Journal of Infection and Public Health (2016) 9, 766—773

Towards an evaluation framework forLaboratory Information Systems

Maryati M. Yusof ∗, Azila Arifin

Centre for Software Technology & Management, Faculty of Information Science andTechnology, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia

Received 24 June 2016; received in revised form 20 July 2016; accepted 24 August 2016

KEYWORDSError;Evaluation;Framework;Total Testing Process;Laboratory InformationSystems

SummaryIntroduction: Laboratory testing and reporting are error-prone and redundant dueto repeated, unnecessary requests and delayed or missed reactions to laboratoryreports. Occurring errors may negatively affect the patient treatment process andclinical decision making. Evaluation on laboratory testing and Laboratory Informa-tion System (LIS) may explain the root cause to improve the testing process andenhance LIS in supporting the process. This paper discusses a new evaluation frame-work for LIS that encompasses the laboratory testing cycle and the socio-technicalpart of LIS.Methodology: Literature review on discourses, dimensions and evaluation methodsof laboratory testing and LIS. A critical appraisal of the Total Testing Process (TTP)and the human, organization, technology-fit factors (HOT-fit) evaluation frameworkswas undertaken in order to identify error incident, its contributing factors andpreventive action pertinent to laboratory testing process and LIS.Result: A new evaluation framework for LIS using a comprehensive and socio-technical approach is outlined. Positive relationship between laboratory and clinicalstaff resulted in a smooth laboratory testing process, reduced errors and increasedprocess efficiency whilst effective use of LIS streamlined the testing processes.

Conclusion: The TTP-LIS framework could serve as an assessment as well as aproblem-solving tool for the laboratory testing process and system.© 2016 King Saud Bin Abdulaziz University for Health Sciences. Published by ElsevierLimited. All rights reserved.

∗ Corresponding author.E-mail addresses: Maryati.Yusof@ukm.edu.my (M.M. Yusof),

azilaarifin@ymail.com (A. Arifin).

I

Lo

http://dx.doi.org/10.1016/j.jiph.2016.08.0141876-0341/© 2016 King Saud Bin Abdulaziz University for Health Scie

ntroduction

aboratory testing errors can happen at any stagef the testing process, from the pre-analytic

nces. Published by Elsevier Limited. All rights reserved.

T stem

si(ni(pthweriaoLmritpoHaai[wtc

ewasowTrtbbia

T

Lgrtfwd

mtete(tpiroaitewsTfL

T

TpfnonotFao

LcpftLntTearacaam

owards a framework for Laboratory Information Sy

teps (for example, test selection and order-ng, specimen collection) to post-analytic stepsfor example, reporting and interpreting results,otifying patients) [1]. Errors in laboratory test-ng generally include unreasonable testing ordere.g.: additional test copy, duplicate test); wrongatient identification/specimen/labelling; uniden-ified failure in quality control; problems inandling, storing and transporting test sample;rong validation of data analysis; and data entryrror [1—5]. Various strategies have been used toeduce error and monitor workflow performancen laboratory including quality control programmend Information Systems/Technology [2,3]. The usef Health Information Systems (HIS), particularlyaboratory Information Systems (LIS) to validate,anage, deliver, process, and store data should

educe problems and ease process implementationn the laboratory testing workflow [6]. LIS facili-ate smooth and fast interaction between medicalractitioners and laboratory staff, specifically inrdering tests and delivering test reports [7—9].owever, numerous error factors related to LIS havelso been reported including wrong data entry andccess; poor system interface and reporting; lim-ted system functionality; and incompetent users1,10,11]. The involvement of multiple units in aorkflow requires effective methods to monitor the

ask performance as the method would ensure pro-ess smoothness and ease error detection.

The paper aims to discuss influencing factors forrrors in laboratory testing processes and LIS asell as to present our proposed framework knowns TTP-LIS which combined laboratory process andocio-technical factors. The framework makes usef the original Total Testing Process (TTP) frame-ork and combines it with Human, Organization,

echnology-fit (HOT-fit) framework [12]. Laboratoryelated errors are briefly described in this introduc-ion section. Section two discusses the theoreticalackground of TTP and HOT-fit frameworks; theasis of our proposed framework. The third sectionllustrates the new framework whilst the discussionnd conclusion are included in the last section.

