Attending physicians ’ attitudes towards physical exam and bedside teaching

Background: Medical education has experienced a gradual shift away from traditional bedside attending rounds, from 75% of rounds occurring at bedside in the 1960s to about 30% today. Aim: To examine attending physicians’ attitudes towards bedside teaching and physical exam. Methods: Anonymous survey of medical attendings in six academic hospitals. Results: 77% of respondents (n=97) completed the survey. The vast majority (89%) of attendings concurred that physical diagnosis skills are essential, felt that more emphasis on bedside teaching is needed (77%), and believed that bedside teaching should be a priority (71%). Additionally, 87% reported that bedside rounds are important to patient care. Yet, only 31% reported conducting teaching rounds at bedside. Finally, only 5% of attendings trained outside the US expressed fear of poor teaching performance in front of house staff, compared with US trained attendings (28%, p=0.023). Conclusion: Physicians are spending less and less time at the bedside, particularly those trained within the United States. At a time when the U.S. healthcare system is struggling to meet the increasing demand of escalating costs and declining patient satisfaction, the return to bedside teaching may be a surprisingly simple and untapped solution.


Introduction
Within the last 50 years, medical education has experienced a gradual shift from traditional bedside attending rounds to the conference room and hallway.In the 1960s, it was reported that 75% of attending rounds occurred at the bedside (Reichsman, Browning, & Hinshaw, 1964), while recent studies estimate this percentage to be between 8% and 28% (Gonzalo, Masters, Simons, & Chuang, 2009;Williams, Ramani, Fraser, & Orlander, 2008).This decline seems to have shifted the focus away from the patient, depriving the new generation of physicians of an opportunity to observe and learn physical exam techniques, interpersonal skills, and professionalism from senior role models at the bedside (Williams et al. 2008).Commonly cited reasons for this decline include time constraints, noisy wards, greater work demands, logistics of hospital organization, overreliance on technology, and perceived patient discomfort (Crumlish, Yialamas, & McMahon, 2009;Dewji, Dewji, & Gnanappiragasam, 2015;Gonzalo et al. 2009;LaCombe, 1997;Nair, Coughlan, & Hensley, 1997;Nair, Coughlan, & Hensley, 1998;Rogers, Carline, & Paauw, 2003;Williams et al. 2008).One study utilizing qualitative methods to explore these barriers, reported that physicians who are performing bedside evaluation are expected to possess an almost unrealistic level of diagnostic skill, making it daunting for the average physician to rise to the challenge (Ramani, Orlander, Strunin, & Barber, 2003).It has been suggested that the less time one spends at bedside, the more uncomfortable they are conducting rounds, and therefore the less time they spend doing so (Thibault, 1997).
Despite these barriers, a recent review revealed that patients, learners, and teachers all seem to favor bedside teaching (Peters & ten Cate, 2014).Learners still believe that bedside learning is important for professional development and for learning core clinical skills such as patient-physician communication, physical examination, and clinical reasoning (Crumlish et al. 2009;Gonzalo et al. 2009;Nair et al. 1997;Rogers et al. 2003;Williams et al. 2008).Moreover, the belief that bedside presentations are stressful for patients has not been supported (Simons, Baily, Zelis, & Zwillich, 1989).Studies in both outpatient and inpatient settings reveal that patients exposed to bedside presentations are more likely to report favorable perceptions of their care, perceive greater educational benefit, and would prefer subsequent presentations by the bedside (Fletcher, Rankey, & Stern, 2005;Lehmann, Brancati, Chen, Roter, & Dobs, 1997;Majdan, Berg, Schultz, Schaeffer, & Berg, 2013;Nair et al. 1997;Rogers et al. 2003;Wang-Cheng, Barnas, Sigmann, Riendl, & Young, 1989) .Overall, patients have been found to be very satisfied with bedside teaching (Peters & ten Cate, 2014).In essence, bedside teaching fosters the patient-physician relationship, as it can provide a profound catalytic experience, allowing physicians to immerse themselves in the depths of human illness (Qureshi & Maxwell, 2012).
Today's ubiquitous medical technology further shifts the focus away from the traditional examination skills cultivated at bedside (Dewji et al. 2015;Qureshi & Maxwell, 2012), undermining the diagnostic utility of the physical exam (Crumlish et al. 2009;Gonzalo et al. 2009).
Of additional concern, is the lack of growth in bedside clinical skills over time along with the striking absence of progress in the skill level of medical professionals, including students, fellows, and faculty (Vukanovic-Criley et al. 2006;Mangione, 2001;Mangione, Burdick, & Peitzman, 1995;Mangione & Nieman, 1997;Mangione & Neiman, 1999).Indeed, previous research has shown that half of practicing hospitalists do not feel confident in teaching physical examination skills (Crumlish et al. 2009).Similarly, studies of cardiac examination competency reveal that bedside skills do not improve significantly through different levels of training (Mangione, 2001;Mangione & Nieman, 1997;Vukanovic-Criley et al. 2006).In one such study that included cardiology students, fellows, and faculty, only cardiology fellows tested significantly better than students and residents (Vukanovic-Criley et al. 2006).Likewise, when Mangione et al. (1995) had internal medicine and family medicine residents listen to 12 prerecorded common cardiac events, both groups recognized only 20% of events on average, improving only slightly with training level (Mangione & Nieman, 1997).
Despite the popularity of simulations to supplement learning (Peters & ten Cate, 2014;Qureshi & Maxwell, 2012), certain components of the physical examination have great diagnostic utility and cannot be learned any other way than at the bedside.For example, the presence of a third heart sound (Drazner, Rame, Stevenson, & Dries, 2001), indicative of severe hemodynamic dysfunction (Tribouilloy et al. 2001), is the most important predictor of postoperative complications, and can only be assessed at the bedside (Goldman et al. 1977).Other experiences that cannot be simulated and adequately learned outside of bedside teaching include the tactile sensation of hepatosplenomegaly and joint effusions (Qureshi & Maxwell, 2012).
In addition to declining skills, studies also point to a lower perceived utility of the physical exam.In a study of attending physicians at an academic teaching center, authors found a significant negative correlation between the mean overall perceived utility of the physical exam and increased training level (Wu, Fagan, Reinert, & Diaz, 2007).They also reported a positive correlation between self confidence in performing the exam and increased training levels -with attending physicians reporting a mean level of confidence 3.9 out of 5 (Wu et al. 2007).When Fagan and colleagues (2006) used the same methodology to survey fourth-year medical students (MS4s) at United States and Dominican Republic medical schools, they found that students at the Dominican school reported significantly greater confidence in their overall physical examination skill as compared to US students.The students at the Dominican school also had more positive views about the diagnostic utility of the physical examination (Fagan, Lucero, Wu, Diaz, & Reinert, 2006).These findings could be related to the increased availability of diagnostic technology in the US, which results in decreased emphasis on the physical examination as a tool.
Such data is troublesome, given the fact that a well-performed physical examination can provide over 20% of the data necessary for patient diagnosis (Campbell & Lynn, 1990).With the rising costs of healthcare and the relatively low cost of a physical examination, compared to imaging and laboratory studies, perhaps going back to the bedside would not only be prudent, but ultimately more economical (Peixoto, 2001).In an effort to contribute to a greater understanding of these issues, we decided to investigate whether there is an association between faculty attitudes toward the physical exam and frequency of bedside teaching during attending rounds.

