Article Text
PDF
Abstract
Background Despite recommendations and the need to accelerate redesign of delivery models to be team-based and patient-centred, professional silos and cultural and structural barriers that inhibit working together and communicating effectively still predominate in the hospital setting. Aiming to improve team-based rounding, we developed, implemented and evaluated the Interprofessional Teamwork Innovation Model (ITIM).
Methods This quality improvement (QI) study was conducted at an academic medical centre. We followed the system’s QI framework, FOCUS-PDSA, with Lean as guiding principles. Primary outcomes included 30-day all-cause same-hospital readmissions and 30-day emergency department (ED) visits. The intervention group consisted of patients receiving care on two hospitalist ITIM teams, and patients receiving care from other hospitalist teams were matched with a control group. Outcomes were assessed using difference-in-difference analysis.
Results Team members reported enhanced communication and overall time savings. In multivariate modelling, patients discharged from hospitalist teams using the ITIM approach were associated with reduced 30-day same-hospital readmissions with an estimated point OR of 0.56 (95% CI 0.34 to 0.92), but there was no impact on 30-day same-hospital ED visits. Difference-in-difference analysis showed that ITIM was not associated with changes in average total direct costs nor average cost per patient day, after adjusting for all other covariates in the models, despite the addition of staff resources in the ITIM model.
Conclusion The ITIM approach facilitates a collaborative environment in which patients and their family caregivers, physicians, nurses, pharmacists, case managers and others work and share in the process of care.
- communication
- continuous quality improvement
- patient-centred Care
- teamwork
- hospital medicine
Statistics from Altmetric.com
Introduction
Hospital settings present important challenges to teamwork. Medical teams can be large and diverse, and team members are seldom in the same place at the same time, resulting in suboptimal communication practices.1 2 Communication problems range from misunderstanding messages, problems interpreting clinical information due to channel issues (eg, not correctly hearing phone messages, jumbled text messages), problems coordinating different disciplines, and omissions or lack of crucial information.3–5 Moreover, professionals from different departments traditionally have acted independently, in isolation, without consistent efforts to communicate their efforts to other members of the team; for example, a physical therapist might evaluate a hospitalised patient and document the findings and recommendations, but not communicate to the patient’s physician. Multiple studies in settings ranging from operating rooms, intensive care units (ICUs) and hospital medical floors show that nurses and physicians widely differ in their views on the quality of collaboration and communication.2 6 7 Despite recommendations and the need to accelerate redesign of delivery models to be team-based and patient-centred, professional silos and cultural and structural barriers that inhibit working together and communicating effectively still predominate in the hospital setting.
In a climate constantly demanding efficiency and quality improvement (QI), the triple aim8 directs hospitals and health systems across the country to develop and test ways to improve care quality and patient outcomes while lowering or maintaining costs. Importantly, research documents that clinician and staff burnout are associated with overuse of resources, prescription errors, poor clinical outcomes, lowered patient satisfaction and higher costs.9 Consequently, many organisations and agencies began endorsing a shift from the triple aim to a quadruple one—adding clinician and staff satisfaction, to achieve the overarching goal of higher quality care with optimal efficiency.9–12 Teamwork and communication are essential elements of achieving the quadruple aim through safe and effective healthcare systems.3 4 The Institute of Medicine suggested that healthcare professionals can best communicate and address patients’ complex and challenging needs by working in interprofessional teams.3 Interprofessional collaborative care is the process through which providers of different professions work as a team to promote improved communication, coordination of care and patient-centred, shared decision-making.13–16 The ‘quality-of-relationships’ between and among healthcare providers, patients and families is the key principle of interprofessional collaboration, yet this is frequently disregarded.17 Patient-centred care is interpreted differently by institutions, teams and even patients, but research documents that patients and their family should be included as part of the interprofessional team to add another layer of safety.18–21 Additionally, pharmacists are integral members given the frequency of medication errors. Bedside rounding has been a historical clinical model that brings together care providers and the patient to discuss the plan of care (POC), treatment adjustments and discharge planning goals. Interprofessional bedside rounding has been suggested as a primary method of promoting collaboration and patient-centred care in hospital-based settings.22 23
