OBJECTIVES: Establishing a high-value care (HVC) culture within an institution requires a multidisciplinary commitment and participation. Bedside rounds provide an ideal environment for role modeling and learning behaviors that promote an HVC culture. However, little is understood regarding the types of HVC discussions that take place at the bedside and who participates in those discussions.
METHODS: A prospective observational study at a tertiary-care, university-affiliated, free-standing children’s hospital. The prevalence of HVC discussions was captured by using the HVC Rounding Tool, a previously developed instrument with established validity evidence. For each observed HVC discussion, raters recorded who initiated the discussion and a description of the topic.
RESULTS: Raters observed 660 patient encounters over 59 separate dates. Of all patient encounters, 29% (191 of 660; 95% confidence interval: 26%–33%) included at least 1 observed HVC discussion. The attending physician or fellow initiated 41% of all HVC discussions, followed by residents or medical students (31%), families (12%), and nurses (7%).
CONCLUSIONS: Despite a recent focus on improving health care value and educating trainees in the practice of HVC, our study demonstrated that bedside discussions of HVC are occurring with a limited frequency at our institution and that attending physicians initiate the majority of discussions. The capacity of the nonphysician team members to contribute to establishing and sustaining an HVC culture may be underused. Multi-institutional studies are necessary to determine if this is a national trend and whether discussions have an impact on patient outcomes and hospital costs.
Although health care spending in the United States far outstrips that of other industrialized nations, Americans generally do not enjoy better health.1,2 Unnecessary tests and treatments contribute significantly to wasted expenditure.3,4 Accordingly, proponents, including consumer groups and specialty societies, have called on physicians and hospitals to respond by emphasizing high-value care (HVC) principles in medical decision-making.5,6 Following Kleinert’s7 description, we define HVC as care that is focused on patient outcomes by using evidence-based medicine to deliver individualized care that is economically responsible.
Establishing an HVC culture within an institution, with a sustained effect on practice and behavior, requires multidisciplinary commitment and participation. Although physicians play a critical role in the rise of health care use,8 implementing and sustaining an HVC culture also requires multidisciplinary collaboration and education with role modeling of HVC practices.9,10 Pharmacists,11,12 nurses and nurse practitioners,13 trainees,14–17 and patients18,19 may all influence and model HVC practices.
Bedside rounds and discussion represent a crucial environment for learning clinical and communication skills that promote HVC practices20 and provide an opportunity to role model the culture of patient centered care.21 Despite this, little is understood regarding the types of HVC discussions that take place at the bedside and, importantly, who participates in those discussions. Given that multidisciplinary collaboration and education are crucial to driving organizational change, understanding these discussions is critical to establishing this culture within an institution.
We previously developed a tool to capture HVC topics discussed during bedside rounds.22 For this study, our primary objective was to use this tool to quantify the prevalence of HVC discussions during bedside rounds. On the basis of trainee perceptions that HVC principles are rarely role modeled,23,24 we hypothesized that HVC discussions would be infrequently observed during bedside rounds. Our secondary objectives were to identify who initiated the HVC discussion and describe the specific topics discussed. Historically, the attending physician directed patient care discussion and educational conversation during bedside rounds. As such, we hypothesized that most conversations around HVC would be initiated by the attending physician and that other members of the multidisciplinary team would not initiate a significant number of HVC discussions. Characterizing practice variation could inform future efforts to incorporate more HVC discussions during rounds.
From August 2016 to December 2016, we conducted a cross-sectional observational study of multidisciplinary bedside rounds at Seattle Children’s Hospital (SCH), a tertiary-care, university-affiliated, free-standing children’s hospital. Since 2010, SCH has developed and implemented >60 clinical standard work pathways for a range of conditions.25 Clinical standard work uses a standardized approach to develop evidence-based, cost-effective clinical pathways. This approach is integrated into the workflow via tools such as electronic order sets, nursing documentation, respiratory therapy protocols, and treatment algorithms. Providers incorporate the pathways into clinical management, although they retain the ability for a nuanced approach in the context of individual patient care.
