Journal of the American College of Nutrition, Vol. 27, No. 1, 59-64 (2008)
Published by the American College of Nutrition
Survey Correlations: Proficiency and Adequacy of Nutrition Training of Medical Students
Tanis V. Mihalynuk, PhD, RD,
John B. Coombs, MD, MS,
Michael E. Rosenfeld, PhD,
Craig S. Scott, PhD and
Robert H. Knopp, MD
Interdisciplinary Nutritional Sciences Program (T.V.M.)
Departments of Family Medicine (J.B.C.)
Pathobiology and Pathology (M.E.R.)
Medical Education (C.S.S.)
Medicine (R.H.K.), University of Washington, Seattle, Washington
Address correspondence to: Dr. Tanis Mihalynuk, Program Research Leader, Nutrition, Alberta Cancer Board, Holy Cross Site, Box ACB, 2210 - 2 St SW Calgary AB T2S 3C3 CANADA. E-mail: tanismih{at}cancerboard.ab.ca or tanisvye1{at}yahoo.com
 |
ABSTRACT
|
|---|
Objective: The majority of graduating US medical students reported inadequate nutrition training over the past decade. This trend could in part be due to the lack of valid measures to assess the relationship between adequacy of nutrition training and proficiency on nutrition topics deemed essential. The study's objective was to test the hypothesis that self-reported nutrition proficiency is positively correlated with the perceived adequacy (quality, quantity, coverage and importance) of nutrition training of University of Washington medical students.
Method: Cross-sectional e-mail survey of 1st to 4th year medical students (n = 708), including a survey prompt and three e-mail follow-up measures. To reduce and interpret the survey data, principal components analysis was employed, followed by Varimax rotation with Kaiser normalization. To assess internal consistency reliability, alpha (
) of nutrition proficiency items and factors was determined.
Results: A 44.5% response rate was achieved (n = 315 respondents). The 31-item questionnaire was reduced to 6 factors, explaining 60.2% of the total variance ( = 0.947). Self reported nutrition proficiency was positively correlated with the perceived quality, quantity and coverage of nutrition training in all 6 essential nutrition factors or topics determined after factor analysis (P < 0.01).
Conclusion: Quality and coverage may be effective gauges of adequacy of nutrition training and related nutrition proficiency in medical education. Current national medical education evaluation measures focus on the quantity of nutrition instruction. The lowest reported proficiency topics; nutrition and disease management, micronutrients and complementary and alternative medicine are recommended for particular curricular emphasis.
Key words: nutrition proficiency, self report, adequacy of nutrition training, medical students
|
INTRODUCTION
|
|---|
Physicians surveyed persist in reported lack of confidence in basic nutrition counseling due to perceived inadequate nutrition training in medical school [1–2]. Furthermore, over the past decade, the majority of US medical student graduates reported inadequate time devoted to nutrition training, as summarized by aggregate evaluation reports by the Association of American Medical Colleges’ Graduation Questionnaire (AAMC-GQ) [3].
There are many possible explanations for these reports. First, medical students may not be receiving the essential information required to achieve basic nutrition competencies. This is despite the creation of several expert-committee-derived essential nutrition lists, including those developed by the American Medical Students Association [4], National Institutes of Health's Nutrition Academic Award (NIH-NAA) [5], and American Society of Clinical Nutrition (ASCN) [6]. Considerations include whether these expert committee-derived essential nutrition lists include the consensus ‘voice’ of the future and practicing physicians, as well as temporal and geographic variation in essential nutrition information.
Second, ideal gauges of the adequacy of nutrition training remain elusive. The AAMC-GQ focuses on time devoted to nutrition instruction or quantity as a national gauge of adequacy of nutrition instruction. Yet many medical schools have not quantified their nutrition curriculum. Also, curricular content is often modified in response to evolving knowledge, research trends and changing priority education topics. As such, it is possible that other variables of adequacy may have more generalized relevance.
