Wireless Sensor Network based System for the Prevention of Hospital

Acquired Infections

Iuliana Bocicor

, Maria Dasc

alu

, Agnieszka Gaczowska

, Sorin Hostiuc

, Alin Moldoveanu

Antonio Molina

, Arthur-Jozsef Molnar

, Ionut

Negoi

and Vlad Racovit

SC Info World SRL, Bucharest, Romania

NZOZ Eskulap, Skierniewice, Poland

Carol Davila University of Medicine and Pharmacy, Bucharest

Polytechnic University of Bucharest, Romania

Innovatec Sensing&Communication, Alcoi, Spain

Keywords:

Hospital Acquired Infection, Nosocomial Infection, Clinical Workﬂow Monitoring, Cyber-physical System,

Wireless Sensor Network.

Abstract:

Hospital acquired infections are a serious threat to the health and well-being of patients and medical staff

within clinical units. Many of these infections arise as a consequence of medical personnel that come into

contact with contaminated persons, surfaces or equipment and then with patients, without following proper

hygiene procedures. In this paper we present our ongoing efforts in the development of a wireless sensor

network based cyber-physical system which aims to prevent hospital infections by increasing compliance to

established hygiene guidelines. The solution, currently developed under European Union funding integrates a

network of sensors for monitoring clinical workﬂows and ambient conditions, a workﬂow engine that executes

encoded workﬂow instances and monitoring software that provides real-time information in case of infection

risk detection. As a motivating example, we employ the workﬂow in the general practitioner’s ofﬁce in or-

der to comprehensively present types of sensors and their positioning in the monitored location. Using the

information collected by deployed sensors, the system is capable of immediately detecting infection risks and

taking action to prevent the spread of infections.

1 INTRODUCTION

Hospital acquired infections (HAI) are a serious threat

to the health and well-being of both patients and

medical staff within clinical units. Various micro-

organisms, especially multi drug resistant bacteria

may lead to signiﬁcant hospital related morbidity and

mortality. These infections have high related costs

and represent a direct occupational hazard for clini-

cal personnel. Hospital infections are a worldwide

problem, regardless of geographical, political, so-

cial or economic factors (World Health Organization,

2002), (World Health Organization, 2010). Further-

more, technological development and sophistication

of medical care does not automatically result in lowe-

red infection rates (Tikhomirov, 1987), (Coello et al.,

1993), (World Health Organization, 2010), (European

Centre for Disease Prevention and Control, 2015).

According to the ﬁndings of the World Health Orga-

nization, average HAI prevalence in Europe is 7.1%,

in the United States it is 4.5%. In low- and middle-

income countries infection rates vary between 5.7%

and 19.1% (World Health Organization, 2011). In

intensive care units located in high-income countries

the proportion of infected patients can be as high as

51%, while in low- and middle-income countries it

can reach 88.9% (World Health Organization, 2011).

Unfortunately, many of these infections lead to pa-

tient deaths. Annually, infections are accountable for

37 000 deaths in Europe and 99 000 deaths in the USA

(World Health Organization, 2011). While measures

and precautions are being taken to successfully reduce

these rates (Centers for Disease Control and Preven-

tion, 2016), there is still much room for improvement.

Many hospital infections arise as a consequence

of medical personnel that come into contact with con-

taminated surfaces or equipment, relatives coming

in contact with patients or as auto infections, which

158

Bocicor, I., Dasc

alu, M., Gaczowska, A., Hostiuc, S., Moldoveanu, A., Molina, A., Molnar, A-J., Negoi, I. and Racovi¸t

a, V.

Wireless Sensor Network based System for the Prevention of Hospital Acquired Infections.

DOI: 10.5220/0006357801580167

In Proceedings of the 12th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2017), pages 158-167

ISBN: 978-989-758-250-9

occur by touching sensitive body parts, such as the

face, with the hands. The most common target sites

for hospital infection are the urinary and respiratory

tracts, which are often involved in minimally-invasive

procedures such as catheter-related procedures.

While the methodology for prevention exists, it

is often ignored due to lack of time, unavailability

of appropriate equipment or because of inadequate

staff training. Research shows that the most impor-

tant transmission route are staff members who come

into contact with patients or contaminated equipment

without following proper hygiene procedures (Ham-

mer, 2013). More particularly, most often transmis-

sion is made by touching a patient or contaminated

equipment and then touching another patient without

proper hand hygiene (Pittet, 2001). Recent guideli-

nes, such as ”Five moments for hand-hygiene” (World

Health Organization, 2015) provide concise and well-

structured information on efﬁcient disinfection means

to signiﬁcantly reduce the risk of infection.

In this paper we present our progress in the de-

velopment of a cyber-physical system based on a wi-

reless sensor network (WSN) that is targeted towards

HAI prevention. Our solution employs a sensor net-

work that will monitor clinical workﬂows and am-

bient conditions, integrated with conﬁgurable soft-

ware to detect deviation from established hygiene

practices. The sensor network collects information

in real-time about substance and material presence.