Methods
Design and participants: A cross-sectional anonymous survey study was conducted.Surveys were distributed via interoffice mail and during grand rounds to all attendings taking part in the inpatient internal medicine teaching services of six hospitals in the New York metropolitan area during the 2009-2010 academic years.Attendings without inpatient teaching responsibilities during the year were excluded from the analysis.
Attendings targeted for participation received a cover letter in inter-office mailbox introducing the survey, with the questionnaire attached.To ensure anonymity and unbiased responses, surveys were administered before grand rounds, during which attendings were instructed to drop off completed surveys into a box located right outside the grand rounds hall.Those physicians receiving the survey via inter-office mail also received an instruction letter detailing where to drop off completed surveys at their convenience.Attendings were given one month to return the surveys.
The study was approved by the Northwell Health Institutional Review Board.
Measures: A survey was developed for the purpose of this study, with questions about self-confidence adapted from Wu et al. (2007), and attitude questions adapted from Gonzalo et al. (2009) and Crumlish et al. (2009).Data collected included demographics (gender, ethnicity, field of expertise, whether they were trained in the US or abroad, number of years working as an attending physician), number of months per year working on a teaching service, average length and frequency of rounds, and time spent at different locations during rounds.
In addition, participant self-rated confidence in performing an overall physical exam and eleven specific skills were assessed, along with Likert type items rating the importance of bedside teaching and the diagnostic utility of physical exam.
Data analysis: Chi-square and Fisher's Exact Tests were used for hypothesis testing, using a 2-tailed analysis with an alpha of 0.05 as the criterion for significance.Attendings who did not report "length of service per year" (n=14) were excluded from the analysis.

