The Accreditation Council for Graduate Medical Education requires that residents demonstrate increasing autonomy during their training. Although residents report a better educational environment with hospitalists present during family-centered rounds (FCRs), there is a concern that attending presence may reduce resident autonomy. We aim to determine the effect of FCRs without an attending during rounds on senior residents’ sense of autonomy.
We conducted a multicenter, retrospective, preintervention-postintervention study at 5 children’s hospitals to evaluate the effect of rounding without an attending on senior residents’ self-efficacy, using a questionnaire developed by using Bandura’s principles of self-efficacy and Accreditation Council for Graduate Medical Education milestones. Questions addressed skills of diagnosis and/or management, communication, teaching, and team management. We compared preintervention and postintervention results using paired t tests and Wilcoxon rank tests. One-way analysis of variance tests were used to compare means among >2 groups.
116 (82% response rate) of 142 eligible senior residents completed the questionnaire, which yielded a high reliability (α = 0.80) with a 1-factor score. The average composite score of self-efficacy significantly improved after intervention compared with the preintervention score (66.71 ± 6.95 vs 60.91 ± 6.82; P < .001). Additional analyses revealed meaningful improvement of each individual item postintervention. The highest gain was reported in directing bedside teaching (71.8% vs 42.5%; P < .001) and answering learner questions on rounds (70.7% vs 47.0%; P < .001).
Conducting FCRs without an attending increases resident reported self-efficacy regarding core elements of patient care and team leadership. In future studies, researchers should examine the impact of rounding without the attending on other stakeholders, such as students, interns, patients and/or families.
Resident autonomy, defined as progressive independence from supervision, is a core principle of clinical training. The Accreditation Council for Graduate Medical Education requires that residents show progressive independence during training, developing increasing responsibility under the supervision of attending physicians.1 Although studies have revealed that residents report a better educational environment with supervision by an attending,2–5 there are concerns that increased attending presence might reduce resident autonomy and opportunities to make independent decisions.6–12
In clinical teaching settings, attending physicians must balance safe and effective patient care while allowing residents to develop self-confidence, skills, and competencies to become independent and competent physicians. However, one study revealed discordance between attending and resident perceptions of resident autonomy,13 suggesting that current practices may not allow trainees to be independent to the degree that faculty believe they should by the end of residency. Ultimately, residency programs must ensure physicians in training master skills necessary to competently care for patients independently.14 As such, progressive independence for residents and optimal health outcomes for patients are compatible goals.15 Family-centered rounds (FCRs) allow attending physicians to observe trainees’ interactions with patients and families and provides an opportunity for attending physicians to gauge residents’ ability to practice independently.16
An individual’s belief in their capacity to produce results plays a major role in attaining new skills.17 Self-efficacy, defined as an individual’s judgment about their capabilities, is an important determinant of skill performance and a sense of autonomy.17 Measures of self-efficacy regarding specific behaviors reflect the capacity to perform those behaviors autonomously and have been used to assess residents’ abilities in decision-making in pediatrics.18 Autonomy, along with competence and relatedness, is the basis for optimal function and growth and development according to self-determination theory (SDT),19–21 and, therefore, the environment should nurture these needs. Using SDT as a conceptual framework, we aimed to determine if conducting FCRs without the attending physician (attendingless rounds) during pediatric hospital medicine (PHM) rotations impacts senior residents’ sense of autonomy and self-efficacy in decision-making.
Methods
Study Design
We conducted a multicenter, retrospective, preintervention-postintervention study using self-efficacy scales to study the effect of an educational initiative on senior residents’ self-perception of autonomy and confidence in decision-making. This study was approved by the institutional review board at the coordinating site and participating sites.
Site Selection
Five geographically diverse sites in New York (2); Washington, District of Columbia; Illinois; and Texas were included (Table 1). Sites were selected on the basis of identification of a local champion with interest in promoting resident autonomy, the preexistence of (or ability to implement) attendingless FCRs, and the ability to adhere to the study protocol and collect the necessary data at their institution for at least 1 year.