Availability of antimicrobial agents and sterile glo-

ves, as well as environmental conditions that affect

pathogen spread, such as oxygen level, airﬂow, and

temperature will be monitored. To complete the pic-

ture, the system will facilitate monitoring of complex

processes such as management of indwelling urinary

catheters, postoperative care, intubation and mecha-

nical ventilation. Given the complexity of these pro-

cesses, as well as the diversity of hospital regulati-

ons, these processes will be described using software

workﬂows. Each clinical process will be modelled

using one workﬂow instance executed by a software

workﬂow engine. When the sequence of transitions

inferred by the system from sensor data presents de-

viations from the expected ﬂow, the system will alert

responsible personnel.

The current stage of development represents a

system proof of concept, including multifunctional

smart sensors for monitoring the use of soap, antimi-

crobial gel and water sink together with a ﬁrst clini-

cal workﬂow that describes the required hygiene pro-

cedures in the general practitioner’s ofﬁce. This pa-

per details the smart devices employed, the hardware-

software integration as well as the software compo-

nents that ensure the cyber-physical system achieves

its objective of lowering the number and severity of

hospital infections.

The present paper is structured as follows. The

following section presents some of the last decade’s

advancements, from an Information and Communica-

tion Technologies (ICT) perspective. This includes

several software-based or cyber-physical systems de-

veloped to increase compliance to guidelines and de-

crease infection rates. Section 2 offers a detailed des-

cription of the workﬂow in the general practitioner’s

ofﬁce, as well as the main challenges faced when en-

coding this process using a software-based workﬂow

model. The wireless sensor network custom designed

for the system is depicted in Section 3, where all as-

pects related to its hardware and software components

are covered. Section 4 focuses on presenting the soft-

ware workﬂow engine that interprets events and mo-

nitors execution of conﬁgured workﬂows. Moreover,

the means by which the hardware part of the system is

integrated with the software workﬂow engine, as well

as the communication infrastructure employed are il-

lustrated in this section. Finally, we outline our con-

clusions and further work in the last section.

1.1 Existing Solutions

Because HAI have signiﬁcant detrimental effects, in

recent decades health care facilities have started to

implement prevention programmes for patients and

medical staff. There are even practical guides devi-

sed by specialised agencies (World Health Organi-

zation, 2012), which can be used as starting points

for the development of good practice plans concer-

ning workplace and patient safety. Following the cur-

rent technological developments in all medical areas,

technology is also present for monitoring and preven-

tion of HAI. This is illustrated by the development

of various software-based or cyber-physical solutions

that monitor and ensure compliance. In this section,

we present some of the most popular such systems.

1.1.1 Monitoring Hand Hygiene

Inadequate hand hygiene is responsible for a large

proportion of infections (Pittet, 2001). There are se-

veral automated solutions to reduce infections cau-

sed by improper hand hygiene, most of which use

continuous surveillance and immediate notiﬁcation

in case non-compliance is detected (Shhedi et al.,

2015). IntelligentM (Ryan, 2013) and Hyginex (Hy-

ginex, 2015) are two solutions that monitor employ-

ees using bracelet-like devices equipped with Radio

Frequency Identiﬁcation (RFID) technology and mo-

tion sensors. Whenever a hygiene event has been

omitted, the device alerts them either using vibration

Wireless Sensor Network based System for the Prevention of Hospital Acquired Infections

159

(IntelligentM) or coloured lights (Hyginex). Biovi-

gil technology (BIOVIGIL Healthcare Systems, Inc.,

2015) and MedSense (General Sensing, 2014) are de-

signed having the same purpose, only in these cases

bracelets are replaced with badges worn by healthcare

workers. The Biovigil device uses chemical sensors

to detect whether hand hygiene is observed according

to established standards. The system can be conﬁgu-

red to remind clinicians to disinfect their hands be-

fore entering patient wards, or before administering

procedures such as intravenous drips or catheter in-

sertion. Furthermore, these systems record hygiene

events, centralise them and enable analysis, visua-

lisation and report generation. SwipeSense (Swipe

Sense, 2015) employs small, alcohol-based devices

and wearable gel dispensers. This allows medical per-

sonnel to perform hand hygiene without interrupting

their activities to go to a sink or disinfectant dispen-

ser (Simonette, 2013). In opposition to the systems

mentioned so far, which use sensors placed at patient

ward entrances, UltraClenz’s Patient Safeguard Sy-

stem (UltraClenz, 2016) is ”bed-centric” and prompts

workers to sanitize before and after every patient con-

tact. The DebMed system (DebMed - The Hand Hy-

giene Compliance and Skin Care Experts, 2016) does

not use RFID technology, nor any devices for the me-

dical personnel, but instead estimates the number of

hand hygiene opportunities per patient-day and com-

pares this number with the actual hand hygiene events

that were performed, which are determined using a

network of wireless-enabled dispensers.

