The Front Line

News from the field

For fewer unhappy returns, use a hospital readmission model

Penn State, Geisinger predict patient’s risk for needing more medical care
Cheng-Bang Chen and Soundar Kumara (below) hope their team’s patient readmission model can benefit hospitals nationwide.

One thing worse than going to a hospital is having to be readmitted – or even dying – after you leave.

But according to the Institute of Medicine, patients who have been discharged are at a high risk of adverse effects that include readmission, going to the emergency room or death. So researchers from the Pennsylvania State University and Geisinger Health System teamed up to figure out how to predict a patient’s risk for needing additional medical care three days after being discharged from the hospital. They used two years of clinical, administrative and socioeconomic data from Geisinger patients to develop REDD, which stands for re-admission, emergency department or death.

REDD is a machine learning model, said Deepak Agrawal, one of the Penn State researchers and a dual doctoral student in industrial engineering and operations research at the time of the research study.

“Using the REDD model, we were able to leverage large amounts of data to identify these high-risk patients at the point of discharge, which helps physicians target interventions to effectively reduce adverse events,” Agrawal said. Cheng-Bang Chen and Soundar Kumara (inset) hope their team’s patient readmission model can benefit hospitals nationwide.

Beyond the adverse health effects, U.S. hospital readmissions cost $41 billion as of 2011, according to a 2014 Agency for Healthcare Research and Quality report. Medicare’s tab alone was $26 billion annually, with $17 billion attributable to avoidable re-hospitalizations. This was before the Centers for Medicare and Medicaid Services started financially penalizing participating hospitals for 30-day readmissions.

The researchers concluded that readmissions after 30 days tend to be related to factors such as poor home environment, limited access to service and little social support.

“However, readmissions closer to discharge are more likely to be related to factors that are actually present but are not identified at the time the patient is discharged,” said Soundar Kumara, Allen E. Pearce and Allen M. Pearce Professor of Industrial Engineering, who led the Penn State research team.

Researcher Cheng-Bang Chen, an industrial engineering doctoral candidate, added that the more time that passes after a patient is discharged, the longer it may take the treating physician to sort out what may have landed the patient back to the hospital.

“After 30 days, physicians must read through the patient’s records, determine the after-care treatment they may or may not have received, find out if medications were taken, determine if any other drugs were in their system, see if they worked with their primary care physician on a new course of treatment, etc. There are many more factors that need to be known to identify what could have gone wrong with the initial treatment,” said Chen. “Our model focuses on just three days after discharge, which gives physicians a better chance to improve their processes when treating the patient in the hospital setting.”

For six months, Geisinger tracked high-risk patients and added services to try to reduce their chances of being readmitted, visiting the emergency room or dying, said Eric Reich, manager of healthcare re-engineering at Geisinger. Interventions ranged from:

  • Scheduling a return appointment with the patient’s primary care provider
  • Educating patients about their prescriptions and post-discharge care plans
  • Having the inpatient clinical pharmacist review the discharge medication list after reviewing the provider’s recommendations
  • Filling the patient’s prescriptions filled before discharge
  • Having the hospital check on any patient who has been discharged to a skilled nursing facility the day after discharge.

Although limited nursing resources prevented reliable estimates on the reduction in readmissions, Reich said the model can be introduced in hospitals.

“If the REDD model was fully implemented and aligned with clinical workflows, it has the potential to dramatically reduce hospital readmissions,” he said. “Through these data-driven predictive models, hospitals will be able to provide better quality care while efficiently allocating their scarce resources at the same time.”

Kumara said the work gave him immense satisfaction.

“It’s not just a published piece of research, but a real model that potentially is useful to hospitals and patients across the country,” he said.

Colombia gets solar-powered 3-D printing

Rochester Institute of Technology students develop system to boost developing economies
Rochester Institute of Technology students and faculty traveled in May to Cali, Colombia, to deliver and show residents how to operate this solar-powered 3-D printing system they developed for their multidisciplinary senior design project. (Photo by Kate Melton)A team of industrial, electrical and mechanical engineering students have developed a solar-powered 3-D printing system for use in developing economies.

The students, along with faculty from Rochester Institute of Technology, traveled in May to Cali, Colombia, to deliver the system and show residents how to use it. The team worked with partners from the Universidad Autónoma de Occidente (UAO) during the spring semester to create the product. Organizers hope the venture helps residents of a Cali neighborhood develop marketable skills in advanced manufacturing. Many youths in the area are at risk of getting involved in gang activities.

