Simulation Study of Using a Radiofrequency Identification System for Linens Delivery and Tracking Process in a Hospital

Organizations reengineer their business processes to improve efficiency, to contain costs, and to stay competitive in the marketplace. With escalating healthcare costs, hospitals also seek ways to provide quality healthcare services while containing costs. Hospitals traditionally have emphasized breakthroughs in healthcare technology to stay competitive. However, as competition among hospitals continues to intensify, patients may perceive little difference in the healthcare technology used by different hospitals. Consequently, hospitals are beginning to understand that process reengineering can be a better solution to achieve competitive advantage. Just as many businesses successfully reduce costs and gain competitive advantage by reengineering their business processes, hospitals can reengineer the way certain healthcare processes are carried out to achieve efficiency and cost containment. The same types of computer simulation that have proven successful in improving various business processes can also serve as an effective tool in the search for more efficient processes in hospitals.

This article describes a case study undertaken at a hospital (referred to as the Hospital hereafter for brevity and anonymity). Using computer simulation, the study assesses the efficiency of the linens delivery process in terms of the time taken and the visibility (i.e., inventory and location) of linens in the process. The results of the study will prove helpful to those who are considering reengineering and improving the linens delivery and tracking processes or other similar processes using radiofrequency identification (RFID) technology in hospitals.

Radiofrequency Identification Technology in Healthcare

RFID is emerging as a viable technology solution for hospitals to identify and track various mobile assets in medical facilities, identify and locate patients, and manage healthcare staff.(1) In the case of linens tracking, for example, an RFID tag can be attached to linens, allowing their location to be tracked by the RFID system. An RFID system usually consists of a reader and a tag that communicate with each other over a certain radiofrequency. An RFID tag is made up of an integrated circuit and an antenna. The antenna allows the integrated circuit to receive power and communicate, enabling the RFID tag to exchange data with the reader. The global market for RFID systems in healthcare is anticipated to reach $4.1 billion by 2021, increasing at a compound annual growth rate of 21.64% during the forecast period between 2017 and 2021.(2) Although most studies on RFID technology in healthcare have explored the tracking of various physical assets, few have studied the specific use of RFID for tracking linens in hospitals.

Computer Simulation in Healthcare

Computer simulation involves modeling processes. These models enable analysts to study how a system reacts to conditions that are not easily or safely applied in real-world situations and to examine how the working of an entire system can be altered by changing individual parts of that system.(3) The real power of simulation is fully realized when it is used to study complex systems.(4) Healthcare encompasses a dynamic system with complex interactions among various components and processes. Furthermore, healthcare management operates in an environment of aggressive pricing, tough competition, and rapidly changing guidelines. To meet these challenges, healthcare management must respond quickly to identify critical system processes, recognize all relevant resources, access real-time information, and analyze “what if?” cases.(5)

Although there are many applications of computer simulation to healthcare management and operations,(6) they may be broadly classified into two groups: (1) applications to healthcare systems at the various levels of communities, regions, or the nation; and (2) applications to specific operations, processes, or services in healthcare institutions. The first group includes applications intended to study the provisions of mental health, public health, health reform, or healthcare workforce, often with policy implications. For example, Anderson et al.,(7) Jacobson and Sewell,(8) Rauner,(9) and Zaric(10) illustrate the use of simulation for various health policy analyses. The second group, which is relevant to our case study, includes applications intended to improve facility design, staffing, and scheduling, and to reduce patient wait times and operating costs.(11) The case study described in this article attempts to extend this line of studies by considering the linens delivery and tracking process with an RFID system implemented in a hospital.

Modeling the Linens Delivery Process

The main objective of the simulation in this study is to model the linens delivery process on both the current scenario, without the RFID system, and the new scenario with the RFID system, and to assess the process in terms of two efficiency measures: (1) the times taken in the process; and (2) the visibility of linens in the process.