1.1.2 Disinfection Robots

Dangerous pathogens can remain in the air or on dif-

ferent types of surfaces in a hospital room for long

periods after the infection source was removed. To

tackle this issue, which cannot always be resolved

using traditional cleaning and disinfection procedu-

res, several types of disinfection robots have been de-

veloped. Generally, they are able to perform thorough

disinfection using ultraviolet (UV) light or chemi-

cal substances. The Xenex ”Germ-Zapping Robot”

(Xenex, 2015) can disinfect a room using pulses of

high-intensity, high-energy ultraviolet light. The ro-

bot must be taken inside the room to be disinfected

and in most cases, the deactivation of pathogens ta-

kes place in ﬁve minutes. Tru-D Smart UVC (Tru-D

Smart UVC, 2016) scans the room to be disinfected

using eight sensors and computes the optimal short

wavelength ultraviolet light dose required for disin-

fection according to the size, geometry, surface re-

ﬂectivity as well as the amount and location of equip-

ment found in the room. The robot performs disin-

fection of the entire room, from top to bottom in one

cycle and from one location, ensuring that the ultravi-

olet light reaches even shadowed areas. The Bioquell

Q-10 robots (Bioquell, 2016) emit a powerful anti-

bacterial bleaching agent, called hydrogen peroxide

to kill multi-drug resistant organisms. As hydrogen

peroxide is toxic to humans, after disinfection the Q-

10 uses another solution to ensure that it is safe for

humans to enter the room.

1.1.3 Managing Infection and Outbreaks

A different procedure in the ﬁght against infection

is implemented by the Protocol Watch decision sup-

port system for prevention and management of sepsis

(Philips, 2015). Protocol Watch regularly checks cer-

tain medical parameters of patients, to reduce the time

elapsed between the moment sepsis is ﬁrst detected

and beginning of treatment. If the system detects that

certain conditions indicative of sepsis are met, it alerts

medical staff and indicates which tests, observations

and interventions must be performed, according to es-

tablished prevention and treatment protocols.

Another goal pursued by clinicians when dealing

with hospital infection is the identiﬁcation of control

policies and optimal treatment in infection outbreaks.

A comprehensive approach that uses electronic health

records to build healthcare worker contact networks is

described in (Curtis et al., 2013). Its main goal con-

cerns putting efﬁcient vaccination policies into place

in case of infection outbreaks.

Among other relevant software systems developed

to enhance treatment policy in case of infection outb-

reak or epidemics are RL6:Infection (RL Solutions,

2015) and Accreditrack (Excelion Technology Inc.,

2013). RL6:Infection is a software solution developed

to assist hospitals in the processes of controlling and

monitoring infections and outbreaks, while Accre-

ditrack is designed to ensure compliance with hand

hygiene guidelines, verify infection management pro-

cesses as well as to provide procedural visibility and

transparency.

1.2 The HAI-OPS Platform

The proposed platform is developed within the Hos-

pital Acquired Infection and Outbreak Prevention Sy-

stem (HAI-OPS) research project (HAI-OPS, 2017).

Its main objective is to decrease overall mortality and

morbidity associated with hospital infection. It is

designed to handle both singular infection cases as

well as outbreaks, by targeting most common sour-

ces and transmission pathways. Operationally, the

platform will leverage advances in computing po-

wer and availability of custom-developed, afforda-

ble hardware that will be combined with a conﬁgura-

ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering

160

ble, workﬂow-based software system (Bocicor et al.,

2016).

Existing solutions, such as those detailed in the

section above (Ryan, 2013; Hyginex, 2015; BIOVI-

GIL Healthcare Systems, Inc., 2015; General Sen-

sing, 2014) can be successfully employed to monitor

a single process, such as hand hygiene, or equipment

and room disinfection (Xenex, 2015; Tru-D Smart

UVC, 2016; Bioquell, 2016). While these processes

are important for keeping patients and staff safe from

infection, there are many other processes that can lead

to hospital infection. Among the most prevalent, we

mention catheter management, mechanical ventila-

tion, invasive procedures and surgical site care (World

Health Organization, 2002; Coello et al., 1993). One

solution for monitoring multiple processes would be

to deploy several such systems in parallel. Howe-

ver, given that eHealth interoperability is currently

an open issue, this is not only cost-ineffective, but

technologically infeasible. We believe that monito-

ring several clinical and maintenance workﬂows can

be successfully addressed using a single system. Such

a system must be conﬁgurable so that it covers diffe-

rences between clinical unit location and layout, dif-

ferences in types and speciﬁcs of undertaken proce-

dures, as well as variation between hygiene guideli-

nes that must be observed by staff. The HAI-OPS

platform is designed to address these issues in both

hardware as well as software. First of all, using cu-

stomized, but affordable hardware allows sensors to

be deployed in key locations in cost effective manner.

Workﬂow engines allow researchers to create custom

BPMN-encoded (Object Management Group, 2015)

workﬂows that encode key events in monitored pro-

cesses. Furthermore, implementation of a user inter-

face for workﬂow management will allow epidemio-

logists to further customize the monitored workﬂows.

To the best of our knowledge, our proposed system is

the ﬁrst of its kind to combine a sensor network and

software in a cyber-physical system of the proposed

versatility.

2 GENERAL PRACTITIONER’S

OFFICE WORKFLOW

The cyber-physical system depicted in this paper em-

ploys pre-deﬁned workﬂows that describe the proces-

ses that the system will monitor. They allow the sy-

stem to take real-time action in case an infection risk

is detected. Our development approach is bottom-up,

and starts with modelling some of the less complex

workﬂows, which involve only medical staff and pa-

tients. The more complex workﬂows, that also in-

volve equipment, such as endoscopic or surgical pro-

cedures will be addressed at a later time. Thus, the

ﬁrst workﬂow we approach for the system prototype,

which is also the subject of the present paper’s moti-

vating example, is the workﬂow of the general practi-

tioner’s ofﬁce.