In a 2016 systematic review that looked at the structure and outcomes of interdisciplinary rounds, the authors found wide variation in design and team composition.24 Although data supported an association with length of stay (LOS) and staff satisfaction, little research showed any change in patient outcomes or satisfaction. One cluster randomised controlled trial (RCT) of interprofesional bedside rounds indicated no impact on patients’ perceptions of shared decision-making, activation or satisfaction with care.25 More recent studies of teamwork on medical units and team-based rounding reveal discrepant findings. One pragmatic cluster RCT demonstrated that standardised attending rounds yielded higher patient satisfaction, but resident trainees reported lower satisfaction.26 Another study of bedside interdisciplinary rounds at an academic medical centre showed no reduction in LOS nor frequency of clinical deterioration, compared with a control group of typical physician team rounding.27
There are several models for interprofessional rounds; however, these models either include only physicians and nurses as routine rounding members, for example, the accountable care unit approach by Stein et al,28 or occur away from the patient bedside, such as the structured interdisciplinary rounds described by O’Leary et al.29 Aiming to build on past research and evaluate a novel version of team-based rounding, our health system developed and implemented the Interprofessional Teamwork Innovation Model (ITIM) to undertake bedside clinically focused rounds that included the patient and/or family caregiver, pharmacist, and case manager or social worker, in addition to a bedside nurse and hospitalist. The need for ITIM arose from efforts to implement Project BOOST (Better Outcomes by Optimizing Safe Transitions).30 31 In this study, we sought to (1) incorporate identification and description of the POC into interprofessional bedside rounds using QI methods and tools, (2) examine patient outcomes and experience, and (3) assess efficiency and cost savings.
Methods
Study setting
This study was conducted at a 302-bed, community-based, acute-care hospital—Good Samaritan (GS), one of two hospitals in an academic health system. With the opening of a renovated 30-bed medical unit on the seventh floor (GS7) staffed by hospitalists from the Division of Hospital Medicine, the system designated it as an innovation unit to trial novel approaches to care delivery design. In preparation, the nursing leadership and staff of the unit, hospitalists and pharmacists were informed about this designation and the plan for rapid Plan-Do-Study-Act (PDSA) cycles of change.
In December 2014, we kicked off implementation of Project BOOST.32 Many components of the BOOST toolkit (eg, patient needs assessment, medication reconciliation, postdischarge placement planning) are essential aspects of hospital care delivery. Our team conducted a current status assessment for local adaptation, and identified medication management and reconciliation, postdischarge placement and needs, and addressing social needs as major contributing factors to readmission at our institute. ITIM was developed as an approach to interdisciplinary bedside rounding that incorporated aspects of Project BOOST. Given that geographical cohorting is integral to success, we cohorted two hospitalist-led teams, averaging 12–14 patients, on the new GS7 medical unit in preparation for ITIM implementation. Each hospitalist was joined by a bedside nurse, a case manager and a pharmacist to form an ITIM team collaborating on care delivery and making medical evaluation and management decisions together with the patient, and their family caregivers if available. Two designated professionals beyond existing staff on the floor, a pharmacist and case manager, were added resources for the ITIM model. Starting April 2015, a QI specialist from the Office for Value & Innovation in Healthcare Delivery and a data analyst began working with these teams to develop and implement ITIM bedside rounds.
Study design
The study was developed and implemented as a QI project with statistical process control (SPC) charts for performance monitoring. We followed our systematic QI framework, FOCUS-PDSA (F ind a process to improve, O rganize a team to improve the process, C larify current knowledge of the process, U nderstand sources or causes of variation, S elect the improvement or intervention – P lan, D o, S tudy, A ct),33 34 with Lean as guiding principles (online supplementary appendix 1). Because GS7 is the sole designated innovation test unit, we were not able to identify a comparable control unit for an RCT. Thus, we planned to evaluate the impact of ITIM on patient outcomes and efficiency using a matched-cohorts difference-in-difference analysis.
