Analysis of Nursing Workload in Primary Care University of Michigan Health System Final Report Client: Candia B. Laughlin, MS, RN Director of Nursing Ambulatory Care Coordinator: Laura Mittendorf Management Engineer Fellow Program and Operations Analysis Department IOE 481 Team: Paul Beekman Mike Duboe Derek Gould Daniel Weiser November 27, 2006
Executive Summary This report presents a detailed account of our analysis of nursing workloads in one primary care unit of the University of Michigan Health System. The study was conducted by a group of four IOE 481 students. One primary care unit at Brighton Health Center, which consisted of 7 nurses, was studied. After reading this report, one can understand the goals and objectives of the nursing workload study, a detailed background of the current situation, our approach to observing and measuring the activities of nursing staff, the expected impact of the project, and findings and conclusions from the three weeks of data collection. In addition, our team has made several recommendations based on our experiences to help the client conduct future studies and determine appropriate staffing levels. Background Currently, doctors performance is compared to a set of workload standards. However, nurses are not evaluated under any set of standards. Some doctors and others feel that such workload standards would also be useful to ensure best performance from nursing units. To help develop such standards and identify any inefficiency in nursing time allocation, it is useful to gain a quantitative understanding of current nursing workloads and the time spent on various phone activities. Our team is part of a larger lean initiative involving fifteen primary health care centers. This group includes nursing staff and other representatives of primary care centers. The ultimate goal of the larger initiative is to develop a set of standards for nursing activities and phone calls and to eliminate waste in the process. Through this initiative, it is the intent that nurse staffing levels can more accurately be determined based on data and demand. The following key issues provide reasons that necessitate this project: The establishment of workload standards for doctors but none for nurses A lack of data regarding nursing workload tasks and approximate times Inefficient use of RN time by completing tasks meant for other nurses or technicians Non-optimal time management on phone calls Differences in nursing workload between large versus small clinics Variability of nursing workload based on the day of the week or the time of day Patient dissatisfaction related to waiting times Goals and Objectives To help ambulatory care clinics determine appropriate nursing levels, develop time and procedural standards for nursing workload and phone calls, and determine a methodology for ongoing measurement on nursing units, the student team has performed the following tasks: Developed a list of all nursing activities, as well as one for specific phone activities Performed a workload study using random time sampling on one ambulatory care clinic Quantified the proportion of time spent on various activities and phone calls Gained information to identify waste in the phone calls nurses receive and handle 1
With this information, the client will be able to: Determine appropriate staffing levels for nursing units Develop performance benchmarks for nursing workloads Develop time standards for nursing phone calls Identify areas for improvement through appropriate time allocation Determine a methodology for ongoing measurement on units Methodology To complete the project, our team has utilized the following resources: Literature Search: To discover any previous related projects and identify current trends in nursing workload studies, we performed an extensive literature search. Interviews: To gain a qualitative understanding of the current scenario and analyze nurses overall feelings with their current workloads, we built two questionnaires to estimate the proportion of time nurses spend on various activities. We also recorded qualitative feedback from discussions with the nurses. Data Collection Workload Study: To gain a quantitative insight into the proportion of time nurses spend on various activities, we performed a beeper study on 7 nurses from all departments at Brighton Health Center. When beeped at random time intervals, nurses were asked to place a check next to the current activity they were performing. This method of data collection was minimally intrusive and gave sufficient data to perform extensive statistical analyses. Data Collection Phone Study: To gain insight into the types of activities associated with each phone call, we spent one week performing a similar beeper study on various phone activities. Data Collection Phone Log: To analyze the average call length and gain a quantitative understanding of how nurses multi-task while on phone calls, we devoted a third week of data collection to a more thorough phone log study. This study asked nurses to record the length of time and various activities associated with each call received during the day. Observation: To verify the results obtained from the data collection and gain a better understanding of overall nursing workload, we performed 24 hours of observation in which we measured the time nurses devoted to various workload activities. Data Analyses: To form presentable results and conclusions from our data collection so that the client can begin to develop a set of workload standards and analyze staffing levels, we used our statistics training to perform various data analyses. The results of these analyses can be seen below. Findings and Conclusions After analyzing the data from phase 1 of data collection, we developed the following key findings: For all departments (excluding specialty), nurses allocate 34% of their time to speaking on the phone with patients or families of patients, 21% of their time documenting, and 11% of their time looking up patient and care information. 2