Much of the current process for the delivery of soiled linens back to the linens department is automated by using air-pressured pipes that connect the linens department to all the wards throughout the hospital. Soiled linens are first packed into bundles and then delivered into air-pressured pipes. Disposal of soiled linens is performed every day throughout the day. When soiled linens arrive at the linens department, they first are stacked up and held in the chute opening. Once the volume of soiled linens reaches a certain level preset by the hospital, the chute door opens and soiled linens are released into a metal cage mounted on a rail. When the sensor that monitors the level of soiled linens in the chute is activated, the metal cage starts traveling on a rail until it reaches the staging area. There, the bundles of soiled linens are handled by four employees, who count them and transfer the bundles of soiled linens into the metal cage.

The cleaning process of soiled linens is subcontracted out. The subcontractor’s responsibilities include transporting soiled linens from the hospital, cleaning the soiled linens, and delivering clean linens back to the Hospital. The subcontractor sends a delivery truck to and from the linens department of the Hospital every day. When the delivery truck arrives at the linens department at 8:00 AM, Hospital employees first unload clean linens from the delivery truck and then load soiled linens onto the delivery truck by 8:40 AM.

When clean linens are delivered, they packed on the storage shelves in the linens department. A predetermined level of linens is assigned to each ward. Employees in the linens department pack linens into metal trolleys according to the daily requirements. Metal trolleys are parked at the designated automatic guided vehicle (AGV) loading bay and are delivered to the wards at 1:30 PM. Metal trolleys with clean linens are left at the designated spot (the linens pick-up point) outside each ward while metal trolleys with unused clean linens from the previous day are replaced with new metal trolleys. Finally, the AGVs pick up the metal trolleys with unused clean linens from the previous day’s allotment and return them to the linens department.

In addition to the annual stock tracking, the linens department performs a circulation count every four months to account for the total inventory of linens on the storage shelves as well as linens in circulation in the wards of the Hospital.

Problems in the Current Process

In the current linens delivery process, the linens department experiences poor visibility of the inventory level of soiled linens, largely due to the hygiene standard that requires minimal human contact with soiled linens. When soiled linens are delivered by air-pressured pipes, they are packed into bundles for easy collection and accounting. Opening up bundles to count individual linens, however, would be time-consuming and non-hygienic. Instead, employees use estimates in counting soiled linens that are to be sent to the subcontractor for cleaning. Consequently, there is no exact count of soiled linens sent out to the subcontractor and, hence, it is impossible to track for any losses during the cleaning process at the subcontractor.

AGVs deliver clean linens according to the daily requirements of each ward. This is a replacement system rather than a replenishment system. That is, the AGVs do not count unused clean linens from the wards; instead, they bring metal trolleys with unused clean linens back to the linens department for accounting. Because of this replacement system, the linens department tends to keep a higher inventory level of clean linens than what is actually needed in the wards. Low visibility of linens being used in the wards makes it impossible to track for any losses during usage in the wards.

The New Process with the Radiofrequency Identification System

The problems we have discussed suggest three points where RFID stations would be useful:

1. Soiled linens are collected through centralized air-pressured pipes and consolidated into bundles in the linens department. The Hospital management considers establishing an RFID station where a reader and antenna are mounted on a conveyor belt, which can automate the process of loading bundles of soiled linens onto the delivery truck and at the same time provide real-time visibility of outgoing soiled linens.

2. The Hospital management considers establishing another RFID station where a reader and antenna are mounted on a conveyor belt, which can automate the process of moving clean linens onto the storage shelves and at the same time provide real-time visibility of incoming clean linens.

3. The problem of low visibility of linens being used in the wards can be handled by deploying another RFID station at the linens pick-up point in each ward, because the pick-up point is a common choke point to collect unused clean linens from the ward. This RFID station can allow real-time visibility of clean linens and can further allow an automated decision rule, such as formulating the linens replenishment quantity to be carried out.

The Hospital management also expects the RFID system to help meet the hygiene standard regarding soiled linens, because employees do not need to touch the soiled linens in order to count them when an RFID system is in place.

