University of Michigan Health System Programs and Operations Analysis. Order Entry Clerical Process Analysis Final Report

University of Michigan Health System Programs and Operations Analysis Order Entry Clerical Process Analysis Final Report To: Richard J. Coffey: Director, Programs and Operations Analysis Bruce Chaffee: Clinical Pharmacist, Pharmacy Services From: IOE 481 Project Team, Programs and Operations Analysis Lindsey Beauchamp John Frank Michael Miller Shireen Palsson Date: December 15, 2004

1 TABLE OF CONTENTS Executive Summary 2 Overview 2 Goals 2 Approach 2 Findings 2 Recommendations and Implementation 3 Introduction 4 Project Goals and Objectives 4 Key Issues 4 Project Scope 4 Background 5 Current Study 5 University of Michigan Hospital Study 5 C.S. Mott Children s Hospital Study 5 Industry Trends 6 Approach 6 Collected Data 6 Analyzed Data 6 Findings 7 Clerical Interview Findings 7 Clerical Time Study Findings 7 Projects for Monthly and Yearly Time Savings 12 Conclusions 14 Project Limitations 14 Recommendations for Future Analysis 14 Appendix A: Clerical Interview Findings 16 Appendix B: Process Flowcharts 17 Appendix C: Order Form Samples 22

2 EXECUTIVE SUMMARY Overview This project is a micro-level analysis of the clerical order entry process. The purpose was to determine the profile of work and time requirements for unit clerk order entry. The University of Michigan Hospital and C.S. Mott Children's Hospital plan to implement an Order Management Process (OMP) that will significantly change the unit clerk's job. The Order Management Process will lead to an electronic order entry process as opposed to the current paper order entry system. This OMP is the result of problems with accuracy and legibility of orders and is projected to improve both of these areas as well as decrease order-processing times. Data was collected via clerical interviews and time studies in order to evaluate the expected impact of this implementation. Goals The following goals were associated with this project: Document how much clerical time is spent processing orders. Determine the time commitment required for each type of order. Determine areas for future order entry analysis. Approach The project was conducted in three phases: data collection, data analysis, and data presentation. During the data collection phase, the team observed the inpatient clerical staff in both the University Hospital and Mott Children's Hospital. We conducted interviews with float clerks and flowcharted the order entry process for each of the different order types (Blood, Respiratory, Diet, Medicine, Nursing, Patient Equipment, Materials Services (MSC), Diagnostic and Testing, and Laboratory). We conducted time studies on each of these order types and collected processing times for each of the ten different order types. Additionally, we collected service code information for each of the orders that were processed. In the data analysis phase, results from the data collected provided information regarding the frequency of each order type along with service code information. The work-study data was investigated to determine the average processing times for each order type. Findings Data was collected on frequency and average processing times for each of the order types, shown below in Table 1. Table 1: Time Study Data - Summary by Type Order Type Frequency Average Time St.Dev. Admissions 36 0:04:02 0:02:26 Blood 4 0:01:37 0:00:14 Diet 63 0:02:04 0:01:36 D&T 9 0:01:58 0:02:28 Lab 104 0:01:41 0:01:01 Medication 309 0:00:22 0:00:10 MSC 4 0:02:36 0:02:03 Nursing 143 0:00:37 0:00:33 PT Equip 4 0:01:48 0:00:26 Respiratory 22 0:01:42 0:01:52

3 The most frequent order types were found to be medicine and nursing orders. The order type with the longest average processing time was found to be admission and medicine was found to have the shortest average time. We also collected service code, shift, and unit information for each order processed by the unit clerks. This information was compared to the frequency of the respective order types. Recommendations and Implementation After analysis of clerical time study data, order entry flow charts, order frequency data, and clerical interview findings, we recommend the following to facilitate the implementation of the Order Management Process. Further Study Low Volume Orders Our study was unable to collect adequate sample sizes for blood, diagnostic and testing, MSC, and patient equipment orders. We therefore recommend further analysis of these order types. This additional data will add to the robustness of a narrower base line study on these specific order types. This information could be gathered from the respective departments' billing services and tracked to specified units to find floors with high frequencies of these order types. Further Study Faxing Floor Orders A pilot program to fax pharmacy (medication) orders is currently in place on units 6B and 6C of the University Hospital. We were unable to collect an adequate amount of orders in these units due to time and scheduling constraints. We therefore recommend future studies and analysis in addition to our preliminary findings. Although data from the faxing units was gathered during our project, a more comprehensive study should be conducted in order to compare the differences and similarities between the current faxing order entry process to that of the future Order Management Process. This data will be helpful in providing a baseline comparison of the Order Management Process and the pilot faxing program.