Third, there is a dearth of valid measures to assess relationships between nutrition training and proficiency. Earlier studies focused on nutrition knowledge [7–10], as well as the relationship between nutrition knowledge and the perceived quantity and quality of nutrition training [7]. However, knowledge does not necessarily predict competence or proficiency in basic physician nutrition counseling, which is at the heart of physician-related health promotion and disease prevention strategies. As such, examining the relationships between nutrition confidence, self-efficacy, or clinical competence and perceived adequacy of nutrition training is of increasing import. There is growing evidence that nutrition training enhances student and physician proficiency and related patient care. Carson et al [11] observed a two-fold increase in nutrition self-efficacy of the experimental group of 4th year medical students receiving an education intervention over the control group. The intervention included provision of nutrition-related resources, web-based case studies and class discussions. The authors also observed that students with greater self-efficacy in cardiovascular nutrition were more apt to counsel patients on nutrition-related issues. In terms of physician nutrition proficiency, Levine et al [12] conducted a national survey to determine the extent to which primary-care physicians were practicing core nutrition competencies as outlined by Young et al [13]. Although respondents felt nutrition played an important role in health promotion and disease prevention, most physicians surveyed did not practice the core nutritional competencies due to lack of perceived nutrition proficiency. Predictors of medical-nutrition practice in this survey sample included personal attention to dietary habits and use of nutrition-specific resources. Moreover, Kolasa [14] argues that nutrition education focus on confidence and skill building (proficiency) over knowledge in an effort to provide effective nutrition services to primary care patients.
In an earlier study, we examined relationships between self-reported nutrition proficiency and perceived adequacy (quality, quantity) of nutrition training of Washington Academy of Family Physicians (WAFP) members [1]. Self-reported nutrition proficiency was positively correlated with the perceived quality of nutrition training in all five essential nutrition factors or topics determined after factor analysis. This correlation was also observed with a dichotomous measure of perceived quantity (presence/absence) of nutrition training in some factors. Lowest proficiency factors or topics included: micronutrients in health, including complementary and alternative medicines (CAM); and nutrition & disease management.
At the University of Washington (UW) where this study was based, a 12-hour core nutrition course is offered in 2nd year of medical school. The content of the course has undergone many revisions in response to student and faculty feedback. A recent addition included a practical component in which students conduct a peer-based analysis of diet and physical activity patterns, and subsequently provided their student partner recommendations for lifestyle modification. A nutrition course elective is also offered to interested students in clinical years of training. To further examine the relationship between nutrition training and proficiency, we tested the hypothesis that self-reported nutrition proficiency is positively correlated with several variables of perceived adequacy of nutrition training, including scaled measures of quality, quantity, coverage and importance of essential nutrition information of UW medical students.
|
METHODS
|
|---|
A cross-sectional e-mail nutrition proficiency and training survey of 1st to 4th year medical students was developed, administered and analyzed following approval by the UW Human Subjects Committee under minimal risk guidelines from April, 2002 (ongoing). Data collection ended August, 2002. The survey was administered to all available student e-mail addresses (n = 708) using UW's "Catalyst Web Q Survey Tools’ [15]. This system confidentially tracks respondents, providing ease of follow-up to non-respondents. To enhance response rates, the survey methodology was based on Dillman's Tailored Design Method [16], including three e-mail follow-up measures and an e-mail survey prompt sent previous to survey administration. The e-mail prompt informed medical students that they would receive a survey in their e-mail account in one week's time. Although extensive methodological details have been published earlier [1], a brief synopsis of research methods follows.
As in the study of family physicians, the survey designed for medical students assessed self-reported nutrition proficiency scores on 31 essential nutrition items derived from ASCN [6] using a 3-point modified likert rating scale (totally proficient, partially proficient, not proficient). Variables of perceived adequacy included modified likert scales of quality (5-point—poor, fair, good, very good, excellent), quantity—other than the 12 hour, core nutrition course (6-point—none—0 hours, 1–6 hours, 7–17 hours, 18–28 hours, 29–39 hours, 40 or more hours), coverage on essential topics derived from the WAFP study (5-point not at all covered, partial coverage-inadequate for future practice, partial coverage, adequate for future practice, complete coverage, excessive coverage) and importance (5-point—not at all important, minimally important, moderately important, very important, extremely important). A-priori proposed factors (6 factor model) included 5 factors derived from the WAFP study and an added Obesity factor. We also asked a dichotomous (yes, no) question regarding whether the quantity of nutrition training received in medical school was adequate.