The OB/GYN department spends significantly more time on the phone with patients or families than all other departments. There is little variation across departments from the mean values of 21% and 11% for documenting and looking up patient and care information. After analyzing the data from phase 2 of data collection, we developed the following key findings: For all departments (excluding specialty), 3 of phone-related activities is devoted to the actual process of talking to patients or families, 21% is devoted to documenting, and 12% is devoted to looking up patient and care information. The OB/GYN department allocates significantly more time than all other departments to the actual process of talking on the phone with patients or families. The specialty department allocates significantly less time than all other departments to the actual process of talking on the phone with patients or families. After analyzing the data from phase 3 of data collection, we developed the following key findings: For all departments (excluding specialty), the mean call time is 5.4 minutes and the median is 4 minutes. This shows that there are a relatively small proportion of calls that consume a high amount of time. For the specialty department, the mean call time is 6.9 minutes and the median is 6 minutes. Based on the proportion of time spent on the phone from phase 1 results and the mean call lengths, the projected number of calls that can be dealt with each day is 30.5 calls for all nurses (excluding specialty) and 15.2 calls for specialty nurses. With our analysis, the director of nursing can more effectively determine staffing levels, develop performance benchmarks, develop time standards for phone calls, and identify any future deviations from appropriate time allocation. Based on our experience, our team recommends a more robust methodology for ongoing measurement on units. Our team suggests the development of an electronic data collection tool that would eliminate the need to transcribe all results from paper to electronic format. This method would also make the collection process less intrusive to nurses. 3
Table of Contents Introduction and Background......5 Background..........5 Purpose of Project........6 Goals and Objectives.......6 Project Scope.......6 Approach and Methodology.... 7 Approach..... 7 Methodology........7 Expected Impact....... 9 Findings and Conclusions...9 Phase 1 Findings..9 Phase 1 Findings....11 Findings From Nursing Surveys....13 Workload Observations.....15 Phase 3 Findings....17 Conclusions 19 Appendix A. Data Collection Tools for Project Phases 1 and 2...........21 Appendix B. Nursing Workload Surveys.....25 Appendix C. Additional Figures from Phases 1-3 Data Analysis...28 4
Introduction and Background Recent initiatives at the University of Michigan Health System have been put in place to develop productivity standards for doctors. These standards create incentives to provide the best possible patient care. However, no set of standards are in place for nursing workloads in nurses outpatient practice. Specifically, concerns exist that there may be waste in nursing time spent on various phone calls. Therefore, the Director of Nursing would like to examine how nurses in ambulatory care units are currently allocating their time to activities throughout the work day. In addition, she would like to gain insight into the proportion of time devoted to various phone activities. This report provides a summary of work that has been completed throughout the course of our project. First, this report states the goals and objectives of the nursing workload study, explains the background of the current situation, presents our approach to observe and measure the activities of nursing staff, and identifies the expected impact of the project as per the project proposal. Then, this report analyzes the findings from the three weeks of data collection and presents various conclusions and recommendations from these findings. Background Recent initiatives have developed workload standards for doctors. Some doctors and others feel that such workload standards would also be useful to ensure best performance from nursing units. Currently, no clear quantification of how nurses allocate their time while working is available. Additionally, there may be some waste in the nursing phone call allocation. For example, some calls may be directed to nurses that may not require the help of a nurse. This type of activity leads to an allocation of time that may be more effectively spent on other phone calls or other activities. Moreover, some physicians ask nurses to perform tasks that should not be included in nursing activities, potentially due to a misunderstanding of nurses workloads and defined duties. This type of request may make nurses feel that they are performing work that does not utilize their specialized training. In the past, similar studies have been performed on nursing workloads, but no quantifiable performance and phone call standards have been developed. In addition, duties of nurses are constantly changing over time. For these reasons, it is useful to understand current nursing workloads. Our team is part of a larger lean initiative involving fifteen primary health care centers. This group includes nursing staff and other representatives of primary care centers. The ultimate goal of the larger initiative is to develop a set of standards for nursing activities and phone calls and to eliminate waste in the process. Through this initiative, it is the intent that nurse staffing levels can more accurately be determined based on data and demand. 5
