A Collaborative Staffing Model: Utilizing Nurse Practitioners and Physicians on the Hospitalist Service

The purpose of this needs assessment is to determine an effective solution to provide additional coverage to the hospitalist service. It also serves to identify what patient factors are responsible for increases in workload for the hospitalist service. Studies indicate that an increase in patient age, acuity, comorbidity and requests for co-management lead to increased lengths of stay and workload for providers, which leads to poorer patient outcomes (Elliot, Young, Brice, Aguiar, Kolm 2014). An evaluation of the current staffing model to determine resources was conducted and used as a baseline template for staffing needs. General admission tends to the service were tracked for the previous 3 years and found an increase in patient admission volumes by 5.85%. A retrospective chart review of 7096 patients was performed on all admissions to the service for the previous year to determine how age, acuity, co-morbidity and length of stay (LOS) affect workload. It was found that a patient age of 61-65 significantly affects both length of stay and patient acuity, which could explain increases in workload due to the average admission age of 62. Co-management requests were also identified to determine their impact on workload and were found to increase daily patient census by 12 patients. Utilizing the data from chart review and comparing it to the current staffing model allowed us to project future staffing needs and create a staffing model. It was found that the service could benefit from hiring 6 additional full-time equivalent providers. Evaluation of current salaries for both physicians and advance practice clinicians (APCs) found a significant cost savings (25%) for a staffing model that incorporates APCs. Studies have also shown that use of APCs have comparable outcomes as resident teams making them an affordable option that will still meet the needs of the patient (Wall, Scudamore, Chin, Rannie, Tong, Reese, & Wilson 2014).

Running Head: A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE A Collaborative Staffing Model: Utilizing Nurse Practitioners and Physicians on the Hospitalist Service. Amber Lovell, BSN, AGACNP, DNP Student Project Chair: Clint Child Content Experts: Megan O'Hara, Nathan Starr, Douglas Philpot. University of Utah In Partial Fulfillment of the Requirements for the Doctor of Nursing Practice A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 2 Abstract The purpose of this needs assessment is to determine an effective solution to provide additional coverage to the hospitalist service. It also serves to identify what patient factors are responsible for increases in workload for the hospitalist service. Studies indicate that an increase in patient age, acuity, comorbidity and requests for co-management lead to increased lengths of stay and workload for providers, which leads to poorer patient outcomes (Elliot, Young, Brice, Aguiar, Kolm 2014). An evaluation of the current staffing model to determine resources was conducted and used as a baseline template for staffing needs. General admission tends to the service were tracked for the previous 3 years and found an increase in patient admission volumes by 5.85%. A retrospective chart review of 7096 patients was performed on all admissions to the service for the previous year to determine how age, acuity, co-morbidity and length of stay (LOS) affect workload. It was found that a patient age of 61-65 significantly affects both length of stay and patient acuity, which could explain increases in workload due to the average admission age of 62. Co-management requests were also identified to determine their impact on workload and were found to increase daily patient census by 12 patients. Utilizing the data from chart review and comparing it to the current staffing model allowed us to project future staffing needs and create a staffing model. It was found that the service could benefit from hiring 6 additional full-time equivalent providers. Evaluation of current salaries for both physicians and advance practice clinicians (APCs) found a significant cost savings (25%) for a staffing model that incorporates APCs. Studies have also shown that use of APCs have comparable outcomes as resident teams making A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE them an affordable option that will still meet the needs of the patient (Wall, Scudamore, Chin, Rannie, Tong, Reese, & Wilson 2014). 3 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 4 A Collaborative Staffing Model: Utilizing Nurse Practitioners and Physicians on the Hospitalist Service Introduction Hospitalists have been an integral part of the United States (US) healthcare system since their inception about two decades ago. These groups of highly trained medical specialists are responsible for the care of patients admitted to the hospital and perform an array of crucial services, including; coordinating and providing medical care, overseeing transitions from the inpatient to outpatient setting, and co-managing patients with multiple specialty services. Although hospitalist medicine (HM) is the fastest-growing medical specialty, several factors are increasing workload, decreasing reimbursement and leading to staffing shortages (Radler, 2016). Obesity and Co-Morbidity For the first time in several centuries, the life expectancy for the current generation of children in the US may be shorter than their parents. This is due, in part, to the unyielding rise of obesity, a condition associated