heoretical background

IS supports laboratory requirements [13] and inte-rates multiple laboratories [8]. However, the LISole in preventing recurring error in laboratory

esting process is still a work in progress. Plat-orm heterogeneity in lab-clinical settings [14,15]hich involve system development, software use,iscrepancy in technology management and infor-

s

fia

evaluation 767

ation systems used in both settings, contributeo error incidents. In order to identify root causes,rror incident in laboratory testing processes needso be evaluated rigorously. Laboratory testing gen-rally consists of nine steps: (1) test request;2) sample collection; (3) sample labelling; (4)ransportation of labelled sample to the lab; (5)reparation of raw specimen; (6) analysis of spec-men testing; (7) interpretation of test results; (8)eporting of test interpretation; and (9) archivingf test results [16]. These steps are represented in

framework known as Total Testing Process (TTP);t can be used to evaluate laboratory process whilehe socio-technical aspects of LIS require anothervaluation framework called the HOT-fit frameworkhich takes a socio-technical approach to repre-

ent the interaction of social and technical in IS.he following section elaborates TTP and HOT-fitramework and their relationship that formed TTP-IS.

otal Testing Process (TTP) framework

TP is used as a basic guideline in the testingrocess of medical laboratories. It is a uniqueramework for analysing and minimising error riskot only in laboratory test centre but also inther clinical units [17]. TTP encompasses inter-al and external laboratory activities that comprisef one or more procedures and require interac-ion between internal and external laboratory staff.ailure in any TTP activity can affect patient cares doctors make decisions based on clinical resultsbtained from laboratory [18].

The original TTP framework was introduced byundberg [19], known as brain-to-brain loop con-ept (Fig. 1). The concept has been used by medicalractitioners in conducting lab testing processes;rom a triggered idea to testing patient samples toaking action in treating a patient. The simplifiedundberg concept in Fig. 1 illustrates the thorough-ess of laboratory processes, from ordering testso generating and utilising laboratory test results.he evidence of implementation effectiveness inach step indicates error reduction in patient carend treatment. TTP workflow in medical laborato-ies also focuses on process smoothness as smoothnd systematic process yield to effective qualityontrol. Process thoroughness based on productivend ethical work culture is critical in maintainingnd improving workflow quality as it contributes toinimising error and subsequently ensuring patient

afety [20].The testing steps in Fig. 1 can be aggregated into

ve phases namely pre-pre-analytic, pre-analytic,nalytic, post-analytic and post-post-analytic, as

768 M.M. Yusof, A. Arifin

Discussion for laboratory tests requirement

Specimen / pati ent identi fica tion

Laboratory tests selection

Laboratory test requ est

Analyiss of the laboratory test result

Spec imen coll ecti on

Laboratory test preparati on

Spec imen transportati on

Laboratory test results

Repo rts interpretati on

Action / treatment

Figure 1 Brain-to-brain

Pre-pre-analytical Pre- analyti cal

Analytica l

Post- analytica l Post-post- analytica l

ahmt

H

Tfo(fiItvpsi(msitaciadpaif

T

Timised laboratory testing process and LIS flow as well

Figure 2 TTP workflow [3].

modelled in Hawkins [3] (Fig. 2). Most studies didnot introduce the first and last phases in the TTPframework to classify activities based on the brain-to-brain loop concept. Pre-pre-analytic phase takesplace outside clinical laboratory and post-post-analytic phase involved activities within laboratory[17,21]. Phase based activity approach can be usedto identify whether an error initiates before, dur-ing, or after the laboratory test [3]. Early errordetection could potentially prevent the same errorfrom recurring.

The brain-to-brain loop concept and phase chainfrom related, disparate studies were combined andillustrated in Fig. 3. Each activity can be identifiedbased on these phases; this structure eased erroridentification and classification as well as facili-tated doctor, clinical specialist, and lab staff toperform and monitor lab activity smoothly and thor-oughly [3].

Most studies on TTP identified error incidentsthat occur in all TTP phases; however, the firstand last phase have the highest error percent-age [3,22] due to the absence of monitoring onexternal laboratory processes. An error that occursin TTP is regarded as a laboratory error althoughit happens outside the laboratory control. Con-ditions that contributed to those errors includepoor communication; action taken by individualinvolved in laboratory testing process (doctor,nurse, and phlebotomists) such as role confusion;

and ineffective process flow such as incompleteand redundant process steps. Therefore, the Inter-national Standardisation Organisation recommends

tad

loop concept [19].

wider definition for laboratory error. To ensureigh quality laboratory service, error risk must beinimised, particularly before and after laboratory

esting process [17].

