Www.umich.edu



Float Pool Staffing Needs 19F7-Final-Report To:Mrs. Balqis Elhaddi, Assistant Director, Michigan MedicineMr. Abdulsallam Alnajjar, Continuous Improvement Fellow, Michigan MedicineMr. Zachary Costello, Continuous Improvement, Michigan MedicineDr. Mark Van Oyen, Professor, Industrial and Operations EngineeringMs. Mary Duck, Staff Specialist, UMH QI Michigan Quality System, Michigan Medicine From: IOE 481 Project Team #7Sajay SrivastavaShyam SudheendraEli StrausAddison Zhou Date: December 10, 2019 Table of Contents Executive Summary 2Introduction4Background and Key Issues Staffing Process Key Issues within the Float Pool557Goals, Objectives, and Impact7Project Scope8Design Process Engineering Challenges Literature SearchDeliverables and Design Tasks Qualitative and Quantitative Current State Assessment Float Pool Staffing Model Recommendation Hiring Process Recommendation Design RequirementsDesign ConstraintsEngineering StandardsData Collection and Analysis, Findings, and Recommendations Data Collection PSA Float Pool Data Interview Results Human Resources Data Data Analysis & Findings Alternatives Considered Conclusions Recommendations Expected Impact References Appendix A. Requirements, Constraints, and Standards MatrixAppendix B. Pugh Decision Matrix89910101111121213131313141414202222232425 28Executive Summary IntroductionThe University of Michigan Medical Group (UMMG) consists of three float pools, registered nurses (RN), medical assistants (MA), and patient service assistants (PSA). Float pools consist of groups of employees that are temporarily deployed to outpatient clinics when such clinics have a staffing need (typically due to a clinic employee taking leave). Within Michigan Medicine, there are 166 Ambulatory Care Unit (ACU) outpatient clinics. The primary issue is that the float pools are having difficulty utilizing staff in a manner that fulfills the requests submitted by the clinics. The clinics are submitting a high volume of requests for float staff to fill-in temporarily, and the float pools have neither the proper allocation nor number of current staff on payroll to satisfy the requests. The main metric used to assess how well the requests are satisfied is fill rate, the number of fulfilled requests divided by the number of total requests submitted. A team from the College of Engineering at the University of Michigan’s IOE 481 Senior Design class identified the current staffing level shortage and the factors that are contributing to that shortage. The project was conducted over the course of the Fall 2019 University of Michigan semester, from September to December 2019.MethodsAfter determining the current fill rates for the float pools, the following goals and objectives have been completed: Creation of a current state map showing how the requests are submittedAssessment of current state performance and current staffing levelsDevelopment of float pool staffing recommendations that helped find the gaps in the staffing levels for the PSA float poolDetermination of the frequency of the requests for the various types of leave (family and medical leave, paid time off, EpicCare Training, temporarily open positions) for the PSA float poolGain of insight into the current hiring process to understand the high turnover ratesDue to data availability limitations, many of the objectives had to be narrowed to focus on the PSA float pool from the MA and RN float pools. Qualitative and quantitative current state assessments were conducted, which included: Qualitative and quantitative current state assessmentConstruction of a current state map of how requests are submitted and processed to the ACUCreation of time plots showing the variation in fill rate for the PSA float pool Assessment of normality and seasonality in the number of requests submitted per week for the PSA float poolCreation of time plots showing the variation in staffing levels (FTEs) across all float poolsCreation of bar charts showing which reasons for requests are the most frequent in the PSA float poolFloat pool staffing recommendation that will show the increase in FTE values needed to fulfill requests at a higher fill rate in the PSA float poolRecommendations for improving the hiring process within all float pools after an analysis of root causes for turnover and related hiring practicesRecommendationsUsing the PSA data as a proxy for the other float pools, several key recommendations were made for the process. Based on needed FTE calculations, scheduling of PSAs and planning of PTO for float pool staff should be further investigated to improve utilization of current hours of work. Also, given the goal of 80%, onboarding of three or more PSA FTEs should be strongly considered. In consideration of all three float pools, recommendations included strengthening the hiring process via stricter background and reference checks to reduce the number of terminations in the post probationary period. Another recommendation showed the importance that should be given to the different reasons for requests while scheduling. Overall, these recommendations will allow better utilization of current float pool staff, as well as improved fill rate through onboarding of additional staff. However, the main risk associated with these recommendations is its basis on PSA data which may not be as applicable to the other float pools as assumed.Expected ImpactInvestigation of the scheduling of PSAs and planning of PTO for float pool staff will improve utilization of current available work hours, theoretically increasing fill rate from 40% closer to 71%. Increasing PSA FTEs will also help to further increase fill rate towards the goal of 80%. Strengthening the hiring process via stricter background/reference checks will decrease underutilization and reduce turnover rates. Submitting requests as short or long term rather than on a weekly basis will help to streamline the request submission process. Overall, these changes will allow the float pool system to better utilize its current staff, as well as provide for an improved fill rate through onboarding of additional staff. Introduction The University of Michigan Medical Group (UMMG), which manages the 166 outpatient clinics across the University of Michigan health system, is having difficulty meeting fill-in requests when they have temporary staffing