with many complications, including: an increased risk for developing type 2 diabetes, coronary heart disease, cancer, and other complications (Must, Coakley, Field, & Colditz, 2001). Not only do these complications lead to higher costs, increased length of stay, and poorer outcomes for the patient, but they also place strain on the hospitalists through increased workload (Elliot, Young, Brice, & Kolm, 2014). Aging population As new technologies are introduced, the growth of outpatient care expands and hospital stays are reserved for the sickest and most acutely ill. Unfortunately, in 2009, researchers found that the elderly comprise 61% of admissions to the hospitalist service: a 20% increase from the A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE previous 20 years (Yong-Fang, Gulshan, Freeman, & Goodwin, 2009). As these generations continue to age the nation is faced with physician shortages that affect care provided in both the inpatient and outpatient settings leaving these frail patients without the help they need. In turn the hospitalist service sees more complex hospital admissions in a setting where rapid assessment and precise decision-making are commonplace. These patients add an additional layer of stress to the already overstretched and burned-out hospitalist service. Co-management As patient care grows more complex through shifts in demographics and regulatory trends, co-management has been a proposed solution to improve patient care and satisfaction (Beresford, 2011). Although, not listed by the Society of Hospitalist Medicine as a core competency of a hospitalist, it is estimated that 85% of hospitalist teams have provided some kind of co-management. "The concept involves shared responsibility, authority, and accountability for the care of hospitalized patients, typically with orthopedic surgeons or other specialties, and with the hospitalist managing the patient's medical concerns, such as diabetes, congestive heart failure, or DVT" (Beresford, 2011). Unfortunately, if not well-planned and executed, co-management can lead to care "dumping," which consumes staff time and raises concerns about scope of practice and utilization of appropriate skill sets. Obesity, an aging population, and increases in co-management have led to surges in hospitalist workload and physician burnout. This burnout leads to turnover leaving the service understaffed and overworked. This has led to greater costs and less reimbursement for both physicians and hospitals, requiring innovative solutions to remedy the staffing shortage. Implementing advance practice clinicians (APC), such as Nurse Practitioners, could be 5 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 6 incorporated into the hospitalist-staffing model to bridge staffing gaps and to work with physicians to improve patient outcomes. Rationale Using the Workforce Planning Model (Craig, Byrick, & Carli, 2002) as a guide, one can define the areas in which staffing shortages pose the greatest needs and how to appropriately fill those needs. This is done through analysis of the current workforce plan, forecasting of needs, exploring gaps, and developing a collaborative staffing model. This can be identified in the study objectives, which are as follows: 1. Assess current hospitalist staffing model, projected trends in admission rates and population increases. 2. Evaluate current staffing in relation to projected admission trends and diagnosis. 3. Develop a collaborative staffing model that incorporates APCs co-managing patents with the hospitalist that is based on analysis of the current admission trends, projected increases, and research. 4. Propose the staffing model to stakeholders (Figure 1). Specific Aims The aim of this needs assessment is to identify an effective solution to provide additional coverage to the hospitalist service. It is projected that the additional coverage for the service will decrease physician workload, increase hospital reimbursement, and increase patient satisfaction. A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 7 Methods Context This project is being conducted at a large, urban, level-one trauma center located in the central region of the Salt Lake Valley. The inpatient hospitalist team admits and cares for thousands of patients of various ages, ethnicities, cultures and socioeconomic backgrounds. This group of highly trained specialists consists of forty-seven physicians that staff 4, thirty-two bed medical floors. The team also employs 4 nursing coordinators and provides a teaching environment for medical students, interns and residents. Current daily staffing of the hospitalist service includes 7 teams: 4 of which are teaching teams, including residents and interns, and 3 that are non-teaching (Table 1). Each team rotates taking admissions to distribute patient loads with a few guidelines. The max number of patient encounters per house staff team per day is 20 and each intern may only have 10 encounters per day. Non-house staff teams have no patient encounter max and therefore take higher patient loads when patient volume is increased. Currently, the teams employ no advanced practice clinicians. Intervention(s) In order to develop a collaborative staffing model to eliminate gaps in the hospitalist schedule the current staffing model was evaluated to determine what current provider resources are available. While accomplishing this objective it was noted that the non-hospitalist physician teams were disproportionately caring for more patients than the housestaff teams, increasing workload for several members of the