The following key issues provide reasons that necessitate this project: The establishment of workload standards for doctors but none for nurses A lack of data regarding nursing workload tasks and approximate times Inefficient use of RN time by completing tasks meant for other nurses or technicians Non-optimal time management on phone calls Differences in nursing workload between large versus small clinics Variability of nursing workload based on the day of the week or the time of day Patient dissatisfaction related to waiting times Purpose of Project Our team has conducted a series of workload studies to quantify the proportion of time spent on common nursing activities. We have also looked more in-depth at the proportion of time spent on different types of phone calls. From our study, the Director of Nursing will develop a set of performance benchmarks that can be used to analyze the activities of nursing staff and identify any opportunities to remove waste or improve time allocation. Goals and Objectives To help ambulatory care clinics determine appropriate nursing levels, develop time and procedural standards for nursing workload and phone calls, and determine a methodology for ongoing measurement on nursing units, the student team has performed the following tasks: Developed a list of all nursing activities, as well as one for specific phone activities Performed a workload study using random time sampling on one ambulatory care clinic Quantified the proportion of time spent on various activities and phone calls Gained information to identify waste in the phone calls nurses receive and handle With this information, the client will be able to: Determine appropriate staffing levels for nursing units Develop performance benchmarks for nursing workloads Develop time standards for nursing phone calls Identify areas for improvement through appropriate time allocation Determine a methodology for ongoing measurement on units Project Scope As proposed, the scope of project includes seven Registered Nurses (RN s) in one ambulatory care nursing unit located at the Brighton Health Center. Brighton Health Center is one of fifteen primary care health centers associated with the University of Michigan Health System. The first study estimated the proportion of time nurses spend on various general nursing tasks. The second study focused more specifically on the proportion of time spent on various types of phone calls. The third study went further in-depth to examine how nurses multi-task while on the phone. 6
The Brighton Health Center has three divisions: Pediatrics, Obstetrics and Gynecology, and Internal Medicine, all of which are included in the project scope. Brighton Health Center employs a total of seven nurses between all three divisions, who work for a total of 6.45 full time equivalents (FTEs). Ambulatory care nursing units located at the remaining fourteen UMHS primary care centers are excluded from the scope of this project. However, results from this project may be generalized and applied to other nursing units in the future. All employees other than nurses at Brighton Health Center will also be excluded. Finally, the project scope originally included 3 nurses at West Ann Arbor Health Center. However, this health center has been excluded from this study due to beeper availability and time constraints. Approach and Methodology Our first task was to develop a set methodology for analyzing nurses workloads. We then proceeded to study more in-depth activity regarding time spent on phone calls. Finally, our team quantified the proportion of time nurses spent on various phone activities, and how much of this time was spent multi-tasking. This study involved the use of a phone log, which we developed and distributed to the nurses on November 1, 2006. Both data collection tools and the phone log can be found in Appendix B. In order to verify our results and gain qualitative information on nursing workloads, our team also conducted a total of 24 hours of observation. A description of this observation is included as a subsection in the methodology below. Approach Our team has addressed this project through the following major steps: Finalized the two data collection tools for the workload study and phone study Interviewed nursing staff and introduced nurses to the data collection tools Observed general nursing workload Observed phone call allocation to various classifications Quantified total phone calls received throughout the day Analyzed data collected from the studies Developed a model of nursing workload for general clinics Our team has completed the above work according to the proposed project timeline. We have progressed according to schedule and did not encounter any significant delays. Methodology To complete the project, our team has utilized the following resources: literature search, interviews, data collection, and data analyses. Literature Search 7
We have conducted an initial literature search to gain insight from previous projects. In particular, our coordinator directed us to one previous study, entitled Pediatric Multi-Specialty Clinic Nursing Staff Workload Study that used a similar random beeper system. This study examined the clinical time distribution of the multi-specialty clinic Pediatric Nursing Staff in the Taubman Center. Nursing staff workload was examined by quantifying the proportion of time spent on clinical activities versus non-clinical activities and the proportion of time spent performing each separate clinical activity. In addition, the team gathered some qualitative information regarding the phone calls handled by nursing staff. By carefully analyzing this report, we gained some insight into methods of workload study, including random beeper systems, time allocation statistics, and Pareto analyses. However, our project will focus on quantifying the phone activity allocation, whereas the previous project only gathered qualitative information. Our team has also performing a more detailed literature search to identify current trends and progresses in workload study in the health care industry. By keeping up-to-date with current projects, we gained a better understanding of how our project fits with new health care trends and initiatives. Interviews Throughout the project, we have spent time interviewing the seven nurses involved in the study to finalize our data collection tool, introduce ourselves, and gain qualitative insight into workload issues. Through these interviews, we were able to better understand nurses perception of their work versus the actual amount of time that is spent on a given task. In addition, our team developed and distributed two short questionnaires to ask the nurses their estimated time allocation to various workload activities and phone activities. After finishing data collection, we analyzed how the actual collected data compared to nurse expectations. Data Collection Workload Study and Phone Task Study We have worked alongside the client, nursing staff, and the Program and Operations Analysis department at UMHS to develop two simple, comprehensive data