shortages. Outpatient clinics only treat patients who do not need to stay overnight in a hospital. These requests are sent to the three main float pools within outpatient care: registered nurses (RN), medical assistants (MA), and patient service assistants (PSA). RNs specialize in various areas of medical practice, MAs perform administrative work within the clinics and assists physicians, and PSAs perform clerical services. Float pools are systematic models that can help provide temporary or long-term staffing, through which staff can work at different clinics and in different roles on a daily basis. Currently, the float pool staff systems are not able to meet all of the requests sent to UMMG. The project client, an Associate Director within UMMG, has noticed that the lack of staff may be due to a few factors, including family and medical leave (FML), paid time off (PTO), temporarily open positions (OP), and EpicCare Training (ECT). Staffing requests to fill temporarily vacant positions requires the assignment of a float staff member to a clinic. The Associate Director within UMMG has asked a team from the College of Engineering at the University of Michigan’s IOE 481 Senior Design class to identify the current staffing level shortage and the factors that are contributing to that shortage. The project was conducted over the course of the Fall 2019 University of Michigan semester, from September to December 2019. This was done by reviewing anecdotal evidence given by Human Resources and the float pool managers and analyzing several sets of data consisting of information about the three float pools. One data set only contained information on the PSA float pool and it included the total number of FTE shifts requested, the total number of FTE shifts filled, and the total number of FTE shifts requested by FML, PTO, OP, ECT, and other reasons (OTH). Another data set contained the total FTE (Full-Time Equivalent Workload) values, which indicated the number of full-time shifts with 40 hours per week being equivalent to one FTE, and turnover rates across all three float pools. All of these metrics are still currently associated with fill rate, a key performance metric for the float pools that indicates the proportion of staffing requests that are fulfilled. The ideal fill rate for all three float pools is 80%. Regarding the current state analysis and design tasks, the total number of requests within the PSA pool was obtained in the form of aggregated FTE shift values from FY14 to FY19. These values were compared to the average available FTE value for the PSA pool to identify the staffing gap within that pool. A frequency analysis was conducted for the PSA float pool to determine which reasons for requests are most prevalent.This led to a recommendation describing which types of requests the PSA pool needs to account for more than others. The project deliverables also consist of a review of the turnover process within all the three float pools, and recommendations to how the ACU can counteract the most prevalent reasons for turnover. This final report goes into detail about the background and key issues observed; the goals and completed project objectives; the design process, which includes a literature search, deliverables and design tasks; data collection methods and analysis; findings and recommendations; and alternatives considered prior to developing the recommendation, the final recommendations, the expected impact, and the project conclusions. Background and Key Issues In order to move forward with the requisite analysis, gaining a better understanding of the staffing process and identifying key issues within the float pools was of paramount importance.Staffing ProcessCurrently, all three float pools lack the staff needed to fulfill the volume of requests sent in by the clinics. To better understand the relationship between staffing requests and available staff members, the high-level process through which a request is made and fulfilled had to be examined. There are two main types of requests that can be submitted: temporary requests completed through call-ins, or long-term standing requests completed online. The steps below, determined through interviews of the float pool managers, detail the process for the call-in requests submitted on a daily basis:A clinic calls the ACU automated line.The caller from the clinic states their name, the clinic they are from, whether they are requesting for RN, MA, or PSA, and their priority for coverage (how urgently they need staffing assistance). Additional information collected include the number of staff members that are needed, the number of many members currently on staff, and the reason for submission.The scheduling team receives and logs the information. The scheduling team reviews the current allocation of staff, and decides whether or not to move the staff from certain clinics to others based on priority and/or urgency (e.g. from an occupational clinic to a surgical unit).The scheduling team sends staff to areas in which they specialize or are oriented with. For example, call-ins indicate an emergency due to a sudden vacancy, and because the clinic expects there to be no disruption in the flow of work, the scheduling team may need to adjust the allocation of the current float staff to satisfy these conditions.The scheduling team pages the manager to notify them of the call-in request. This is done so that changes in staff allocation have upper-level supportThe scheduling team contacts the clinic to inform them of the staffing changeFor standing requests, such as ones for FML or open positions, requests are submitted in advance through an online portal and are not monitored on a daily basis:The clinic submits an online submission that asks for similar information as the call-in request but requires some additional information (such as the name of the staff member that they are covering for, the number of weeks needed, the number of days per week needed, and the number of hours per day needed) The scheduling team places the request on the float pool’s calendar and notifies the applicable float pool managerThe scheduling team notifies the float pool manager of the request The scheduling team notifies the clinic of the staffing assignment A current state map for call-in and online requests is shown in Figure 1 below. Figure 1. Current