team. Next, objectives 2-4 were accomplished through analysis of the current admission, length of stay, patient acuity, and co-management requests to determine if and where additional staff was needed. A staffing model was then created utilizing A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE the regression statistics of the data and was presented to stakeholders at the hospitalist staff meeting. The methods utilized to accomplish these objectives are discussed below. Study of the Intervention(s) Research for this needs assessment began with a retrospective patient chart review to determine if the hospitalist service has an increased need for additional providers. All patients age 18 and older who were admitted to the hospital under physician's with the service line title "hospitalist" from September 2016 through September 2017 were included (n=7096). All patients admitted to other services (non-hospitalist) were excluded. General hospital-wide admissions, by specialty group, from January 2015-September 2017 were also obtained to determine admission trends over a larger period of time. Review of the chart provided the patient's admission date and APR-DRG score, which was utilized to evaluate the current trends in admission and acuity. The All Patient Refined Diagnosis Related Group Scores (APR-DRG) are used to measure resource intensity of the inpatient stay by evaluating six dimensions of the patient's care: 1) severity of illness; 2) risk of mortality; 3) prognosis; 4) treatment difficulty; 5) need for intervention; and 6) resource intensity. After determination of these factors a score of minor, moderate, major or extreme risk of mortality is assigned to the patient. In addition to admission and acuity, chart review identified certain co-morbid conditions that may increase the complexity and risk for readmission of the patient. Co-morbid conditions were identified using ICD-10 including hypertension (I50.9, I50.1, I50.20-I50.23), congestive heart failure (I50.30-I50.33, I50.40-I50.43), kidney disease (N18.9-N17-N19), liver disease (K76.9), and diabetes (E11.9, E08, E09, E10, E11, E13). It is hypothesized that the 8 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 9 aforementioned factors are indicators of the time spent caring for a patient and helps define the provider's workload for the day to determine additional provider needs. Identification of gaps in care in which additional provider help was needed has been used to build a staffing template. Cost analysis using current pay scales for both APCs and physicians in the Salt Lake Valley were compared and a model incorporating Nurse Practitioners was developed. The study data was then complied into a PowerPoint presentation and was presented to the medical director and lead physician team of the hospitalist service. Data Analysis Simple descriptive statistics and regression analysis were calculated from data on patient admissions, co-morbidities, acuity, and length of stay. This information was used to determine relationships between the variables and how they affected the provider workflow. Ethical Considerations This study was determined to be a non-human subjects research by the University of Utah and Intermountain Healthcare Institutional Review Boards. All data collected was coded and kept on a double password, encrypted computer in which only the study chair and author have access to. Results Admission data by specialty for all inpatient and observation patients were collected from January 1, 2015 to September 1, 2017. This data revealed that the number of adult admissions to the hospitalist service increased by 5.85% between 2015 (25.92%) and 2017 (27.53%) (Table 2). In the first quarter of 2017 (Jan-March), the average daily census for the group was 73 patients, which allocates approximately 10.4 patients per physician team per day. By the second quarter (Apr-June), the average daily census had increased to 105 patients per day, which A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 10 averages to 15 patients per team per day, prompting a request for additional providers for the service (Starr, N., personal communication January 28, 2018). While discussing the possibility of hiring more providers, hospital administration decided that the hospitalist service would begin to see patients at a new telehealth site and admit patients for the oncology service. This would make the hospitalists responsible for all medical management of the oncology patient's acute and chronic conditions during their inpatient stay. The oncology team would follow the patient loosely and defer all treatments, with the exception of chemotherapy administration, to the hospitalist team. Through admissions data review, it was determined that the oncology service averages approximately 12 patients per day. When we add the average daily census of oncology to the average hospitalist daily census we will have 117 patient contacts or 16.7 patients per physician team per day, which recent studies suggest has a negative effect on patient LOS (Elliot, D.J., Young, R., Brice, J. Aguiar, R., Kolm, P. 2014). 