4 INTRODUCTION The University of Michigan Health Systems (UMHS) unit clerks currently process handwritten prescriber orders and either input them into ancillary systems or write the orders onto ancillary departmental requisitions. This process is inefficient due to problems in completeness and legibility of the handwritten forms and the duplicate effort required by the unit clerk. Therefore, University of Michigan Health Systems is implementing the Order Management Process (OMP), which comprises prescriber order entry with nursing work lists and medication administration documentation. We were asked to perform a study on the current clerical order entry process for comparison with the Order Management Process. The purpose of this project was to determine the profile of work and the time requirements that unit clerks have in processing and entering prescriber orders. Project Goals and Objectives The goals and objectives of this project changed during the study period due to time limitations and to the developing scope of the project. The final goals and objectives were to: Document how much clerical time is spent processing orders. Determine the time commitment required for each type of order. Determine areas for future order entry analysis. This project did not include developing specific recommendations for process improvement, although recommendations for further studies were noted. Information taken from the current order process will be used for comparison by UMHS for the future implementation of the Order Management Process. Key Issues The following key issues were the driving need for this project: Time spent processing orders are unacceptable within current order entry system. A new data processing system (Order Management Process) is being implemented throughout the University of Michigan Health System. Project Scope As originally proposed, this project includes a micro-level analysis of all inpatient prescriber orders. The most frequent and time-consuming orders were time studied. These orders were selected after interviewing the Patient Unit Services Supervisor and the Inpatient Unit Clerk Trainer, and are as follows: Dietary Respiratory Laboratory Patient Equipment Diagnostic and Testing Nursing Admissions MSC Medications Blood The prescriber order entry process begins when the order is filled out and ends when the clerk inputs the data. The scope of this project included studying various units, based on the frequency of orders processed and availability of participating clerical staff. Five float clerks volunteered for the study, one of which was also the Inpatient Unit Clerk Trainer. The expected data collected was set at roughly ¾ from the University of Michigan Hospital and ¼ from C.S. Mott Children s Hospital. Any prescriber orders dealing with outpatient care were not studied in this project.

5 BACKGROUND Current Study This project is a micro-level analysis of the clerical order entry process. The purpose is to determine the profile of work and time requirements for unit clerk order entry. The Order Management Process (OMP) will significantly change the unit clerk s job. Implementation is scheduled for early 2006. The hope is that accuracy and legibility will be improved, as well as time savings for order processing. Also, any discipline will have the ability to access and use patient information concurrently, which is impossible with the current system. Data is needed to evaluate the expected impact of this change. This data was collected through clerical interviews and time studies to capture process from beginning to end, as well as information on order type, volume, service, etc. A literature search was performed for material related to this study. Two prior studies have been conducted regarding the manual order entry process on a macroscopic level. These were done using beepers to time the different processes of inpatient unit clerks at the University of Michigan Hospital and C.S. Mott Children s Hospital. These studies correlate with ours in that they recommended the implementation of a computerized order entry system for the unit clerks. The implementation of this computerized order entry system drives our study. University of Michigan Hospital Study The study which was limited to units 5A and 7C of the University of Michigan Hospital found that 31% of a unit clerk s time was spent processing flagged physician orders, with another 30% spent on admissions, transfers, and discharges. This study was limited to two units, 5A and 7C. The major recommendation made to reduce the time spent on admission, physician, and supply orders was to automate the process. The computer automation is being put into place by the Order Management Process (OMP) scheduled for implementation for early 2006. Another major area of concern expressed through clerical interviews by 60% of the interviewees was the duration to process physician s orders due to illegibility. This was examined in our clerical interviews and determined to be irrelevant for our study, as discussed later in this report. Further study detail is available in the following document: University of Michigan Hospital: Analysis of Inpatient Unit Clerk Workload. Kimika Edwards, Pramod Karkarala, Kerry Raschke, Priya Sehgal. April 15, 2004. C.S. Mott Children s Hospital Study The study conducted within C.S. Mott Children s Hospital found that the specific tasks performed most often were Receptionist, Processing Physician s Orders, and Paging. The study identified all inefficiencies associated with these tasks and determined the impact on the overall workload of the unit clerks. The study was limited to floors 6 and 7 in Mott and 5E and 5W in the University Hospital. The study provided several recommendations to improve the ineffective unit clerk tasks. These recommendations included designating a receptionist for the unit clerks, allocating tasks between clerks and standardizing the layout of the inpatient unit clerk station.