State Map for Staffing Request ProcessBeyond this process, each float pool has its own unique methodology of allocating staff. RN staff typically perform long-term staff coverage and need specific technical skills to be assigned to a specific clinic or unit. For this reason, the entire RN float pool is split into smaller, clustered pools that specialize by task. In these smaller float pools, the same float nurses are guaranteed a certain number of working hours on an annual basis.For the PSA float pool, the staffing method is slightly different. While nursing focuses on clustering staff, PSA focuses on giving opportunities for staff to move between clinics and gain various skills. The manager of the PSA pool believes that doing this increases the versatility of staff, and thus their ability to meet emergency call-in requests. Unlike nurses, who are required to be proficient in the clinic they are filling-in for, PSA staff only need basic training to perform their tasks. However, each clinic is known to have its own workflow, and the staff often take time to get accustomed to the clinic’s work environment.Like PSA, the MA staffing model does not revolve around clusters that are organized by job specialty. MA staff are instead divided into multiple pods that are organized geographically. Key Issues within the Float PoolsFloat pool managers experience three main issues. The first issue is that their float staff are underutilized. The specialization of nurses is a key cause of this issue because nurses are only sent to clinics for requests requiring their particular skills. In the MA, PSA, and RN float pools, staffing for a standing request can also lead to underutilization, since the clinics often do not notify the float pool managers if the previously absent clinic staff member returns to work earlier than expected. Float pools also do not monitor long-term staffing assignments, and as a result, do not notice whether a clinic employee has returned earlier than expected. Because the float staff member is staying at the clinic unnecessarily, the float pool is now understaffed while the clinic is overstaffed. Another issue is the high turnover rates. One reason for this is that during long-term assignments, float staff often develop a desire to transfer permanently to the clinic or unit to which they were assigned, sometimes only giving the float pool managers a notice of one week. When this occurs, the manager must start looking for a new hire to fill the vacancy, along with allotting time for training. As a result, these situations adversely impact the availability of current staff to respond to emergency call-in requests.Another reason for high turnover rates is Probationary Period Discharge (PPD), the 6-month period in which float pool staff are required to meet certain expectations for their positions. If a staff member does not satisfactorily meet the conditions, then they are terminated due to PPD. Goals, Objectives, and Expected Impact The primary goals of the project were to quantify the staffing shortage and identify the root factors causing unsatisfactory fill rates in the float pools. Since the data sets were restricted to the PSA float pool, finding the staffing shortage was limited to only that pool. Identifying the root causes related to the unsatisfactory fill rates was completed using the data sets provided by HR, which consisted of information for all three pools. The primary goals guided the creation of a staffing model proposal that informs the client of the staffing gap in the PSA float pool. Recommendations on how improvements can be made to float pool hiring process, given various constraints, has been provided later in the report. In order to accomplish the primary goals, the following objectives were completed:Creation of a current state map showing how the requests are submittedAssessment of current state performance and current staffing levelsDevelopment of float pool staffing recommendations that helped find the gaps in the staffing levels for the PSA float poolDetermination of the frequency of the requests for the various types of leave (family and medical leave, paid time off, EpicCare Training, temporarily open positions) for the PSA float poolGain of insight into the current hiring process to understand the high turnover ratesCreating the current state map helped to understand the relationship between the 166 clinics and the ACU float pools. Assessing the current state performance and staffing levels helped to explain the fill rate for the PSA float pool. Developing a float pool staffing recommendation informed the client of the number of FTEs needed to meet the number of FTE shifts requested in the PSA float pool. Determining the frequency of the different reasons for requests helped the PSA float pool know which types of requests should be filled over others. Gaining insight into the current float pool hiring process helped highlight certain issues within the hiring process to see if turnover could be reduced. The key impact from accomplishing and implementing these objectives will the increase in fill rates across all three float pools.Project ScopeThe scope consisted of reviewing staff requests submitted to the ACU by all 166 outpatient clinics. Within outpatient care, the only three float pools that were analyzed are the RN, MA, and PSA float pools. The departments that may be impacted are units deploying the float staff and the clinics that RNs, MAs, and PSAs.Design Process The following sections go into more detail about the project and the factors that could be used to increase the fill rate to 80%.Engineering Challenges An engineering challenge encountered over the course of the project was the collection of data regarding the current state fill rates and the frequency of the various types of requests (FML, PTO, etc.) for all three float pools. The MA and RN float pool managers were not electronically transferring the data inputted from the call-in and online requests. Instead, they had been handwriting the data onto sheets of paper and could not provide the current state data for their respective float pools. The data consisted of the clinic name, the department name, the time when float staff were needed, the number of float staff needed, the number of staff that the clinic has available, the reason for the request, and the