collection tools for the general workload study and the phone task study. The data collection tools are in simple checklist format to ease the quantification of data and make the recording less time-consuming for nurses. These data collection tools can be found in Appendix B. The first two stages of data collection utilized a random beeper system that required nursing staff to record their current task when paged at random time intervals. The beepers were set to page the nurses an average of four times per hour, at which time the nurses record their current task on the checklist. This study was less intrusive and time-consuming than those which require nurses to record every activity completed throughout the day. One week of data collection was devoted to the workload study and one week was devoted to the phone task study. The data was collected from one larger nursing unit at Brighton Health Center, which includes 7 nurses. Data Collection Phone Log Study The third phase of data collection examined in further detail how nurses allocate their time to other activities during phone calls. Our team developed a phone call log that included all classifications of phone calls that nurses may receive throughout a typical workday. The log was separated by hours of the workday, similar to the previous data collection tools. This phone log can be found in Appendix B. When nurses responded to a phone call, they placed a check in the appropriate category and time slot. After this phase of data collection, we gained a more precise 8
quantification of time allocated to various phone calls. We also quantified the proportion of time nurses spent multi-tasking and determined which tasks are most frequently combined with phone calls. Data Collection - Observation To provide validation for our recorded data, we spent a total of 24 hours observing the general workload activities of the seven nurses at Brighton Health Center. During observation hours, each student studied the activities of one specific nurse and recorded (to the nearest minute) how much time was spent on various workload activities. These activities were encompassed in the data collection tool which was used in the general workload study. Data Analyses Our team first developed summary statistics for the time spent on various nursing activities and on phone activities. From this, we computed standards for how nurses allocate their time throughout the day. To account for differences between data collection days, time intervals, and test subjects, we have performed various statistical quality control measures. This statistical analysis can be seen in the Findings and Conclusions section below. We have used similar analysis techniques for the phone study. By analyzing the overall phone activity across all nurses, we can help the Director of Nursing identify areas of opportunity for improving time allocation. Currently, there are no quantifications of phone activity. By developing quantitative measures, the client can implement methods to eliminate waste in call allocation. In addition, we have compared our observed results with nurses expectations for workload and phone call allocation. We have presented any significant discrepancies to the client to gain awareness of the current scenario. Expected Impact The desired outcome of our team s work is to determine the proportion of time that nurses at one specific ambulatory care unit spend on various daily tasks and specifically on phone activities. This information will allow the client to take various actions: Develop a method for ongoing measurement of nursing workload and phone activity Build a model for appropriate staffing levels Determine benchmarks for time allocation to various nursing activities Findings and Conclusions Phase 1 Findings General Nursing Workload The first phase of data analysis focused on all activities that may be included in a day s work for nurses. From this study, a quantitative proportion of time devoted to various general workload activities can be determined. Figure below shows a distribution of the proportions of various workload activities for all nurses except the specialty nurse. As per discussion with the director of nursing, the specialty nurse is to be left out of the overall proportions, leaving a sample size of 6 nurses. 9
Phase 1 Time Proportions - All RN's 4 3 34% 3 2 21% 1 11% 8% 7% 4% 4% 3% 2% 1% On phone with patient/ family Documenting (electronically or on paper) Looking up patient and care information Coordinating for patients Coordinating prescriptions (pre/post patient visit) Performing Caring for walk- Consulting with administrative responsibilities Ins physicians Taking personal time Performing nurse visit Figure 1. Overall proportion of workload activities for all nurses As seen in Figure 1, 34% of all workload activities is dedicated to the talking on the phone with patients or families of patients. The following two activities, in decreasing order of proportion, are documenting and looking up patient care information. The following two activities, which each comprise less than of overall workload, are coordinating for patients and coordinating prescriptions. After analyzing the overall workload proportions for the entire group of nurses, the data is segmented to compare the proportions across departments. In addition, it is useful to see how these departmental values compare with the averages computed above. A tabular comparison of this data is seen in Table 1 below. Note that the top 10 activities from the first phase of data collection are shown. Table 1. Comparison of overall workload proportions by department Phase 1 - Comparison of Proportions by Dept Caring for walk-ins 6% 1% 3% 2% 1% On phone with patient/ family 52% 22% 29% 31% 21% Documenting (electronically or on paper) 14% 23% 22% 26% 18% Coordinating for patients 2% 13% 4% Consulting with physicians 4% 6% 4% 1% 4% Looking up patient and care information 12% 14% 9% 1 6% Coordinating prescriptions (pre/post patient visit) 1% 13% 1% 8% Performing administrative responsibilities 4% 1% 1% 2% 17% Other meetings 4% 6% Taking personal time 1% 1% 1% 8% Total 9 96% 94% 99% 89% OBG SPEC IM MULTI PED 10