Description of Participating Programs
Site . | Residents and/or Class . | Rotation Duration, wk . | Team Structure . | Intervention Implementation . |
---|---|---|---|---|
1 | 29 | 2 | 1 PGY2 senior; 2–3 PGY1 residents and/or sub-I; and 2–3 MS3s | Once a week × 1 or 2 wk |
2 | 40 | 4 | 1 PGY2 or PGY3 senior; 1–2 PGY1 residents and/or sub-Is; and 2–4 MS3s | Once a week × 4 wk |
3 | 27 | 4 | 1–2 PGY2 or PGY3 seniors; 4 PGY1 residents and/or sub-Is; and 2–3 MS3s | Once per 2-wk period |
4 | 13 | 4 | 1 PGY2 or PGY3 senior; 1–2 PGY1 residents and/or sub-Is; and 1–2 MS3s | Once per 2-wk period |
5a | 24 | 4 | 2 PGY3 or PGY4 seniors; 3 PGY1 residents; and 3 MS3s | Once per 2-wk period |
Site . | Residents and/or Class . | Rotation Duration, wk . | Team Structure . | Intervention Implementation . |
---|---|---|---|---|
1 | 29 | 2 | 1 PGY2 senior; 2–3 PGY1 residents and/or sub-I; and 2–3 MS3s | Once a week × 1 or 2 wk |
2 | 40 | 4 | 1 PGY2 or PGY3 senior; 1–2 PGY1 residents and/or sub-Is; and 2–4 MS3s | Once a week × 4 wk |
3 | 27 | 4 | 1–2 PGY2 or PGY3 seniors; 4 PGY1 residents and/or sub-Is; and 2–3 MS3s | Once per 2-wk period |
4 | 13 | 4 | 1 PGY2 or PGY3 senior; 1–2 PGY1 residents and/or sub-Is; and 1–2 MS3s | Once per 2-wk period |
5a | 24 | 4 | 2 PGY3 or PGY4 seniors; 3 PGY1 residents; and 3 MS3s | Once per 2-wk period |
MS3, 3rd year medical student; Sub-I, subintern.
Before the study period, team and/or attending rounded outside the room and only variably had families on round. All other sites had the full team (including attending) in the room on FCRs.
Participants
We recruited senior residents on a PHM rotation from January 2018 to October 2019. Pediatric residents in postgraduate year (PGY) 2 or 3 who provided verbal consent to participate in attendingless rounds were considered eligible to participate in this study. Residents from other training specialties were excluded. Residents were contacted via e-mail or in-person to participate. Residents were given unique identifiers so that responses were deidentified, return of questionnaires was blinded, and faculty members did not see the self-assessments.
Intervention
Senior residents on select PHM rotations led and conducted FCRs without the attending present at least once during their inpatient PHM rotation. At 3 sites, attendingless rounds were already being practiced, although not formally evaluated, and at 2 sites, attendingless rounds were instituted at the beginning of this research project. The frequency of attendingless rounds varied among sites. They occurred once per week over a 2-week or 4-week rotation at some sites or once per 2-week period in a 4-week rotation at other sites. Team composition at each site (medical students, interns, and nurses) remained the same as preintervention. The attending rounded on patients independently, before or after the resident team, on the basis of attending preference. Attending and/or senior resident dyads decided when to discuss the patients and to what degree. Some chose to discuss the patients before rounds; however, the discussion was limited to patients with a high degree of acuity, or medical and/or social complexity, to preserve senior resident autonomy. Before the first attendingless rounds session for each senior resident, the senior resident received instruction regarding the workflow of attendingless rounds in the form of either an information sheet or discussion with the attending. The manner in which the information was conveyed to the resident was at the discretion of each attending. There was no formal faculty development; however, faculty were advised to limit prerounds discussion.
Instrument Development and Validity Evidence
We followed the framework for survey development by Gehlbach et al22 and established validity evidence of our survey, as described by Messick.23 To establish content validity, we began with a literature review, including the Accreditation Council for Graduate Medical Education milestones used for evaluations24 and existing instruments used to assess learner self-efficacy. Synthesizing the current literature, we developed a de novo survey using Bandura’s principles of self-efficacy and scale development.17 Two survey methodology experts critically reviewed the initial survey design. To establish response process validity, we conducted think-aloud cognitive interviews25 with residents who were not study participants, and the results informed edits to the instrument, which was iteratively revised for clarity. The instrument (Supplemental Tables 7 and 8) was an 11-item, self-administered questionnaire that measured resident’s ratings of self-efficacy across 4 domains: diagnosis and/or management; communication; teaching; and team organization and/or management, including organization of rounds, as well as role as supervisor. Responses were based on a 7-point Likert scale ranging from 1 (with significant supervisor input) to 7 (fully autonomous); not applicable was also an option.
