit the increase in moral reasoning of medical students that otherwise would be expected in young adults of this age and education level²⁶.

This observation contrasts with Piaget’s and Kohlberg’s propositions, for which, as Bataglia, Morais and Lepre ⁷ point out, moral development should follow educational levels. Similarly, according to Lind, several studies on adolescent students show moderate to high correlations between levels of education and moral development²⁷. Therefore, our study proves the absence of progression (or even decline) in the development of moral competence, characterized by the non-occurrence of expected C score (MCT) progression during academic education.

Importantly, the observed decline or stagnation data start from low (or very low) levels of moral scores, contrary to what Feitosa and collaborators ¹⁴ and Lind ⁸ discussed, with reservations, about a possible “ceiling effect” to justify the non-progression of moral competence in European medical students, since they already presented very high scores at the start of the course, which could impair the perception of moral development.

In the present study, however, the opposite can be argued, that is, whether such low levels of C score since freshmen year could not be impairing the perception of a more significant moral regression, thus suggesting a possible “floor effect.”

Finally, other extrinsic aspects should be considered, such as: the potential impact of recent changes in the National Curriculum Guidelines for Undergraduate Medical Education ³, which could be influencing a smaller decline in moral competence scores than those cited in other national studies; or the influence of the studied institution’s curriculum model, which adopts a PBL methodology, but without evidence that this curricular model actually interferes in attitudinal issues, such as moral behavior.

Applicability of MCT and the “segmentation phenomenon”

This study observed a distinct performance of the dilemmas, based on the discordant behavior of the doctor’s dilemma in relation to the others, confirming the “segmentation phenomenon” ⁷,¹² well described in Latin America. The multivariate analysis showed that the worker’s and judge’s dilemmas behaved equivalently, regardless of course period evaluated, and tend to present similar scores to each other, but significantly higher than those of the doctor’s dilemma: (worker=judge)>doctor.

In turn, the univariate analyses showed that the C segmented scores for the worker’s and the judge’s dilemmas behaved similarly, whereas the doctor’s dilemma presented very distinct and always lower values (see Table 1 and Figure 1). The inclusion of the doctor’s dilemma seems to limit the perceived decline of the C score, since it reduces the scores that include it.

Considering that internal consistency is based on the principle that each part of the test should be consistent with the other parts¹¹, we highlight the fact that the divergence between the dilemma scores can compromise the internal consistency or reliability of the test. Thus, suppressing the doctor’s dilemma could contribute to improve its consistency and validity.

The C score is also negatively influenced by the number of dilemmas used (the C score calculated with three dilemmas is always smaller than the one calculated with two dilemmas, which is smaller than each dilemma separately). As Lind argues, if the C-score is calculated only for each dilemma separately, it will be higher, because the variation due to the dilemma’s context is omitted¹².

Thus, the C segmented score with the best performing MCTxt dilemmas in Brazil (worker and judge) could reproduce a score level more comparable to the MCT C score in Europe. Thus, both scores with two dilemmas would support more equivalent values, higher than the MCTxt (with three dilemmas) and without the negative influence of the doctor’s dilemma in this context.

Classification by categories is also impaired with MCTxt, as these are not appropriated by calculations with three dilemmas, as already described by Lind ¹², nor by the MCT score, which is distorted by the negative influence of the doctor’s dilemma. In both situations, classification would bring negative falsified results. In Brazil, therefore, the use of the C segmented score with the worker’s and the judge’s dilemmas would allow for a categorization more comparable with that of the original MCT in Europe.

Sociodemographic and academic aspects

To better understand the reports of “ethical erosion” in medical education, cited by several authors, we sought to identify variables related to the students’ life history that could explain them. The linear regression model developed showed a correlation of moral competence scores only with the variable “living with family.” Despite being associated with higher scores, this variable does not sufficiently explain the variability of the scores but may directly or indirectly reflect other variables that better explain it.