Supplementary material 1
Overall, BOOST/ITIM implementation received support from physician, nursing and pharmacy leaders. The champion team included a nurse unit manager, a hospitalist attending, a pharmacist and a case management director. Importantly, everyone on the floor was informed that failures were to be expected, that is, a ‘normal’ part of QI, but would trigger adjustment and revised attempts. One essential process in our PDSA cycles was listening to front-line staff and patients for intervention development and refinement. Weekly successes and opportunities were solicited from clinicians and staff through an anonymous REDCap survey link, and the perception of patients and families was sought through informal enquiries by bedside nurses. The implementation team met every week to report on pilot status, examine metrics and progress, review feedback from staff and patients, share observations on problems, discuss strategies and advice on changes that could be made to aid progress, and revise the rounding structure as needed, for example, adjusting rounding time for optimal family involvement. The team piloted different rounding structures, team member roles, start and finish time goals, and POC documentation, before arriving at the current ITIM model in August 2015.
ITIM intervention
ITIM rounds for each hospitalist were scheduled for a consistent start time each day, to promote participation by all attendees and ensure that the nurse case manager would not have more than one physician rounding at the same time (table 1). The daily rounds follow an organised, structured approach, with each team member responsible for explicit aspects of care (box 1). All members of the collaborative team inclusive of the patient, and family caregivers if present during the rounds, then discuss and agree on the POC and discharge plan. The goal of this effort was to foster mutual trust among team members and the patient and their family caregivers. Targeted efforts solicit input from every participant. For example, the physician normally ends the discussion by enquiring if there are any questions or if anyone has anything else to add. Special emphasis also encourages questions from patients and family caregivers, and then teach-back is used to ensure understanding of responses—the bedside nurse normally assumes responsibility for reinforcing throughout the day. Through this process, the bedside nurse is made aware of patient and family caregiver issues and concerns, and makes certain that these are or have been addressed. Group dialogue through the team rounds can identify necessary tasks, for example, consultation with other services such as nutrition, chaplain and/or pain management.
ITIM rounding time guideline
ITIM rounds structure
Goals/Purpose
Involve patients, families and team members (physician-MD, nurse-RN, case manager-CM, PharmD and other services) in the discussion about plan of care and daily goals.
Keep patients and families well informed of progress towards daily goals and the discharge plan, update the plan of care as needed as the patient progresses.
Use teach-back with every patient encounter to promote enhanced understanding of care and decrease preventable readmissions.
Provide document for the patient to follow their progress and take notes.
Improve interprofessional communication.
Intervention guideline
Physician/APP (Advanced Practice Provider)
At beginning of rounds, inform team of rounding schedule.
Lead team into room to greet patient/family and introduce team.
Lead plan of care discussion and provide update on patient status: reason for hospitalisation, active problems, response to treatment, test results and consults.
Make nurse, case manager and/or pharmacist aware of new orders.
Discuss potential discharge date and realistic time with team and patient/family.
Redirect to stay on time, as needed.
Review and reiterate daily plan of care goals with team, patient and family.
Pharmacist (PharmD)
Provide brief update of medications, current medication and/or any changes to medications.
Discuss any issues or concerns at this time.
Update medication orders if needed and able.
Case manager (CM)
Ensure next bedside nurse ready for the team.
Provide update regarding discharge plan: next site of care, what has been done, what needs to be completed, any potential issues delaying discharge.
Together with the physician, should discuss potential discharge date and realistic time with team and patient/family.
Update physician orders if needed or provide guidance to residents on what they will need to order pertaining to discharge planning (OT, PT, intravenous antibiotic recommendations and others).
Nurse (RN)
Make patient aware of rounding process and time during shift change pass-off.
Provide brief update: overnight events and goals for the day.
RN checklist: vital signs, pain control, bladder, bowel, lines, drains, airway, wounds, mental status, falls, safety and nutrition.
Discuss and/or request needed orders.
Patient care associate/unit clerk
Provide team with RN assignment list and phone numbers at beginning of shift.
Print RN bedside report tool at beginning of each shift.
Make follow-up appointments for patients as needed.
ITIM, Interprofessional Teamwork Innovation Model.