Several key takeaways can be drawn from Table 1 above. First, the OB/GYN department spends significantly more time than all other departments on the phone with families or patients of families. More than half of the workday (52%) is spent on the phone, while the average across all departments is 34%. The rest of the departments are closer to the average, with Pediatrics spending the lowest proportion of time (21%) on the phone. Second, for the second most prominent activity, documenting, the variation between departments is not very high. Most departments are close to the average of 21%, with OB/GYN spending the least time documenting. Third, few departments deviate far from the mean value of 11% for looking up patient and care information. It should be noted, however, that the Pediatrics department spends the least time looking up patient and care information. Fourth, the Internal Medicine department spends 13% of its overall workload coordinating prescriptions. This value is significantly higher than the values from all other departments. Fifth, the Pediatrics department spends 17% of its time performing administrative responsibilities, which is significantly higher than all other departments. Phase 2 Findings Phone Activity Study The second phase of data analysis focused specifically on phone-related activities included in nursing workload. From this study, a quantitative proportion of time devoted to various phonerelated activities can be determined. Figure 2 below shows a distribution of the proportions of various phone-related activities for all nurses except the specialty nurse. As per discussion with the director of nursing, the specialty nurse is to be left out of the overall proportions, leaving a sample size of 6 nurses. 11
Phase 2 Proportions - All RN's 3 3 3 2 21% 1 12% 8% 6% 2% 1% In process of talking to patient/ family Documenting (electronically or on paper) Looking up patient and care information After call coordinating for patients (non-phone) Taking personal time/self development Performing administrative responsibilities In-clinic care Consulting with providers (nonphone) Leaving a message for pt/family Talking to pharmacy or DME Figure 2. Overall proportion of phone-related activities for all nurses As seen in Figure 2, 3 of phone-related activities is dedicated to the actual process of talking to the patient or family of the patient. The following two activities, in decreasing order of proportion, are documenting and looking up patient care information. This is very similar to the order observed in the first phase of data collection, which calculated nearly identical proportions of these top 3 activities. After analyzing the phone activity proportions for the entire group of nurses, the data is segmented to compare the proportions across departments. A tabular comparison of this data is seen in Table 2 below. Note that the top 10 activities from the second phase of data collection are shown. Table 2. Comparison of phone activity proportions by department Phase 2 - Comparison of Proportions by Dept In process of talking to patient/ family 48% 18% 26% 29% 27% Talking to provider (on phone) 7% 2% 1% Talking to external agency 3% 2% 2% 1% Documenting (electronically or on paper) 18% 27% 2 2 12% After call coordinating for patients (non-phone) 26% 13% 4% 4% Consulting with providers (non-phone) 2% 4% 4% Looking up patient and care information 17% 4% 6% Performing administrative responsibilities 2% 1% 2% 4% 1 In-clinic care 9% 4% 9% 7% Taking personal time/self development 2% 2% 2% 17% Total 98% 93% 9 97% 94% Several key takeaways can be drawn from Table 2 above. First, it is seen that the OB/GYN department allocates 48% of its time on phone-related activities to the actual process of talking to the patient or family. This is significantly higher than the average of 3 seen across all OBG SPEC IM MULTI PED 12
departments. On the other end of the spectrum, the Specialty department allocates 18% of its time to the actual process of talking to patients or families of patients. This value is 12% less than the average across all departments. Second, the process of documenting (electronically or on paper) shows some variation from the average across all departments. Primarily, the Pediatrics department devotes only 12% of its time to documenting, which is 9% less than the average value of 21%. Other departments are within reasonable range. Third, the process of looking up patient and care information shows significant variation both from the average and between departments. The average across all departments is 12%. Some departments, such as the Specialty and Pediatrics departments are significantly less than this average, at 4% and 6% respectively. On the other end, the Multi department is significantly higher than the average, at. Aside from varying from the average, these values have a fairly large range, from 4% to. This significant area of variation will be brought to the attention of the director of the nursing. Fourth, the Specialty department spends significantly more time than all other departments coordinating for patients after the call. The specialty department allocates 1 more than the average across all departments to this activity. Fifth, the Pediatrics department has allocates significantly more time to performing administrative responsibilities and taking personal time (or self-development) than all other departments. Compared with the average across all departments at 6% and, respectively, the Pediatrics department allocates 1 and 17%. Findings from Nursing Surveys To ensure valid and representative data collection from phases 1 and 2 and to compare nurses estimations of their workload with collected data we developed and administered a series of surveys to all nurses. These surveys, as seen in Appendix B, asked the nurses to approximate the percentage of time they spend performing both general workload tasks and phone-specific tasks. After compiling this data, we verified that the actual data collected from the nurses matched the data obtained from the survey. The resulting comparison for phase 1 can be seen in Figure 3 below. 13