Thus, our study signals a variable that can be broken down or that allows to identify correlated conditions that explain the variation in scores. This result was probably not more elucidative as a result of the homogeneous socioeconomic profile of the sample and the low participation in extracurricular activities. Therefore, gaps remain in the explanation of factors that influence inadequate moral development during medical education.

Factors can still be related to the students’ life history, including socioeconomic, cultural, and academic aspects, especially variables concerning living with family and/or distance from the family structure. Further research should expand these analyzes by comparing samples of students (and institutions) with greater heterogeneity, which better represents the diversity of the Brazilian population, and seek other approaches related to family ties/dynamics, student mental health, self-esteem, spirituality, metacognition, among others.

Given the importance of the educational environment, cited by Lind ²⁶ as a determinant of students’ moral development, studies should focus on other issues concerning educational organizations that provide environments conducive to moral development, including moral evaluation of the student body and its correlation with students’ moral status.

Professors are, admittedly, seen as role models with the potential to influence students in the development of attitudes and values, especially regarding ethical and moral aspects, which are still relegated to the hidden curriculum. But the challenge remains to understand how these institutions and their faculty have acted to meet the demands arising from generational changes and the impact of new curricular platforms being implemented.

Final considerations

This study identified, among the students evaluated, low moral competence scores in all course periods from admission to medical school, with values lower than those of previous studies described in the literature, suggesting a possible generational trend related to contemporary aspects.

In comparing the course periods, we observed a significant drop in scores, with a reduction both in the C total score and in some segmented sub-analyses, showing a regression of moral competence between the first and final years.

The classification by categories showed no significant difference between course periods, with most students in all periods presenting low or very low levels of moral competence. Similarly, the multivariate analysis did not confirm the influence of course period on moral competence scores. Both situations indicated moral stagnation, understood as the non-occurrence of the expected moral development during undergraduate education.

In evaluating the performance of the moral competence assessment tool, we observed that the MCTxt C segmented scores of the worker’s and the judge’s dilemmas do not differ significantly from each other. But the mean scores for these two dilemmas are significantly higher than the C segmented scores of the doctor’s dilemma—(worker=judge)>doctor—, regardless of course period. These findings point to the “segmentation phenomenon” of the MCTxt, already well reported in Latin America, observed based on the discordant behavior of the “doctor’s dilemma,” with mean scores lower than the other two, and negatively influencing the global score.

Regarding the sociodemographic, academic, and extracurricular variables, analyzed by linear regression model, only the variable “living with family” showed a statistically significant association with the scores. However, the model failed to sufficiently explain the variability of the scores and determine a relevant correlation between these variables and the moral competence score.