In this retrospective, preintervention-postintervention design, senior residents completed both their preintervention and postintervention surveys, reporting their perceived self-efficacy before and after the intervention by the end of the attendingless rounds day. The retrospective, preintervention-postintervention design was used to reduce response-shift bias, in which the frame of reference for abilities changes on the basis of the intervention rendering pretest scores inaccurate.26
Statistical Analysis
Study data were collected and managed by using Research Electronic Database Capture, a secure Web-based electronic data capture software hosted at University of Chicago.27,28 Data from all sites were compiled and analyzed as 1 cohort. Cronbach’s α was used to assess the internal consistency of survey responses. An exploratory factor analysis was performed before developing scores for the survey instrument by using the baseline survey data. After factor analysis, we developed a composite score of self-efficacy that was compared between preintervention and postintervention. On the basis of the baseline distribution of responses of each item, individual item responses were grouped into low autonomy (1–4), moderate autonomy (5), or high autonomy (6–7), and changes in categorized item responses from preintervention to postintervention were also analyzed. Quantitative data were summarized with mean ± SD of the participants’ score.29 Categorical data were expressed with frequencies and percentages. Preintervention and postintervention results were compared by using paired t tests, whereas categorized responses were compared by using the nonparametric Wilcoxon rank test. Changes in the composite scores were compared by using either unpaired t tests or analysis of variance according to the baseline cofactors. Within each group, paired t test was used to assess significant changes. The effect size for the composite score was summarized with either Cohen’s d or mean difference, along with their 95% confidence intervals (CIs). Sensitivity analyses, after restricting to 1-time participants and excluding some participating sites, were performed to assess the validity of findings obtained in the study. P < .05 was considered statistically significant. All analyses were conducted by using Stata version 15 (Stata Corp, College Station, TX).
Results
There were 148 residents who participated in the study across all 5 clinical training sites. Six residents were excluded because of inadvertent data loss. Of those included, 116 completed the self-efficacy scales (82%). There were 47 PGY2s and 69 PGY3s (Fig 1). Seventy-six participants completed the questionnaire after their first experience leading attendingless rounds. Sixty-three residents completed the questionnaire more than once, so there were 181 encounters included in the analysis. The analyses were repeated on residents who responded once (1-time participants, n = 53) after excluding repeat participants.
The reliability of the developed self-efficacy instrument was high from baseline data (Cronbach’s α = 0.89). The exploratory factor analysis including all 11 items revealed that 1 factor (factor 1) had the highest variance (4.83) relative to the rest of the factors and explained the highest proportion of variability (83%) (Supplemental Table 3). This indicates 1 composite score is necessary for explaining the variability in the self-efficacy responses (Supplemental Table 3). The weights assigned to each item in factor 1 were higher than other factors, indicating that 1 factor was sufficient to combine scores from the 11 items (Supplemental Table 4). The weights were similar across items, and thus, for simplicity, we used a simple sum of all item scores (Supplemental Table 4). Thus, an individual may receive a low score of 7 or the highest score of 77 points.
The average composite self-efficacy score was significantly increased (P < .001) after intervention (66.04, SD = 7.66) compared with preintervention (60.78, SD = 6.81). The effect size of intervention on self-efficacy score was found to be large (Cohen’s d = 0.74, mean difference = 5.3; 95% CI: 4.23 to 6.3). All sites, except site 5, showed a significant increase in self-efficacy score postintervention (Table 2). The sensitivity analyses yielded similar findings even after restricting analysis to 1-time participants or excluding data from site 4 (maximum improvement) and 5 (least improvement).