priority level (high, low, etc.). The only float pool manager that had been transferring the data to Excel sheets was the PSA float manager. Another challenge with the data was collecting information showing the total number of hours each of the 166 clinics requested. Instead of identifying the clinic name that requested the staff, Human Resources and the float pool managers only had the capability to pull the department names within the clinics. This was a large issue, since the data could not be aggregated to only include employee leave data pertaining to the 166 ACU clinics. Due to the lack of information, the data set containing this information was not used for the analysis.Literature SearchSeveral sources were found in the form of past reports and articles relevant to this study. The first source was Google Scholar, where the first article, Demand for Temporary Agency Nurses and Nurse Shortages was found. The article was authored by researchers at the University of California, Berkeley and the Korea Institute for Health and Social Affairs, and contained an analysis of temporary nurses and whether they could help hospital systems to meet their demand in a more cost-effective manner [1]. Based on the research conducted, utilizing temporary workers had a positive impact. Another article that was reviewed was Managing a Recruitment and Retention Issue: Floating, which was published in the Journal of Nursing Administration. This report included the results of a charter team of staff nurses assembled to address negative factors within the float staffing system at West Virginia University Hospital. This was accomplished by defining methods to increase the benefits of floating by maximizing flexibility and minimizing staff dissatisfaction [2]. By following their design methods of floating on a rotational basis, providing adequate departmental resources, and offering a float pay differential, nursing satisfaction increased while RN vacancies decreased.Finally, the last article reviewed was Cost-Effectiveness of Clustered Unit vs. Unclustered Nurse Floating, which was published in the November 1997 edition of Nursing Economics [3]. This report contained research comparing the cost and staffing balance performances of unclustered unit floating and clustered unit floating patterns. This included a computer simulation to isolate staffing model performance from human and program variables. The article also stated that while there was a negligible difference in absolute costs, there was significant differentiation in understaffing, absolute availability, and orientation.Deliverables and Design Tasks The final deliverables consisted of the following:Qualitative and quantitative current state assessmentConstruction of a current state map of how requests are submitted and processed to the ACUCreation of time plots showing variation in fill rate for the PSA float pool Assessment of normality and seasonality in the number of requests submitted per week for the PSA float poolCreation of time plots showing variation in staffing levels (FTEs) across all float poolsCreation of bar charts showing which reasons for requests are the most frequent in the PSA float poolFloat pool staffing recommendation that will show the increase in FTE values needed to fulfill requests at a higher fill rate in the PSA float poolRecommendations for improving the hiring process within all float pools after an analysis of root causes for turnover and related hiring practicesThe deliverables indicate better ways to handle float staff both externally and internally as they are assigned to the clinics. The external aspect, which was exclusive to the PSA float pool, dealt with the staffing level increase needed to significantly raise the fill rate. The internal aspect dealt with analyzing the turnover rates of float staff in the three float pools and findings improvements to the current hiring process to reduce turnover. Completion of these deliverables and implementation of the associated strategies is expected to increase the fill rate towards 80%.Qualitative and Quantitative Current State Assessment The current state process map shows how the information flows from submitting the staffing requests to the ACU to the scheduling of the staff across all three float pools. Understanding this process provides insight into the factors that should be considered when considering float pool staffing. A generic process map was created based on anecdotal information provided by the float pool managers. Regarding the quantitative current state assessment, a bar chart was created that illustrated the most frequent reasons for requests within the PSA float pool. The data analyzed was over a 52-week period from November 2018 to November 2019, and is plotted in terms of labor days requested off. This dataset provided a look into which problems were plaguing the PSA float pool in terms of understaffing, and which causes were most important to look into first when analyzing why requests were made over the past year. The distribution of the number of requests submitted by week was assessed for normality as well. Time plots were also created to show the changes in fill rates for the PSA float pool since November 2018. The current staffing levels within each float pool (in units of FTE shifts) were visualized to show year-to-year trends, and further analysis on the PSA dataset was conducted to show how the fill rate and number of requests submitted varied on a weekly basis over the course of FY19. Float Pool Staffing RecommendationThe float pool staffing model recommendation will help the PSA float pool manager understand the staffing gap in FY19, which is from July 1, 2018 to June 30, 2019. The first step in calculating the gap was summing the total number of requests over the fiscal year. These requests were obtained in the form of FTE values. After obtaining the total number of FTEs requested for FY19, the average FTE value for PSA was multiplied by 52 weeks per year and five days per week. The days per week was five in order to not include weekends. The difference was found between the total number of FTE shifts requested and the total number of corresponding FTEs available over FY19 not including weekends. The result was then divided by 52 weeks per year and then divided again by five days per week to convert the increase in