2.1 Process Description

The general practitioner (GP) is a medical doctor

whose practice is not limited to a certain speciality

and who provides treatment and preventive care to pa-

tients. As opposed to physicians working with inpa-

tients admitted to hospital for certain procedures, the

general practitioner works with outpatients, who re-

quire consultation or treatments which do not necessi-

tate hospital admission. All information regarding the

GP ofﬁce, as well as the consultation workﬂow des-

cribed were supplied by NZOZ ESKULAP (NZOZ

Eskulap, 2016), an outpatient clinic from Poland that

is targeted for the ﬁrst pilot deployment of our system.

Figure 3 illustrates the general practitioner’s ofﬁce la-

yout from the Polish clinic. The ofﬁce contains a desk

for the physician, a consultation bed and, most impor-

tantly for our use case, an area with several elements

for ensuring hygienic conditions: a sink, a waste bin

and an area dedicated to disinfectants and disinfectant

dispensers. The same ﬁgure also depicts the planned

layout of the wireless sensor network used for mo-

nitoring the workﬂow. These are described in more

detail in Section 3.

In order to ensure compliance with the infection-

prevention guidelines in the Polish clinic, the ﬁrst step

was identifying the hygiene practices to which the ge-

neral practitioner must adhere to before, during and

after patient consultation. The conventional workﬂow

for an outpatient consultation, including all required

actions for ensuring conformity with hygiene stan-

dards are depicted within the following sequence of

steps:

1. Patient enters the ofﬁce.

2. The GP starts a conversation with the patient, in

order to learn about their medical history, current

treatment and reason for the visit. Generally, the

physician uses pen and paper or a hospital infor-

mation system to record information to the patient

ﬁle.

3. The GP prepares to examine the patient. The

preparation process is crucial with regards to in-

fection prevention. According to current regulati-

ons within the target clinic, the doctor must sani-

tize their hands according to 10 steps for effective

hygiene. These are:

Wireless Sensor Network based System for the Prevention of Hospital Acquired Infections

161

(a) Wet hands thoroughly.

(b) Soap up, using the liquid soap dispenser. The

used tap must be elbow or wrist operated. The

physician must rub palms.

(d) Massage between ﬁngers, right palm over the

left hand and then vice-versa.

(e) Scrub with ﬁngers locked, including ﬁngertips.

(f) Rub rotationally, with thumbs locked.

(g) Rinse thoroughly.

(h) Dry hands using a paper towel that must be pla-

ced in proximity to the hand washing facility.

(i) Work towel between ﬁngers.

(j) Dry around and under the nails.

4. The GP throws the wet towel to a special waste

bin.

5. The GP starts patient examination.

6. After the examination, the GP uses an alcohol-

based sanitizer for hand disinfection.

7. The GP goes back to the desk and records exami-

nation results using pen and paper or the hospital

information system.

8. Patient leaves the ofﬁce.

The procedure described above concerns a regular

examination. However, for special cases such as ex-

aminations involving the head, eyes, ears, nose and

throat (HEENT), or when the patient presents with

skin infection, the doctor must also employ nitrile or

latex disposable gloves. Gloves should also be worn

whenever there might be contact with blood, body

ﬂuids, mucous membranes or non-intact skin. Glo-

ves must be put on immediately before the task to be

performed, and removed and discarded as soon as the

procedure is completed.

The BPMN workﬂow for the consultation process

is illustrated in Figure 1. Sections 3 and 4 describe

how the wireless sensor network is used for monito-

ring and how the workﬂow engine monitors the exe-

cution of hygiene-relevant events. In the case a devi-

ation from the expected steps of the workﬂow is de-

tected, a real-time alert is generated and sent to the

GP using a mobile device in their possession.

2.2 Workﬂow Description

The general practitioner workﬂow is shown using

BPMN speciﬁcation in Figure 1. For the description

of the workﬂow we use both Figures 1 and 3, as the

events speciﬁed in the workﬂow are detected by har-

dware devices placed in different locations in the of-

ﬁce. As soon as the patient enters the ofﬁce, this is

detected by the infrared array sensor element placed

near the entrance (Goga et al., 2016). The system re-

cords and interprets the received data and a workﬂow

instance is started. As illustrated in Figure 1, the ﬁrst

steps required from the GP is to start the water sink,

use the soap dispenser and then stop the sink. The sy-

stem interprets this as hand hygiene being performed.

These events are detected by the sensor elements in

the sink and those in the disinfectant dispenser area,

which are all connected to smart nodes placed near the

physician’s desk. This enables transmitting the data to

the software server via wireless network. While cur-

rent regulations described in the previous section re-

quire a speciﬁc sequence of actions to be undertaken

for hand hygiene to be considered effective, our sy-

stem only checks that the sink and disinfectants were

operated. The main reason for this is that the system

is envisaged as an additional aid for medical person-

nel that ensures their safety from possible infection.

The system is designed on the principle that medical

personnel are responsible and aware of the detailed

actions they must undertake to ensure their own, as

well as their patients’ safety.

