Process Mining in Frail Elderly Care: A Literature Review

Nik F. Farid

1,2

, Marc De Kamps

and Owen A. Johnson

School of Computing, College of Arts and Sciences, Universiti Utara Malaysia, Kedah, Malaysia

School of Computing, University of Leeds, Leeds, U.K.

Keywords: Frail Elderly, Frailty, Healthcare Service, Process Mining.

Abstract: Process mining has proved to be a valuable technique for extracting process knowledge from data within

information systems. Much work has been conducted in applying process mining to domains such as logistics,

banking, transportation and many areas of the government, including healthcare. Frail elderly people who

have an increased risk of adverse outcomes are amongst the main users of healthcare services and

understanding healthcare processes for the frail elderly is challenging because of their diverse and complex

needs combined with an often high number of co-morbidities. This paper aims to provide an overview of work

applying process mining techniques to improving the care of frail elderly people. We conducted a literature

search using broad criteria to identify 1,047 potential papers followed by a review of titles, abstract and

content which identified eight papers where process mining techniques have been successfully applied to the

care of frail elderly people. Our review shows that, to date, there has been limited application of process

mining to support this important segment of the population. We summarise the results based on five themes

that emerged: types of source data and process; geographical location; analysis methodology; medical

domain; and challenges. Our paper concludes with a discussion on the issues and opportunities for process

mining to improve the care pathways for frail elderly people.

1 INTRODUCTION

The over 60s are the main users of health and social

care (Oliver, 2009) and the number of people over 60

is expected to more than double from 962 million in

2017 to 2.1 billion by 2050 (UN, 2017). While many

adults remain in good health well over 60 there is an

increasing risk of frailty associated with aging. Frailty

is a common clinical condition among the elderly and

is often associated with stress caused by a cumulative

decline in organ and clinical functions over time

(Clegg et al, 2013). The progression of frailty can be

seen as a continuous sequence from normal ageing, to

pre-frail, frailty and finally to severe frailty (Lekan et

al, 2017). An inability to maintain normal body

functions may result in difficulty in managing with

everyday activities independently (Xue, 2011) and

increases the chance of hospitalizations,

institutionalization, and adverse health outcomes

such as falls, delirium and even mortality (Fried et al,

2001; Crandall et al, 2016; Minitski et al, 2001; Eeles

et al, 2012). Frailty progression over time is poorly

understood and there is growing interest in using

electronic health record data to understand and

identify the factors that influence this progression.

One approach can be to visualize frailty progression

using multi-dimensional data including patient

characteristics, diagnoses and medication history

from electronic health records (Chamberlain et al,

2016). Our interest is in the extent to which process

mining of these records could help improve the

understanding of frailty and the pathways of care

designed to support the needs of the frail elderly

population.

Electronic health records (EHRs) and other health

information systems store data associated with the

highly complex processes involved in delivering

health care services to patients and this data can be

used in process mining (Mans et al, 2008). Process

mining is an emerging approach that combines

business process management methods with data

mining technologies (Aalst, 2011). Process mining

aims to explore sequence of logged events over time

and to abstract generalisations of the underlying

process as process models. The approach can help

analysts discover how processes are followed in

practice, measure the conformance of real event logs

to the ideal process to identify deviations, recommend

332

Farid, N., De Kamps, M. and Johnson, O.

Process Mining in Frail Elderly Care: A Literature Review.

DOI: 10.5220/0007392903320339

In Proceedings of the 12th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2019), pages 332-339

ISBN: 978-989-758-353-7

improvements to the process and monitor the

effectiveness of interventions (Aalst, 2011). When

applying process mining to electronic healthcare

record data we treat the pathways of care as a type of

business process (Mans et al, 2008).

The phrase process mining started to appear in the

literature around 2006 based on the work of van der

Aalst (Aalst et al, 2006) which applied data science

to businesses process improvement efforts. A wide

range of tools and approaches have subsequently been

developed and applied to corporate organizations

(Aalst et al, 2007; Aalst, 2015) and to healthcare

(Partington et al, 2015; Weber et al, 2018). Process

mining is generally based on data extracted from

information systems but related work has used data

from sensor devices that record daily activities to

develop interventions (Fatima et al, 2013) and

improve personalized care (Sztyler et al, 2015).