7096 patient charts from September 2016 to September 2017 were reviewed with an average age of 61, LOS of 3.5 days and acuity of moderate. Regression analysis determined that age, LOS, congestive heart failure and hypertension had a positive correlation in acuity to a pvalue of <0.01. Analysis also shows that as the patient's age increase the risk for mortality increased for all co-morbid conditions except hypertension. Likewise, as the patient's acuity score increases their length of stay also increases, indicating that these patients require more time and resources to care for (Tables 4,5). In order to safely staff the service and account for the increase in patient contacts the hospitalist group has requested 6 full time equivalent (FTE) providers. This would come at a significant cost to the hospital at approximately 1.32 million dollars per year ($220,000 per A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 11 physician/year, plus benefits). However, if the group were to hire six nurse practitioners ($90,000/year, plus benefits) and two physicians ($220,000/year, plus benefits) to off-load the teams and staff a new oncology service the average cost would be $980,000 per year. This would be a cost savings of $340,000 or 25.76% savings (Starr, N., personal communication January 28, 2018). The new staffing model would add an APC to each of the current non-house staff teams and they would see an average of 8 patients per day. This would reduce the patient contact hours to 12.6 patients per team per day (including the added oncology patients) (Table 6). This data was presented to several stakeholders at a hospitalist staff meeting and was well received. Stakeholders presented the study data to hospital administration and a plan to hire APCs was approved. Discussion Summary Our results demonstrate that the volume of patients admitted and cared for by the hospitalist team has increased since 5.85% since 2015. Additional increases were seen in patient contact hours during the first 6 months of 2017. Retrospective chart review data suggests could be due to the increases in patient APR-DRG scores and LOS found to correlate with the average age (61) of patients admitted to the service reaffirming the need for additional providers in order to keep up with provider workload demands. In addition, the decision by administration for the team to staff a new telehealth site and admit for the oncology service has prompted the request for an additional 6 FTE providers. It was found that incorporating nurse practitioners into the A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 12 model would result in a 25.76% cost reduction and increased provider pool than hiring physicians alone. Interpretation Data review from hospitalist service admissions for the past 3 years do account for an increase in patient volume. Data obtained on patient age, acuity and co-morbidity could be responsible for this increase as regression analysis does show an increase in both ARP-DRG score and LOS for patients 61-65 years of age, which includes the average age of the hospitalist patient. This increase in provider workload does place significant strain on the service and the need for additional providers to alleviate the workload is apparent based on the patient to physician contact hours and benchmark data supporting the reduction of patient contacts to less than 15 patients per provider (Elliot, Young, Brice, Aguiar, Kolm, 2014). If the service were to opt not to hire more providers it is projected that the LOS would continue to increase leading to increased patient and provider cost, decreased reimbursement, provider burnout and increased turnover costs. A staffing model that includes APCs and physicians was proposed due to the ability to hire more providers at a cost efficient price with similar patient satisfaction and quality ratings as non-APC teams (Collins, N., Miller R., Kapu, A., Martin, R., Morton, M., Forrester, M., Atkinson, S., Evans, B., Wilkinson, L. 2014). Reduction in workload is beneficial to providers through increased productivity. Patients and insurance companies will also see benefits through decreased LOS and increased satisfaction. Limitations This study is limited in a few ways. Study data was only obtained from one hospital and one service so it may not be representative for all other specialties and hospitals in the region. A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 13 Furthermore, data collection through chart review has been a considerable challenge throughout the duration of this project due to the change in electronic health records. Data from retrospective chart review was only available for the previous year making it challenging to determine if increases in length of stay, age, acuity and co-morbid conditions were correlated with the increased in admissions from 2015-2017. Data specific to obesity not obtained and could have offered further information about the effect of obesity on patient admissions. Future research could focus on the implementation and evaluation of the staffing model to determine its effectiveness in workflow distribution. Conclusions Results of this study show that the increase in patient volume and workload due to comanagement warrants the need for additional providers to be hired. Evaluations in current salary trends for the group have determined that a collaborative model that employs both APCs and physicians would allow for more full time equivalent providers to be hired at a reduced cost than hiring physicians alone. The utilization of APCs would help alleviate workflow burden and is estimated to decrease annual salary costs by 25% compared to non-APC models. Further studies could examine the implementation of this model to determine its sustainability, benefits to patient satisfaction, LOS and cost. If the model is found to be sustainable and cost effective it could serve as a guide for other hospitals or services looking to employ APCs. Acknowledgements Clint Child, DNP, MBA, CENP, RN Denise Ward, DNP, FNP, ACNP, RN Megan O'Hara, MD A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE Nathan Starr, MD Douglas Philpot, MHA, MBA, MRH, FACHE Benjamin Peterson, Statistician Paul Sperry, Data Analyst 14 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 15 References Ashbrener, G. (2005). Issues in Determining Appropriate Levels of Hospitalist Staffing. The Hospitalist 2005, January. 