• Approval CEP-Conep-CAAE 82290017.0.0000.5143

Referências

• ¹
  Cooke M, Irby DM, Sullivan W, Ludmerer KM. American medical education 100 years after the Flexner report. N Engl J Med [Internet]. 2006 [acesso 9 maio 2022];355(13):1339-44. DOI: 10.1056/NEJMra055445
  » https://doi.org/10.1056/NEJMra055445
• ²
  Eckles RE, Meslin EM, Gaffney M, Helft PR. Medical ethics education: where are we? Where should we be going? A review. Acad Med [Internet]. 2005 [acesso 9 maio 2022];80(12):1143-52. p. 1144-5. Tradução livre. DOI: 10.1097/00001888-200512000-00020
  » https://doi.org/10.1097/00001888-200512000-00020
• ³
  Brasil. Ministério da Educação. Resolução CNE/CES n° 3 de 20 de junho de 2014. Institui Diretrizes Curriculares Nacionais do Curso de Graduação em Medicina e dá outras providências. Diário Oficial da União [Internet]. Brasília, 23 jun 2014 [acesso 20 maio 2022]. Disponível: https://bit.ly/37G86uH
  » https://bit.ly/37G86uH
• ⁴
  Doukas DJ, McCullough LB, Wear S. Perspective: medical education in medical ethics and humanities as the foundation for developing medical professionalism. Acad Med [Internet]. 2012 [acesso 9 maio 2022];87(3):334-41. DOI: 10.1097/ACM.0b013e318244728c
  » https://doi.org/10.1097/ACM.0b013e318244728c
• ⁵
  Mahajan R, Aruldhas BW, Sharma M, Badyal DK, Singh T. Professionalism and ethics: a proposed curriculum for undergraduates. Int J Appl Basic Med Res [Internet]. 2016 [acesso 9 maio 2022];6(3):157-163. Disponível: https://bit.ly/3SXGMup
  » https://bit.ly/3SXGMup
• ⁶
  Castro MRC, Pereira AA, Bataglia PUR. Ética do cuidado: revisitando a ética na educação médica. Rev Bras Bioét. 2020;16(e9):1-20.
• ⁷
  Bataglia PUR, Morais A, Lepre RM. A teoria de Kohlberg sobre o desenvolvimento do raciocínio moral e os instrumentos de avaliação de juízo e competência moral em uso no Brasil. Estud Psicol (Natal) [Internet]. 2010 [acesso 9 maio 2022];15(1):25-32. DOI: 10.1590/S1413-294X2010000100004
  » https://doi.org/10.1590/S1413-294X2010000100004
• ⁸
  Lind G. Moral regression in medical students and their learning environment. Rev Bras Educ Méd [Internet]. 2000 [acesso 9 maio 2022];24(3):24-33. Disponível: https://bit.ly/3QIHpGY
  » https://bit.ly/3QIHpGY
• ⁹
  Lind G. How to teach morality: promoting deliberation and discussion, reducing violence and deceit. Berlin: Logos; 2016. p. 73.
• ¹⁰
  Rego S. A formação ética dos médicos: saindo da adolescência com a vida (dos outros) nas mãos [Internet]. Rio de Janeiro: Fiocruz; 2003 [acesso 22 ago 2022]. Disponível: https://bit.ly/3QGzD0f
  » https://bit.ly/3QGzD0f
• ¹¹
  Bataglia PUR. A validação do Teste de Juízo Moral (MJT) para diferentes culturas: o caso brasileiro. Psicol Reflex Crit [Internet]. 2010 [acesso 24 ago 2022];23(1), 83-91. p. 89. DOI: 10.1590/S0102-79722010000100011
  » https://doi.org/10.1590/S0102-79722010000100011
• ¹²
  Lind G. The meaning and measurement of moral judgment competence: a dual-aspect model. In: Fasko D Jr, Willis W, editors. Contemporary philosophical and psychological perspectives on moral development and education [Internet]. Cresskill: Hampton Press; 2008 [acesso 9 maio 2022]. p. 185-220. p. 200. Tradução livre. Disponível: https://bit.ly/3wmziYh
  » https://bit.ly/3wmziYh
• ¹³
  Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira. Sinopse estatística da educação superior 2017 [Internet]. Brasília: Inep; 2018. Disponível: https://bit.ly/3RIFhi5
  » https://bit.ly/3RIFhi5
• ¹⁴
  Feitosa HN, Rego S, Bataglia P, Rego G, Nunes R. Competência de juízo moral dos estudantes de medicina: um estudo piloto. Rev Bras Educ Méd [Internet]. 2013 [acesso 9 maio 2022];37(1):5-14. DOI: 10.1590/S0100-55022013000100002
  » https://doi.org/10.1590/S0100-55022013000100002
• ¹⁵