We observed 4 combined general pediatric medical and subspecialty teams during bedside rounds, excluding the pediatric units and NICUs. Each rounding team included a general medicine attending physician as well as 1 to 2 subspecialty attending physicians who each conducted rounds on their respective patients. The teams also included a senior resident (postgraduate year [PGY] 2 or PGY 3), at least 1 intern (PGY-1), one third-year medical student, and/or a fourth-year medical student. For the subspecialty services, a fellow would often join rounds and serve as the supervising physician. Rounds typically included bedside nurses, pharmacists, registered dieticians, nurse case managers, and an interpreter, if necessary.
Family-centered rounds (FCR) are defined as interdisciplinary work rounds at the bedside in which the patient and family share control of the management plan as well as the evaluation of the process itself.26 FCR are the standard model for inpatient general medicine and subspecialty rounds at SCH except on weekends. Accordingly, we observed rounds on weekdays to capture the maximum multidisciplinary interactions. Observations were conducted by 6 attending hospitalists. At the start of FCR, observers described the study to the patient, family, and all team members as an observational study of “bedside teaching during rounds” but did not reveal the focus on HVC topics to minimize the likelihood that participants would modify their typical behaviors.27
All aspects of this study were approved by the Institutional Review Board at SCH. The study was exempt, and participation was voluntary.
HVC Rounding Tool
As members of a multidisciplinary research team, we previously used a modified Delphi approach to develop the HVC Rounding Tool, which is an instrument to measure the frequency and content of HVC discussions (Table 1). The development, piloting, and interrater reliability of the tool are described in McDaniel et al.22 Nineteen national HVC experts representing a spectrum of clinical experience years, regions of the country, and subspecialties agreed to participate in the modified Delphi process. A total of 10 topics were ultimately chosen by the panel and then classified into 3 domains (Quality, Cost, and Patient Values) representing critical areas for HVC role modeling and bedside discussion. Instrument piloting ultimately demonstrated weighted kappas for each of the 3 domains ranging from 0.96 to 1.0 and percent positive agreement measures ranging from 95.7% to 100%.
Observers used the HVC Rounding Tool to capture HVC discussions for each individual patient encounter. Raters recorded observed topics dichotomously as “discussed” or “not discussed.” For each HVC topic discussion, the user recorded who initiated the discussion and a description of the topic. On the basis of published descriptions of the factors that influence bedside discussions,28 we collected the following information for each patient encounter: (1) Was the patient a new admission (admitted within the previous 24 hours) or an established patient? (2) Was a parent, guardian, and/or caretaker present during rounds? (3) Was an interpreter used? (4) Did rounds occur in the patient room (versus the hallway)? We also recorded the entire duration of rounds in minutes and the total number of patients observed, allowing for the calculation of average time spent per patient.
Observation dates were chosen on the basis of a convenience sample of the raters’ schedules. The 6 raters were part of the development of the HVC Rounding Tool and had previously participated in a 3-part rater training and instrument piloting.22 Some dates included observations by multiple raters, although each rater observed different teams. Individual raters were assigned to 1 of the 4 teams for observation by using a simple block randomization allocation stored in REDCap (research electronic data capture).
Data were deidentified before analysis. We calculated the frequencies and proportions of patient encounters with at least 1 HVC discussion and 95% confidence intervals (CIs). Additionally, we analyzed the prevalence of HVC topics by domain, rounding team specialty, attending physician academic rank, and encounter characteristics (new versus established patient, parent and/or guardian present, interpreter used, and rounds in patient room), reporting the frequency and percent of recorded HVC discussions on the basis of who initiated the discussion. We calculated descriptive summaries for observation time, including the median, range, and interquartile range (IQR) of individual rounding sessions and the length of time per patient (total session time divided by the number of patients). Given that this was a descriptive study at a single institution to determine initial prevalence estimates, we did not a priori conduct a power analysis.