Simulation of the Process

Data and Assumptions

For the simulation in this study, we used data obtained from the Hospital. When necessary data for the simulation model were unavailable, we used the best estimates provided by the Hospital staff familiar with the linens delivery process. The simulation model involves entities, resources, and locations, as described in the following section.

An entity refers to an object or person that a simulation model processes. The simulation model includes two types of entities: soiled linens and clean linens. Bundles of soiled linens arrive continuously at the linens department through air-pressured pipes from midnight to 8:00 AM, and they are loaded onto the delivery truck beginning at 8:00 AM. The linens department handles about 450 bundles of soiled linens a day, each one of which contains about 14 linens. Loading all bundles of soiled linens onto the delivery truck takes about 40 minutes. The linens department receives about the same number of clean linens as of soiled linens each day. Unloading clean linens from the delivery truck takes about 30 minutes.

A resource is a person, piece of equipment, or some other device used for one or more of the following functions: treating and moving entities; assisting in performing tasks for entities at locations; and performing maintenance on locations or other resources. The simulation model includes one type of resource: employees in the linens department. There are four employees who handle soiled linens and five employees who handle clean linens.

A location is a fixed place in the system where entities are routed for processing or some other activity or decision. The simulation model has five types of locations: chutes, metal cages, open cages, metal trolleys, and storage shelves. The capacity of a location is the maximum number of entities that the location can hold at any given time. All locations of the simulation model are multicapacity locations:

• A chute is capable of accommodating 700 entities (i.e., 50 bundles of soiled linens with 14 linens in each bundle);

• A metal cage is capable of accommodating 700 entities (i.e., 50 bundles of soiled linens with 14 linens in each bundle);

• An open cage is capable of accommodating 1260 entities (i.e., 90 bundles of soiled linens with 14 linens in each bundle); and

• A metal trolley is capable of accommodating 300 entities (i.e., clean linens).

Storage shelves for clean linens are capable of accommodating enough entities.

We made several assumptions in the simulation model for the new linens delivery and tracking process, in consultation with the hospital staff. First, the travel time for the bundles of soiled linens from the metal cage to the open cage on the conveyor is assumed to be 30 seconds. Second, the travel time for the packs of clean linens from the delivery truck to the storage shelf on the conveyor is assumed to be 30 seconds. Third, the time to read the RFID tag on each linen moving on the conveyor is assumed to be 0.34 seconds. Forth, both the conveyor from the metal cage to the open cage and the conveyor from the delivery truck to the storage shelf are 2.50 meters long.

Results

We constructed the simulation model using Arena simulation software. The simulation model ran for 156 independent replications (26 weeks × 6 days), as the linens department operates from Monday to Saturday. The simulation results presented in the following paragraphs are based on the average results of the 156 independent replications.

The main objective of the simulation model is to find out the time needed to complete the linens delivery process with the RFID system while providing visibility of linens. The results of simulation show that the process of moving soiled linens from the chute to the open cage takes, on average, 39.19 minutes without the RFID system, whereas it takes 44.23 minutes with the RFID system. The increased time of the process using the RFID system is due mainly to the activities added in order to implement the RFID system, including moving soiled linens from the metal cage to the conveyor, moving soiled linens from the conveyor to the open cage, and reading RFID tags on soiled linens on the conveyor. The increased time of 5.04 minutes seems tolerable, because the new process with the RFID system provides visibility of soiled linens.

The simulation also shows that the process of moving clean linens from the delivery truck to the linen pickup points in the wards takes, on average, 339.75 minutes without the RFID system, while it takes 371.23 minutes with the RFID system. Collecting clean linens from storage shelves and loading them onto metal trolleys take as much as 225.00 minutes in the process, with or without the RFID system. Due to the activities added in order to implement the RFID system, including moving clean linens from the delivery truck to the conveyor, reading RFID tags on linens on the conveyor, moving clean linens from the conveyor to the storage shelf, and reading RFID tags on linens at the linen pickup points, the process with the RFID system takes longer. The increased time of 31.48 minutes also seems tolerable, because the new process with the RFID system provides visibility of clean linens.