Further study detail is available in the following document: C.S. Mott Children s Hospital: Analysis of Inpatient Unit Clerk Workload. Dorothy Leung, Patrick Rich, David Risely, Abbie Schultz. April 12, 2004. Industry Trends A search was conducted for related industry trends outside the University of Michigan Health System as well. Currently hospitals across the nation are implementing electronic medical records to prevent illegibility, inaccuracy, inaccessibility and incompleteness. 6 Montefiore Medical Center in New York is one of the many hospitals to implement a computerized program. Montefiore has successfully implemented the electronic medical record and 100% computerized physician order entry in three acute care facilities totaling approximately 1,100 beds. However, not all hospitals found the paperless system to be completely successful. Cedars-Sinai Medical Center in Los Angeles turned off its computerized physician order entry system in 2003, after hundreds of physicians complained that rather than speeding up and improving patient care, it actually slowed down the process of filling their orders. APPROACH This project was completed in three phases: data collection, data analysis, and data presentation. Collected Data To collect data, the following steps were followed. We: Gathered all existing information from related projects on order entries. Interviewed the Inpatient Unit Clerk(s), Management Engineer, Clinical Pharmacist and Patient Unit Services Supervisor in order to coordinate efforts and information from previous projects. Flowcharted steps of the prescriber order entry process by order type. Conducted time studies of inpatient unit clerks to gather prescriber order entry data. Utilized time study data to illustrate the length of the prescriber order entry process, the distribution of prescriber orders, and the general order entry flow. Analyzed Data To properly analyze the data, the following steps were taken. We: Developed and analyzed a flowchart of the prescriber order entry process. Analyzed prior studies pertaining to order entry. Determined criterion for prescriber order completion. Determined whether further data collection and analysis will be necessary.

7 FINDINGS Clerical Interview Findings Clerical interviews of the float clerk trainer and four float clerks were conducted to clearly document the order entry process. Their responses were used to determine the types of orders processed and the manner in which they would be timed, which were: Dietary Respiratory Laboratory Patient Equipment Diagnostic and Testing Nursing Admissions MSC Medication Blood There was consistency in the clerks responses with regards to the University Hospital order system and that at Mott Children s Hospital. The clerks agreed that the UH discharge and dietary orders, specifically tube feeding diets, took the longest to process due to the quantity of steps associated with each of these types (i.e. database input for dietary orders). Additionally, the clerks expressed time issues with lab and dietary orders at Mott. These timing issues were due to researching correct lab requisitions and inputting formula requests for infant dietary orders. A complete list of questions and responses can be found in Appendix A. Clerical Time Study Findings Clerical time studies were conducted to document the current time taken by unit clerks to process each type of order. The data collected comprised of 473 orders from the AM shift and 228 orders from the PM shift. There were 556 orders collected at the University Hospital and 120 orders taken from Mott Children s Hospital. Sample order forms can be found in Appendix C. A summary of the time study data by type is shown below in Table 2 broken down by the frequency of orders observed, average time, and standard deviation of time. Table 2: Time Study Data - Summary by Type Order Type Frequency Average Time St.Dev. Admissions 36 0:04:02 0:02:26 Blood 4 0:01:37 0:00:14 Diet 63 0:02:04 0:01:36 D&T 9 0:01:58 0:02:28 Lab 104 0:01:41 0:01:01 Medication 309 0:00:22 0:00:10 MSC 4 0:02:36 0:02:03 Nursing 143 0:00:37 0:00:33 PT Equip 4 0:01:48 0:00:26 Respiratory 22 0:01:42 0:01:52 The most frequent orders were medication orders (309), nursing orders (143) and lab orders (104). The most infrequent order types were patient equipment, MSC, blood, and diagnostic and testing orders. Figures 1 and 2 below represent the frequency of each of the order types as a percentage of total orders processed and average time per order, respectively.

