Executive summaries of IIE Transactions and IIE Transactions on Healthcare Systems Engineering
Edited by Susan Albin and John Fowler
This month our research highlights focus on reliability and warrantees. The first summary presents a method to determine whether it is worthwhile, and if so how much, to upgrade secondhand products before they are sold and the warranty contract begins. The second summary estimates system reliability by allocating testing resources to various component types. These articles will appear in the November issue of IIE Transactions (Volume 44, No. 11).
Organizations, such as hospitals that deal with a large number of devices, depend on periodic inspections to ensure that these devices are safe and reliable. Most follow the manufacturers’ recommendations for inspection and maintenance, but these are not necessarily the best possible practices in their specific operating context.
Professor Sharareh Taghipour of Ryerson University and visiting professor Dragan Banjevic of the University of Toronto are trying to close the gap between optimal and actual practices in periodic inspection of repairable systems by using available data, formulating models that best describe the data and recommending evidence-based policies based on the outcomes of the models. In “Optimum Inspection Interval for a System under Periodic and Opportunistic Inspections,” they develop models to find the optimal inspection frequency for a multicomponent system subject to different types of failure. Their work is motivated by a real problem of maintaining medical devices, but it can be applied more generally to any repairable system, including packaging lines, weaving looms or spool thread winders.
The models are developed for a system with components that are subject to either hard or soft failures. For example, in a medical infusion pump, failures of some components such as indicators and switches stop the system, and the failed component is fixed immediately. These are hard failures. Other components have hidden or soft failures, such as audible signals. The pump can continue to operate if these components fail, even though failures can have serious, even catastrophic consequences if left unattended. In hospitals, clinical engineering departments have a checklist to inspect each device's major components/features and detect soft failures. However, when a device fails between periodic inspections due to a hard failure, the components prone to soft failure are inspected also (opportunistic inspection), regardless of any scheduled inspection plan.
Taghipour and Banjevic develop a model to find the optimal inspection interval for such a system under periodic and opportunistic inspections. Applying the proposed model to a hospital’s medical infusion pump, they obtain the optimal inspection interval and find it differs from the interval recommended by the manufacturer.
CONTACT: Sharareh Taghipour; email@example.com; (416) 416-979-5000, ext. 7693; Department of Mechanical and Industrial Engineering, Ryerson University, 350 Victoria St., Toronto, Ontario, Canada, M5B 2K3
Functional relationships in manufacturing industries
Due to recent progress in sensing and information technology, the massive amount of data readily available provides us with opportunities to integrate advanced statistical methods with system knowledge for better understanding, modeling and controlling of the quality of manufacturing processes. Often, the quality of a process could be characterized and summarized better by functional relationships between a response variable and several explanatory variables. That is, the focus would be on monitoring the profiles that represent such relationships, instead of on monitoring a single quality characteristic.
Consider an aluminum electrolytic capacitor manufacturing process provided by ENW Electronics Ltd. in Hong Kong. The whole manufacturing process, which is a typical multistage process, includes a sequence of operations, such as clenching, rolling, soaking, assembling, cleaning, aging and classifying. The relationship between the characteristics of an aluminum electrolytic capacitor from one stage to another stage often can be described by linear models. Usually, the engineers are concerned about significant changes in such relationships, which may indicate that some assignable causes in the process have occurred. With a well-designed monitoring system for those relationships, we will be able to target the cause of changes in the process efficiently and make correction and improvement accordingly.
With this in mind, in “A Distribution-Free Robust Method for Monitoring Linear Profiles Using Rank-Based Regression,” Xuemin Zi from Tianjin University of Technology and Education, Changliang Zou from Nankai University and professor Fugee Tsung from Hong Kong University of Science and Technology developed robust profile monitoring techniques to detect abnormalities in such functional relationships. Most of the existing work on profile monitoring relies on an unambiguous specification of the error distribution, e.g., a normal distribution, and the corresponding control schemes are constructed based on the least-squares estimates of the parameters of each profile. It is well-recognized that the underlying process distribution in many manufacturing applications is not normal. To this end, the authors suggest obtaining the estimated profile parameters by using a rank-based regression approach. Then, a novel multivariate sign, exponentially weighted, moving average control scheme is applied to achieve robust performances.
CONTACT: Changliang Zou; firstname.lastname@example.org; (86) 22-28212510; School of Mathematical Sciences, Nankai University, Tianjin, China, 300071
The most recent issue of IIE Transactions on Healthcare Systems Engineering (Volume 2, No. 3) contains five articles that cover a wide range of healthcare system problems and solution methods. The first one summarized below describes an approach to optimize dispatching and relocating emergency vehicles. The other provides a survey of nursing contributions to the management of medications.