The intermediate step of the workﬂow concerning

patient examination starts when the system has de-

tected that hand hygiene compliance is achieved. Ot-

herwise, the system generates and stores a hygiene

alert, which is immediately sent to the general practi-

tioner. The two activities are exclusive: if an alert is

generated, the workﬂow instance is stopped and the

recorded hygiene breach is recorded. In case of an

alert, the GP must perform hand hygiene, after which

the system initiates a new workﬂow instance. In case

initial hand hygiene and patient consultation are car-

ried out according to the workﬂow, the GP must dis-

infect their hands using antimicrobial gel after the last

contact with the patient. This event is again recorded

by the system using the same sensors situated in the

disinfectant dispenser area and the smart nodes near

the physician’s desk. The workﬂow is thus comple-

ted. All the information related to patient entry/exit,

hygiene compliance and alerts is saved to persistent

storage for further reuse, including statistics and ad-

vanced analyses for ﬁnding the source or propagation

of an outbreak.

2.3 Current Challenges

Although seemingly straightforward, the process des-

cribed above can become quite complicated, mainly

due to various types of constraints and interferences

that may occur. Below we present the main challenges

to the system, with regard to the GP ofﬁce workﬂow

and the methods we use to approach and overcome

ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering

162

Figure 1: General Practitioner Workﬂow.

some of them. Others are still open to discussion and

solutions are currently being investigated.

First and foremost, one key aspect to consider is

achieving minimal overhead on the clinical process

and minimal intrusive interaction, from the user ex-

perience point of view. It is important that the system

does not impose any constraints and does not restrict

the doctor’s movements. In many clinical units, in-

cluding the one targeted for pilot deployment, hospi-

tal regulations specify that personnel are not allowed

to wear jewellery, watches or bracelets, as these can

hamper their freedom of movement and spread bacte-

ria, especially if these wearables are difﬁcult to dis-

infect. To tackle this, the proposed system does not

require the use of additional wearables. Monitoring is

done using the deployed wireless sensor network no-

des, which are placed in key locations within the GP

ofﬁce, as described in Section 3. In addition, medi-

cal personnel already employ chest-mounted badges

to which radio-frequency tags can be easily added.

The placement of the wireless sensors is an essen-

tial challenge in itself, as locations must be chosen in

a manner that allows a complete and preferably op-

timal surveillance of monitored workﬂows. The ar-

rangement of sensors in the ofﬁce must be adjusted

to the process, but should also be sufﬁciently gene-

ral in order to allow monitoring several workﬂows: in

this case, both the regular consultation workﬂow as

well as HEENT examinations. Thus, in addition to

placing wireless sensors at the ofﬁce entrance, sink,

soap or disinfectant dispenser, in order to ensure com-

plete process monitoring, a device is also placed on

the waste bin, to detect when gloves are thrown away.

Device positioning in the GP ofﬁce is discussed in

more detail within Section 3.

One of the remaining challenges for cyber-

physical systems such as the proposed one concerns

short-term human interactions that are difﬁcult to de-

tect. In the case of the GP workﬂow, how should the

system detect and react to a person entering the GP of-

ﬁce during an examination? In this case, the hygiene

event performed by the GP before patient examination

is considered cancelled, as the third person can conta-

minate the physician or patient with micro-organisms.

Medical staff wear badges that can be used to identify

them using the sensors deployed near the entrance;

however, if the person is not part of the medical staff,

they cannot be identiﬁed. A potential solution is that

once the system detects someone entering the ofﬁce,

regardless of whether the person is medical staff or

not, the system triggers the execution of a new work-

ﬂow, including the necessary hygiene events. In case

this is not performed, the GP is alerted to take imme-

diate corrective action.

3 THE SENSOR NETWORK

A wireless sensor network consists of a group of elec-

tronic devices in which every node controls one or

more sensors that measure physical phenomena such

as light, heat or proper acceleration. All collected me-

asurements are sent using a wireless network protocol

to another device featuring more powerful processing

capabilities. Depending on their functionality, nodes

are classiﬁed into dummy and smart nodes. As the

name suggests, dummy nodes consist of small devi-

ces that have to effectuate just one simple task: de-

tect a generated event and pass the information to a

smart node. A dummy node is particularly charac-

terized by its small size (35x35mm) and low power

consumption. Some sensors, like RFID readers, do

not generate events by themselves and require a pre-

processing stage, which can be exclusively carried out

by a powered device. Smart nodes must be able to

Wireless Sensor Network based System for the Prevention of Hospital Acquired Infections

163

collect key actions detected from dummy nodes and

generate more complex events comprising informa-

tion regarding four relative clauses: who is the person

involved, what was the action generated, when it hap-

pened and where it happened.

3.1 Sensor Types

Required sensors were selected to enable monitoring

the clinical workﬂow detailed in Section 2. From the

mentioned steps, sensors in dummy nodes should be

applied mainly to detect key actions, such as the uti-

lization of hygienic elements that can be found in the

GPs ofﬁce: water sink, soap dispenser, waste bin, al-

cohol sanitizer and glove dispenser. The main sensor

types required to ensure effective monitoring of the

general practitioner ofﬁce workﬂow are as follows:

1. Accelerometer. These sensors measure changes

in gravitational acceleration on two or three axes,

allowing to detect changes in motion and orien-

tation. Accelerometers may be attached to water

taps, which regulate water ﬂow on the vertical axis

and temperature on the horizontal. They may also

be applied to sanitizer or glove dispensers, where

detected motion implies that they have been used

by a practitioner or checked by cleaning staff.