FTEs into an average value. This value represents the average FTE increase required for the PSA float pool in order to have an average fill rate of 80%. Hiring Process Recommendation The data provided by UMMG Human Resources (HR) showed how the float pools internally manage their staff. The main metric under review was turnover rates across the three float pools. The reasons for turnover included discharge (firing), probationary period discharge, quitting without notice, finding another position elsewhere, returning to school, finding a different career opportunity, retirement, relocation, family responsibility, their future plans unknown, and other. These reasons for turnover were analyzed from FY14 to FY19 to determine which were the most prevalent. They were also used to explore float staff hiring and retention to identify room for improvement. A deeper analysis has also been conducted splitting the number occurrences for PPD by job position (Patient Service Assistant, Medical Assistant etc.). This helped determine which float pools experienced the most PPD.Design Requirements All deliverables were assessed by their capability to increase the fill rates. The department-wide fill rate requirement is 80%, and the deliverables will help the ACU move closer towards that goal.Another requirement is an in-significant increase in costs. In conjunction with moving towards an 80% fill rate, any improvement measures implemented must not be cost-ineffective for the ACU. Design Constraints The constraints focused on the availability of the data that was obtained from the float pool managers and HR. The initial data request to HR was made to initiate the analysis regarding the movement of float pool staff between different clinics and the reasons for which they were moving (FML, PTO, etc.). This would have helped analyze how the float pools can improve their ability to fill clinic requests. However, the data received from HR contained information regarding the FTE values within each pool, the turnover rates within each pool, the annual salary of pool staff per position, and the number of new applicants in each pool. Even though this was not the data that had been requested, it provided insight into the internal operations of each pool rather than external or clinic-related factors. The conclusions drawn from the data can still be useful in producing recommendations to reduce the staffing shortage and increase the overall fill rate.Another design constraint was the limitation of another data set provided by HR. This data set consisted of the number of total hours taken for different types of leave in all the departments within the U-M health system. However, the source database did not have information pertaining to which clinic each of the departments were grouped within. As a result, this database did not have the capability to show the total hours of leave by clinic location and name. Since the purpose was to relate the occurrences of leave by clinic, the dataset could not be used for the analysis.A third design constraint with the data was the availability of the current state data (number of requests, reasons for requests, and fill rates) being limited to only the PSA float pool. As explained earlier, the PSA float pool manager was the only float pool manager to digitally input the current state data into Excel sheets. Thus, the analysis on the current state and the staffing shortage was strictly limited to the PSA float pool. Engineering Standards The U.S. Department of Labor states that under Guidance on the Protection of Personal Identifiable Information, the following is stated:Disclaimer: “Any representation of information that permits the identity of an individual to whom the information applies to be reasonably inferred by either direct or indirect means. Further, PII is defined as information: (i) that directly identifies an individual (e.g., name, address, social security number or other identifying number or code, telephone number, email address, etc.) or (ii) by which an agency intends to identify specific individuals in conjunction with other data elements, i.e., indirect identification. (These data elements may include a combination of gender, race, birth date, geographic indicator, and other descriptors). Additionally, information permitting the physical or online contacting of a specific individual is the same as personally identifiable information. This information can be maintained in either paper, electronic or other media.” [4]This law specifically limited the information that was given within the data sets. No personally identifiable information regarding the float staff was given to the team. The data sets collected only included the specific job titles in each float pool, the number of hours worked per week, and the current FTE values.A summary of the Design Requirements, Design Constraints, and Engineering Standards is shown in Appendix A.Data Collection and Analysis, Findings, and Recommendations The following section of the report outlines the process through which the results of data collection and analysis were used to formulate final recommendations.Data Collection Over the course of the project, three different sources of information were obtained.PSA Float Pool Data The PSA float pool manager provided a Qualtrics survey that showed the opinions of various ACU managers regarding the frequency of reasons for requests. The PSA float pool manager also provided a data set showing fill rates and number of FTE shift requests from FY15 to FY19. This data was used to determine various current state measures within the PSA float pool, such as fill rate over time, number of requests aggregated by reason for the request, and the current FTE shortage gap. Interview ResultsTwo sets of entities were interviewed and provided anecdotal evidence on various topics. The three float pool managers provided information pertaining to the long-term and short-term request processes, the staffing method within each pool, and any float pool specific issues. The HR representative provided information regarding the purpose of PPD and how it impacted the float pool hiring process.Human Resources DataThe HR department provided a data set showing the turnover rates and reasons for termination within all three float pools from FY15 to FY19. The staffing levels and headcounts