Phase 1-- Survey Comparison All RN's 3 33% 3 27% 24% 2 1 12% 12% 9% 7% 6% 6% 4% 4% 2% 3% 3% 2% 2% On phone with patient Documenting Looking up patient information Coordinating prescriptions Caring for walk-ins Coordinating for patients Performing nurse visit Nursing Tasks Consulting with physicians Transcribing from voicemail Delegating tasks/ supervising Survey Actual Figure 3. Comparison of survey results vs. actual results for phase 1 study Figure 3 shows the comparison between the actual results found in phase 1 and the data found in the phase 1 surveys. Please note that the percentages do not add up to 10 because the figure only shows 10 of the top 17 nursing workload tasks, which account for 97% of the actual data. As seen in Figure 3, there are few serious differences between nurses estimations of their current workloads and the actual data collected. The most serious differences were that nurses estimated spending only 27% of their time on the phone with patients, while they actually spent 33% of their time on the phone. To verify the accuracy of phase 2 data, we formed a similar figure for phone-specific tasks. Figure 4 below shows the comparison between the results found in phase 2 and the data found in the phase 2 surveys. 14
Phase 2-- Survey Comparison All RN's 3 3 3 2 22% 1 Talking to patient 21% 21% 12% 11% 8% 6% 6% 6% 4% 2% 1% 2% 1% 1% Documenting Looking up patient information After call coordinating for patients Taking personal time/self development Consulting with providers Performing administrative responsibilities Nursing Tasks In-clinic care Leaving a message for pt/family Talking to provider (on phone) Survey Actual Figure 4. Comparison of survey results vs. actual results for phase 2 study Figure 4 shows the comparison between the results found in phase 2 and the data found in the phase 2 surveys. Please note that these percentages do not add up to 10 because the figure only shows 10 of the top 17 nursing tasks, which account for 9 of the actual data. As seen in Figure 4, the data received from the surveys is quite similar to the data received in phases 1 and 2 of the project. The most serious difference is that nurses only estimated spending 22% of their time on the phone with patients, while they actually spent 3 of their time talking with patients. From this analysis, we feel that the data collected from Brighton Health Center s nurses may be used to develop a representative workload standard. However, to further verify these findings, our group performed 24 total hours of observation. Workload Observations In order to validate the data collected during the first two phases of the project, we conducted 24 total hours of workload observations. This is equivalent to of the total time over which the study was conducted. All of the observations were made in two-hour intervals from Nov 3, 2006 to Nov 13, 2006, during normal business hours. Observations consisted of sitting in on the nurses in their normal work environment, and documenting the amount of time spent on each task they performed. If the nurse performed more than one task at once, we recorded the primary task for that period of time. For example, if the nurse was looking up patient or care information once the call had already started, only the time of the phone call was being logged. This is consistent with the manner in which nurses recorded their data during phases 1 and 2. Figure 5 below shows the proportion of time the nurses were observed doing each task, compared to the actual proportion that found from phase 1 and 2 data collection. 15
Phase 1 Observations - All RN's 4 3 34% 33% 3 2 24% 21% Observed Actual 1 12% 12% 11% 7% 9% 8% 3% 3% 4% 2% 1% 2% 2% 2% 4% On phone with patient/ family Documenting (electronically or on paper) Coordinating prescriptions (pre/post patient visit) Looking up patient and care information Coordinating for patients Consulting with physicians Transcribing from voicemail Caring for walk-ins Taking personal time Performing administrative responsibilities Figure 5. Comparison of observed vs. actual results for phase 1 data collection. As seen in Figure 5, the observed proportions are very close to the proportions calculated from phases one and two. This confirms the validity of the data from the phase 1 study. The observed proportion and the calculated proportion are in fact within 2% of each other for eight of the ten most frequently performed nursing tasks. The most significant difference was between the two proportions for coordinating prescriptions; while the value calculated from the nurse data was 7%, the observed value was 12%. This difference can be attributed to one nurse spending a full hour setting up a narcotics contract, which we were told is atypical. Since we observed for 24 hours total, that one hour represents approximately 4% on the chart above, which accounts for most of the difference. To determine the level of variation within subgroups, which is another measure of validity, we also analyzed the observations by individual department. It is important to note that 7 of the observations were conducted on internal medicine (IM) nurses, and thus the combined nurse data mostly reflects the numbers calculated for IM. Thus, when comparing the subgroups, it makes sense to compare them to each other, as opposed to comparing them to the all RN data. The data shows that the most commonly performed tasks for all nurses overall (on the phone, documenting, and looking up information) also tended to be the most common tasks for each department, with the approximate proportion of time spent being the same. There are, however, a few notable differences in the OB/GYN and Specialty departments. While IM nurses spent 11% of their time looking up patient information and about coordinating for patients and prescriptions, the OB/GYN nurse spent 23% of the time looking up patient information and only 3% coordinating for patients and prescriptions. 16