Study outcomes
The primary outcomes included 30-day all-cause same-hospital readmissions and emergency department (ED) visits. Secondary outcome measures included LOS, total direct cost, average direct cost per patient day and patient satisfaction as measured by the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) through the responses to survey questions dealing with communication, care transitions, discharge planning and overall hospital rating.
Data collection and analysis
Data sources included the Enterprise Data Warehouse, which contains data from inpatient and outpatient care, coding, case management electronic health record data, financial department’s databases, and documentation of ongoing feedback from the ITIM team members as to the perceived effects of ITIM and how it might be enhanced. Given limited resources and available previous findings by other research teams, we did not conduct rigorous interviews and surveys to assess patients’ and healthcare providers’ experience. Instead, we used ongoing feedback to test and refine the intervention. HCAHPS measures were received monthly from the Patient Experience Office.
During ITIM development and implementation, 30-day readmissions and LOS SPC charts were generated and monitored monthly. In order to assess the impact of the ITIM programme on the outcomes of interest, patients receiving care from hospitalists on the two teams assigned to GS7 were considered as the intervention group and patients receiving care from other hospital medicine (HM) teams in the same time frame were matched with a control group, with 1:2 matching ratio. Of note, no specific or general criteria were followed to determine assignment of medicine patients to the two GS7 HM teams versus the other 14 HM teams; that is, patients were assigned to teams based purely on bed availability and team capacity. Of note, there were no differences in individual hospitalist’s patient team size. The matching pool of non-ITIM patient visits included internal medicine adult (age 18+) discharges from either GS or Chandler hospitals (CH), and excluded cancer comanagement and observation teams as they were believed to be markedly discrepant from the general medicine teams. Separate baseline and postintervention patient matching was conducted with a baseline time period of 1 July 2017 through 31 December 2014, and postintervention implementation time period of 1 September 2015 through 30 June 2016. The 1 January 2015 through 31 August 2015 time period was used as an intervention development and refinement period. Discharge exclusion criteria included inhospital death, outlier LOS (>30 days), having ICU stay(s) during the hospitalisation, left against medical advice and discharged to hospice.
Matching was done using Mahalanobis distance35 between patient discharges. The Mahalanobis distance is the ‘distance’ between observations/cases, based on the standardised measurements of the variables used in the distance calculation. The standardisation of variables prior to the distance calculation ensures that variables with relatively large values and ranges compared with other variables do not dominate the subsequent matching algorithm. An optimal matching algorithm was used to minimise the overall sum of case–control distances. Variables used in matching algorithm include age (continuous), gender, admitted as transfer (yes/no), patient resident location (rural/urban cluster/urban area), Diagnosis-Related Group (DRG) weight, discharge disposition, CH campus (yes/no), payor and selected Elixhauser Comorbidity Index groups.
Difference-in-difference analysis was the main analytical approach, with three variables of interest—GS7, POST and ITIM—defined a priori. The GS7 variable represents the effect of the study cohort on the outcome at the preimplementation period, POST represents the effect of time on the control group, and ITIM represents the effect of the study cohort on the outcomes in the postimplementation period. Total cost, cost per patient day and LOS were modelled using linear regression, while readmission and ED visits were modelled by conditional logistic regression modelling. Plots were used to check for equality of variance and appropriate functional form of covariates used in modelling. Data were analysed using SAS V.9.4.
Results
Feedback from both hospitalists and bedside nurses indicated that they initially viewed ITIM with scepticism, and thought it would increase their rounding time and workload. However, with implementation of the ITIM, participants reported enhanced communication among team members and overall time savings at the project meetings. The overall HCAHPS survey response rate is less than 30%.36 In our matched cohorts, only small numbers of patients discharged from GS7 floor completed the HCAHPS survey; therefore, we were not able to conduct difference-in-difference analysis for patient experience.
The 30-day readmission rate SPC chart from July 2014 to June 2017 (figure 1) demonstrated six consecutive downtrending points starting August 2015 and a run of ten points in a row starting September 2016 below the centreline, both indicating that special cause variation (ie, ITIM intervention) occurred. An analysis (30-day readmission rate SPC chart) including all medicine discharges, that is, without exclusion criteria applied, can be found in online supplementary appendix 2. This showed a similar special cause variation with run of 10 points. The LOS SPC chart did not demonstrate special cause variation.