8 Lab 15% Diet 9% Admissions 5% Nursing 20% Respiratory 3% D&T 1% Blood 1% MSC 1% PT Equip 1% Medication 44% Figure 1. Frequency by Order Type 0:04:19 0:03:36 0:02:53 Average Time 0:02:10 0:01:26 0:00:43 0:00:00 Admissions MSC Diet D&T PT Equip 1 Respiratory Lab Blood Nursing Medication Order Type Figure 2. Average Time by Order Type

9 The longest average time was found to be 4 minutes 2 seconds and was associated with admissions orders. This was due to the length of time it took the clerks to process a patient discharge. The shortest average time was found to be the medication orders (22 seconds), which were also the most frequently processed orders for the unit clerks. Service Code Breakdown Service codes were also collected for the processed orders observed. The service code relates to the specific treatment a patient is receiving. We observed 24 different service codes and placed them into one of two categories, medicine or surgery. By separating the medicine and surgery orders we were able to see if there was a difference in average time of an order or if the percentage breakdown of order types was significantly different. A summary of the data by service code is shown in Tables 3 and 4 below. Table 3. Percentage of Orders for Med/Surgery Service Codes Order Type Meds %Meds Surgery %Surgery Admissions 6 4% 17 5% Blood 3 2% 1 0.50% Dietary 4 3% 27 8% D&T 6 4% 3 1% Labs 31 20% 31 9% Medication 75 48% 155 46% MSC 1 1% 2 1% Nursing 26 17% 90 26% PT Equip 0 0% 4 1% Respiratory 4 3% 11 3% Total 156 100% 341 100% Table 3 consists of a breakdown of how many times different orders occurred in each service type. In an attempt to compare the two different service types the percentage make-up was placed next to the actual number of orders that occurred. When looking at the percentage breakdown of order types there does not seem to be much of a difference between medicine or surgery orders. The only major difference seems to be in lab and nursing orders. Lab orders make up about 20% of medicine orders whereas for surgery orders labs only make up about 9%. The difference in the labs (about 11%) is mostly explained by a similar difference in nursing orders (about 9%). Nursing orders make up about 17% of medicine orders and 26% of surgery orders. The difference in the percentage breakdown of the nursing orders could be explained by the fact that surgery patients need more services to be completed by the nurses.

10 Table 4: Order Processing Times for Med/Surgery Service Codes Order Type Meds Med Std Surgery Surgery Std Admissions 4:22 3:40 1:33 1:22 Blood 1:44 0:01 1:16 n/a Dietary 2:08 2:11 2:17 1:44 D&T 2:13 3:05 1:26 0:06 Labs 1:32 0:56 1:43 0:41 MSC 1:40 n/a 1:33 0:45 Respiratory 3:22 3:49 1:20 1:11 Table 4 shows the average time each order type takes for both medicine and surgery services. The only major differences appear to be for admissions, D&T and respiratory orders. All of these differences could be explained by the fact that for all three of these order types one of the service types, medicine or surgery, had such a small sample of data that the average time derived does not reflect the true average time. The largest discrepancy seen is between admission orders. It must have been the sampling that explains this difference as it does not make sense that an admission order in a medicine unit would vastly differ from an admission order in a surgery unit. It is also evident that table lacks three order types; medicine, nursing and patient equipment orders. When we started collecting service code information medicine orders and nursing orders had already been found as they are fairly standard in all units. The explanation for the nonexistence of an average time for patient equipment orders is due to the fact that this order was never observed in correlation with a medicine service code. University Hospital vs. Mott Children s Hospital Breakdown The data was broken down by Mott Children s Hospital orders and University Hospital orders. Tables 5 and 6 below show the frequency percentages, average times, and standard deviations for the different order types. Table 5: Mott Order Data Order Type Frequency % of Total Avg. Time St. Dev. Admissions 10 3% 0:02:57 0:01:29 Blood 1 1% 0:01:45 n/a Diet 13 10% 0:02:06 0:01:41 D&T 4 3% 0:02:46 0:03:49 Laboratory 11 8% 0:02:58 0:02:22 Medication 61 45% 0:00:30 0:00:16 MSC 1 1% 0:01:40 n/a Nursing 30 22% 0:01:25 0:01:14 PT Equipment n/a n/a n/a n/a Respiratory 4 3% 0:03:23 0:03:49 TOTAL 135 100%