Data were coded and analyzed in a SPSS 10 for Windows database [17]. For data reduction purposes and to test study hypotheses, principal components analysis was used to extract the factors, followed by Varimax rotation with Kaiser normalization to rotate the factors. This method creates new variables or factors that are exact mathematical representations of the original data. Labels were attached to the identified subscales or factors, according to content. Alpha (internal consistency reliability) of the entire instrument and individual factors was determined, whereby survey items were statistically grouped together that measure the same concept. We looked for significant correlations at the P < 0.01 level (two-tailed). To assess non-response bias, we compared basic demographic characteristics of respondents and non-respondents.
|
RESULTS
|
|---|
A 44.5% response rate (n = 315) was achieved from this study design. Survey respondents were similar in sex, but older in mean age than the total UW medical student population (P < 0.001). The majority of survey responses were received from students in pre-clinical years (59.8%, n = 186) (Table 1). The perceived adequacy of quantity of nutrition training was positively correlated with self reported nutrition proficiency in all six factors determined after factor analysis (P < 0.001). First year students reported lower proficiency than all other medical school years (P < 0.001). Second year students recently completing the core nutrition course reported highest proficiency levels (P < 0.001). About one third (33.6% or n = 226) of respondents who had been exposed to the core nutrition course reported the quantity of their nutrition training as adequate. Reports of adequacy of quantity of nutrition training decreased in 4th year student respondents (27.5% or n = 70) near completion of their training (Table 2).
Alpha and Factor Analysis
The alpha of the 31-item nutrition proficiency and related training needs instrument was 0.947 (n = 292). Alpha of individual factors ranged from 0.738 to 0.857. Factors listed in descending order of self-reported nutrition proficiency (Table 3) included: i) women, infants, children & obesity (WIC); ii) macronutrients (MAC); iii) nutrition in prevention & wellness (NPW); iv) lifecycle nutrition (LIF); v) micronutrients /CAM (MIC); and vi) nutrition and disease management (NDM). The six-factor model explained 60.2% of the total variance.
Construct validity was supported by a-priori hypotheses, all positive, moderate or high factor loadings (>0.3) and normal distributions of composite scores for self-reported nutrition proficiency factors and for perceived coverage and importance of nutrition information, and scaled scores for perceived quality and quantity of nutrition training. Moderate factor loading items (less than 0.4) included: advice for feeding a colic infant (Factor 3); nutrition strategies to reduce common digestive complaints (Factor 6); and overall benefits of aerobic exercise on health and well being (Factor 3).
Person Correlations (PC)
Nutrition proficiency factor scores and scaled or composite scores of perceived adequacy of nutrition training (quality, quantity, coverage and importance) were normally distributed, allowing for study hypotheses to be tested. Positive PC were observed between self-reported nutrition proficiency and perceived quality, quantity and coverage of nutrition information across all six factors (P < 0.01). The most robust correlation was between perceived coverage and second quality of nutrition training and nutrition proficiency self-report. Correlations between self-reported nutrition proficiency and perceived importance of nutrition information were also observed in two factors: Lifecycle Nutrition and Micronutrients/CAM factors (P < 0.05) (Table 4).
View this table:
[in this window]
[in a new window]
|
Table 4. Pearson's Correlation Coefficients (PC) between Composite Scores of Perceived Quality, Quantity, Coverage and Importance of Nutrition Information in Medical School and Self-Reported Nutrition Proficiency (NP) factors (F1 to 6)
|
|
|
DISCUSSION
|
|---|
We did not expect coverage to provide the most robust correlation with nutrition proficiency self-report, followed by quality. In addition, it is possible that a stronger correlation may have existed across proficiency factors with perceived importance of nutrition training using more equidistant likert scale labels that better approximated a continuous random variable. We used a right-packed scale given overwhelming medical school graduate reports of inadequacy of nutrition training in medical school [3].
In terms of related studies, Weinsier et al [7] assessed the nutrition knowledge of senior medical students at 10 southeastern US medical schools in the Southeastern Regional Medical-Nutrition Education Network (SERMEN). The average school response rate was 21% (range = 12% to 53%). Knowledge scores were positively correlated with the perceived quality (r = .35, p < 0.001) and quantity (r = .28, p < 0.001) of nutrition training. The relationship between the perception of importance of nutrition to their career and their knowledge was not statistically significant (r = 0.11). The relationship between perceived coverage of nutrition information and nutrition knowledge was not assessed.