The Process Mining Manifesto (Aalst et al, 2011)

proposed the L* life-cycle methodology for process

mining projects. This approach consists of five

different stages (labelled 0 to 4) which are: 0) plan

and justify; 1) data extraction; 2) creating a control

flow model and connecting an event log; 3) creating

an integrated process model; and 4) providing

operational support. More recently, van Eck et al

(2015) introduced an enhanced approach called

Process Mining Project Methodology (PM

There are a number of recent literature reviews of

process mining in healthcare (Rojas, 2016; Erdogan

and Tarhan, 2018; Williams et al, 2018) and other

reviews which focus on specific healthcare specialties

such as cancer (Kurniati et al, 2016) and

cardiovascular disease (Kusuma et al, 2018).

However, until now there has been no literature

review specifically examining process mining for the

care of the frail elderly. This paper describes the

approach we adopted to identifying literature relevant

to process mining for frail elderly patients. Eight

papers were found and are discussed here.

To date there has been limited application of

process mining approaches to support this important

segment of the population. Our paper aims to initiate

discussion on the value and potential of process

mining of frail elderly care pathways and identify

opportunities to work in this field of study.

2 METHODOLOGY

The literature review was conducted in October 2018

to identify papers which describe the application of

process mining to care involving frail elderly people.

2.1 Search Process

A four stage approach for search and selection was

used (Figure 1). The first stage covered the search of

papers from medicine, technology and engineering

databases; PubMed, Medline, British Medical Journal

Open, ACM DL, Elsevier, ScienceDirect, database

systems and logic programming (DBLP), Web of

Science and Google Scholar.

Figure 1: Summary of search process.

Keywords for process mining replicated those

used in previous literature reviews (Kurniati et al,

2016; Kusuma et al, 2018). To ensure the search

obtained papers that cover all relevant conditions

related to older people a broad selection of keyword

terms was applied following the Medical Subject

Headings (MeSH) terms used in the PubMed and

Medline databases. MeSH terms provide a

comprehensive vocabulary for journal and articles

indexing in medical studies used to facilitate

searching. An additional eight keywords that are

synonyms of the initial six MeSH terms keywords

were obtained from the thesaurus website

(https://www.thesaurus.com/browse).

The following keywords were used:

("Process mining" OR "workflow

mining" OR "pathway mining") AND

("Frailty" OR "Elderly" OR “Older

Process Mining in Frail Elderly Care: A Literature Review

333

Adults” OR "Ageing" OR "Geriatric" OR

"Palliative" OR “Debility” OR

“Decrepit” OR “Deteriorate” OR

“Vulnerable” OR “Senile” OR

“Impairment” OR “Fallibility” OR

“Senescence”)

We followed the review process adopted by

Kurniati et al, (2016) where paper selections was

conducted based on title, abstract and content

checking. We checked our search results to make sure

that they included the list of papers identified on

processmining.org, the main process mining research

community web site. A careful filtering approach was

performed at each stage to ensure all potentially

relevant papers were identified. For example, papers

were passed to the next stage if insufficient

information was provided in either the title or the

abstract. Finally, an in-depth ancestor search was

conducted to identify additional relevant papers from

references in the final set.

2.2 Selection Process

The initial extracted papers were selected according

to the set of inclusion and exclusion criteria outlined.

2.2.1 Inclusion Criteria

The following were the inclusion criteria when

selecting papers for analysis has been applied to the

frail elderly domain:

IC1: Articles published from year 1998

IC2: Publication language is English

IC3: Articles published are peer-reviewed or

conference proceeding articles

IC4: Articles should include case studies where

process mining technique has been applied into

frail elderly domain

2.2.2 Exclusion Criteria

The following were the exclusion criteria applied

when choosing extracted papers:

EC1: Duplicate publication of initial searched

EC2: Books

EC3: Articles discuss case studies other than in the

frail elderly domain

2.2.3 Quality Assessment Process

The following activities were undertaken to ensure

the quality of the search process. The paper

extraction, analysis and evaluation were done

manually by the first author. Google Scholar searches

were performed in incognito mode to eliminate any

bias that might arise from browsing history that might

influence the search results. The review and the

verification of the selected publications in the final

stage was supervised by all co-authors.