This project marks the first step in an agreement RIT and UAO reached last fall. The agreement aims to develop student and faculty project and research exchanges focused on international product design theory and commercialization for developing economies. The project was spearheaded by associate professor Marcos Esterman and his former student Alvaro Rojas Arciniegas. Arciniegas, now a faculty member at UAO, has degrees in industrial engineering and imaging science.

The versatile 3-D printer power system can seamlessly switch power sources for use in developing economies. The group collaborated remotely with UAO students who worked on ways to use recycled plastic bottles as reinforcement in the 3-D printing filament.

“The goal is to implement this printing system in Colombia because their electricity isn’t that reliable and because 3-D printers need a constant flow of electricity to function,” said Josh Cohen, a fifth-year student who worked on the project. “Having those backup power sources like the solar panels or the battery or also being able to plug into the grid are all things that will keep this printer up and running in the community.”

Smart bridges cross the communication divide

Sandia sensors allow real-time structural health monitoring to detect cracks
Sandia National Laboratories engineer Stephen Neidigk positions a Comparative Vacuum Monitoring sensor on a bridge. In his other hand is the control system that periodically checks the sensor and a wireless transmitting device to autonomously alert the maintenance engineers if it detects a crack. (Photos courtesy Randy Montoya)

What if a bridge could tell its engineers that it needs a hug – or at least a repair job?

With the industrial internet of things helping factory machines talk to their maintenance crews, Sandia National Laboratories is working to bring similar technology to bridge repair. The company teamed with Structural Monitoring Systems PLC, a U.K.-based manufacturer of structural health monitoring sensors, to outfit a U.S. bridge with a network of eight real-time sensors to alert maintenance engineers when they detect a crack or when a crack reaches a length that requires repair. Dennis Roach shows off the Comparative Vacuum Monitoring sensor.

Like its manufacturing counterparts, the Comparative Vacuum Monitoring sensors are designed to extend the lifetime of structures, reduce operating costs and improve safety. Engineers mount the sensors on a structure and analyze the data to assess the condition of a bridge or another kind of transportation infrastructure.

In 2016, the U.S. government classified more than 54,000 bridges in the U.S. as “structurally deficient,” so about 9 percent of U.S. bridges need regular monitoring. Sandia senior scientist Dennis Roach said bridges, pipelines and other critical structures are difficult to monitor because of remote locations and difficulty of access. A network of structural health monitoring sensors could be a solution.

Roach has worked with airlines on the sensor system since 2001 to detect cracks on commercial aircraft. Real-time, remote structural health monitoring could be used in a variety of applications, including railcars and rail lines, ships, wind turbines, power plants, remote pipelines, storage tanks, vehicles and even buildings.

BOOK OF THE MONTH: Save the world with lean Six Sigma

IISE’s Hurley writes how process improvement can alleviate social problems
Lean Six Sigma for GoodFor decades, industrial and systems engineers have used the continuous process improvement methodologies of lean and Six Sigma to enhance their organizations’ quality, reduce costs, increase profits and upgrade customer satisfaction.

IISE member Brion Hurley, also a contributing writer for ISE magazine, believes that these skills can help reduce pollution, poverty, homelessness, domestic violence or almost any problem. So he penned Lean Six Sigma for Good, a short primer designed to motivate you to do two things: Help your current company get more engaged in community issues and help you spend more of your free time helping nonprofit organizations become more efficient.

However, don’t look for a technical primer, as this short tome is only 50 pages. Instead, Hurley wrote about various instances where lean Six Sigma decreased time to deliver food to hurricane victims, cut costs for mosquito nets and office supplies in East Africa and reduced volunteer sign-up times for Big Brothers Big Sisters of Edmonton, Canada. Hurley discusses his first volunteer project that applied lean Six Sigma, a drive to increase recycling options at the University of Iowa football stadium, where he played while attending college. His first sustainability project at work netted his company $300,000 a year and saved around 3 million kilowatt hours of electricity.

Then he details how others, including IISE members Kate Groot and Joel Brock, have applied process improvement to everything from animal shelters to beach cleanups.

Lean Six Sigma for Good is available for free online at Audible sells an MP3 version for $2.99, or you can buy a paperback copy for $9.99 from or Amazon.

Making better decisions is an untapped competitive advantage

Boards of directors could benefit from evaluating processes as well as results
A new report details how corporate boards can tackle barriers to good decision- making, including data overload and a tendency toward groupthink.