  Feitosa HN, Rego S, Bataglia PUR, Sancho KFCB, Rego G, Nunes R. Moral judgment competence of medical students: a transcultural study. Adv Health Sci Educ Theory Pract [Internet]. 2013 [acesso 9 maio 2022];18(5):1067-85. DOI: 10.1007/s10459-013-9449-5
  » https://doi.org/10.1007/s10459-013-9449-5
• ¹⁶
  Serodio A, Kopelman BI, Bataglia PUR. The promotion of medical students’ moral development: a comparison between a traditional course on bioethics and a course complemented with the Konstanz method of dilemma discussion. International Journal of Ethics Education [Internet]. 2016 [acesso 9 maio 2022];1(1):81-9. DOI: 10.1007/s40889-016-0009-8
  » https://doi.org/10.1007/s40889-016-0009-8
• ¹⁷
  Otani DON. Avaliação do desenvolvimento da competência moral de estudantes de medicina – duas escolas, dois mundos [dissertação] [Internet]. Curitiba: Faculdades Pequeno Príncipe; 2018 [acesso 10 maio 2022]. p. 56. Disponível: https://bit.ly/3AEaRZ3
  » https://bit.ly/3AEaRZ3
• ¹⁸
  Castro MR. Avaliação da competência moral de estudantes de medicina [dissertação]. Belo Horizonte: Universidade José do Rosário Vellano; 2019.
• ¹⁹
  Calliari MP. Importância e urgência de uma revisão na abordagem geracional na atualidade: uma proposta. [dissertação] [Internet]. São Paulo: Universidade Presbiteriana Mackenzie; 2018 [acesso 10 maio 2022]. Disponível: https://bit.ly/3c7iG04
  » https://bit.ly/3c7iG04
• ²⁰
  Sandars J, Morrison C. What is the net generation? The challenge for future medical education. Med Teach [Internet]. 2007 [acesso 9 maio 2022];29(2-3):85-8. DOI: 10.1080/01421590601176380
  » https://doi.org/10.1080/01421590601176380
• ²¹
  Borges NJ, Manuel RS, Elam CL, Jones BJ. Differences in motives between millennial and generation x medical students. Med Educ [Internet]. 2010 [acesso 9 maio 2022];44(6):570-6. DOI: 10.1111/j.1365-2923.2010.03633.x
  » https://doi.org/10.1111/j.1365-2923.2010.03633.x
• ²²
  Patenaude J, Niyonsenga T, Fafard, D. Changes in students’ moral development during medical school: a cohort study. CMAJ [Internet]. 2003 [acesso 9 maio 2022];168(7):840-4. Disponível: https://bit.ly/3PGOmqJ
  » https://bit.ly/3PGOmqJ
• ²³
  Self DJ, Baldwin DC Jr. Does medical education inhibit the development of moral reasoning in medical students? A cross-sectional study. Acad Med [Internet]. 1998 [acesso 9 maio 2022];73(10):S91-3. DOI: 10.1097/00001888-199810000-00056
  » https://doi.org/10.1097/00001888-199810000-00056
• ²⁴
  Feudtner C, Christakis DA, Christakis NA. Do clinical clerks suffer ethical erosion? Students’ perceptions of their ethical environment and personal development. Acad Med [Internet]. 1994 [acesso 9 maio 2022];69(8):670-9. DOI: 10.1097/00001888-199408000-00017
  » https://doi.org/10.1097/00001888-199408000-00017
• ²⁵
  Hojat M, Vergare MJ, Maxwell K, Brainard G, Herrine SK, Isenberg GA et al The devil is in the third year: a longitudinal study of erosion of empathy in medical school. Acad Med [Internet]. 2009 [acesso 9 maio 2022];84(9):1182-91. DOI: 10.1097/ACM.0b013e3181b17e55
  » https://doi.org/10.1097/ACM.0b013e3181b17e55
• ²⁶
  Self DJ, Baldwin DC Jr, Wolinsky FD. Further exploration of the relationship between medical education and moral development. Camb Q Healthc Ethics [Internet]. 1996 [acesso 9 maio 2022];5(3):444-9. p. 446. Tradução livre. DOI: 10.1017/S0963180100007271
  » https://doi.org/10.1017/S0963180100007271
• ²⁷
  Lind G. Favorable learning environments for moral competence development: a multiple intervention study with 3.000 students in a higher education context. Int J Univ Teach Fac Dev [Internet]. 2015 [acesso 9 maio 2022];4(4):173. p. 12. Tradução livre. Disponível: https://bit.ly/3T8Yprh
  » https://bit.ly/3T8Yprh

Publication Dates

• Publication in this collection
  28 Nov 2022
• Date of issue
  Jul-Sep 2022

History

• Received
  16 Nov 2021
• Reviewed
  10 Aug 2022
• Accepted
  15 Aug 2022