Comparison of Self-Efficacy Between Preintervention and Postintervention
. | Preintervention, Mean (SD) . | Postintervention, Mean (SD) . | Mean Difference (95% CI) . | P . |
---|---|---|---|---|
Entire cohort (N = 181) | 60.78 (6.81) | 66.04 (7.66) | 5.27 (4.23 to 6.3) | <.001 |
Institution | ||||
Site 1 (n = 78) | 59 (7.04) | 64.4 (6.7) | 5.4 (4.17 to 6.62) | <.001 |
Site 2 (n = 55) | 64.38 (6.48) | 70.4 (5.69) | 6.02 (4.2 to 7.84) | <.001 |
Site 3 (n = 13) | 57.77 (4.57) | 64.23 (7.14) | 6.46 (2.2 to 10.72) | .006 |
Site 4 (n = 7) | 59.14 (6.31) | 69.29 (3.35) | 10.14 (3.53 to 16.76) | .01 |
Site 5 (n = 28) | 60.46 (4.99) | 62.11 (10.32) | 1.64 (−2.31 to 5.6) | .4 |
PGY level | ||||
PGY2 (n = 79) | 58.94 (7.02) | 64.33 (6.69) | 5.39 (4.18 to 6.6) | <.001 |
PGY3 (n = 102) | 62.21 (6.31) | 67.37 (8.13) | 5.17 (3.56 to 6.77) | <.001 |
Previous experience with attending less rounds | ||||
No (n = 105) | 62.1 (7.11) | 68.14 (6.58) | 6.04 (4.82 to 7.26) | <.001 |
Yes (n = 76) | 58.95 (5.93) | 63.14 (8.13) | 4.2 (2.38 to 6.02) | <.001 |
Quarter of enrollment | ||||
January to March (n = 43) | 59.05 (5.51) | 65.07 (9.64) | 6.02 (3.06 to 8.98) | <.001 |
April to June (n = 60) | 59.68 (9.09) | 65.19 (7.92) | 5.15 (3.7 to 6.6) | <.001 |
July to September (n = 38) | 63.11 (5.04) | 66.82 (6.03) | 4.58 (2.11 to 7.05) | .001 |
October to December (n = 40) | 60.08 (6.97) | 65.35 (7.14) | 5.28 (3.54 to 7.01) | <.001 |
Sensitivity analysis | ||||
After excluding data from site 4 (n = 174) | 60.84 (6.84) | 65.91 (7.76) | 5.07 (4.02 to 6.12) | <.001 |
After excluding data from site 4 and 5 (n = 146) | 60.92 (7.15) | 66.64 (6.98) | 5.73 (4.73 to 6.72) | <.001 |
. | Preintervention, Mean (SD) . | Postintervention, Mean (SD) . | Mean Difference (95% CI) . | P . |
---|---|---|---|---|
Entire cohort (N = 181) | 60.78 (6.81) | 66.04 (7.66) | 5.27 (4.23 to 6.3) | <.001 |
Institution | ||||
Site 1 (n = 78) | 59 (7.04) | 64.4 (6.7) | 5.4 (4.17 to 6.62) | <.001 |
Site 2 (n = 55) | 64.38 (6.48) | 70.4 (5.69) | 6.02 (4.2 to 7.84) | <.001 |
Site 3 (n = 13) | 57.77 (4.57) | 64.23 (7.14) | 6.46 (2.2 to 10.72) | .006 |
Site 4 (n = 7) | 59.14 (6.31) | 69.29 (3.35) | 10.14 (3.53 to 16.76) | .01 |
Site 5 (n = 28) | 60.46 (4.99) | 62.11 (10.32) | 1.64 (−2.31 to 5.6) | .4 |
PGY level | ||||
PGY2 (n = 79) | 58.94 (7.02) | 64.33 (6.69) | 5.39 (4.18 to 6.6) | <.001 |
PGY3 (n = 102) | 62.21 (6.31) | 67.37 (8.13) | 5.17 (3.56 to 6.77) | <.001 |
Previous experience with attending less rounds | ||||
No (n = 105) | 62.1 (7.11) | 68.14 (6.58) | 6.04 (4.82 to 7.26) | <.001 |
Yes (n = 76) | 58.95 (5.93) | 63.14 (8.13) | 4.2 (2.38 to 6.02) | <.001 |
Quarter of enrollment | ||||
January to March (n = 43) | 59.05 (5.51) | 65.07 (9.64) | 6.02 (3.06 to 8.98) | <.001 |
April to June (n = 60) | 59.68 (9.09) | 65.19 (7.92) | 5.15 (3.7 to 6.6) | <.001 |
July to September (n = 38) | 63.11 (5.04) | 66.82 (6.03) | 4.58 (2.11 to 7.05) | .001 |
October to December (n = 40) | 60.08 (6.97) | 65.35 (7.14) | 5.28 (3.54 to 7.01) | <.001 |
Sensitivity analysis | ||||
After excluding data from site 4 (n = 174) | 60.84 (6.84) | 65.91 (7.76) | 5.07 (4.02 to 6.12) | <.001 |
After excluding data from site 4 and 5 (n = 146) | 60.92 (7.15) | 66.64 (6.98) | 5.73 (4.73 to 6.72) | <.001 |
Between-group comparison.