Results
Out of the 126 surveys distributed, 97 were returned completed, with an overall response rate of 77%.After excluding attendings who did not report length of service, 83 respondents were included in the analysis.A summary of participant demographics can be found in Table 1.
With regard to confidence in physical exam skills, most attendings felt very confident (48%), or somewhat confident (43%), in their overall diagnostic abilities.Respondents reported more confidence in the detection of ascites (93%) and interpretation of systolic murmur (92%), and were least confident with regard to the fundoscopic exam (21%) and distinguishing between mole and melanoma (59%).There was no significant association between physical exam skills and time spent at bedside during rounds.
Table 2 presents overall attending bedside and physical exam attitudes.The majority of attendings reported that bedside rounds are important for teaching purposes (92%) and patient care (87%).The vast majority (89%) of attendings reported that physical diagnosis skills are essential, more emphasis on bedside teaching is needed (77%), and that bedside teaching is a priority (71%).Just under half reported their belief that patients (49%) and house staff (42%) prefer bedside teaching.Additionally, 87% reported that bedside rounds are important to patient care.Notably, 17% feared poor teaching performance in front of patients and house staff (22%).
Interestingly, significant results were found for fear of poor teaching by location of training (Table 3).Only 5% of attendings trained outside the US expressed fear of poor teaching performance in front of house staff, while significantly more attendings trained within the US (28%) reported this fear (p=0.023).A significantly greater proportion of attendings with training outside the US also reported making bedside teaching a priority (91% v. 63%, p=.015).
Significant associations were found between the amount of time spent teaching at the bedside and attitudes towards bedside teaching (Table 4).Attending beliefs that physical exam can only be taught at the patient's bedside was greater for those spending more than 25% of time at the bedside (p=0.027).Similarly, there was a significant association between the amount of time spent teaching at the bedside and attending beliefs that patients prefer bedside rounds (p=0.007), as well as beliefs that bedside rounds are important for teaching purposes (p=0.018).
Attendings spending more time teaching at bedside felt strongly that more emphasis on bedside teaching was needed in the curriculum (p=0.003) and that bedside teaching is a priority (p=0.003).

Discussion
Our study results support well-established data published almost two decades ago which indicated that, from the patient's perspective, bedside case presentations were "at least as good as conference-room presentations, and perhaps preferable" ( Lehmann et al. 1997).Yet the present study is in line with previous research indicating that attendings spend only about 30% of teaching rounds at the bedside (Gonzalo et al. 2009).This percentage is likely to be an overestimate, as other studies found significant differences between estimated and actual times that physicians spend at the bedside (Miller, Johnson, Greene, Baier, & Nowlin, 1992).When comparing studies reporting physician estimates to direct observation, the percent of time devoted to bedside examination during teaching rounds dwindled to 11-17% (Crumlish et al. 2009;Miller et al. 1992).
Among the most interesting of our results, are the differences found between attendings trained within and outside the US.Very few other studies have examined attitudes toward bedside teaching across different countries, often with methodological limitations (Mangione, 2001).In our study, when comparing physicians trained within and outside the US, both groups concur completely with the belief that physical diagnosis skills yield clinically relevant information and are required to make the correct diagnosis.However, the vast majority (91%) of physicians trained outside the US make bedside teaching a priority, compared with less than 65% of physicians trained within the US (p<.01).Furthermore, while both groups feel equally confident in their ability to lead bedside teaching rounds, US trained attendings report a significantly greater level of fear of exhibiting poor teaching performance in front of house staff than non-US trained physicians (28% vs. 5%, p<.03).
These findings provide some context for much of the findings in the medical education literature with regard to international differences between curriculum approaches.For instance, a study examining cardiac auscultation teaching among trainees in the US, UK, and Canada found that British and Canadian trainees received significantly more training of this skill in medical school and residency as compared to those in the US (Mangione, 2001).In addition, both British and Canadian trainees are expected to undergo an objective assessment of physical examination skills.British trainees improved the most in assessing cardiac auscultation and Canadian trainees had the greatest accuracy.
As common with survey studies, the reliance on subject recall represents one limitation of the current study and may have partially accounted for the findings.In addition, while the survey was designed to be anonymous, social desirability bias may have resulted in an overestimation of physician time spent at bedside.Notable strengths of this research lie in the fact that we sampled across six different academic hospitals across two urban and suburban boroughs of New York City, and attained an excellent response rate (77%).

Conclusion
The literature consistently reports, over several decades, that bedside teaching is greatly valued by physicians at all levels, from medical students to attendings.Yet, physicians are spending less and less time at the bedside.This is particularly true of physicians trained in the United States.At a time when the US healthcare system is struggling to meet the increasing demands of escalating costs and declining patient satisfaction, the return to bedside teaching may be a surprisingly simple and untapped solution.

Take Home Messages
The vast majority of attendings report that bedside teaching is important for teaching purposes and patient care.Findings indicate that only 31% of teaching rounds are held at bedside.Attendings trained outside the United States felt more confident in their ability to perform bedside teaching in front of house staff and were more likely to make bedside teaching a priority.

Notes On Contributors
REVEKKA BABAYEV, MD, received her medical doctorate from Albert Einstein College of Medicine and went on to do her Internal Medicine residency and nephrology fellowship at Columbia University Medical Center.She is currently working at Stamford hospital as a clinical nephrologist and is interested in medical education.