2. Proximity and Light Sensor. Proximity sensors

emit infrared radiation and look for changes in the

return signal. This type of sensors are already ap-

plied in some water sinks and soap containers, but

to the best of our knowledge none of them have

communication capabilities to report actions.

3. Switch Detection. A switch is just an electronic

component that interrupts the ﬂow of electric cur-

rent from one conductor to another. It may be

operated by a moving object, which makes it a

great choice for applications such as detecting the

use of waste bins or opening of a door. This is

the least energy consuming element from the list,

because it does not have to expend energy doing

continuous measurement.

Dummy nodes generate action events indicating,

for instance, that someone used the soap dispenser or

the waste bin, but it is the smart node who has to ﬁll

in information and identify who generated the action.

To achieve that goal, it is necessary to process and

combine the output from the following two sensors:

1. Infrared Array Sensor. It is a thermopyle type

infrared sensor which detects the amount of infra-

red rays. It has a built-in lens with a 60 degree

viewing angle. The sensor offers output for ther-

mal presence, direction and temperature values.

Figure 2: Infrared array sensor detecting people inside the

GP ofﬁce.

2. RFID Reader. Radio-frequency identiﬁcation

works using tag-based identiﬁcation. Tags are

small devices similar to stickers that may be car-

ried by people, animals or objects. They can also

be easily attached to wearables such as badges or

mobile equipment. The frequency range and ap-

plied antenna depend on the application and indi-

rectly on the distance between readers and tags. In

some clinics, medical and cleaning staff are used

to carry a badge with an identiﬁcation card based

on this principle.

3.2 Device Positioning

Device positioning and calibration are crucial for the

proper functioning of the system. In the case of

dummy sensors, the proximity sensor may detect false

positives if the distance range is not correctly adjusted

or if the sensor is incorrectly placed. When applied

to water sinks or gel dispensers, the proximity sensor

must be tied to the tap pointing downwards. The sy-

stem registers when someone places their hand under

the tap and when they stop using it. Figure 3 illustra-

tes the positioning of both smart and dummy nodes

within the general practitioner’s ofﬁce.

In addition to the proximity sensor, the RFID an-

tenna and passive infrared array sensor must also

be placed according to their detection range. RFID

readers provide received signal strength indication

(RSSI) levels for detected tags, a measure which is

proportional to the distance between them. Patch an-

tennas consist of a planar dielectric substrate material

with a radiating patch on one side and a ground plane

on the other. The radiating side must point to the GP’s

ofﬁce where elements to be identiﬁed are located and

the ground plane must point to the corridor, ceiling or

to an adjacent room. Radio frequency power output

must be conﬁgured to meet European Union regulati-

ons and to avoid false positive detections as much as

possible.

Passive infrared array sensors complement the

ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering

164

Figure 3: Layout of the general practitioner ofﬁce augmen-

ted with wireless sensor network.

information from RFID readers. If this information is

combined properly, the dummy node is able to locate

people inside the room, identify people wearing an

RFID card (typically clinical staff) and detect people

who are not wearing tags (typically patients). The

passive infrared sensor must be placed on the ceiling,

pointing downwards and centred in the room, as seen

in Figure 2. If the sensor’s angle of view is not enough

and doesn’t ﬁt the whole room, scalability is achieved

using several sensors.

3.3 Communication Protocols

As already stated, the main features of dummy no-

des are their small size and low power consumption.

Both features are very closely correlated, because in

most cases product size is determined by the battery.

In this case, the power consumption in dummy no-

des is so low, that they can be powered using coin

batteries. The reduced power consumption is due to

the integration of a Bluetooth Low Energy (BLE) mo-

dule (Bluetooth SIG, Inc., 2017). Compared to previ-

ous Bluetooth standards, BLE is intended to provide

considerably reduced power consumption and lower

cost, while maintaining a communication range of up

to 150 meters with connected devices.

Smart nodes have more complex processing, com-

munication and thus, higher power requirements than

dummy nodes. They are continuously listening for

input BLE connections and when data is received, the

result is forwarded to a database server for persistent

storage and subsequent analyses. Connection with

this server is carried out using existent network infra-

structure, regardless of whether it is wired or wireless.

Every smart device is identiﬁed within the network

using a unique IP address and has a ﬁxed location in-

side the building.

4 HARDWARE-SOFTWARE

INTEGRATION

The software side of the system implements the

client-server paradigm and employs a software ser-

ver to which an arbitrary number of heterogeneous

clients can have simultaneous connection. The ser-

ver contains components that receive sensor data from

the network, a persistence layer that manages sen-

sor readings as well as a workﬂow engine that execu-

tes workﬂow instances in real-time, generating alerts

when sensor readings indicate deviations from ex-

pected workﬂow transitions. The main software com-

ponents of the system are as follows:

1. Connected Device Controller. The connected

devices, or smart nodes, constitute the principal

hardware component of the platform. They are re-

sponsible for monitoring the clinical environment

using sensors and sending sensor readings to the

software server. By themselves, they cannot de-

cide whether an infection risk is present. Each

connected device includes a software controller,

a generic software component that runs indepen-

dently of the software server. Its objectives are to

ensure the correct functioning of smart nodes and

to send sensor readings to the software server.