The comparison of individual items of self-efficacy between preintervention and postintervention is shown in Fig 2. The number of respondents who reported scores in the high-autonomy category increased postintervention for each item (range: 66.3%–86.7%), with addressing parent and/or patient concerns (86.7%) and prioritizing tasks and triaging patient needs (85.6%) increasing most after the intervention. The range of high level of autonomy was 40.9% to 70.7% preintervention, with the highest in prioritizing tasks and triaging patient needs (70.7%) and addressing parent and/or patient concerns (68.5%). Preintervention, ∼45% of items had <50% of participants reporting a high level of autonomy, whereas postintervention >90% of items had >70% reporting of high level of autonomy. All individual items had a significant increase in the proportion of high level of autonomy after intervention compared with preintervention. The largest improvement in the self-efficacy items was observed in directing bedside teaching (71.8% vs 42.5%; P < .001), answering learner questions on rounds (70.7% vs 47.0%; P < .001), and giving a treatment plan (75.1% vs 43.6%; P < .001), with smallest improvements seen in prioritizing tasks and triaging patient needs (85.6% vs 70.7%; P < .001) and addressing social concerns (75.1% vs 60.2%; P < .001). The results of 1-time participants revealed similar findings, except for addressing social concerns and prioritizing tasks and triaging patient needs (Supplemental Table 5).
Comparison of respondents per autonomy category preintervention and postintervention. Low autonomy (self-efficacy scores 1 to 4), moderate autonomy (self-efficacy score 5), and high autonomy (self-efficacy scores 6 to 7).
Comparison of respondents per autonomy category preintervention and postintervention. Low autonomy (self-efficacy scores 1 to 4), moderate autonomy (self-efficacy score 5), and high autonomy (self-efficacy scores 6 to 7).
There was no difference in the overall changes in efficacy scores postintervention between PGY2 and PGY3 residents (5.39 vs 5.17; P = .83). The increase from preintervention to postintervention scores did not significantly vary by the time of the academic year that the resident participated (P = .84), although a slightly lower improvement in scores was observed in the July to September quarter. Although the group with no previous experience with attendingless rounds had a slightly greater increase in the self-efficacy score (6.04) compared to the group with previous experience (4.2), the difference was not statistically significant (P = .085). The results were further validated after removing repeat participants (Supplemental Table 6).
Discussion
Becoming autonomous during residency training is essential to independent practice, and measurement of self-efficacy is one way to assess a resident’s attainment of this skill. The results of our study reveal that conducting FCRs without an attending physician present can increase resident reported self-efficacy across multiple domains: diagnosis and/or management, communication, teaching, and team organization and/or management. These findings were consistent across year of training, previous exposure to FCRs, time of year, and most hospital sites. Our theory-guided instrument for measuring self-efficacy of residents was found highly reliable and simple to score, with 1 composite score only.
We demonstrated similar results across geographic locations, regardless of size of program or year of training. It is important to note that both PGY2 and PGY3 residents had significant improvement in self-efficacy scores, which suggests that implementation of attendingless rounds could start in the PGY2 year. Increased self-efficacy was noted despite repeating the experience, suggesting there is continued benefit for residents to lead attendingless rounds more than once. Site 5 did not reveal statistically significant results. This might be explained by the fact that before the study period, the team rounded outside the room and only variably had families on rounds. Having the attending not round in the room with the family during the intervention was not a significant departure from previous experience.
In our study, senior residents reported higher preintervention self-efficacy in team organization, including prioritizing tasks, compared with other domains, which reveals an appropriate skill transition from intern to senior. However, after the intervention, we observed an increase in senior residents’ self-reported practicing of bedside teaching, answering learner questions on rounds, and giving a treatment plan. These tasks may be more difficult for a resident to do with an attending present because of the implicit deference to the more senior practitioner in the room. Similarly, Finn et al30 found that residents reported less comfort making independent decisions with an attending present. Answering learner questions and directing bedside teaching may be difficult when the attending is present because the attending tends to be seen as the team leader by medical students and interns and even the family. Researchers have examined strategies used by faculty for promoting resident autonomy during FCRs; however, these often depend on the attending’s behaviors to allow for resident autonomy.31,32
According to SDT, strategies that help residents practice autonomy are likely to foster competence.13 With this theory, it is argued that when autonomy, competence, and relatedness are met, people will function and grow optimally; therefore, the environment should nurture these needs. Crockett et al31 found that when residents were asked to describe what autonomy looks like to them, most of residents’ responses focused on the ability to conduct daily rounds without attending presence. Biondi et al13 showed that attending decisions regarding when to allow a resident increased autonomy were based on the individual attending’s subjective perception of residents’ baseline skills. SDT, however, suggests that residents need practice to acquire skills,13 so rather than solely relying on attending assessment of readiness to be autonomous, a standardized intervention in which all residents are afforded the same independence may be a complementary strategy to promote self-efficacy and autonomy.32 Moriarty et al33 assessed independent rounds and found that trainees described increased motivation to take ownership of their patients and team when rounding without an attending present. We believe that rounding without an attending present provides an opportunity for senior resident self-reflection and an understanding of their progression on the continuum of graded, progressive autonomy because rounds set the stage for a resident’s sense of autonomy throughout the day.