Data were collected on 59 separate dates between August 2016 and December 2016. We requested permission to observe rounds of 72 attending physicians. Sixteen attending physicians requested to be observed on a different day with no refusals for participation. A total of 660 patient encounters were observed during 87 separate rounds. Fifteen attending physicians were observed >1 time. Total observation time by raters for the duration of rounds was recorded for 69 rounds, including 574 patient encounters (87%). The median duration of a single rounding session was 103 minutes (range: 16–171 minutes; IQR: 75–130 minutes). Time per individual patient was not recorded, but the calculated average time per patient (total rounding time divided by number of patients) had a median of 12 minutes (range: 8–24 minutes; IQR: 10–14 minutes). The median number of patients per rounding session observed by raters was 8 (range: 1–15 patients; IQR: 5–10 patients).
Overall, 29% of all patient encounters (191 of 660; 95% CI: 26%–33%) included at least 1 observed HVC discussion from the 10 potential topics. A single topic was discussed in 161 encounters; multiple topics were discussed in 30 encounters. In total, we observed 242 HVC topics, and 16% (30 of 191) of the encounters had >1 HVC domain discussed during a single patient encounter. The frequency of HVC discussions was similar across the 3 HVC domains: 11% of encounters (95% CI: 8%–13%) included at least 1 Quality HVC discussion, 12% of encounters included at least 1 Cost HVC discussion (95% CI: 10%–15%), and 11% included at least 1 Patient Values HVC discussion (95% CI: 9%–13%). We found no statistically significant difference in the prevalence of discussions by month of the year.
The frequencies of individual HVC topics are shown in Table 1. The most frequently observed topics were “avoid or cancel a low-value test (daily complete blood count; erythrocyte sedimentation rate and C-reactive protein) or therapy or monitoring (pulse oximeter)” and “customize care plan to incorporate patient and/or family values and/or goals.” These were recorded in 8% (95% CI: 6%–10%) and 7% (95% CI: 5%–9%) of all patient encounters, respectively.
Rounds were observed for 7 pediatric medical specialties: general medicine, endocrinology, neurology, rheumatology, gastroenterology, nephrology, and craniofacial. General medicine accounted for 61% (403 of 660) of patient encounters; the number of patient encounters in other specialties varied from a low of 2% (13 of 660, rheumatology) to a high of 13% (86 of 660, gastroenterology). The percentage of encounters with any HVC discussion by specialty is shown in Fig 1; small sample sizes are reflected in the wide CIs and limit any interpretation of differences across specialties. We found no statistically significant difference when comparing the percent of encounters with any HVC topic between general medicine (30% [95% CI: 25%–34%]) and all subspecialties combined (29% [95% CI: 23%–35%]).
The majority of patient encounters involved established patients (74%). Parents or guardians were present in 63% of encounters. Only 5% of encounters included an interpreter, and teams conducted rounds in a patient’s room for 52% of encounters. We observed 72 unique attending physicians, 13% with the rank of instructor of pediatrics, 46% with the rank of assistant professor, 26% with the rank of associate professor, and 15% with the rank of full professor. The location of rounds, use of an interpreter, or presence of a parent or guardian were not associated with HVC discussions (Fig 2). We observed a slightly higher percentage of HVC topics for established (31%) versus new patients (23%), although this was not statistically significant.
The attending physician or fellow initiated 41% of all HVC discussions followed by residents or medical students (31%) (Table 2). Whereas families initiated 12% of HVC topics, only 7% of HVC topics were raised by nurses and 6% by pharmacists. The majority of nurse-initiated discussions focused on topics in the Cost domain, such as the discontinuation of isolation precautions or unnecessary monitors. Additionally, most pharmacist comments focused on topics in the Quality domain. Discussions by the attending and/or fellow and medical student and/or resident were distributed throughout all 3 domains.
We present the first comprehensive description of HVC discussions held during multidisciplinary rounds. Despite recent national efforts to raise awareness of waste in health care and to educate trainees about HVC, the bedside discussion of HVC topics occurred in less than one-third of observed encounters. No individual HVC topic arose in >8% of conversations and only 29% of multidisciplinary bedside patient rounds had any form of value discussions. Attending physicians initiated the majority of HVC discussions.