in FTEs were also given for each position in the ACU; however, the main focus of the analysis was the float pool staff that were sent out to clinics. Data Analysis & Findings A current state analysis was conducted from the available data for the PSA float pool. First, a survey of 51 ACU managers that had previously been conducted was used as a proxy for part of the unavailable data. The relevant results of this survey can be seen below in Figure 2. Figure 2: Survey Results show Open Position and Medical Leave key reasons for shortages.Sample Size: 51 respondents; Data gathered from Qualtrics Survey conducted in FY18 These results show that Open Position and Medical Leave are the two most frequent reasons that clinic managers submit requests to the float pool. However, these results are limited in their specificity in that they do not separate by clinic or float pool, and are based purely on subjective recollection rather than quantitative measurement.Nevertheless, the survey results were corroborated by the analysis of the PSA float pool, as shown in Figure 3. Figure 3: Breakdown of PSA Requests for FY18-19 corroborates survey results.Sample size: 52 weeks; Data gathered from PSA Float Pool ManagerThis matches the survey results, and lent initial credence to the assumption that PSA metrics could be generally (albeit cautiously) applied to the other float pools.Analyzing the PSA data further, the fill rate over time was also visualized in Figure 4.Figure 4: Overall Fill Rate by Month demonstrates room for improvement and variation. Sample size: 52 weeks; Data gathered from PSA Float Pool ManagerAs can be seen, the PSA fill rate is far below the goal of 80%, indicating significant measures should be taken to help close this gap–?at minimum (ie. in months with the highest fill rates), the current values are still close to half of the identified objective rate. To identify the root causes of the low fill rate (to the extent possible given the limited data), the overall float pool staff data was analyzed. This yielded insight into the changing staffing levels over time, which can be seen in Figure 5. Figure 5: Changes in Headcount over time across Float Pools shows divergence of PSA from RN, MA trend. Sample size: 5 years; Data provided by UMMG Human ResourcesThis demonstrates that the PSA float pool headcount has a negative trend over the past five years, whereas the other float pools had a positive trend. This suggests a low PSA headcount may be responsible for the low PSA fill rate, but without further data on the MA or RN pools, additional insights were not robust. It also prevents further assumptions regarding similarities in the staffing issues faced for MA and RN staff. That being said, the overall staffing data did include turnover rates and causes, which were aggregated in Table 1.Table 1: Reasons for Year 1 Terminations across Float Pools, Sample Size: ___Cause for Termination in Yr 1CountProportionAnother Position Elsewhere413.3%Discharged13.3%Family Responsibilities620.0%Probationary Period Discharge 1653.3%Relocation310.0%As can be seen, PPDs caused the highest proportion of terminations.After these analyses, the next step was to determine the difference between available full-time staff and the FTE staff required to meet the demands from request submissions. This was done using the average PSA FTE value (from the overall staffing data) and the annual total of requested hours. From there, the daily FTE value was computed (see Equation 1). Daily FTE =ρ × [ w52(FTE requested in week w) ÷ 52 ] ÷ 5Equation 1: Calculation of Necessary FTE per Day, ρ = fill rateGiven that the average PSA FTE in FY19 was 10.7 (corresponding to an estimated FTE value of 2782 available per year) and that there was a total of 5512.5 hours requested, the total annual gap in hours could be estimated to be 1611 hours. Using the same formula, an 80% fill rate (ρ = 0.8) would require 6.3 additional FTEs, or a total of 17.0 FTEs per day. To achieve a theoretical 100% fill rate (ρ = 1), 21.2 FTEs would be required. Given that the raw ratio of available hours to requested hours (2782 / 5512.5 = 50.4%) is higher than the observed fill rate, further analysis was deemed necessary. The next step was to analyze the typical number of requests that would be submitted each week. To this end, a histogram, as shown in Figure 6, was built representing the number of weeks in FY19 a certain quantity of requests were submitted.Figure 6: Weekly Request Quantity and Frequency for FY19 appears to indicate Normal DistributionThe quantity of requests appeared to follow a normal distribution. To test this, an Anderson-Darling normality test was conducted. The results of this test can be seen in Figure 7.Figure 7: Normality Test Plot with Associated Anderson-Darling Metrics demonstrates Normality cannot be rejectedThe p-value of 0.486 indicates that normality cannot be rejected at the significance level a = 0.05. Therefore, there is some validity to using normal approximations. As such, the inverse Normal Distribution function was utilized to determine the point at which the weekly demand would be satisfied 80% of the time (note that this value of 80% is not the same as fill rate). This value was determined to be 126.1 FTEs per week (corresponding to 25.2 FTEs available per day) and can be visualized in Figure 8. Number of FTE shifts requested per weekFigure 8: Normal Distribution of Requests Submitted by Week and Coverage with 126.1 FTEs/weekSample Size: 52 weeks; Data gathered from PSA Float Pool ManagerIn other words, Figure 8 shows that to satisfy the entirety of the demand eight out of every ten weeks, 25.2 FTEs are required. Having completed this analysis, the different strategies were compared. This comparison can be visualized in Table 2.Table 2: Comparison of FTEs based on Strategies and Increase from Current LevelMetricDaily FTEsIncreaseCurrently Available10.7-80% Annual Fill Rate17.06.3100% Annual Fill Rate21.210.5InvNorm for 80% of weeks25.214.5The seasonality of the PSA request data was then visualized to better assess if there were times of the year when the fill rate varied significantly. This can be seen in Figure 9.Figure 9: Fill Rate by week with current FTE value and possible increases.This demonstrates both the variability in submitted requests over the course of a year and the inefficient