11 Table 6: UH Order Data Order Type Frequency % of Total Avg. Time St. Dev. Admissions 21 4% 0:04:48 0:02:45 Blood 3 1% 0:01:35 0:00:16 Diet 39 8% 0:02:08 0:01:43 D&T 5 1% 0:01:19 0:00:19 Laboratory 85 17% 0:01:35 0:00:57 Medication 224 44% 0:00:25 0:00:15 MSC 3 1% 0:02:55 0:02:24 Nursing 105 21% 0:00:51 0:01:04 PT Equipment 4 1% 0:01:48 0:00:26 Respiratory 17 3% 0:01:19 0:00:59 TOTAL 325 100% The frequency differences between the orders are shown by percentage. The frequencies of the majority of order types were comparable for Mott and UH orders. The laboratory orders at Mott made up approximately 8% of the total collected orders and 17% of the total UH orders. These discrepancies can be attributed to the smaller sampling size of Mott orders (135 Mott orders versus 325 UH orders). The main differences in average times between the two hospitals orders were between the admissions and respiratory orders. The respiratory orders at Mott were larger by almost two minutes than the UH respiratory orders. The additional time is attributed to the Mott procedure of paging the Respiratory Therapist per respiratory order. This is not part of the order entry procedure at the University Hospital. There were no patient equipment orders collected at Mott Children s Hospital; therefore, no University Hospital comparison was conducted. Due to the limited sampling of orders in this project, some of the standard deviations were larger than the average order times. More data collection is necessary to correct this problem. AM vs. PM Shift Breakdown The data was broken down by AM and PM shifts. Tables 7 and 8 below show the frequency and percentage of totals for each. Table 7: AM Shift Data Order Type Frequency % of Total Avg. Time St. Dev. Admissions 25 5% 0:03:28 0:01:32 Blood 2 0% 0:01:31 0:00:21 Diet 51 11% 0:02:00 0:01:34 D&T 5 1% 0:01:08 0:00:22 Laboratory 80 17% 0:01:42 0:01:02 Medication 197 42% 0:00:22 0:00:11 MSC 3 1% 0:02:55 0:02:24 Nursing 94 20% 0:00:34 0:00:34 PT Equipment 1 0% 0:02:05 0:00:00 Respiratory 12 3% 0:01:36 0:01:04 TOTAL 470 100%

12 Table 8: PM Shift Data Order Type Frequency % of Total Avg. Time St. Dev. Admissions 11 5% 0:05:21 0:03:31 Blood 2 1% 0:01:44 0:00:00 Diet 12 5% 0:02:19 0:01:48 D&T 4 2% 0:03:00 0:03:40 Laboratory 24 11% 0:01:38 0:00:58 Medication 112 49% 0:00:22 0:00:00 MSC 1 0% 0:01:40 0:00:00 Nursing 49 21% 0:00:56 0:00:22 PT Equipment 3 1% 0:01:42 0:00:28 Respiratory 10 4% 0:01:50 0:02:35 TOTAL 228 100% From this data, the clear differences come from the % of Total column. The largest discrepancy is that Diet orders make up 11% of AM orders and only 5% of PM orders. This is expected, as many physicians write new diet orders for their patients in the mornings for the day. There are also significantly more laboratory orders in the AM (17%) than in the PM (11%), which can be explained due to labs for the day being ordered in the morning. Average time differences for orders with high enough volumes to measure (medication, laboratory, nursing) were not larger than a few seconds between AM and PM, and standard deviations were greatly increased when AM and PM were broken down due to lower frequencies for each. Nursing orders seem to take longer in the afternoon, for which the cause is unknown. Since a standard of 36 seconds was set early on in the project, data collected later from PM shifts where nursing orders were not timed may have skewed this average. Further data collection would be necessary to determine the reliability of this claim. Other percentage differences in less frequent orders also do not yield clear conclusions due to the low numbers collected. Projects for Monthly and Yearly Time Savings Estimates were taken a few years ago for total monthly order volumes. Inpatient pharmacy (medication) orders were estimated at 5,843 daily. Our client informed us that this number was at least double the actual amount of medication orders as pharmacy orders include handwritten orders from physicians and an assumption that active orders at discharge will be individually discontinued by the physician. Currently, these orders are automatically discontinued without the prescriber ever writing the order, or they may be handled by one order ("D/C meds"). As a result of this assumption, in the future state there will probably be an automated process to stop the orders or they may be handled by a single "D/C all" order. Therefore, the estimated daily number of medication orders was reduced by 50% to 2922. We used this estimation, along with the knowledge that approximately 44% of total orders are medications, to approximate the monthly and yearly volume estimates of all other order types. These estimates are shown below in Table 9.