Taken together, the results of simulation suggest that the increased times in the new process with the RFID system would be reasonable to provide improved visibility of linens in the process.

Limitations

A few limitations are recognized in this study. First, the arrival rate of soiled linens to the chute is assumed to be constant in the simulation model, whereas in reality it may be random. Second, the simulation process for soiled linens ends when soiled linens are uploaded onto the delivery truck, but in reality soiled linens are still in the process of being cleaned at the subcontractor. The activities done at the subcontractor are not incorporated in the simulation model, because they make the process a continuous cycle, which is not supported in Arena simulation software. Third, the simulation process for clean linens ends when clean linens are used by patients in the wards, but not all clean linens are used by patients, and in reality unused clean linens are returned to the linens department. Unused clean linens are not incorporated in the simulation model, because they make the process in a continuous cycle, which is not supported in Arena simulation software. These limitations certainly are not exhaustive, but they are important. Obviously, these limitations, in turn, suggest several possibilities for further study.

Conclusion

Our study used computer simulation to model and compare the current linens delivery process without the RFID system and the new process with the RFID system at the Hospital. It assessed the impact of the RFID system on the process in terms of time taken and visibility of linens in the process. The simulation results suggest that the increased times in the new process with the RFID system would be tolerable to provide visibility of linens in the process. The increased times are due mainly to the activities added to implement the RFID system in the process. Based on the simulation results, the Hospital management decided to implement the RFID system for the linens delivery and tracking process. The results of this study demonstrate that computer simulation is an effective tool supporting decisions on business process reengineering in hospitals.

References

1. Fisher JA, Monahan T. Tracking the social dimensions of RFID systems in hospitals. Int J Med Inform. 2008;77(3):176-183.

2. Global radio frequency identification devices (RFID) in healthcare market-analysis and forecast, 2017-2021 (focus on component, RFID tags, country analysis, application and competitive landscape). https://www.prnewswire.com/news-releases/global-radio-frequency-identification-devices-rfid-in-healthcare-market-analysis-and-forecast-2017-2021focus-on-component-rfid-tags-country-analysis-application-and-competitive-landscape-300504318.html .

3. Proctor T. Simulation in healthcare. Health Manpower Management. 1996; 22(5):40-44.

4. Kelton DW, Sadowski PR, Sadowski, AD. Simulation with Arena. New York: WCB/McGraw-Hill; 1998.

5. Stepanovich PL, Uhrig JD. Decision making in high velocity environments: implications for healthcare. Quality Management Health Care. 1999; 5(3):72-79.

6. Mielczarek B, Uzialko-Mydlikowska J. Application of computer simulation modeling in the health care sector: a survey. Simulation. 2010; 88:197-216.

7. Anderson JG, Harshbarger W, Weng HC, Jay SJ, Anderson MM. Modeling the costs and outcomes of cardiovascular surgery. Health Care Management Science. 2002;52;103-111.

8. Jacobson SH, Sewell EC. Using Monte Carlo simulation to determine combination vaccine price distributions for childhood diseases. Health Care Management Science. 2002;5(2):135-145.

9. Rauner MS. Using simulation for AIDS policy modeling: benefits for HIV/AIDS prevention policy makers in Vienna, Austria. Health Care Management Science. 2002;5(2):121-134.

10. Zaric GS. Random vs. nonrandom mixing in network epidemic models. Health Care Management Science. 2002;5(2):147-155.

11. Anderson JG. Preface: Special issue of simulation in health care management. Health Care Management Science. 2002;5(2):73.

Support for travel expenses pertaining to this study was provided by the Institute of International Business of the Stillman School of Business at Seton Hall University for Sung J. Shim. An earlier version of this paper was presented at the International Conference on Information Society, Dublin, Ireland, October 10-13, 2016, and was published in its Proceedings.