- Download figure
- Open in new tab
- Download powerpoint
30-Day readmission rates for discharges from GS7 group, July 2014–June 2017. GS7, Good Samaritan seventh floor medical unit; ICU, intensive care unit.
Online supplementary appendix 3 shows the matching cohorts. To compare similarity in the matched cohort, we used standardised difference, the comparison of the means or medians of continuous covariates and the distribution of categorical covariates between intervention and control groups. Although there is no universally agreed-upon criterion, a standardised difference that is less than 0.1 has been taken to indicate a negligible difference in the mean or prevalence of a covariate between treatment groups.37 In our study, all matching variables’ absolute standardised mean differences are within the recommended criterion of less than 0.1, with the exception of the CH campus discharge indicator variable, which was expected since the patients at CH were more complex with a higher case mix index.
The demographics and comorbidities of patients cared for by the GS7 hospitalists were comparable with the control group at both the preintervention and postintervention period (table 2). Preintervention (table 3), compared with the control group, discharges from the two GS7 teams had similar average total direct cost (P=0.895), lower average direct cost per patient day (P<0.0001) offset by longer LOS (P=0.009), and no difference in 30-day readmission and ED visit rates (P=0.0640 and P=0.1154). Postintervention, patients discharged from the GS7 HM teams had non-significant lower average total direct cost (P=0.097), lower average direct cost per patient day (P<0.0001), similar 30-day readmission rate (P=0.2738) and ED visit rate (P=0.4377), and longer LOS (P=0.0026).
Descriptive statistics for discharges from GS7 group and control group
Descriptive outcomes of discharges from GS7 group and control group
For continuous outcomes (cost and LOS), the model with the highest adjusted R-squared value was selected to report the findings. For the logistic models, the model with lowest Akaike information criterion was selected. In multivariate modelling, patients discharged from hospitalist teams using the ITIM approach were associated with reduced 30-day same-hospital readmissions with an estimated point OR of 0.58 (95% CI 0.36 to 0.94), but there was no impact on 30-day same-hospital ED visits (table 4). Difference-in-difference analysis showed that ITIM was not associated with changes in average total direct costs nor average cost per patient day, after adjusting for all other covariates in the models. Importantly, cost calculations included salaries for the additional staff of pharmacist and case manager staffing. ITIM also was not associated with a change in LOS.
Difference-in-difference analysis: effect of GS7 on outcomes
Discussion
Implementation of the ITIM yielded a number of salutary effects. Buy-in from all disciplines to participate occurred as they quickly experienced value through gained efficiencies and improved communication and information sharing. Despite initial trepidation and concerns by the hospitalists and nurses, they subsequently reported qualitatively that overall the model saved substantial time through reduction in the need and occurrence of subsequent pages, reflected in continued sustainability of the model through the end of 2017 and beyond. The integration of ITIM into implementation of Project BOOST yielded results similar to prior experience with a reduction in the same-hospital readmission rate.30
However, we did not observe significant reduction in 30-day ED visits. This discrepancy in postdischarge utilisation by patients may be explained by a couple of factors. First, the ITIM intervention focuses on the inpatient setting and unit-level patient-centred care and teamwork. Although it addresses some of the contributing factors for ED visits, such as patient education for better understanding of medical conditions/care plan/medications and making sure to address postdischarge needs, it does not impact several root causes of avoidable ED visits, for example, limited access to timely primary care services, EDs offering patients immediate reassurance about their medical conditions, patient factors (homeless, mental illness, substance abuse, multiple chronic conditions and others), and practice culture and pattern (primary care providers referring patients to the ED). Second, studies38 39 show that patients with multiple visits to the ED are often burdened by multiple chronic diseases, substance abuse issues and mental illness. As demonstrated in table 2, comorbidity from mental illness, substance use disorders and complex chronic conditions is a significant challenge facing our health system. Care coordination strategies beyond ITIM and the hospital are needed to address these issues.