Dispatching emergency medical service vehicles
The primary goal of emergency medical services (EMS) is to use scarce resources – physical and human – to save lives. During a pandemic, the stakes are especially high because of the large degree of uncertainty relative to the spread of the disease and numerous requests for EMS. Thus, managing EMS resources efficiently is crucial to operations in a pandemic.
Transportation of patients not only accounts for a significant portion of the EMS expenditure, but contributes largely to patients’ safety. Therefore, optimally dispatching EMS vehicles during a pandemic to meet increasing patient demand, satisfy stringent response times, and ensure that all census tracts are covered in anticipation of future requests becomes a logistics challenge.
In “An Optimization Approach for Dispatching and Relocating EMS Vehicles,” Farshad Majzoubi, a Ph.D. candidate at the University of Louisville, and his advisors, Lihui Bai and Sunderesh Heragu, explore the solution to this problem using mathematical programs. The EMS vehicle dispatch solution is implemented in a larger real-time decision support system for public health planning during a pandemic as well as during normal operations.
The mathematical models (integer linear and nonlinear programs) employed in the vehicle dispatch solution minimize the sum of the total travel costs, the penalty cost for not meeting the required response time window for patients, and the penalty cost for not covering census tracts. The authors classify patients into high and low priority groups to get the best vehicle capacity. Consequently, a vehicle is allowed to carry more than one patient when appropriate.
In particular, a vehicle transporting a low priority patient can be rerouted to pick up another patient. This significantly reduces patient waiting times, according to extensive numerical studies in the paper.
CONTACT: Sunderesh Heragu; email@example.com; (502) 852-7463; Mary Lee and George F. Duthie Chair in Engineering Logistics, Director, Logistics and Distribution Institute (LoDI), Department of Industrial Engineering, University of Louisville, Louisville, KY 40292
What healthcare system designers need to know about nurses
In hospital settings, and especially in intensive care units (ICUs), nurses often are a patient’s “last line of defense” against medical errors. Therefore, it is important to understand all that nurses do in order to design healthcare systems that support nurses’ physical and cognitive work needs.
This recognition motivated Joy Rodriguez and her colleagues to write “A Survey Study of Nursing Contributions to Medication Management with Special Attention to Health Information Technology.” Rodriguez’s colleagues include Pascale Carayon and Ben-Tzion Karsh, both from the University of Wisconsin-Madison; Hélène Faye from the Institut de Radioprotection et de Sûreté Nucléaire in France; Christine Baker from St. Mary’s Hospital in Madison, Wis.; and Matthew Scanlon from the Medical College of Wisconsin. After conducting 120 hours of observations, 18 one-hour interviews and two two-hour focus groups with ICU nurses, a survey listing 86 medication management-related activities was developed, pilot tested and administered to ICU nurses to understand how frequently they perform each activity during their shift and how important they think each activity is to the patient’s well-being.
Nurses perform prescribed nursing tasks (e.g., assessing the patient, administering medications, etc.) that have a clear beginning and end. However, the researchers found that nurses often perform complex cognitive activities to support these prescribed tasks. These cognitive activities, such as prioritizing tasks and patients and advocating for patients, have not received as much attention in the literature. But practitioners must consider these activities when designing healthcare work systems that keep patients safe and workers efficient and happy.
Hospitals and practitioners can use this survey to understand nursing work better and help them design or redesign their work systems. The survey helps target value-added versus non-value-added activities. Depending on whether activities are completed often or are perceived as benefitting the patient’s well-being, there are different design considerations to keep in mind.
For example, if an activity is done frequently and deemed important to the patient’s well-being, then the system needs to be designed to facilitate efficient and effective completion of this activity. If an activity is done infrequently but is deemed important to the patient’s well-being, then tools such as a checklist could support the nurse’s memory. However, if an activity is done frequently but not deemed important to the patient’s well-being, then this activity should be designed out of the system as much as possible.
CONTACT: A. Joy Rodriguez; firstname.lastname@example.org; (864) 656-3114; Department of Industrial Engineering, Clemson University, 130-C Freeman Hall, Clemson, S.C. 29631
Susan Albin is a professor at Rutgers University in the Department of Industrial and Systems Engineering. She is editor-in-chief of IIE Transactions and a fellow of IIE.
John W. Fowler is the Motorola Professor and Chair of the Department of Supply Chain Management in the W.P. Carey School of Business and a professor of industrial engineering at Arizona State University. He is editor-in-chief of IIE Transactions on Healthcare Systems Engineering.
IIE Transactions is IIE’s flagship research journal and is published monthly. It aims to foster exchange among researchers and practitioners in the industrial engineering community by publishing papers that are grounded in science and mathematics and motivated by engineering applications.
IIE Transactions on Healthcare Systems Engineering is a quarterly, refereed journal that publishes papers about the application of industrial engineering tools and techniques to healthcare systems.
To subscribe, call (800) 494-0460 or (770) 449-0460.