Our results build on previous studies highlighting not only the importance of role modeling HVC decision-making and behaviors but that HVC discussions are happening infrequently. For example, conversations around test-ordering principles during bedside rounds occurred in only 20% of internal medicine bedside rounds,23 whereas we observed ∼14% during pediatric bedside rounds. Additionally, Patel et al24 reported that residents felt that cost-conscious care was role modeled only 23% of the time. Furthermore, they found no significant difference in the training programs that have a formal HVC curriculum and those that do not, underscoring that HVC behaviors are learned indirectly and informally through other aspects of medical education and are not necessarily learned in a traditional classroom setting. Although we do not know the optimal prevalence of HVC discussions at the bedside to impact trainee behavior, there is a pressing need to equip faculty with time-efficient HVC teaching strategies for role modeling at the bedside.
Many needs compete for time during rounds, including educational priorities, clinical responsibilities, and administrative duties, potentially contributing to the low prevalence of HVC discussions.29–31 Despite these challenges, learning the complex principles of HVC may be most effective when accomplished in a clinical setting,32 and bedside rounds remain a central platform through which trainees are exposed to these concepts.9,33–35 With less than one-third of patient encounters including any form of value discussion at the bedside, it remains unclear whether this is enough role modeling of HVC practices to impact trainee behaviors.
Our work offers insight into the role of nonphysicians in the discussion of HVC topics. Most studies to date focus solely on the perspectives of attending physicians and trainees, although many other stakeholders participate in bedside rounding. Increasingly, the HVC conversation has revolved around promoting shared decision-making by involving patients and families in the discussion of value-based decisions.36,37 However, we found that families initiated only 12% of HVC discussions. Despite the call to empower patients to be more active participants in their care and to make patient centeredness a core aim of our health care system,38 our results suggest that we have room to improve. A previous study reported that there is not a clearly defined role for families and parents during bedside rounds.39 Physicians may need to better introduce the concept of bedside rounds and use language that elicits patient and family goals and preferences on a routine basis.
Nurses and pharmacists also infrequently initiated HVC discussions. Nursing contributions focused primarily on the cost domain because they were often best equipped to identify areas of waste related to direct patient care, such as discontinuing unnecessary continuous pulse oximetry. Pharmacists initiated HVC discussions mostly in the quality domain, often times suggesting the discontinuation of medications given their unique insight into potential side effects. Possible reasons for the limited participation of nurses and pharmacists include the perceived hierarchy within medical teams40 and the lack of clearly defined expectations of participation for nurses and pharmacists. We believe this represents an opportunity to change rounding roles and systems to facilitate sharing value perspectives with all team members, particularly nurses and pharmacists. Blackstone et al41 demonstrated that nurse participation in multidisciplinary, checklist-guided discussions in an ICU, including topics such as discontinuing standing orders or daily studies, lowered hospital charges. Outside of the ICU, organizing nonphysician ownership of targeted HVC practice through checklists could promote opportunities for collaborative HVC.
Lastly, we did not find variation in the prevalence of HVC topics related to the presence of a patient’s family or the rounding location. We had hypothesized that some topics (such as discussing the cost of a test) may be less frequent because of provider discomfort or lack of knowledge of the cost of diagnostic studies,42,43 yet we did not find the presence of patients or families to be associated with the HVC topics discussed. It is possible that our study was not powered to detect this difference given the limited numbers of HVC discussions that we recorded.