allocation of staff that currently exists. It also shows the current FTE level in the PSA float pool and possible alternative staffing levels (“Average” corresponds to the current FTE level, “Recommended” to an 80% fill rate, “Idealized” to a 100% fill rate, and “InvNorm” to a satisfaction of demand in 80% of weeks). It should be noted that these FTE levels would not realistically be held constant over the course of 52 weeks, nor should they be–?some adjustment for the seasonality demonstrated would be advisable (ie. fewer FTEs in the late-December holiday period). Alternatives Considered Contrary to the solutions developed from the analysis, there were a few alternatives that were considered over the course of the project. These alternatives were compared to chosen solutions from the analysis based upon their capability of increasing the fill rates to 80% and whether they would entail insignificant increases in costs. An organization of the ratings for the alternatives and solution is shown in more detail in Appendix B: Pugh Decision Matrix. The current state staffing configuration is causing a 20-45% fill rate within the PSA float pool. Therefore, the current state configuration does not show any capability for an improvement of the PSA fill rate towards 80%, however, since the current state does not involve the addition of new float staff, insignificant costs would not be incurred. As a result, current state was given “-1” for increasing the fill rate towards 80%, but a “+1” for causing no significant increase in overall costs.Using the 80% annual fill rate solution of 17.0 FTEs/day has full capability to increase the overall PSA fill rate to 80%. However, this would include adding 6.3 additional FTEs/Day, which would cause the ACU to incur hiring costs. As a result, the 80% annual fill rate was given a “+1” in for increasing the fill rate towards 80%, but a “-1” for causing an increase in overall costs for the ACU, but not as high as the 100% annual fill rate solution and the InvNorm solution.Using the 100% annual fill rate solution of 21.2 FTEs/day has more than full capability to increase the overall PSA fill rate to 80%. However, this would include adding 10.5 FTEs/Day, which would cause the ACU to incur higher hiring costs than compared to the 80% annual fill rate solution. As a result, the 80% annual fill rate was given a “+1” in for increasing the fill rate towards 80%, but a “-2” for causing a significant increase in overall costs, but not as high of an increase compared to the InvNorm solution.Using the InvNorm solution requires 25.2 FTEs/day, since this solution is looking to satisfy the weekly number of requests 80% of the time. This would comfortably increase the overall PSA fill rate to 80%. However, this would include the addition of 14.5 FTEs/Day, which would cause the ACU to incur higher hiring costs than compared to the 100% annual fill rate solution. As a result, the 80% annual fill rate was given a “+1” in for increasing the fill rate towards 80%, but a “-3” for causing a very significant increase in overall costs.The primary reason that the three solutions involving FTEs were all given the same value was because the requirement of the ACU was to increase the fill rate to 80%, which gave an indication that they would not likely risk incurring more costs to go above 80%.Improving the hiring process would involve reducing turnover by reducing the frequency of PPD. Since this solution can provide a benefit to the float pools in the long-run because they can keep a higher average FTE value since they will not be having open positions and searching for candidates as often. This can increase the overall fill rate towards 80%, and would not cause the ACU to incur more costs because although they would keep more staff on hand at any given point, they are reducing the costs of termination and hiring. As a result, improving the hiring process was given a “+1” for increasing the fill rate towards 80%, and a “0” for causing no increase in significant costs for the ACU.An alternative at the initial stages was to consider solutions proposed in the literature search Cost-Effectiveness of Clustered Unit vs. Unclustered Nurse Floating. This search brought up the idea of potentially forming clustered units of float staff by specialization with the intention of creating a more balanced staffing allocation. However, in interviews with the MA and PSA float pool managers, they stated that their staffing allocation did not heavily rely on focusing on specific skill sets of their staff because their skills were transferable to a majority of positions to which they floated. As a result, focusing on clustering by specialization could not have brought significant improvement to the fill rates in these float pools. In addition, clustering by speciality could not bring upon significant extra costs to the ACU because it would involve simply adjusting the orientation of staff, and not the addition of new staff members. As a result, cluster float pools was given a “-1” value for increasing the fill rate towards 80% and a “+1” value for causing no significant increase in overall costs to the ACU. Another alternative was developed from the literature search suggestions in the article Managing a Recruitment and Retention Issue. This article spoke about decreasing staff dissatisfaction. The issue with this alternative was that within the data set provided by HR, the turnover cases were only labeled by the objective reasons for termination. There was not any quantitative information within the data set to identify whether the workers who left were dissatisfied, and further information could not be collected on those workers through interviews. As a result, a solution regarding the improvement of float staff satisfaction could not have been used to confirm an increase in the overall fill rate. However, the implementation of simple measures to ensure the satisfaction or happiness of the float pool staff would not bring upon on significant extra costs. Therefore, customer satisfaction was given a “0” value for increasing the fill rate towards 80% and a “+1” for not causing a significant increase in overall costs. 