13 Table 9: Monthly and Yearly Order Volume Estimates Order Type % of Total Orders Estimated Monthly Volume Estimated Yearly Volume Admissions 5.16% 10,215 122,580 Blood 0.57% 1140 13,680 Diet 9.03% 17,865 214,380 D&T 1.29% 2550 30,600 Lab 14.90% 29,505 354,060 Medication 44.27% 87,645 1,051,740 MSC 0.57% 1140 13,680 Nursing 20.49% 40,560 486,720 PT Equip 0.57% 1140 13,680 Respiratory 3.15% 6240 74,880 From this data, a monthly and yearly time savings was found, shown below in Table 10. Table 10: Monthly and Yearly Time Savings Estimates Order Type Average Time Estimated Monthly Savings Estimated Yearly Savings Admissions 0:04:02 686:40:30 8240:06:00 Blood 0:01:37 30:43:00 368:36:00 Diet 0:02:04 615:21:00 7384:12:00 D&T 0:01:58 83:35:00 1003:00:00 Lab 0:01:41 827:46:45 9933:21:00 Medication 0:00:22 535:36:30 6427:18:00 MSC 0:02:36 49:24:00 592:48:00 Nursing 0:00:37 416:52:00 5002:24:00 PT Equip 0:01:48 34:12:00 410:24:00 Respiratory 0:01:42 176:48:00 2121:36:00 TOTAL 3456:58:45 41483:45:00 The most time will be saved from laboratory and admission orders, 9933 and 8240 hours annually respectively. Also, over 41,000 estimated total hours yearly (over 3450 monthly) in total will be saved by the implementation of the Orders Management Process through eliminating clerical order writing. This equates to approximately 20 paid full-time equivalents (FTE s) saved.

14 CONCLUSIONS Project Limitations Despite the various strengths within this study, there were also various limitations that need to be mentioned. These weaknesses can be attributed to the data and methods used in the study. Initial time study data obtained for this study posed problems as there were discrepancies in the method of timing the unit clerks. This initial data was omitted from our analysis because it was not consistent with the standard methodology of the rest of the collected data. The test participants were four experienced float clerks that worked on various units throughout the hospital. Because we were only studying these four clerks, our study was limited to the units that they worked in both Mott Children s Hospital and the University Hospital. The experience of the clerks was very consistent. Each of the test participants had at least three years of unit clerk experience. This limited the range and variety of experience for the timing of the order completion. Clerks that volunteered for this project were seasoned. Therefore, clerks that are not as experienced may be slower at processing orders. Upon completion, our study showed low frequency of some order types (i.e. patient equipment, respiratory, diagnostic and testing). The low volume could be attributed to the limited units that we were studied, but may also demonstrate the actuality that these orders are low frequency for inpatient order entry. Recommendations for Future Analysis Through analysis of clerical time studies, order entry flow charts, order frequency data, and clerical interviews, the following recommendations were developed to aid in the future implementation of the Order Management Process. Further Study Low Volume Orders Our study was unable to collect adequate sample sizes for blood, diagnostic and testing, MSC, and patient equipment orders. We therefore recommend further analysis of these order types. The cause for the infrequency of these order types is inconclusive; however, a study much similar to this could be implemented on floors with high frequencies of these orders to collect statistically significant data on these order types. This additional data will add to the robustness of a narrower base line study on these specific order types. This information could be gathered from the respective departments billing services and tracked to specified units to find floors with high frequencies of these order types. Further Study Faxing Floor Orders A pilot program to fax pharmacy orders is currently in place on units 6B and 6C of the University Hospital. We were asked to collect timing data from these units for further comparison with the current process and the future Orders Management Process. We were unable to collect an adequate amount of orders due to time and scheduling constraints. We therefore recommend future areas of study and analysis in addition to our preliminary findings, shown below in Table 11.