In our study, first year students who had not been exposed to the required nutrition course reported significantly lower proficiency scores than other medical school years. The second year medical students who recently completed the core nutrition course reported highest proficiency scores. This "recency effect’ combined with lowest reports of adequacy of nutrition training (quantity) of fourth year students underscores the importance of ongoing exposure to nutrition information in medical school. Weinsier et al [9] also highlighted the importance of ongoing exposure to nutrition to enhance both nutrition knowledge and attitudes of medical students at three points in their training. Furthermore, Morgan et al [10] observed significantly higher nutrition knowledge scores after students were exposed to a required nutrition course compared to a biochemistry course covering nutrition topics, underscoring the importance of exposure and labeling of nutrition in the medical curriculum.
Aside from our earlier study of family physicians, we are not aware of existing research examining correlations between perceived adequacy of nutrition training and self-reported nutrition proficiency on essential topics derived from factor analysis. It is possible that proficiency may be a more comprehensive gauge of adequacy of medical-nutrition training as it includes both knowledge and related competence in counseling. Although a self-report measure of proficiency—a more complex variable than knowledge—was used, paradoxically, higher correlations have been observed between self-report and actual performance and clinical skills compared to self-reported and actual knowledge [18]. This observation merits further study in other study populations and curricular domains to determine its generalizability.
Quality and coverage may be ideal measures of adequacy of nutrition training, particularly given that many medical schools do not offer a required nutrition course, and quantity of instruction may change in response to student evaluations, shifting priorities and scientific updates. In a commentary of quality improvement in medical students’ education, Lockwood et al [19] noted that aggregate responses of the AAMC-GQ provide a national perspective on the quality of medical students’ education. Perhaps inclusion of quality evaluation measures of individual courses would provide further insight into the adequacy of instruction on general and specific medical education topics.
In terms of priority nutrition curricular topics, and as was observed in this study of medical students and our earlier study of family physicians, the lowest proficiency factors or topics, nutrition and disease management and micronutrients/CAM are suggested for particular curricular emphasis. In a survey of a random sample of US physicians, Darer et al [20] recently observed majority physician reports of inadequate medical training in chronic disease care in ten competencies, including nutrition.
Our study's limitations include the moderate response rate, and as such, there is the possibility of non-response bias, particularly in the lower responding years 3 and 4. Non-response causes bias if the probability of response is correlated with survey variables. Yet a higher age of survey respondents did not appear to bias nutrition proficiency total or factor scores and the relationship of these scores to perceived adequacy of nutrition training. Moreover, we were able to observe distinct medical school year reports of nutrition proficiency, regardless of age. Finally, survey response rates were sufficient to provide adequate statistical power, including sufficient number of respondents to items (10:1) [21] to conduct a factor analysis.
Other study limitations included temporal issues underlying a cross-sectional survey methodology. Many senior medical students received the survey at a time when they had other overriding priorities, including clerkship rotations and career choice considerations. The moderate or high positive factor loadings and normal distributions of scaled and composite variable scores, along with conceptually similar factor structures or essential nutrition topics identified in two populations (medical students and physicians) suggest we are measuring a construct of nutrition proficiency and related training needs. Future studies in geographically diverse settings (external validity) which examine correlations between self-reported and actual nutrition proficiency of future and practicing physicians, and how these impact patient outcomes is recommended to address the decades-old question, "What is the essential nutrition information that should be included in medical education?"
|
ACKNOWLEDGMENTS
|
|---|
We gratefully acknowledge Dr Adam Drewnowski for his invaluable support and consultation during various stages of survey instrument development and analysis, and the University of Washington medical students for their interest and participation in this study.
TVM contributed to the design, data collection, data analysis and writing of this study. She is grateful to JBC, MER, CSS and RHK for their review of the manuscript and for their contribution to the study design. There are no conflicts of interest in this study.
|
FOOTNOTES
|
|---|
When this research was conducted, Dr Mihalynuk was a doctoral student in the University of Washington's Interdisciplinary Nutritional Sciences Program, Seattle, Washington.
Received February 22, 2006.
Accepted November 17, 2006.
|
REFERENCES
|
|---|
- Mihalynuk TV, Scott CS, Coombs JB: Self reported nutrition proficiency is positively correlated with the perceived quality of nutrition training of family physicians in Washington State.Am J Clin Nutr77
:1330
–1336,2003
.[Abstract/Free Full Text]
- Kushner RF: Barriers to providing nutrition counseling by physicians: a survey of primary care practitioners.Prev Med24
:546
–552,1995
.[Medline]
- Association of American Medical Colleges (AAMC) Medical School Graduation Questionnaire: All Schools Report. Washington, DC, USA (1995-2005). Online: http://www.aamc.org/data/gq/allschoolsreports/start.htm Accessed 11 November,2006
.