3 RESULT

Eight papers were identified after a comprehensive

search. This section describes the search process and

an analysis of the selected papers. Table 1 provides

an overview of the number of papers initially

extracted and the final selection of published articles

from the different sources. The initial search retrieved

a total number of 1,091 papers from ten different

search engines. We note that zero results were

returned from BMJ Open, Elsevier/Springer, DBLP

and processmining.org sources and most final

selected papers were from the Google Scholar search.

Table 1: The number of retrieved papers.

Sources Initial Extraction Final

Selection

Google Scholar 991 5

PubMed 32 0

BMJ Open 0 0

ScienceDirect 48 1

Elsevier/Springer 0 0

ACM 6 0

Web of Science 11 0

Medline 3 2

DBLP 0 0

processmining.org 0 0

Total 1,091 8

The duplication step removed 44 papers and left

1,047 papers for the next stage. The inclusion and

exclusion criteria were applied to the remaining

papers to carefully select those papers that met with

the aim of the work. Figure 1 details the number of

papers excluded at each stage of the process based on

the inclusion and exclusion criteria.

3.1 Characterisation of Element

Analysis

The review identified eight papers and these are

presented here. A complete list of reviewed papers are

summarised in Appendix 1. A thematic analysis of the

papers identified five themes: (1) data and process

type; (2) geographic analysis; (3) methodology (4)

medical domain; (5) challenges that arose when

conducting the study.

(1) Data and Process Type: the classification of

process and data type followed the approach in Rojas

HEALTHINF 2019 - 12th International Conference on Health Informatics

334

et al, (2016) that divided sources of data extraction by

clinical or administrative healthcare dataset and the

process type categorized as clinical treatment process

or organizational process. However in this literature

review, the most commonly extracted data were from

sensors collected either from elderly behaviour living

in smart environment (Vitali and Pernici, 2015; Tax

et al, 2018); mined process data collected from an

MIT smart home dataset (Tapia et al, 2004) and from

nursing homes as in Llatas et al, (2011), Wolf et al

(2013) and Munstermann et al, (2012) for patients

who require ambulant services. Meanwhile Triki et

al, (2015) analyse data from scenario generators for

elderly people’s daily activities. There are only two

papers that directly study electronic health record

(EHR) data and one related to acute care and simple

one day surgery (Najjar et al, 2018); while Conca et

al, (2018) used administrative data, which identified

different healthcare discipline roles. The nature of the

data will determines the type of analysis possible.

Najjar et al, (2018) investigated the clinical treatment

while Conca et al, (2018) discussed the

organizational process of collaboration between

physicians, nurses and dietician. The other six papers

(Munstermann et al, 2012; Llatas et al, 2011; Tax et

al, 2018; Triki et al, 2015; Wolf et al, 2013; Vitali

and Pernici, 2015) analysed processes which looked

into daily activity of elderly people.

(2) Geographic Analysis: most papers analyse

data from Europe - Wolf et al, (2013) Munstermann

et al, (2012) from Germany; Triki et al, (2015) from

France; Vitali and Pernici, (2015) from Italy and

Llatas et al, (2011) from Spain. The other papers were

Najjar et al, (2018) from Canada, Conca et al, (2018)

from Chile and Tax et al, (2018) from the

Netherlands, but the source of data was from the

USA.

(3) Methodology: none of the papers described

followed the process mining methodologies of the L*

life cycle or PM

. All papers reported that they had

developed their own methodology. It was evident that

each had carried out process mining using clustering

techniques from event logs generated from either

EHRs or from sensor devices.

(4) Medical Domain: three different care

processes within the medical domain have been

investigated. Two papers (Llatas et al, 2011; Wolf et

al, 2013) analysed processes to detect or reduce the

progression of dementia. Najjar et al, (2018) obtained

data from patients who suffered from heart diseases,

whereas Concas et al, (2018) collected data from

patients who had Type 2 diabetes mellitus. The other

papers (Triki et al, 2015; Tax et al, 2018; Vitali and

Pernici, 2015; Munstermann et al, 2012) did not

describe the medical domain associated with their

work.