15 Table 11: Faxing Unit Average Times Order Type Average Time Admission 0:05:43 Blood 0:01:28 Diet 0:01:59 D&T 0:02:02 Lab 0:01:39 Medication 0:00:36 Nursing 0:00:36 PT Equip 0:01:55 Respiratory 0:00:42 From this preliminary data no clear conclusions can be made, although we believe the only order processing time that will be truly affected is medication, as it is the only order required by fax to the pharmacy. Therefore, the time for orders processed along with a medication order may increase slightly as well. As the volume of orders collected is not statistically significant for any of the order types, a more thorough study on these particular units is recommended to provide more conclusive data for future comparison to OMP implementation.

16 APPENDIX A: CLERICAL INTERVIEW FINDINGS 1. What types of orders do you process? Dietary Respiratory Laboratory Patient Equipment Diagnostic and Testing Nursing Admissions MSC Medication 2. Which types of orders are processed with greatest frequency? Medication 3. Which types of orders take the longest time to process? Admissions (UH Discharge specifically) Dietary (UH Tube feed specifically) Labs (MOTT Researching requisitions) Dietary (MOTT Adding correct formula per patient) 4. What do you do with those orders? See Figures 1-9, Appendix B (Process Flowcharts) 5. What defines a complete order? All steps completed from Figures 1-9, Appendix B 6. Which orders will be affected most by the OMP implementation? Those requiring longest processing time. 7. Who is responsible for input? (clerk vs. prescriber) Clerk responsible for duties in Figures 1-9, Appendix B. Nursing staff and prescriber responsibilities vary by order type. 8. How much time spent clarifying order and entering order/completing form? Time study conducted

17 APPENDIX B: PROCESS FLOWCHARTS Pull Chart Note (date/time/initials) Enter diet into computer (Order Entry System) Pink Copy Pharmacy Hard Copy Nurse Replace Chart Figure 1: Diet Orders Pull Chart Note (date/time/initials) Page RT Enter into computer under Respiratory Care Order Management System Pink Copy Pharmacy Hard Copy Nurse Replace Chart Figure 2: Respiratory Orders

18 Pull Chart Note (date/time/initials) Pull lab requisitions Stamp requisition with CPI Card Fill out lab requisition STAT? Page vehipuncture/ nurse (if STAT) Put STAT sticker on form (if STAT) Place requisition in box for venipuncture/nurse Pink Copy Pharmacy Hard Copy Nurse Replace Chart Figure 3: Lab Orders

19 Pull Chart Note (date/time/initials) Pull requisition for diagnostic test Check for completeness Fill out appropriate forms for specific test Note appointment in daily referral log Order test by phone/tube Call transport to setup patient transport (if necessary) Tell nurse time of appointment (verbal/page) Replace Chart Figure 4: D&T Orders

20 Pull Chart Note (date/time/initials) Note on census sheet NO DISCHARGE? YES Set up discharge appt (Mott) Use discharge papers (Main) NO TRANSFER? YES Pink Copy Pharmacy Hard Copy Nurse GENERAL ADMISSION? YES Do CPR Card according to weight (Mott) Replace Chart Make service tag, door tag, attending tag, locator tag Page house officer to tell them they are on floor Process orders Order old chart Figure 5: Admission Orders

21 Pull Chart (if strip order) Pull Chart Note (date/time/initials) Note (date/time/initials) Research MSC stock number in online catalog (if unknown) Call MSC to place order by stock number Call Nurse (If STAT) Put STAT Sticker on pink copy (if STAT and not available on floor) Pink Copy Pharmacy Hard Copy Nurse Pink Copy Pharmacy Hard Copy Nurse Replace Chart Figure 6: MSC Orders Replace Chart Figure 7: Medicine Orders Pull Chart (if strip order) Pull Chart Note date/time/initials (if strip order) Order equipment through computer Supply Chain System Note (date/time/initials) Pink Copy Pharmacy Hard Copy Nurse Pink Copy Pharmacy Hard Copy Nurse Replace Chart Replace Chart Figure 9: Nursing Orders Figure 8: Patient Equipment Orders

22 APPENDIX C: ORDER FORM SAMPLES Figure 10: Medicine Order Form Sample

Figure 11: Admissions Order Form Sample 23