“Decision-Making in the Visionary Boardroom,” released by WomenCorporateDirectors (WCD) and the KPMG Board Leadership Center, discusses how improving the processes boards of directors use to make judgement calls under tight timelines and great scrutiny can be a competitive advantage.

QUOTE, UNQUOTE: Better teamwork through chemicals – or at least caffeine

“We found that increased alertness was what led to the positive results for team performance. Not surprisingly, people who drank caffeinated coffee tended to be more alert. … They’re talking about more relevant things after drinking caffeinated coffee.”

— Amit Singh, a Ph.D. candidate at the Ohio State University and co-author of a study that showed caffeinated coffee helped facilitate teamwork, and that people who drank caffeinated coffee were more likely than those who drank decaf to want to work with their group again. “Coffee with Co-Workers: Role of Caffeine on Evaluations of the Self and Others in Group Settings” appeared in the Journal of Psychopharmacology.

“The decision-making process is an untapped frontier for boards, and improving upon how the board deliberates and makes decisions can enhance the company’s competitive advantage,” said Susan Angele, senior advisor with the KPMG Board Leadership Center. “What the commission found, based on the experiences of a diverse group of corporate directors from around the world, were common blind spots and innate biases that present themselves in almost all decision-making processes. Boards that can spot these early and address them head-on are simply better positioned to make smarter decisions.”

The report addresses multiple factors affecting board decision-making processes – from the biases directors bring to the table to “analysis paralysis” to a sense of “ownership” over an idea that leads board members to overvalue it or hang on too long.

To combat these barriers to good decision-making, the WCD Thought Leadership Commission offers tips for boards to improve their decision-making processes, including, among others:

  • Taking a “clean sheet” approach to optimize the quality and relevance of materials provided to the board
  • Identifying cognitive biases and mental shortcuts – recognize the widespread influence of innate tendencies that lead to overconfidence, groupthink and blind spots
  • When evaluating significant business proposals, probing management’s consideration of contrary views, alternative options and critical risks
  • After a major decision, evaluating the decision-making process as well as the result

When windows are better than walls

Berkeley Lab’s ‘super window’ could save $10 billion annually in energy costs
What if your windows insulated better than your walls? Well, according to Berkeley Lab researchers, that could save about $10 billion every year in energy costs.

Survey says: Self-driving cars must be 75 percent safer to be accepted

So how safe must self-driving cars be before they are accepted by consumers? They must reduce traffic fatalities by at least 75 percent compared to their human-controlled counterparts, according to a new study published in Risk Analysis. Co-author Peng Liu, an assistant professor of industrial engineering at Tianjin University, said the research could help government authorities establish safety requirements for regulating autonomous vehicles.

Current double-glazed windows have two layers of glass with a low-e coating and argon gas between the glass layers. Berkeley’s version inserts a third layer of thin glass sandwiched between the two layers of a double-glazed window, adds a second low-e coating and replaces the argon gas with higher-insulating krypton gas. Unlike other triple-glazed windows on the market, the “super window” is the same width and almost the same weight as existing double-glazed windows, which means builders don’t have to redesign the window sash and frame.

Berkeley actually patented the “super window” more than 20 years ago, back when large sheets of thin glass weren’t available and materials were cost-prohibitive. The lab is working with manufacturers Andersen Corp. and Alpen High Performance Products to bring the windows to market. About $20 billion worth of energy leaks out of windows in the United States each winter.

For the military, AI and 3-D might make for easier maintenance

Army uses machine learning to monitor Bradley fighting vehicles; Marines quickly fix aircraft
The U.S. military is hoping artificial intelligence and 3-D printing can boost its maintenance programs.

First, the Army is using a machine learning algorithm from Uptake Technologies to monitor several dozen Bradley M2A3 vehicles, the technology news site reported. The goal is to reduce unscheduled maintenance and make repairs more efficient by predicting when components will fail. Instead of replacing mechanics, Uptake’s Asset Performance Management application will help them make better decisions on when to service the armored infantry fighting vehicles.

And the 31st Marine Expeditionary Unit (MEU) already has demonstrated how 3-D printing can make for more effective operations, according to, which covers the additive manufacturing industry. Marine Fighter Attack Squadron 121 recently flew an F-35B Lightning II plane that had a replacement part fabricated on a 3-D printer. In another case, part of a landing gear door’s bumper on VMFA-121 F-35B aircraft wore down during the MEU’s spring patrol. Traditional fabrication would have required replacing the entire door assembly, but the team printed just the bumper and had the plane flying in a few days.