(5) Challenges: the challenges could be

categorized as technique, data and team limitations

from sensor devices as in Kurniati et al, (2016). The

papers working with sensor data experienced data

quality issues related to granularity (Triki et al, 2015;

Llatas et al, 2011; Wolf et al, 2013; Vitali and Pernici,

2015; Tax et al, 2018; Munstermann et al, 2012). The

other limitation was data that was incomplete or

inconsistent. This was the main issue in Conca et al,

(2018). Najjar et al, (2018) suggested pre-processing

of the extracted pathway data through multiple

iterations to narrow the model to specific elements of

interest. Conca et al, (2018) used a medical expert to

help address their process mining challenges.

3.2 Evaluation of Experimental Result

Most of the case studied in the papers concerned

traces collected from sensor devices (Wolf et al,

2013; Tax et al, 2016; Vitali and Pernici, 2015;

Munstermann et al, 2012; Llatas et al, 2011). Two

papers conducted experiments using EHR data

(Najjar et al, 2018; Conca et al, 2018) and one used a

scenario generator (Triki et al, 2015). There are two

papers that applied a clustering algorithm to cluster

set of events such as Hidden Markov Model (Najjar

et al, 2018) and a combination of flow disintegration

functionality and measuring dissimilarity based on

heuristic topological editing distance (Conca et al,

2018). Llatas et al, (2011) used a workflow mining

technique based on the Workflow Instance Acceptor

Algorithm. Wolf et al, (2013) employed classification

techniques on sensor data for activity recognition of

traces before labelling the recorded activities to be

modelled as processes. Triki et al, (2015) generated a

variety of simulations to describe different models.

Vitali and Pernici, (2015) constructed a methodology

to understand the connection between process and

events.

There are two papers that performed evaluation on

the proposed algorithms. Wolf et al, (2013) showed

that their subjective logic condition evaluator was

outperformed by both a fuzzy event assignment

method and a method that combines the FlowCon and

FlexCon algorithms. Tax et al, (2016) describes the

effect of proposed algorithms on the computation

time. Munstermann et al, (2012) used a dynamic

threshold for an F

-measure that combines both

precision and recall to differentiate between the

abnormal and normal days of activities. A set of

abnormal traces were artificially created from normal

sensor data by simulating four different types of

Process Mining in Frail Elderly Care: A Literature Review

335

errors - swap, remove, delay and repeated activities.

Both papers (Llatas et al, 2011; Conca et al, 2018)

employed the PALIA process discovery application

to create visualizations of the process model. The

paper by Triki et al, (2015) represented process flow

as a Petri Net. Najjar et al, (2018) created a

hierarchical visualisation of clustered processes

based on frequency analysis using abstraction and

pruning.

4 DISCUSSION

4.1 A Lack of Focus on the Frail

Elderly

Most of the finally selected papers are very recent and

this suggests that work using process mining for the

care of older adults has started to gain attention only

recently. We noted that, during the literature search,

a number of papers on assisting the living of elderly

people were also identified. Although many of these

were excluded due to our focus on process mining

techniques, it is evident that this is an important

theme where process mining could help. Several of

the process mining papers use sensor data to detect

changes in behaviour concerning the daily activities

(Munstermann et al, 2012; Triki et al, 2015; Vitali

and Pernici, 2015). Smart living environments for

elderly people can be fitted with home sensors, and

monitoring using wearable devices can include not

only the elderly but also their carers (Wolf et al, 2013)

and this allows very rich data to be gathered for

process analysis. For example, the data gathered and

analysed in nursing homes conducted by Llatas et al,

(2011) helped to detect abnormal behaviour by

elderly people who suffer from dementia and

cognitive impairment.

Some of the papers presented additional

visualizations of processes. Tax et al, (2018) used a

plot of the events from the log file as a dotted chart

with coloured dots representing events over time.

Conca et al, (2018) created process models from the

collaboration patterns they found. Najjar et al, (2018)

used the percentage of participation, referral and self-

referral to describe significant patterns.

Our search has some limitations. The review was

limited to papers on the recent status of process

mining related to frail elderly patients. The inclusion

and exclusion decisions were made by a single

reviewer. Google Scholar matching for search criteria

can have variability in its results. However the search

was comprehensive and followed well established

methods.