2. Data Acquisition. This is the software compo-

nent that will be responsible with receiving sensor

readings. Received data is stored within the per-

sistence layer, from which it is read and used by

other components. This includes advanced ana-

lyses components yet to be developed which do

not make the object of the present paper. The ser-

ver adopts a REST architecture (Fielding, 2000)

to receive sensor data acquired as presented in

Section 3. The data interchange format employed

Wireless Sensor Network based System for the Prevention of Hospital Acquired Infections

165

is JavaScript Object Notation. Information con-

tained in ﬁles received from sensors includes the

event’s timestamp, a Uniform Resource Locator

that identiﬁes which node generated the infor-

mation as well as sensor reading values. For

instance, a sensor monitoring temperature will

transmit a temperature value in degrees Celsius.

A sensor monitoring the presence of an individual

will transmit a boolean value, according to whet-

her presence was detected. An RFID reader will

transmit the RFID tag identiﬁer and the received

signal strength indicator value.

3. Workﬂow Engine Adapter. This software com-

ponent is a fac¸ade to the workﬂow engine imple-

mentation used by the system. Its main purpose

is to abstract the particularities of the workﬂow

engine. This allows the system to operate with

any major off the shelf workﬂow engine imple-

mentation. This component provides the required

features that allow for the creation, update and de-

letion of clinical workﬂows monitored by the sy-

stem. The workﬂow engine interprets events, such

as inputs from deployed sensors (e.g. hand wa-

shing detected), and acts upon them according to

a predeﬁned process. The actions are conﬁgurable

and can vary from saving a new entry into a data-

base, sending an e-mail or emitting a real-time no-

tiﬁcation via an external application or short mes-

sage service. Its input is represented by process

descriptions. Processes are composed of activi-

ties connected with transitions. Processes repre-

sent an execution ﬂow. Each execution of a pro-

cess deﬁnition is called a process instance. As an

example, hand disinfection in the general practiti-

oner’s ofﬁce can be represented as a process. Each

time a patient enters the ofﬁce a new process in-

stance is started, managed by the business process

management system. Some activities, such as re-

cording an event, or sending an alert are automa-

tic. Others involve waiting for an external event

to occur, such as a sensing device reporting the

physician has disinfected hands. The workﬂow

engine keeps track of the state of process executi-

ons and manages creation and progress of process

executions.

4. Data Store. This component of the software ser-

ver acts as the system’s persistence layer for all

system data. In addition to user, alert, connected

devices and workﬂow data, it includes a com-

plete record of data received from connected sen-

sors. This is required in order to facilitate advan-

ced analyses for outbreak prevention, identiﬁca-

tion and monitoring. The data store will be imple-

mented using an SQL database.

5 CONCLUSIONS

This paper details the ongoing effort in the develop-

ment of a cyber-physical system intended for preven-

ting hospital infections. The HAI-OPS platform will

integrate a wireless sensor network and a software

server that uses a workﬂow execution engine to mo-

nitor key steps within various clinical processes that

were identiﬁed as responsible for a large proportion

of hospital infection. When key steps to ensure pro-

cess hygiene are not taken, the system will generate

real-time alerts. The present paper focuses on descri-

bing and modelling an initial clinical process used as

motivating example: outpatient consultations within

the general practitioner’s ofﬁce. This process is the

one selected for implementation during the system’s

ﬁrst pilot deployment within a Polish outpatient clinic

(NZOZ Eskulap, 2016). Although we directed our at-

tention speciﬁcally towards the motivating workﬂow,

the system is designed to allow for deployment of di-

verse sensor network conﬁgurations, as well as facili-

tate creation and execution of many different clinical

workﬂows.

Upcoming system development will build on exis-

ting achievements. First of all, the system will be used

to model more complex clinical workﬂows, including

endoscopic and minor surgery procedures, which will

be implemented in the pilot site location. Second

of all, as part of the project a graphical component

will be developed to allow management of monitored

workﬂows. In addition, we aim to leverage availa-

ble sensor readings by implementing advanced repor-

ting and evaluation capabilities. These are expected to

help clinical epidemiologists in pinpointing infection

and outbreak sources using visualizations such as risk

maps and healthcare worker contact networks (Hla-

dish et al., 2012).

ACKNOWLEDGEMENT

This work was undertaken as part of the HAI-OPS

project funded by the European Union, under the Eu-

rostars programme

REFERENCES

Bioquell (2016). Bioquell q-10.

http://www.bioquell.com/en-uk/products/life-science-

products/archive-hc-products/bioquell-q10/.

BIOVIGIL Healthcare Systems, Inc. (2015). Biovigil and

our team. http://www.biovigilsystems.com/about/.

https://www.eurostars-eureka.eu/project/id/9831

ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering

166

Bluetooth SIG, Inc. (2017). Bluetooth low energy.

https://www.bluetooth.com/what-is-bluetooth-

technology/how-it-works/low-energy.

Bocicor, M. I., Molnar, A.-J., and Taslitchi, C. (2016).

Preventing hospital acquired infections through a

workﬂow-based cyber-physical system. In Procee-

dings of the 11th International Conference on Evalu-

ation of Novel Software Approaches to Software Engi-

neering, pages 63–68.