2005(1). https://www.the-hospitalist.org/hospitalist/article/122923/issuesdetermining-appropriate-levels-hospitalist-staffing Beresford, L. (2011). The Comanagement Conundrum. The Hospitalist. 2011, April. 45(4). Collins, N., Miller R., Kapu, A., Martin, R., Morton, M., Forrester, M., Atkinson, S., Evans, B., Wilkinson, L. (2014). Outcomes of adding acute care nurse practitioners to a Level I trauma service with the goal of decreased length of stay and improved physician and nursing satisfaction. Journal of Trauma and Acute Care Surgery. 2014, Feb. 76(2): 353-7. Craig, D., Byrick, R., Carli, F. (2002). A physician workforce planning model applied to Canadian anesthesiology: planning a future supply of anesthesiologists. Canadian Journal of Anesthesia, 49(7), 671-677. Elliot, D.J., Young, R., Brice, J. Aguiar, R., Kolm, P. (2014). Effect of the Hospitalist Workload on the Quality and Efficiency of Care. The Journal of Academic Medicine. 174(5):786-793. Koplan JP, Liverman CT, Kraak VI, eds. 2005. Preventing childhood obesity: health in the balance. National Academies Press, 18(2):43-48. Must, A., Spadano, J., Coakley, E.H., Field, A.E., Colditz, G., Dietz, W.H. 1999. The disease burden associated with overweight and obesity. The Journal of the American Medical Association 1999;282:1523-1529. Radler, B. 2016. Year of the hospitalist: Celebrating 20 years of medicine's fastest growing specialty. The Hospitalist. 56(2):21-23. A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE 16 Wall, S., Scudamore, D., Chin, J., Rannie, M., Tong, S., Reese, J., Wilson, K. 2014. The evolving role of the pediatric nurse practitioner in the hospital setting. Journal of Hospitalist Medicine. 2014 Apr;9(4):261-5. Yong-Fang K., Gulshan, S., Freeman, J.L., Goodwin, J. 2009. Growth in the care of older patients by hospitalists in the United States. New England Journal of Medicine 2009 Mar 12; 360(11): 1102- 1112. A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE Appendix Figure 1: Objectives Table Objective Assess current hospitalist staffing model in a large urban hospital including projected trends in admission rates and population increases. Evaluate current staffing in relation to projected admission trends and common diagnosis. Method • Identify a project specialist within the hospitalist group. • Evaluate the current staffing model. Specifically the patient to physician ratio on housestaff vs non-housestaff teams. • • Develop a collaborative staffing model that incorporates APCs comanaging patents with the hospitalist that is based on analysis of the current admission trends, projected increases, and research. • • Outcome/Product Determine what the current and potential future staffing needs for the group are. Evaluate the current hospital admission rates, staffing ratios, and common diagnoses by talking with the physician and looking at data for the past 3 months. Evaluate trends in hospitalist comanagement. Identify which specialty services request hospitalist management most frequently. Determine the common types of admissions, number of admissions and staff needs. This will help identify areas where an APC will provide needed coverage. Utilizing data from staffing ratios, admissions and projected trends I will identify where the hospitalist group could utilize an APC. Develop a sample schedule that This will tell us when the APC will be needed most and what the potential schedule will look like. It should aim to eliminate gaps in staffing ratios to provide more provider coverage. 17 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE incorporates APCs and MDs for the comanagement of patients. Propose the staffing model to stakeholders. A meeting will be held with the stakeholders to present the data. Present data that shows why and how the APC will be helpful in closing the current and projected provider gap. Table 1: Current Hospitalist Staffing Model Current Staffing Model Max 20 patient encounters Team 1 Physician Resident Intern Team 2 Physician Resident Intern Team 3 Physician Resident Intern Team 4 Physician Resident Intern NHS 1 Physician Unlimited patient encounters NHS 2 Physician Unlimited patient encounters NHS 3 Physician Unlimited patient encounters Max 20 patient encounters Max 20 patient encounters Max 20 patient encounters 18 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE Table 2: Hospial Admissions by Specialty (2015-2017) Table 3: Monthly Patient/Co-morbid Count 2017 19 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE Table 4: Effect of patient age on length of stay and APR-DRG (illness severity) Table 5: Effect of Age on APR-DRG (illness severity) 20 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE Table 6: Propsoed Staffing Model Team 1 Proposed Staffing Model Max 20 patient Physician encounters Resident Intern Team 2 Physician Resident Intern Team 3 Physician Resident Intern Team 4 Physician Resident Max 20 patient encounters Max 20 patient encounters Max 20 patient encounters 21 A COLLABORATIVE STAFFING MODEL: UTILIZING NURSE PRACTITIONERS AND PHYSICIANS ON THE HOSPITALIST SERVICE Intern NHS 1 Physician APC NHS 2 Physician APC NHS 3 Physician APC Unlimited patient encounters Unlimited patient encounters Unlimited patient encounters 22