52 ISE Magazine | www.iise.org/ISEmagazine

Inside IISE Journals

research

How tailored clustering approaches reduce

the spread of infectious disease outbreaks

Cluster analysis is one of the most adopted unsupervised

learning methods with uses that arise in various applica-

tions, some with a signiﬁcant impact on human lives. While

several off-the-shelf clustering techniques exist, many of

these techniques perform poorly for speciﬁc problem types.

This is largely attributed to three factors: First, existing ap-

proaches are based on approximate schemes with no guar-

antees. Consequently, the resulting recommendations can

deviate, sometimes substantially, from global optimality.

Second, there is no clear way of selecting between existing

approaches, leaving practitioners in the dark as to which

technique to adopt. Third, these techniques do not account

for speciﬁc problem structure which, if taken advantage of,

can improve solution performance.

In their work, “Optimal Clustering of Frequency Data

with Application to Disease Risk Categorization,” Texas

A&M University assistant professor Hrayer Aprahamian

and George Mason University assistant professor Hadi El-

Amine investigate novel clustering procedures that are tai-

lored to handle a speciﬁc type of data, frequency data, that

commonly arises in numerous applications. The tailored

approach provides many advantages over conventional

procedures by exploiting the structure of the problem. Al-

though the work is speciﬁc to frequency data, their results

are independent of the application, which greatly broadens

its usability.

The authors achieve this by interpreting the process

of clustering a dataset as a procedure that characterizes a

unique discrete distribution. This novel interpretation al-

lows them to extract valuable statistical information, which

they embed within an optimization framework. To solve

the resulting challenging combinatorial optimization prob-

lem, the authors construct a scheme that casts the clustering

problem as a tractable network problem. This reformulation

scheme enables the construction of an efﬁcient polynomial-

time algorithm that guarantees global optimality.

The authors demonstrate the use of the developed meth-

odologies through an application that aims to identify

screening strategies that mitigate the spread of infectious

diseases. In this context, the population needs to be clus-

tered into risk categories in a manner that maximally high-

lights their heterogeneity. Doing so provides practitioners a

tool that not only identiﬁes optimal categories, but one that

enables the construction of custom screening approaches.

Their results reveal that the considered approach con-

This month we highlight two articles in IISE Transactions. The ﬁrst article focuses on how to tailor clustering of

heterogenous data and how that practice helps reduce the spread of infectious disease outbreaks. The authors investigated

a novel clustering procedure for handling frequency data, which is to interpret the process of clustering that characterizes

a unique discrete distribution. This reformulation enables the construction of an efﬁcient polynomial-time algorithm that

guarantees global optimality, and the tailored clustering signiﬁcantly reduces misclassiﬁcations when applied to identifying

screening strategies that mitigate the spread of infectious diseases. The second article attempts to provide an understanding

of the possible effects of competition in terms of consumption and economic welfare for providing water services.

The authors studied two types of competition – one based on an attraction model for water providers within a public-

private partnership and another involving quantity competition within a spot market. The authors found that spot-market

competition does not necessarily result in greater levels of supply, nor in a lower price, and that a public-private partnership

could be preferable. These articles will appear in the August 2022 issue of IISE Transactions (Volume 54, No. 8).

Hrayer Aprahamian Hadi El-Amine

July 2022 | ISE Magazine 53

sistently outperforms current practices by signiﬁcantly re-

ducing misclassiﬁcation while using the same amount of

resources. Such results underscore the value of tailored

optimization-based approaches in addressing complex real-

world challenges.

CONTACT: Hrayer Aprahamian; hrayer@tamu.edu; Wm Michael Barnes

‘64 Department of Industrial and Systems Engineering, Texas A&M

University, College Station, TX 77840; Hadi El-Amine; helamine@gmu.

edu; Systems Engineering and Operations Research Department, George

Mason University, Fairfax, VA 22030

Using a dynamic model to decide if

public-private partnership is preferable to

spot market competition for water services

Water consumption is increasing due to worldwide popula-

tion growth, the urbanization of countries around the globe

and the demand for a higher quality of life. Actually, wa-

ter joined gold, oil and other commodities traded on Wall

Street, highlighting worries that the life-sustaining natural

resource may become scarce across most of the world.

