Establishing a Successful and Cost- Effective Staffing Model Stanley N. Chamallas Melissa Leone Top 5 Things to Know for CE: 1. Make sure your BADGE IS SCANNED each time you enter a session, to record your attendance. 2. Carry the Evaluation Packet you received on registration with you to EVERY session. 3. If you re not applying for CE, we still want to hear from you! Your opinions about our conference are very valuable. 4. Pharmacists and Nurses need to track their hours on the Statement of Continuing Education Certificate form as they go. 5. FOR CE: At your last session, total the hours and sign your Statement of Continuing Education Certificate form. Keep the PINK copy for your records. Place the YELLOW and WHITE copies in your Evaluation packet Make sure an evaluation form from each session you attended is completed and in your Evaluation packet Put your name on the outside of the packet, seal it, and drop it in the drop boxes in the NHIA registration area at the convention center Objectives 1. Describe the components of a successful Staffing Model. 2. Review staffing model development process for nursing and pharmacy, using industry examples. 3. Discuss steps for successful implementation of a Staffing Model. 1
Why Create a Staffing Model? Minimize labor to revenue ratios Justify the need for additional staff Prevent labor cuts Predict and budget for labor costs in a growing/ shrinking marketplace Plan for the ebb and flow of seasonal business Justify purchase of labor-saving tool or equipment The Basics Employee needs are defined in units called FTEs (Full Time Equivalents). 1 FTE represents 40 hours of work per week 1 FTE may equal 1 FT employee or 3 PT employees Office-based work is measured differently from field-based work Different work load measures Plan: 1. Decide on categories of FTEs: a. Core Infusion (in-office staff): -RPhs -Compounding b. Field RNs c. Delivery personnel 2
Plan: 2. Graph these FTE types against various workload measures: Core infusion FTEs vs. Net Revenue Core infusion FTEs vs. Shipments Core infusion FTEs vs. Units Compounded Core infusion FTEs vs. Total Patients Q: How should we decide which measures to use? A: Let statistical methods find them for us! R 2 - measure of relationship (correlation) between two variables An R 2 value of 1.0 indicates a perfect (100%) correlation, whereas a 0 R 2 indicates no correlation. Potential Workload Measures Net Revenue Shipments Units Compounded Deliveries / Visits Miles Driven Admissions / discharges 3
Excellent Relationship! Terrible Relationship! Plan: 3. Select the ones with good R 2 s, reject the others 4. Build a model based on the winners 4
The Winners are: 1. Net Revenue R 2 = 0.85 The Winners are: 2. Units Compounded R 2 = 0.87 The Winners are: 3. Total Shipments R 2 = 0.88 5
A Sample Reject: Patients Serviced R 2 = 0.55 But we can do even better: Multiple Regression Analysis - statistical method that combines individual graphs to obtain even higher correlation Doing this with: -Net Revenue -Units Compounded -Shipments R 2 = 0.96 Excellent! For Core Infusion We repeated this process for all infusion staff: Workload Measures R 2 Core Infusion Field Nurses Delivery Staff Net Revenue Units Compounded Total Shipments 0.96 # Visits Miles per Visit 0.96 # Deliveries Miles per Delivery 0.96 6
Q: Why not just use Net Revenue? A-1: Why use a measure with an R 2 of 0.85 when we have one with an R 2 of 0.96 available every month automatically? A-2: Any single measure can be deceptive Q: Why aren t you looking at gross profit percent? A-1: Because it has nothing to do with how many people are needed to get the job done! Core Pharmacy Productivity 7
Getting Started You can establish a basic time and motion study by therapy, including clinical monitoring activities using the following data elements: 1. Therapy type 2. Units 3. New patients 4. Patient discharges Assumptions Staffing is a function of workload Workload can be determined by time and motion studies It is as much art as science No two home infusion companies are alike What we Know In home care pharmacy, demand is unpredictable. Some work can be done during slow periods Refill work, preparation, cleaning, and environmental testing are schedulable. Volume and timing of new referrals are not 8
Establishing a Benchmark for Pharmacy Labor Compounding time per dose NMC Homecare 1996 Khan, Lawson 1996 NHIA / B.Braun Clinical Pharmacy Duties NAIT Study Texas Study NMC Homecare 9
Starting Slow Every activity in the office may be benchmarked Start with high dollar departments: pharmacy and nursing Select a few metrics that are easy to monitor Use as a learning tool: the inaccurate bathroom scale Diagram work-flow: you may find additions steps where they are not needed Evaluation of Process Flow Number of activities performed The order in which the activities are performed (serial versus parallel path) The content of the activity (does it add value) The staff performing the activity ( are certain individuals required by law or technical skill or expertise?) 10
Activity Tracking Activities of clinical management are primarily dependant on patient type. Compounding activities are dependant on patient type (TPN, ABX, Chemo, PNM), patient volume, and delivery container (syringe, elastomeric device, minibag) Other activities: management, in-services, environmental monitoring) are volume independent. Work Sampling Tracking Tool The pharmacist wore a beeper that would activate randomly 6 times per hour The wearer would place a mark in the box corresponding to activity being performed at that moment Value of Activities Does the activity add value? Is the activity redundant? Is the activity required by law or standard of practice? If the activity created information, is that information used? Does the activity create work for others? 11
How often to look Time-and-motion studies may then be repeated quarterly to identify trends and improve overall productivity. The most important step is to start benchmarking and measuring and a good way to do this is the proven time-andmotion study method. Tools Field Nurse Productivity 12
Potential Workload Measures Number of patient visits Type of patient visit Average length of patient visits Patient visits by therapy type Average miles driven per patient visit Where To Start? Collect at least 30 days of data, including: Number of visits per nurse per day Average length of visits per nurse Average miles per visit per nurse Type of visit activities being performed Drug administration IV insertion Lab sampling Determine Norms Average admission visit length Average follow-up visit length Average miles per visit Average travel time per visit Consider: Variances between nurses Visit length by therapy or visit type Visit equivalents for longer visits 13
Example Average miles per visit: 33 Nurse A: 35 Nurse B: 22 Nurse C: 42 Average travel time per visit: 43.7 mins Nurse A: 42 mins Nurse B: 31 mins Nurse C: 58 mins Average travel time per mile = 1.32 mins Example cont d Average follow-up visit length: 58 mins Nurse A: 66 mins Nurse B: 69 mins Nurse C: 39 mins Average travel time per visit: 44 mins Total time per visit: 102 mins Example cont d 8 hour work day includes 480 minutes 480 102 = 4.7 visits per nurse per day 1 visit/day = 0.21 FTE (102/480) 20 visits/mo = 0.21 FTE visits/wk x 102 (mins) = 14
Nurse Staffing Summary Once you have determined your average time per visit (including travel) you can determine how many nurses you need for any given workload. Next project: Nurse Zoning Use the information you gathered to see how you can decrease mileage and thereby increase productivity in visits/day. Financial Implications Given the cost basis of the pharmacy and nursing areas of your company you can t afford to not know these costs because knowing them is the first step to improving them. Use the information you gathered to ensure adequate staffing and subsequently the quality of care and patient satisfaction. Delegation Project Utilize data to ensure that the right person is doing the right job Delegation is key to managing labor expense Ensure that all work activities add value Analyze variation in time studies per employee to develop best practices for performance Employee A vs. Employee B 15
Staffing Trends When do you react? Monthly? Quarterly? Trends are best acted on each quarter A 90 day trend of over/under staffing by 10% or more shows that something needs to change Shorter reaction time may respond to seasonal trends, vacations, or holiday schedules rather than actual workload changes Questions? 16