Centers for Disease Control and Preven-

tion (2016). HAI data and statistics.

https://www.cdc.gov/hai/surveillance/.

Coello, R., Glenister, H., Fereres, J., Bartlett, C., Leigh, D.,

Sedgwick, J., and Cooke, E. (1993). The cost of in-

fection in surgical patients: a case-control study. Jour-

nal of Hospital Infections, 25:239–250.

Curtis, D., Hlady, C., Kanade, G., Pemmaraju, S., Pol-

green, P., and Segre, A. (2013). Healthcare wor-

ker contact networks and the prevention of hospital-

acquired infections. Plos One. DOI: 10.1371/jour-

nal.pone.0079906.

DebMed - The Hand Hygiene Compliance and Skin Care

Experts (2016). A different approach to hand hygiene

compliance. http://debmed.com/products/electronic-

hand-hygiene-compliance-monitoring/a-different-

approach/.

European Centre for Disease Prevention and Control

(2015). Annual epidemiological report. antimicrobial

resistance and healthcare-associated infections. 2014.

http://ecdc.europa.eu/en/publications/Publications/

antimicrobial-resistance-annual-epidemiological-

report.pdf.

Excelion Technology Inc. (2013). Accreditrack.

http://www.exceliontech.com/accreditrack.html.

Fielding, R. T. (2000). Architectural styles and the design of

network-based software architectures. Doctoral dis-

sertation, University of California.

General Sensing (2014). Medsense clear.

hand hygiene compliance monitoring.

http://www.generalsensing.com/medsenseclear/.

Goga, N., Vasilateanu, A., Mihailescu, M. N., Guta, L.,

Molnar, A.-J., Bocicor, I., Bolea, L., and Stoica, D.

(2016). Evaluating indoor localization using wiﬁ for

patient tracking. In International Symposium on Fun-

damentals of Electrical Engineering (ISFEE).

HAI-OPS (2017). home page. http://haiops.eu.

Hammer, S. (2013). Hand washing:

Reducing nosocomial infections.

http://iwsp.human.cornell.edu/ﬁles/2013/09/Hand-

Washing-Reducing-Nosocomial-Infections-

2j1mlfb.pdf.

Hladish, T., Melamud, E., Barrera, L. A., Galvani, A., and

Meyers, L. A. (2012). Epiﬁre: An open source c++

library and application for contact network epidemio-

logy. BMC Bioinformatics, 13(1):76.

Hyginex (2015). Introducing hyginex generation 3.

http://www.hyginex.com/.

NZOZ Eskulap (2016). NZOZ eskulap.

www.eskulapskierniewice.pl/.

Object Management Group (2015). Business process model

and notation. http://www.bpmn.org/.

Philips (2015). Protocolwatch - SSC Sepsis.

http://www.healthcare.philips.com/main/products/

patient monitoring/products/protocol watch/.

Pittet, D. (2001). Improving adherence to hand hygiene

practice: a multidisciplinary approach. Emerging In-

fectious Diseases, 7:234–240.

RL Solutions (2015). The rl6 suite / infection surveillance.

http://www.rlsolutions.com/rl-products/infection-

surveillance.

Ryan, J. (2013). Medtech proﬁles: Intelligentm

- a simple yet powerful app to dramati-

cally reduce hospital-acquired infections.

https://medtechboston.medstro.com/proﬁles-

intelligentm/.

Shhedi, Z. A., Moldoveanu, A., Moldoveanu, F., and Tas-

litchi, C. (2015). Real-time hand hygiene monitoring

system for hai prevention. In The 5th IEEE Interna-

tional Conference on E-Health and Bioengineering -

EHB 2015.

Simonette, M. (2013). Tech soluti-

ons to hospital acquired infections.

http://www.healthbizdecoded.com/2013/06/tech-

solutions-to-hospital-acquired-infections/.

Swipe Sense (2015). Hand hygiene. redeﬁned.

https://www.swipesense.com/.

Tikhomirov, E. (1987). Who programme for the control of

hospital infections. Chemioterapia, 6:148–151.

Tru-D Smart UVC (2016). About tru-d. http://tru-

d.com/why-uvc-disinfection/.

UltraClenz (2016). Patient safeguard system.

http://www.ultraclenz.com/patient-safeguard-

system/.

World Health Organization (2002). Prevention of

hospital-acquired infections - a practical guide.

http://www.who.int/csr/resources/publications/

whocdscsreph200212.pdf.

World Health Organization (2010). The burden

of health care-associated infection worldwide.

http://www.who.int/gpsc/country

work/

summary 20100430 en.pdf.

World Health Organization (2011). Health

care-associated infections - fact sheet.

http://www.who.int/gpsc/country work/

gpsc ccisc fact sheet en.pdf.

World Health Organization (2012). Prevention of

hospital-acquired infections - a practical guide.

http://apps.who.int/medicinedocs/documents/

s16355e/s16355e.pdf.

World Health Organization (2015). Clean care is

safer care - ﬁve moments for hand hygiene.

http://www.who.int/gpsc/tools/Five moments/en/.

Xenex (2015). Xenex germ-zapping robots.

http://www.xenex.com/.

Wireless Sensor Network based System for the Prevention of Hospital Acquired Infections

167