Our study has several limitations. First, it was performed at a single, academic, tertiary-care, children’s hospital, limiting generalizability to adult and community institutions. However, our average rounding times and daily team census closely mirror those reported in other studies on bedside rounds.44,45 Second, our institution has a strong culture of standardized disease management using evidence-based, cost-conscious pathways to guide management. This environment may have impacted the type of discussions and management decisions discussed by teams. However, we cannot determine if pathway-based patients have more HVC discussions on rounds (given the hospital focus on clinical pathways) or fewer HVC discussions (given the standardized approach to care). Although this limits generalizability, we did not restrict observations to patients who were on or off a clinical pathway. Third, we did not record time per individual patient encounter and are unable to report if patients with HVC discussions had longer rounding times. We also did not collect data on patient complexity, ICU use, or length of stay that might characterize HVC opportunities at a patient level. Fourth, the raters did not track total speaking time per participant, limiting our ability to determine if variation among professions reflects their contributions to rounds in general or specifically around HVC topics. Because we did not capture the frequency of other, non-HVC topics discussed during rounds, we do not know how the inclusion of HVC discussions affects the prevalences of other topics discussed during rounds. Lastly, value discussions happen throughout the day. Although we did not capture the discussions that were had during other times of the day, our specific study aim was to look at those that were observed during bedside rounds. Additionally, we observed FCR on any given patient only once, and given the competing demands of rounds (patient needs and learner needs), it is possible that we underrepresented the prevalence of HVC discussions that may have taken place on other dates. However, the HVC Rounding Tool captures discussions held with patients and families regarding their personal values and preferences and whether those are incorporated into the decision-making for a patient. As such, we believe an integrated approach with the incorporation of the family’s goals of care and values should be more prevalent and explicit in most, if not all, patient interactions in which testing and/or treatment decisions are made.
Future studies are needed to confirm these findings in a multi-institution sample and identify potential barriers to and facilitators of effective HVC discussions during bedside rounds. As providing HVC increasingly becomes expected as the way to practice medicine, the call to action of a multidisciplinary approach becomes increasingly fundamental. Essential to this, the capacity of the nonphysician team members to sustain HVC culture remains underused, and empowering discussion from all participants in bedside rounds surrounding HVC is critical to increasing the prevalence of these topics at the bedside. This multidisciplinary, bedside approach to role modeling may be a key to affecting long-term HVC behaviors. It is also likely that faculty development programs that are designed to improve attending role modeling of HVC behaviors, education on leading HVC discussions involving multiple stakeholders, and engaging nonphysician participants may lead to HVC culture change.
Our study demonstrated that bedside discussions of HVC are occurring during a minority of rounds at our institution. Multi-institutional studies are necessary to determine if this is a representative finding. We plan to use these results to inform faculty development efforts, empower nonphysician team members and families to engage in HVC discussions at our institution, and measure whether these discussions have an impact on patient outcomes and hospital costs.
We thank Jack Percelay, MD, Alan Schroeder, MD, and Jonathan Ilgen, MD, MCR, for their help in editing the article.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Provided through a Seattle Children’s Hospital Academic Enrichment Fund grant (Lawson Activity Number 24080047, Academic Enrichment Fund, Beck 2015) and supported by the National Center for Advancing Translational Sciences of the National Institutes of Health (award UL1 TR000423). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Woolf S,
- Aron L
- Yong PL,
- Saunders RS,
- Olsen LA
- Davis K,
- Stremikis K,
- Squires D,
- Schoen C
- Levinson W,
- Kallewaard M,
- Bhatia RS,
- Wolfson D,
- Shortt S,
- Kerr EA
- Kleinert S,
- Horton R
- Gupta R,
- Moriates C,
- Harrison JD,
- et al
- Stinnett-Donnelly JM,
- Stevens PG,
- Hood VL
- Ashok A,
- Combs B
- Gleason M
- ↵Society for Participatory Medicine. Participatory medicine resources for professionals. Available at: https://participatorymedicine.org/what-is-participatory-medicine/participatory-medicine-resources-for-professionals/. Accessed April 11, 2017
- Gonzalo JD,
- Heist BS,
- Duffy BL,
- et al
- McDaniel CE,
- White AA,
- Bradford MC,
- et al
- Pierce C,
- Keniston A,
- Stickrath C
- ↵Seattle Children’s Hospital. Clinical standard work pathways and tools. 2017. Available at: www.seattlechildrens.org/healthcare-professionals/gateway/pathways/. Accessed March 1, 2017
- Paradis E,
- Sutkin G
- Satterfield JM,
- Bereknyei S,
- Hilton JF,
- et al
- Stickrath C,
- Noble M,
- Prochazka A,
- et al
- Korenstein D,
- Smith CD
- Sedrak MS,
- Patel MS,
- Ziemba JB,
- et al
- Baicker K,
- Rosenthal M
- ↵Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press; 2001
- Rock TA,
- Xiao R,
- Fieldston E
- Copyright © 2018 by the American Academy of Pediatrics