- Anonymous: Essentials of nutrition education in medical schools: a national consensus. American Medical Student Association's Nutrition Curriculum Project.Acad Med71
:969
–971,1996
.[Medline]
- Pearson TA, Stone EJ, Grundy SM, McBride PE, Van Horn L, Tobin BW: Translation of nutritional sciences into medical education: the Nutrition Academic Award Program.Am J Clin Nutr74
:164
–170,2001
.[Abstract/Free Full Text]
- Weinsier RL: National dairy award for excellence in medical/dental nutrition education lecture, 1995. Medical-nutrition education—factors important for developing a successful program.Am J Clin Nutr62
:837
–840,1995
.[Abstract/Free Full Text]
- Weinsier RL, Boker JR, Feldman EB, Read MS, Brooks CM: Nutrition knowledge of senior medical students: a collaborative study of southeastern medical schools.Am J Clin Nutr43
:959
–968,1986
.[Abstract/Free Full Text]
- Morgan SL, Weinsier RL, Boker JR, Brooks CM, Feldman EB, Read MS: A comparison of nutrition knowledge of freshmen and senior medical students: a collaborative study of southeastern medical schools.J Am Coll Nutr7
:193
–197,1988
.[Abstract]
- Weinsier RL Boker JR, Morgan SL, Feldman EB, Moinuddin JF, Mamel JJ, DiGirolamo M, Borum PR, Read MS, Brooks CM: Cross-sectional study of nutrition knowledge and attitudes of medical students at three points in their medical training at 11 southeastern schools.Am J Clin Nutr48
:1
–6,1988
.[Abstract/Free Full Text]
- Morgan SL, Weinsier RL, Boker JR, Brooks CM: Nutrition education for medical students: evaluation of the relative contribution of freshman courses in biochemistry and nutrition to performance on a standardized examination in nutrition.Nutrition5
:31
–36,1989
.[Medline]
- Carson JS, Gillman B, Kirk LM, Reddy ST, Battles JB: Enhancing self-efficacy and patient care with cardiovascular nutrition education.Am J Prev Med23
:296
–302,2002
.[Medline]
- Levine BS, Wigren MW, Chapman DS, Kerner JF, Bergman RL, Rivlin RS: A national survey of attitudes and practices of primary-care physicians relating to nutrition: strategies for enhancing the use of clinical nutrition in medical practice.Am J Clin Nutr57
:115
–119,1993
.[Abstract/Free Full Text]
- Young EA, Weser E, McBride HM, Page CP, Littlefield JH: Development of core competencies in clinical nutrition.Am J Clin Nutr38
:800
–810,1983
.[Abstract/Free Full Text]
- Kolasa KM: Dannon Institute award for excellence in medical/dental nutrition education lecture, 2000. "Images" of nutrition in medical education and primary care.Am J Clin Nutr73
:1006
–1009,2001
.[Abstract/Free Full Text]
- Catalyst Web Tools: The Catalyst Tool- Box. Educational Technology Development Group, the University of Washington.Educational Partnerships and Learning Technology
, 1998–2003. Online: http://catalyst.washington.edu/tools. Accessed 11 November2006
.
- Dillman D: "Mail and Internet Surveys: the Tailored Design Method," 2nd ed. New York: John Wiley & Sons,2000
.
- SPSS Inc: SPSS 10 for Windows Statistical Software Package, release 10.0.07 (1 June 2000) Chicago: SPSS, Inc, 1989–1999.
- Gordon MJ: A review of the validity and accuracy of self-assessments in health professions training.Acad Med66
:762
–769,1991
.[Medline]
- Lockwood JH, Sabharwal RK, Danoff D, Whitcomb ME: Quality improvement in medical students’ education: the AAMC medical school graduation questionnaire.Med Educ38
:232
–238,2004
.[Medline]
- Darer JD, Hwang W, Pham HH, Bass EB, Anderson G: More training needed in chronic care: a survey of US physicians.Acad Med79
:541
–548,2004
.[Medline]
- Hazard Munro E: "Statistical Methods for Healthcare Research." Philadelphia: Lippincott Williams and Wilkins,2001
.
TOP
ABSTRACT
INTRODUCTION