A reduction in costs was one of the major goals for ITIM, but this was not realised. Nonetheless, there was no overall increase in costs despite the added expenses of a pharmacist and a case manager beyond existing staff. A major component of offsetting cost savings was a reduction in pharmacy costs driven by the addition of a pharmacist to the team. Thus, the ITIM model generated sufficient savings to be cost-neutral. ITIM appears to represent a successful model of team-based rounding on medical services that engages patients and family caregivers, with seeming cost-effectiveness and facilitates Project BOOST with reduced readmissions. ITIM may represent a cost-effective approach to improving hospital care that can be generalised successfully by implementation on HM units.
The ITIM approach to teamwork with bedside rounding and facilitated communication allows sharing of expertise and perspectives, and yields a shared mental model to form a common goal of achieving patient goals.40 Collaboration is a complex process that requires intentional knowledge sharing and joint responsibility for patient care.41 The roles of healthcare providers, patients and family caregivers are becoming increasingly interdependent requiring mutual trust and understanding among all parties. The team aimed to integrate the practice of interprofessional teamwork during bedside rounds with the patient’s and/or family caregiver’s involvement in POC goal setting. This innovative approach represents true ‘interprofessional practice’ with shared communication and decision-making among the team at the point of care. Healthcare organisations are complex systems that comprised subsystems that are interconnected and interdependent. The classic use of systems theory in healthcare research is the structure-process-outcome model, which has been used more recently to describe the manner in which team care is delivered.5 42 Researchers and clinicians should recognise that structures are inputs such as hospitals, clinics, building design or staffing levels, and elements of systems such as team composition, time and/or location for communication, and communication channels.42 The improvements in outcome measures found in this study indicate that geographically based team rounding can succeed on adult hospital units.
Based on the success of ITIM, our system spread this model to other medical floors at GS and two 36-bed units at the new university hospital, although in the same enterprise the implementation on other floors required adaptation to unit staffing, and patient population characteristics. Based on our experience, developing and implementing team-based practice require purposeful planning to ensure active engagement of the participants so they learn about, from and with each other for person-centred collaborative practice. The key elements of interprofessional practice include responsibility, accountability, coordination, communication, cooperation, assertiveness, autonomy, and mutual trust and respect.43 The current ITIM bedside rounding structure development followed these key components. QI experts at our health system are guiding other teams going through the adoption and adaption processes, and analysis of ITIM continues with this implementation.
There are several limitations to this study. First, we did not track daily start time and duration of ITIM bedside rounds, and percentage of ITIM rounds with all core team members present. Later, we discovered that these are important process measures guiding the implementation of ITIM. We plan to collect these data during future implementations. Second, we did not administer a separate ITIM patient experience survey to avoid conflict with HCAHPS and used unit-level data from those patients’ satisfaction surveys. In our final matched analytical file, because of the small numbers of patients who completed HCAHPS survey, we were not able to conduct difference-in-difference analysis on patient experience. Although preliminary, HCAHPS results suggest an improvement trend in patient satisfaction, but a larger sample size is needed for rigorous evaluation to confirm; HCAHPS response rates were poor. Of note, on implementation of ITIM on two new medical floors, each received the institution’s quarterly patient satisfaction award—heretofore an unprecedented event for a medical unit. Third, we did not survey providers who were involved in implementing ITIM because we thought the feedback collected and discussed at weekly meetings served the purpose of monitoring and adapting the intervention. Future study should undertake a mixed methods approach evaluating the implementation process and outcomes.
Conclusions
The health system of the future will be more dependent than ever on effective and efficient teamwork to coordinate care.44 A healthcare organisation’s success will depend on its ability to effectively foster and coordinate a spirit of teamwork, collaboration and coordination among many professional disciplines. The ITIM approach facilitates a collaborative environment in which patients and their family caregivers, physicians, nurses, pharmacists, case managers and others work and share in the process and systems of care.
References
Footnotes
Competing interests None declared.
Ethics approval According to the policy of activities that constitute research at our university, this study met the criteria for QI activities exempt from the University of Kentucky IRB review.
Provenance and peer review Not commissioned; externally peer reviewed.