In many countries, policymakers face the dilemma of

how to respond to increasing physical water scarcity. There-

fore, considerable effort has been expended in improving

the productive efﬁciency of the water sector using a variety

of methods, including concentration, competition for the

market, competition in the market and cooperation. Thus

far, however, these strategies have only been implemented

in a limited way and the consequences of introducing them

remain unclear.

In their work, “Water scarcity and welfare: Regulated

Public-Private Supply Chain Versus Spot-Market Compe-

tition,” chaired professor Konstantin Kogan, postdoctoral

student Dmitry Tsadikovich and associate professor Tal

Avinadav, all from Bar-Ilan University in Israel, attempt

to understand the possible effects of competition in terms

of consumption and economic welfare. They contrast the

standard monopolistic approach to supplying water with

two different types of competition. One type is based on an

attraction model for water providers within a public-private

partnership, while the second type involves quantity com-

petition within a spot market.

Unlike the case of quan-

tity competition, which im-

plies dynamic pricing, the

public-private supply chain

is characterized by a state-

regulated price that is ﬁxed

over the period of the con-

tract and that is set at a level

to just recover the associated

distribution costs of the pub-

lic entity.

The authors derive dynamic equilibrium replenishment

and inventory policies to show that, contrary to expecta-

tions, spot-market competition does not necessarily result

in greater levels of supply, nor in a lower price, than does

a regulated supply chain. In particular, if the distribution

cost is relatively low, then by encouraging a public-private

partnership, the public entity (municipality or state utility

company) can ensure higher consumer surplus compared to

a spot market.

On the other hand, increasing the distribution cost and

hence the regulated price is likely to diminish the differ-

ences between the two market types. In such a case, man-

aging water supplies for the municipality by means of a spot

market would become advantageous. Speciﬁcally, it would

help the public entity to manage the risk of shortages and

better align supply and demand.

CONTACT: Konstantin Kogan; Konstantin.kogan@biu.ac.il; Bar-Ilan

University, Faculty of Social Sciences, Department of Management, Max

ve-Anna Webb Street, Ramat-Gan, ISRAEL 5290002

This month we highlight two articles from IISE

Transactions on Healthcare Systems Engineering

(Volume 12, No. 2). The first article measures

the total patient value received from creating

a mix of cases for surgeons of varying levels

of experience. The goal is to maintain a high

level of patient outcomes but counterbalance

caseloads to allow less experienced surgeons to

improve their skills. The second article gathers

information from surveys with patients and

caregivers to aid in the design of a wearable

dialysis device. Such a device could provide

increased quality of life for patients suffering

from renal disease if user-centered design

principles are applied to take patients’ needs

and perspectives into account throughout the

product development process.

Konstantin Kogan

Dmitry Tsadikovich Tal Avinadav

54 ISE Magazine | www.iise.org/ISEmagazine

research

How surgical units can leverage

the diversity in surgeons’ experience

levels to improve patient outcomes

While the observance of hos-

pital treatment guidelines and

protocols is critical for good

medical outcomes that di-

rectly offer value to patients, a

surgeon’s case mix could also

be a critical factor impacting

patient outcomes. When dif-

ﬁcult cases are mostly seen

by experienced surgeons, less

experienced surgeons miss

the opportunity to improve

their skills. However, when

less experienced surgeons see

too many difﬁcult cases, it could produce poor patient out-

comes.

Renato de Matta of the University of Iowa College of

Business examines this trade-off in his paper, “Patient-

Centric Surgeons Case Mix Problem.” He shows how the

diversity in experience level of surgeons complements a sur-

gical unit in a way in which good patient outcomes from

experienced surgeons counterbalance the average but im-

proving patient outcomes from less experienced surgeons,

while less experienced surgeons gain more experience.

He formulates and solves an optimization model that

ﬁnds the case mix of surgeons in a surgical unit that maxi-

mizes the unit’s total patient value (TPV) per dollar expen-

diture on healthcare resources. Numerical experiments us-

ing laparoscopic colectomy data show workload balancing

and patient self-selection of surgeon could reduce total pa-

tient value; if the number of patients seen by each surgeon is

restricted, achieving excellent outcomes for high value sur-

geries by managing the learning that takes place improves

TPV; a balanced mix of easy and hard cases produces higher

TPV; and high demand uncertainty increases the mismatch

between the value surgeons can deliver to patients and the

value demanded by patients which lowers TPV.