Phase 3 Findings Phone Log Study The third phase of data collection focused on logging nurse phone activity throughout the day. From this study, quantitative measures on call length can be obtained for overall nursing and by department. The phone log data collected also provides a methodology to determine which nursing tasks are commonly associated with calls. Figure 6 below illustrates the distribution of call length by nursing department from the Phase 3 data collected. The sample sizes ranged between departments due to variability in the actual number of call each sample represents one call. 30 Boxplot of Call Length by Nursing Department 25 20 Data 15 10 5 0 Internal Medicine Multi OBGYN Pediatrics Total (No specialty) Specialty Figure 6. Box plot of Call Length by Nursing Department Table 3 below shows descriptive statistics for call length by nursing department. As per the client s request, the total (across departments) does not include the specialty department. From the box plot above and the table below, the median call length for internal medicine and multi was 4.0 minutes with corresponding means of 4.4 and 5.0 minutes. The pediatric department s mean call length of 3.6 minutes and median call length of 3.0 minutes were slightly lower than the previous two departments, yet close in range. Mean call length for the OBGYN department was 9.5 minutes. While a high mean for OBGYN calls imply that calls in this department take longer time, the call data collected for this department was noticed (5, 10, 15 min increments.) For all individual departments (besides OBGYN) and the total, the mean call length was greater than the median which provides evidence of a non-normal, skewed distribution. The long tail on the distributions suggests the existence of a few, significantly longer calls than average. As expected, specialty calls were notably longer (6.9 minutes) than the sum across all remaining 17
departments (5.4 minutes) due to the more in-depth knowledge and explanations necessary when dealing with these types of patients. Table 3. Descriptive Statistics for Call Length by Nursing Department Department Internal Medicine Multi OBGYN Pediatrics Specialty Total (No specialty) Specialty Mean Minimum Maximum Median N 4.4 1.0 25.0 4.0 434 5.0 1.0 20.0 4.0 124 9.5 1.0 30.0 10.0 207 3.6 1.0 13.0 3.0 230 6.9 1.0 18.0 6.0 81 5.4 1.0 30.0 4.0 995 6.9 1.0 18.0 6.0 81 Figure 7 below provides a graphical display of the distribution of call lengths by nursing departments. The most notable deviation from the overall mean call length of 5.4 minutes is the OB/GYN department, which has a mean call length of about 9.5 minutes. It should be noted that the actual distributions of these call lengths is skewed right rather than simply a normal distribution. This means that most calls are less than 5 or 6 minutes in length, but there are a select few calls that are quite long and increase the mean call length. Frequency 160 120 80 40 0 80 60 40 20 0 9.0 4.5 0.0 Histogram of Call Length by Department Normal Internal Medicine Multi OBGYN 40 60 13.5 18.0 22.5 0.0 2.4 4.8 7.2 9.6 12.0 30 20 10 0-3 -4.8 0 3 0.0 4.8 6 9.6 18 15 12 9 14.4 19.2 24.0 28.8 45 30 15 0 0 6 12 18 24 30 Pediatrics Total (No specialty ) Specialty 20 240 15 180 120 60 0 10 5 0 0 4 8 12 16 Figure 7. Histogram of call length by nursing department Internal Medicine Mean 4.440 StDev 3.557 N 434 Multi Mean 4.960 StDev 3.365 N 124 OBGYN Mean 9.464 StDev 6.084 N 207 Pediatrics Mean 3.591 StDev 2.043 N 230 Total (No specialty) Mean 5.354 StDev 4.490 N 995 Specialty Mean 6.938 StDev 4.069 N 81 While the mean call length data for different departments is useful, a more appropriate measure for call length is to estimate a confidence interval on the mean. In order to calculate a 9 confidence interval on the mean, the call length standard deviation for each department was 18
determined. Using the mean, standard deviation, and normal distribution, 9 confidence intervals were constructed, as seen in Table 4 below. Despite the lack of normality for call length data collected, possibly due to small sample size, it is necessary to assume a distribution in order to calculate a confidence interval. Table 4. Confidence Intervals for Call Length by Nursing Department Department Mean StDev SE Mean 9 CI Internal Medicine 4.4 3.56 0.17 4.1 4.8 Multi 5.0 3.37 0.30 4.4 5.6 OBGYN 9.5 6.08 0.42 8.6 10.3 Pediatrics 3.6 2.04 0.13 3.3 3.9 Total (No specialty) 5.4 4.49 0.14 5.1 5.6 Specialty 6.9 4.07 0.45 6.0 7.8 From the proportion of time spent on phone calls, as found in Phase 1, along with the estimated call length from above, a mean and confidence interval can be projected for the number of calls that different nursing departments can handle per day, as seen in Table 5 below. The mean number of calls that each nurse (excluding specialty) can handle per day is 30.5 with a 9 confidence interval of 29.0 to 32.2 calls. The mean number of calls that each specialty nurse can handle per day is 15.2 with a 9 confidence interval of 13.5 to 17.5 calls. The average actual number of calls handled per day, as found in Phase 3, was 36.9 calls for all nurses (excluding specialty) and 16.2 calls for specialty. While these data points exhibit a trend of overworked nurses, many factors could account for the disparity between estimated and actual nursing calls. Table 5. Projected Nursing Calls Based on Time Proportion Data Time on Phone Number of Calls % Day Hours Minutes Mean 9 CI Actual Internal Medicine 0.29 2.3 139.2 31.4 29.1 33.9 33.4 Multi 0.31 2.5 148.8 30.0 26.8 34.1 31.0 OBGYN 0.52 4.2 249.6 26.4 24.2 28.9 41.4 Pediatrics 0.21 1.7 100.8 28.1 26.1 30.3 46.0 Total (No specialty) 0.34 2.7 163.2 30.5 29.0 32.2 36.9 Specialty 0.22 1.8 105.6 15.2 13.5 17.5 16.2 Conclusions As stated in the goals and objectives of the project proposal, the end result of this project is for the client to be able to: Determine appropriate staffing levels for nursing units Develop performance benchmarks for nursing workloads Develop time standards for nursing phone calls Identify areas for improvement through appropriate time allocation Determine a methodology for ongoing measurement on units 19