To date there has been no work that directly

addresses the care and management of the condition

of frailty in elderly patients. This despite the growing

number of older adults globally and the growing

recognition in the medical world that frailty demands

specific attention as a complex set of diseases and

needs. Kim and Jang (2018) have argued that,

historically, medical attention has often focused on

single diseases and the study and management of

frailty introduces a more holistic approach centred on

the patients, their experience of a range of often

interconnected diseases and the specific needs that

frail individuals have to maintain the best possible

quality of life.

5 CONCLUSION

Process mining is an emerging field within data

science and presents a fresh set of methods for

process improvement in healthcare. Our group are

developing methods for meaningful care pathway

analysis using clinical reference groups and

multidisciplinary domain experts to iteratively

improve understanding of current pathways and

identify potential improvements.

This paper has reviewed the small number of

papers that use process mining of both sensor data and

electronic health records for older adults and those

with frailty. These demonstrate the potential for

process mining to play an important role in improving

our understanding of how best to manage the care of

the frail elderly. However the opportunities for

researchers to apply process mining to improve frailty

care has not yet been explored in any detail. Our

review of the literature shows that novel approaches

are just starting to emerge. Our next step is to use

primary care data from the UK NHS to examine how

care pathways vary between different categories of

frailty over time and the relationship this has to

prescribing patterns.

ACKNOWLEDGEMENTS

The first author would like to thank the Ministry of

Higher Education of Malaysia for the sponsorship and

all the support given throughout the research. All

authors gratefully acknowledge support from the

Connected Health Cities (CHC) programme

commissioned by the Northern Health Science

Alliance (NHSA) and funded by the UK Department

of Health.

HEALTHINF 2019 - 12th International Conference on Health Informatics

336

REFERENCES

Aalst, W.M. van der., Günther, C., Recker, J.C. and

Reichert, M., 2006. Using process mining to analyze

and improve process flexibility-position paper. In The

18th International Conference on Advanced

Information Systems Engineering. Proceedings of

Workshops and Doctoral Consortium (pp. 168-177).

Namur University Press.

Aalst, W.M. van der, Reijers, H.A., Weijters, A.J., Dongen,

B.F. van, De Medeiros, A.A., Song, M. and Verbeek,

H.M.W., 2007. Business process mining: An industrial

application. Information Systems, 32(5), pp.713-732.

Aalst, W.M. van der., 2011. Process Mining: Discovery,

Conformance and Enhancement of Business Processes.

Springer-Verlag, Berlin.

Aalst, W.M. van der, 2011. Process Discovery: An

Introduction in Process Mining, Springer Berlin

Heidelberg, pp. 125-156.

Aalst, W.M. van der, Adriansyah, A., De Medeiros, A.K.A.,

Arcieri, F., Baier, T., Blickle, T., Bose, J.C., Van Den

Brand, P., Brandtjen, R., Buijs, J. and Burattin, A.,

2011. Process mining manifesto. In International

Conference on Business Process Management (pp.

169-194). Springer, Berlin, Heidelberg.

Aalst, W.M. van der., 2015. Extracting event data from

databases to unleash process mining. In BPM-Driving

innovation in a digital world (pp. 105-128). Springer,

Cham.

Chamberlain A.M., Sauver J.L., Jacobson D.J., Manemann

S.M., Fan C., Roger V.L., Yawn B.P. & Rutten L.J.F.,

2016. Social and behavioural factors associated with

frailty trajectories in a population-based cohort of older

adults, British Medical Journal Open, 6(5), e011410.

Clegg A., Young J., Iliffe S., Rikkert M.O., Rockwood K.,

2013. Frailty in elderly people, The Lancet, 381(9868),

pp. 752-762.

Conca, T., Saint-Pierre, C., Herskovic, V., Sepúlveda, M.,

Capurro, D., Prieto, F. and Fernandez-Llatas, C., 2018.

Multidisciplinary Collaboration in the Treatment of

Patients with Type 2 Diabetes in Primary Care:

Analysis Using Process Mining. Journal of medical

Internet research, 20(4).

Crandall M, Duncan T., Mallat A., Greene W., Violano P.,

Christmas A.B., Barraco R., 2016. Prevention of fall-

related injuries in the elderly: An eastern association for

the surgery of trauma practice management guideline,

The Journal of Trauma and Acute Care Surgery.