70% of the total weightage was assigned to increasing the fill rate towards 80%, and 30% was assigned to no significant increase in overall costs. The main reason was that the 80% requirement was emphasized much more than costs by the client and float pool managers. The two solutions that received the highest sum scores were the improvement of the hiring process and the 80% annual fill rate solution and the. Conclusions The interpretations of the findings and their expected impact are detailed in the sections below.RecommendationsBecause the trends in headcount differed for the PSA float pool compared to the MA and RN pools, most of the recommendations were limited to PSAs. Based on the Pugh Matrix (Appendix B), the ACU should increase its PSA float pool staff by 6.3 FTEs to 17.0 total FTEs. This would be a potential method to increase the average fill rate of the PSA float pool to an average of 80% while causing the ACU to incur less costs compared to its alternatives. Additionally, based on historical variations in the week-to-week data across the year, planned adjustments should be made for when PSA staff are available. Although Family and Medical Leave is difficult to plan for, PTO requests are submitted far in advance and can be approved or denied in accordance with these adjustments.In addition to increasing the staffing level, the float pools should give more importance to the reasons for requests while scheduling. Since open positions and medical leave are much more prevalent than any other type of leave, the float pools should account for the fact that they need to allocate more of their staff in to accommodate for those two types of leave. Similarly to how priorities are given importance while scheduling float staff, it may be beneficial to potentially give the reasons for requests just as much importance. In the long run, the benefit would be moving towards an 80% fill rate across all float pools since it will be ensured that the most frequent types of requests are filled the most.Outside of the PSA float pool, the MA float pool hiring process should be strengthened via stricter background/reference checks to reduce the number of terminations after the probationary period. Based on interviews, the standards for a successful probationary period are already quite low, so any identifiable causes for concern for a candidate should be investigated more thoroughly.Expected Impact Investigation of the scheduling of PSAs and planning of PTO for float pool staff will improve utilization of current available work hours, which should theoretically yield a fill rate closer to 80% rather than 40%. Increasing PSA FTEs will also help to increase fill rate towards the goal of 80%, especially in regards to the difficulty of anticipating medical leave. Strengthening the hiring process via stricter background/reference checks to reduce the number of terminations for post-probationary period will decrease underutilization and reduce turnover rates. Submitting requests as short or long term rather than on a weekly basis will help to streamline the request submission process, improving the ability of the float pools to meet demand in different and unexpected situations. Overall, these changes will allow the float pool system to better utilize its current staff, as well as provide for an improved fill rate through onboarding of additional staff. References[1] S. Seo and J. Spetz, “Demand for Temporary Agency Nurses and Nursing Shortages,” INQUIRY: The Journal of Health Care Organization, Provision, and Financing, vol. 50, no. 3, pp. 216–228, 2013.[2] S. Rudy and J. Sions, “Managing a Recruitment and Retention Issue,” JONA: The Journal of Nursing Administration, vol. 33, no. 4, pp. 196–198, 2003.[3] McHugh ML 1997, ‘Cost-effectiveness of clustered unit vs. unclustered nurse floating’, Nursing Economic$, vol. 15, no. 6, pp. 294–300[4] “Guidance on the Protection of Personal Identifiable Information,” U.S. Department of Labor. [Online]. Available: . [Accessed: 10-Dec-2019].Appendix A. Requirements, Constraints, and Standards Matrix Entry #1234…RequirementsR-A. Organizational Policy(R-A-1) R-B. Ethical R-C. Health & Safety R-D. Economic R-E. Implementability R-F. User Acceptance R-G. Patient Acceptance R-H. Task Duration (please add others as needed) Entry #12345Constraints C-A. Organizational Policy C-B. Ethical C-C. Health & Safety C-D. Economic C-E. Implementability(C-E-1) C-F. User Acceptance C-G. Patient Acceptance C-H. Task Duration Entry #12345StandardsS-1. HIPPA S-2. Organization's Std.(S-2-1) S-3. Best Practice S-4. ASTM S-5. Code Standards that are not applicableANSINIOSHOSHA ASEASTMRequirements in Detail(R-A-1) The organizational wide policy across all float pools for fill rate of staffing requests is 80%. The pools are not achieving this requirement currently, and would like for the solution provided in this project to find the staffing level and allocation that will increase the fill rate up to 80%.Constraints in Detail (C-E-1) The solution provided for this project relies on how much data and the type of data received by Human Resources and the float pool managers. If given the full data that was requested for, the findings will need to adjust and the basis for the solution may change.Standards in Detail (S-2-1) Guidance on the Protection of Personal Identifiable Information from the U.S. Department of states that the organization needs to prevent the distribution of data that “(i) that directly identifies an individual (e.g., name, address, social security number or other identifying number or code, telephone number, email address, etc.) or (ii) by which an agency intends to identify specific individuals in conjunction with other data elements, i.e., indirect identification.” [4]Michigan Medicine requires the protection of employee information through employee privacy rights. These laws will specifically limit the information that is given to the project team within the data sets. No personal information should be given to the team. The data sets collected so far have only included the current FTE values per job title, and the sum of each position’s early earnings. Appendix B. Pugh Decision Matrix Proposed StrategyIncrease Fill Rate towards 80%No significant increase in costsTotalCurrent State-1+1080% Annual Fill Rate+1-10.7100% Annual Fill Rate+1-20InvNorm for 80% of weeks+1-30Improve Hiring Process+100.7Cluster Float Pools-1+10Improve Staff Satisfaction0+10.3Weight70%30%- ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download