Throughout the course of the project, our student team has developed several recommendations to help the client act upon our findings. These recommendations are organized in the same structure as the above goals. First, to determine appropriate staffing levels for nursing units, we first analyzed the overall workloads of each specific nursing unit. While it is outside the scope of our project to determine the actual staffing levels, the director of nursing can use our analysis to more appropriately determine these levels. Based on our projections from phase 1 and 3 of data analysis, the nurses can handle 30.5 calls per day. While this figure is merely a projection, it can be used as a benchmark to determine any serious deviations or problems. Second, to develop performance benchmarks for nursing workloads, the director of nursing can use the same projection of 30.5 calls per day. If nurses can handle 30.5 calls per day, efforts should be made to ensure that the actual calls received do not vary much from this target value. Third, to develop time standards for nursing phone calls, we analyzed the average call times as found from the phase 3 phone logs. For all departments (excluding specialty), the mean call time is 5.4 minutes and the median is 4 minutes. This shows that there are a relatively small proportion of calls that consume a high amount of time. In addition, the mean call time for the specialty department is 6.9 minutes and the median is 6 minutes. The director of nursing can use these values to estimate time standards for nursing phone calls. Fourth, to identify areas for improvement through appropriate time allocation, the director of nursing can take measures to record data from nurses and departments that vary far from the target values we calculated. By implementing a robust time tracking system, any deviations can be spotted well before any serious problems develop. This proactive measure will ensure efficient productivity of nurses while improving the job satisfaction of all nurses. Fifth, to determine a methodology for ongoing measurement on units, we reflected on our experiences during the three weeks of data collection and observation. In order to ease the data collection process, and make it possible to extend the data collection format to other primary care units in the UMHS, it would be best to develop an electronic data collection tool. This would eliminate the need to transcribe all results from paper to electronic format, and would also make the collection process less intrusive to nurses. While our team can develop a preliminary data collection model, it would be worthwhile for the director of nursing to contact a software development team to build a robust and efficient electronic collection tool. In addition, based on our experiences, we recommend two additional measures to ensure that ongoing measurement of nursing workloads runs smoothly in the future. First, nurses should ensure to keep beepers on them at all times rather than on their desk. This will ensure that nurses do not miss a beep when they leave their desk to perform other tasks. Second, nurses should be provided with a stopwatch or timer to improve accuracy for the phone log study. Nurses expressed that there was some approximation involved in estimating the length of each call. The more accurately we can collect this data, more accurate time standards can be developed. 20
Appendix A. Data Collection Tools for Project Phases 1-3 21
Part 1: General Study Primary Care Nursing Workload Study Staff: Department: Date: Instructions: When the random reminder beeper sounds, please place a check mark in the box corresponding to the task performed at the time of the beep, under the time that the beep occurred. Also, please note start, lunch, and end times below. Nursing Tasks: Patient Care: Assisting with provider visit Performing nurse visit Caring for walk-ins On phone with patient/ family Indirect Patient Care Documenting (electronically or on paper) Delegating tasks/ supervising Coordinating for patients Consulting with physicians Looking up patient and care information Coordinating prescriptions (pre/post patient visit) Transcribing from voicemail Non-Patient Performing administrative responsibilities Participating in quality improvement (QI) projects Other meetings Staff education/competency Participating in self professional development Taking personal time 7:00-8:00 8:00-9:00 9:00-10:00 10:00-11:00 11:00-12:00 12:00-1:00 1:00-2:00 2:00-3:00 3:00-4:00 4:00-5:00 5:00-6:00 Start Time: Lunch: - End Time: Source: Program Operations Analysis, UMHS 22
Part 2: Phone Study Primary Care Nursing Workload Study Staff: Department: Date: Instructions: When the random reminder beeper sounds, please place a check mark in the box corresponding to the task performed at the time of the beep, under the time that the beep occurred. Also, please note start, lunch, and end times below. Nursing Tasks: On Phone Activities In process of talking to patient/ family Waiting for next patient call On hold Talking to provider (on phone) Talking to another internal department (on phone) Leaving a message for pt/family Talking to pharmacy or DME Talking to external agency Transcribing from voicemail Work related to call Documenting (electronically or on paper) Faxing related to the call After call coordinating for patients (non-phone) Consulting with providers (non-phone) Looking up patient and care information Non-Call-Related Performing administrative responsibilities In-clinic care Supervising/delegating/precepting/orienting Taking personal time/self development 7:00-8:00 8:00-9:00 9:00-10:00 10:00-11:00 11:00-12:00 12:00-1:00 Start Time: Lunch: - End Time: Source: Program Operations Analysis, UMHS 23 1:00-2:00 2:00-3:00 3:00-4:00 4:00-5:00 5:00-6:00
Part 3: Call Log Primary Care Nursing Workload Study Staff: Department: Date: Instructions: For each incoming or outgoing call, please place check marks in the boxes corresponding to the task associated with the given call. If more than 20 calls are received during the day, please continue data collection on another form. Nursing Tasks: #1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 #13 #14 #15 #16 #17 #18 #19 On Phone Activities In process of talking to patient/ family On hold Talking to provider (on phone) Talking to another internal department (on phone) Leaving a message for patientt/family Talking to pharmacy or DME Talking to external agency Transcribing from voicemail Work related to call Documenting (electronically or on paper) Faxing related to the call Coordinating for patients Consulting with providers Looking up patient and care information Other Estimated Call Length (min) #20 Number of messages upon arrival in the morning: Number of messages left unanswered before departure: Start Time: Lunch: - End Time: 24