Eck, M.L. van, Lu, X., Leemans, S.J. and Aalst, W.M. van

der, 2015. PM^2: A Process Mining Project

Methodology. In International Conference on

Advanced Information Systems Engineering (pp. 297-

313). Springer, Cham.

Eeles E.M., White S.V., O’Mahony S.M., Bayer A.J.,

Hubbard R.E., 2012. The impact of frailty and delirium

on mortality in older patients, Age and Ageing, 41(3),

pp. 412-416.

Erdogan, T.G. and Tarhan, A., 2018. Systematic Mapping

of Process Mining Studies in Healthcare. IEEE

ACCESS, 6, pp.24543-24567.

Fatima, I., Fahim, M., Lee, Y.K. and Lee, S., 2013. A

unified framework for activity recognition-based

behavior analysis and action prediction in smart

homes. Sensors, 13(2), pp.2682-2699.

Fernández-Llatas, C., Garcia-Gomez, J.M., Vicente, J.,

Naranjo, J.C., Robles, M., Benedi, J.M. and Traver, V.,

2011, August. Behaviour patterns detection for

persuasive design in nursing homes to help dementia

patients. In Engineering in medicine and biology

society, EMBC, 2011 annual international conference

of the IEEE (pp. 6413-6417). IEEE.

Fried L.P., Tangen C.M., Walston J., Newman A.B.,

Hirsch, C., Gottdiener J., Seeman T., Tracy R., Kop

W.J., Burke G., McBurnie M.A., 2001. Frailty in older

adults evidence for a phenotype, The Journals of

Gerontology Series A: Biological Sciences and Medical

Science, 56(3), pp. M146-M157.

Kim, Y. and Jang, S.N., 2018. Mapping the Knowledge

Structure of Frailty in Journal Articles by Text Network

Analysis. PloS One, 13(4), p.e0196104.

Kurniati, A.P., Johnson, O., Hogg, D. and Hall, G., 2016.

Process mining in oncology: A literature review.

In Information Communication and Management

(ICICM), International Conference on (pp. 291-297).

IEEE.

Kusuma, G.P., Hall, M., Gale, C. and Johnson, O., 2018.

Process Mining in Cardiology: A Literature

Review. International Journal of Bioscience,

Biochemistry and Bioinformatics, 8(4), pp.226-236.

Lekan D.A, Wallance D.C., McCoy T.P., Hu J., Silva S.G.,

Whitson H.E., 2017. Frailty Assessment in

Hospitalized Older Adults Using the Electronic Health

Record, Biological Research for Nursing, 19(2), 213-

228.

Mans, R.S., Schonenberg, M.H., Song, M., van der Aalst,

W.M. and Bakker, P.J., 2008, January. Application of

Process Mining in Healthcare – a Case Study in a Dutch

Hospital. In International Joint Conference on

Biomedical Engineering Systems and

Technologies (pp. 425-438). Springer, Berlin,

Heidelberg.

Mitnitski A.B., Mogilner A.J., Rockwood K., 2001.

Accumulation of deficits as a proxy measure of aging,

The Scientific World Journal, pp. 323-336.

Munstermann, M., Stevens, T. and Luther, W., 2012. A

Novel Human Autonomy Assessment

System. Sensors, 12(6), pp.7828-7854.

Najjar, A., Reinharz, D., Girouard, C. and Gagné, C., 2018.

A Two-step Approach for Mining Patient Treatment

Pathways in Administrative Healthcare

Databases. Artificial Intelligence in Medicine, 87,

pp.34-48.

Oliver D., 2009. Age Based Discrimination in Health and

Social Care Services”. British Medical Journal, 339:

b3378,.

Partington, A., Wynn, M., Suriadi, S., Ouyang, C. and

Karnon, J., 2015. Process Mining for Clinical

Processes: a Comparative Analysis of Four Australian

Hospitals. ACM Transactions on Management

Information Systems (TMIS), 5(4), p.19.

Process Mining in Frail Elderly Care: A Literature Review

337

Rojas, E., Munoz-Gama, J., Sepúlveda, M. and Capurro, D.,

2016. Process Mining in Healthcare: A Literature

Review. Journal of Biomedical Informatics, 61,

pp.224-236.

Sztyler, T., Völker, J., Carmona Vargas, J., Meier, O. and

Stuckenschmidt, H., 2015. Discovery of personal

processes from labeled sensor data: An application of

process mining to personalized health care.