Studies have revealed conflicting findings regarding the perceived educational value of having an attending physician present for rounds. Ways et al34 surveyed internal medicine residents, who reported attending physician presence at work rounds was not useful. Gonzalo et al35 found that internal medicine residents performing bedside rounds with an attending were less likely to believe bedside rounds were educational. Mittal et al36 found that well-conducted FCRs can improve resident education, but attending teaching variability is a barrier to quality. Given the mixed evidence for effectiveness of an attending to improve education on rounds, it may be worthwhile to consider other skills and educational benefits that can be achieved without the attending present, as our study suggests.
This study has some limitations. We used self-evaluation, which can be biased and may not correlate to actual performance; however, we attempted to mitigate response-shift bias through a retrospective preintervention-postintervention design.26 Although observation of residents could have provided more information, it would have required staffing and funding that was not available. It should be considered in future studies. Although we did not assess the effect of attendingless rounds on other learners, in other studies, researchers suggest that interns may benefit from near-peer collaboration achieved by not having the attending present on rounds37 or may even be negatively affected by attending presence.30 We also did not directly assess whether attendingless rounds affect patient care, length of rounds, or errors. However, Finn et al30 did not show an increased error rate when attending physicians were not present on rounds. Another potential limitation is that, although statistically significant, the consequence of the increase in preintervention-postintervention scores’ difference is unknown because the authors developed the survey de novo and do not have established benchmarks. In the absence of a control group, it is unknown the extent to which residents would have experienced increased self-efficacy with an attending present. Residents who experience attendingless rounds more or less often may have a different experience than residents who experience it once per week. With this frequency of data collection (1–2 times per rotation), we may not have captured nuanced differences in the experiences. An additional potential limitation is that attendingless rounds were not implemented uniformly across all sites; some sites did the intervention weekly, whereas others did it every other week. Although this was intentional to reduce site-specific barriers for implementation of this intervention, the variability may constrain our ability to interpret the impact of the intervention or to determine which elements were important to its success.
Finally, we did not explore the ideal frequency of residents’ rounding without an attending present; further study is needed in this regard. Intermittent absence may be the ideal frequency because it allows for benefits of attending role modeling36 and opportunities for direct observation and feedback. Additional next steps include examining the effect on interns, medical students, and other stakeholders when rounding without the attending, as well as replication of the study on subspecialty teams to determine if there is similar increase in self-efficacy in different settings.
Conclusions
Conducting FCRs without an attending increases resident reported self-efficacy regarding core elements of patient care and team leadership. Rounding without the attending may provide a good opportunity for senior resident self-reflection and increased self-efficacy of behaviors for independent practice. Further study is needed to determine effects on patients and families and on health outcomes.
Acknowledgments
We thank the division of hospital medicine at Children’s Hospital at Montefiore and chief residents at Children’s Hospital at Montefiore.
FUNDING: No NIH funding or funding from other sources.
Dr Jain conceptualized and designed the study, collected data, conducted the initial analyses, drafted the initial manuscript, and reviewed and revised the manuscript; Dr Hametz conceptualized and designed the study and reviewed and revised the manuscript; Dr Banker collected data, conducted the initial analyses, and reviewed and revised the manuscript; Drs Escalante and Gutierrez collected data and reviewed and revised the manuscript; Mr Dodoo and Dr Dwivedi conducted the data analyses and reviewed and revised the manuscript; Dr Beck designed the study and reviewed and revised the manuscript; Dr Fromme designed the study, collected data, and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
References
Competing Interests
POTENTIAL CONFLICTS OF INTEREST: The authors indicate they have no potential conflicts of interest disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
Comments