Examples of practical applications of the model are: Plan-

ning the mix of patients seen by junior surgeons in residen-

cy programs at teaching hospitals; U.S. hospitals that accept

diagnosis-related groups (DRG) based on reimbursements

could plan their budget using TPV. An actual TPV that is

higher than projected is an indication of good patient out-

comes, which could result in cost savings and produce good

ﬁnancial performance for hospitals.

CONTACT: Renato de Matta, Ph.D.; renato-dematta@uiowa.edu; Busi-

ness Analytics and Information Systems Department, 108 Pappajohn

Business Building, University of Iowa College of Business, Iowa City IA

52242-1994

Realizing the dream of a wearable

dialysis device: What do patients

and their care partners want?

Imagine that your schedule is structured around visiting a

dialysis center two or three times a week for the rest of your

life. Each visit requires you to spend four to six hours at

the center connected to a stationary dialysis machine that

cleans your blood.

This is the reality for most patients with end-stage re-

nal disease (ESRD), the condition of irreversible kidney

failure.

The in-center mode of dialysis therapy negatively im-

pacts ESRD patients’ quality of life such as impairing their

mobility, independence and mental health. Because of these

challenges and other difﬁculties resulting from the disease,

many ESRD patients need the support of a care partner to

assists with their healthcare needs and daily activities.

The Center for Dialysis Innovation (CDI) at the Univer-

sity of Washington (UW) is developing a wearable dialysis

device to transform current dialysis therapies by allowing

ESRD patients to undergo continuous dialysis wherever

they are.

To design a wearable dialysis device patients are willing

to use, it is important that researchers apply user-centered

design principles. The user-centered design takes users’

needs and perspectives into account throughout the prod-

Renato de Matta of the

University of Iowa College

of Business.

Auður Anna Jónsdóttir Larry G. Kessler

Seung-Yeon Rim Ji-Eun Kim

July 2022 | ISE Magazine 55

IIISE Transactions (link.iise.org/iisetransactions) is IISE’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.

IISE Transactions on Healthcare Systems Engineering (link.iise.org/

iisetransactions_healthcare) 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.

About the journals

uct development process.

In “What Patients and Care Partners Want in Designing

a Wearable Dialysis Device: A Mixed-Methods Study,” the

human factors team at the CDI joins forces with Auður

Anna Jónsdóttir, a doctoral student in industrial and sys-

tems engineering at UW; Larry G. Kessler, a professor in

the department of health systems and population health at

UW; Seung-Yeon Rim, a UW industrial and systems engi-

neering graduate; and Ji-Eun Kim, an ISE assistant profes-

sor at UW.

By interviewing 24 ESRD patients and 12 care partners,

the interdisciplinary team of researchers identiﬁed similari-

ties and differences in ESRD patients’ and care partners’

preferences for the designs of a wearable dialysis device.

The ﬁndings from this study can help guide researchers

in designing a patient- and care partner-centered wearable

dialysis device.

CONTACT: Ji-Eun Kim, jikim@uw.edu; Industrial & Systems Engineering,

University of Washington, Box 352650, Seattle, WA 98195

Yu Ding is the Mike and Sugar Barnes Professor of Industrial

and Systems Engineering at Texas A&M University and Associ-

ate Director for Research Engagement at the Texas A&M Institute

of Data Science. He is editor-in-chief of IISE Transactions and

a fellow of IISE.

Oguzhan Alagoz is a professor in the Department of Industrial and

Systems Engineering at the University of Wisconsin-Madison. He

is editor-in-chief of IISE Transactions on Healthcare Systems

Engineering and a fellow of IISE.

Share & Like IISE

Follow IISE pages, like posts and share them with

your network of contacts on social media. Watch

and enjoy IISE videos at www.youtube.com/c/

iisechannel. Go to www.iise.org/social for details on

IISE’s social networks. If you’re a member of IISE,

chat with fellow members on IISE Connect (connect.

iise.org) and its communities speciﬁc to your area of

interest.

On-campus, online

and in Manhattan

binghamton.edu/ssie

Systems Science and

Industrial Engineering at

Binghamton University

WATSON COLLEGE

Systems Science and Industrial Engineering

SUNY Distinguished

Professors

tenured/tenure-

track faculty

SUNY Chancellor’s

Award recipients

$10

million

in research

expenditures

in 2020–21

Degree

programs

in Industrial and

Systems Engineering

(BS, MS, PhD), Systems

Science (MS, PhD) and

Healthcare Systems

Engineering (MS)

280

undergraduates

210

master’s

180

doctoral students