In Proceedings of the International Workshop on

Algorithms & Theories for the Analysis of Event Data:

Brussels, Belgium, June 22-23, 2015 (pp. 31-46).

CEUR-WS. org.

Tapia, E.M., Intille, S.S. and Larson, K., 2004. Activity

Recognition in the Home using Simple and Ubiquitous

Sensors. In International Conference on Pervasive

Computing (pp. 158-175). Springer, Berlin,

Heidelberg.

Tax, N., Sidorova, N., Haakma, R. and van der Aalst, W.M.,

2018. Mining Local Process Models with Constraints

Efficiently: Applications to the Analysis of Smart

Home Data. Medicine, 24, p.48.

Triki, S., Hanachi, C., Gleizes, M.P., Glize, P. and Rouyer,

A., 2015. Modelling and Simulating Collaborative

Scenarios for Designing an Assistant Ambient System

that Supports Daily Activities. In Computational

Collective Intelligence (pp. 191-202). Springer, Cham.

United Nations, Department of Economic and Social

Affairs, Population Division, 2017. World Population

Ageing 2017 – Highlights (ST/ESA/SER.A/397).

Vitali, M. and Pernici, B., 2015. Pie-processes in Events:

Interconnections in Ambient Assisted Living. In OTM

Confederated International Conferences" On the Move

to Meaningful Internet Systems" (pp. 157-166).

Springer, Cham.

Weber, P., Backman, R., Litchfield, I. and Lee, M., 2018.

A Process Mining and Text Analysis Approach to

Analyse the Extent of Polypharmacy in Medical

Prescribing. In 2018 IEEE International Conference on

Healthcare Informatics (ICHI) (pp. 1-11). IEEE.

Williams, R., Rojas, E., Peek, N. and Johnson, O.A., 2018.

Process Mining in Primary Care: A Literature

Review. Studies in Health Technology and

Informatics, 247, pp.376-380.

Wolf, H., Herrmann, K. and Rothermel, K., 2013. Dealing

with uncertainty: Robust workflow navigation in the

healthcare domain. ACM Transactions on Intelligent

Systems and Technology (TIST), 4(4), p.65.

Xue Q.L, 2011. “The frailty syndrome: definition and

natural history,” Clinics in Geriatric Medicine, 27(1),

pp. 1-15.

HEALTHINF 2019 - 12th International Conference on Health Informatics

338

APPENDIX

Table 2.

Paper Summary

Authors Year Setting Domain Data Sources Case

Persp

ective

Challenges No of

population

Najjar et

al.

2018 Quebec,

Canada

Heart

failure

EHR of acute

care and one

day surgery

Yes Data (pre-

processing) and

team limitation

180,027

Conca et

al.

2018 Chile,

Santiago

Type 2

Diabetes

EHR of

administrative

data

No Data

(inconsistent

and/or

incomplete) and

team limitation

2,843

Llatas et

al.

2018 Spain Dementia Nursing home Yes Data limitation

- (high

granularity)

One

Tax et al. 2018 Eindhoven,

Netherlands

Elderly

behaviour

Secondary

data (Tapia &

et al., 2004)

of elderly

living in smart

environment

Yes

Data limitation

- (high

granularity)

Four

different

datasets

with

number of

events

ranging

from 220 to

1,962

Vitali and

Pernici

2016 Milano,

Italy

Elderly

behaviour

Sensor

collected from

smart living

environment

of home care

Yes Data limitation

- (high

granularity)

One person

with nine

classificatio

n of events

Triki et

al.

2015 Toulouse,

France

Elderly

behaviour

(outdoors)

Scenario

generator of

elderly e.g.

outdoor

activities

Yes Data limitation

- (high

granularity)

One person

Several

scenarios

consist of

three actors

Wolf et

al.

2013 Mainkofen,

Germany

Dementia Geriatric ward

of nursing

home

Yes Data limitation

- (high

granularity)

135

randomly

selected

person

Munsterm

ann et al.

2012 Duisburg,

Germany

Elderly

behaviour

Patient who

uses ambulant

service

Yes Data limitation

- (high

granularity)

Five

Process Mining in Frail Elderly Care: A Literature Review

339