OT-fit framework

he HOT-fit evaluation framework [12,23] for HISeatured comprehensive dimensions and measuresf technology, human, and organisation factorsFig. 4). The adaption of two IS models in HOT-t framework, namely IS Success Model [24] and

T-Organization Fit Model [25] enables HOT-fito become a comprehensive evaluation tool forarious HIS, including LIS. The framework is com-rised of nine interrelated dimensions, namelyystem quality (information processing quality),nformation quality (IS output), service qualitytechnical and service support), system develop-ent, system use, user satisfaction, organisational

tructure (related to management, strategy, organ-sation plan), organisation environment (relatedo politics, finance, inter organisation systems)nd net benefits (overall IS impact). The fit con-ept between technology, human and organisationn the HOT-fit framework is complex, subjective,nd abstract [12,23]. Based on its comprehensiveimension, HOT-fit is not only used to evaluate HISerformance, efficiency, and HIS impact. It couldlso guide error evaluation systematically accord-ng to process phase and level from the threeactors.

he proposed TTP-LIS framework

he proposed framework aims to provide betterllustration of systematic, coordinated, and opti-

o facilitate a rigorous error evaluation. The evalu-tion dimension, process and their relationships areepicted in Fig. 5. The study focused only on pre-

Towards a framework for Laboratory Information System evaluation 769

Figure 3 Combination of brain-to-brain loop concept and TTP phase.

t fra

pT

•

•

•

•

•

•

Figure 4 HOT-fi

re-analytic and post-post-analytic phases of theTP framework.

Pre-pre-analytic phase

The first two processes, namely ‘discussion oflaboratory test requirement’ and ‘laboratory testselection’ involved a ‘human’ role that dependson knowledge, training, commitment, credibility,and patient condition.

A ‘Laboratory test order’ process is made bya doctor/medical specialist/nurse through ‘sys-tem use’ by entering the applicant information,patient information and specimen record.

The ‘Identify laboratory test’ process should becarefully performed by a doctor/nurse beforelaboratory test information and specimen arebrought to the laboratory.

mework [12,26].

Pre-analytic phase

Laboratory test order information is accessed bylaboratory staff through ‘system use’ for furtheraction.

The ‘Identify/check laboratory test’ process isperformed by laboratory staff involved in speci-men collection, specimen identification creationusing bar code, management of laboratory testuse, and monitoring duration.

Post-analytic phase

Laboratory staff enter laboratory test results

through ‘system use’ based on a matching lab-oratory test code.

Post-post-analytic phase

770 M.M. Yusof, A. Arifin

sed

•

•

use yields to negative ‘net benefits’.

Figure 5 The propo

• Doctor/medical specialist accesses laboratorytest result from ‘system use’ and interpret theresults in a report form.

• Doctor reports laboratory test and uses it todetermine further ‘treatment’ on patient.

Human and technology category

• ‘System quality’, ‘information quality’ and ‘ser-vice quality’ influence ‘system use’ and ‘usersatisfaction’.

• ‘System use’ and ‘information quality’ influenceeach other as the generation of laboratory test

result, report, and image from system dependson user knowledge, skill, and training [23].

• Level of ‘information quality’ also influences‘user satisfaction’ and vice versa. Feedbacks t

TTP-LIS framework.

from users pertinent to information qualityshould improve the level of information quality.

‘System use’ and ‘user satisfaction’ influenceeach other. Effective ‘system use’ that includeconditions such as LIS-task fit, low error rate,and user friendly interface could encourage usersto optimise system use, which subsequentlyincreases ‘user satisfaction’.

‘System use’ and ‘user satisfaction’ result indirect or indirect ‘net benefits’ negatively or pos-itively. Likewise, intensive ‘system use’ results inpositive ‘net benefits’ while ineffective system

The enhancement in the TTP-LIS framework aimso facilitate basic laboratory test procedures; coor-

T stem

dhdeposial

T

SaTtsltpprt

tmcofiriultt

othrsepsp

dria[ua

aM(utramcrt

H

TefoaRuolusMoufaa

O

Cimpactt

eKiicipb

owards a framework for Laboratory Information Sy

inating the process with system use to reduceuman and technological error as well as to pro-uce smoother workflow. Therefore, evaluation ofrror incidents in laboratory test processes can beerformed by analyzing factors related to human,rganisation and technology—–such as the ease ofystem use and learning, system flexibility, relevantnformation, user attitude, planning, strategy, man-gement and communication between doctor andaboratory staff.

echnology

ystem quality is related to system performancend interface [6]. Elements of system quality inTP-LIS measure system performance from sys-em design, system function, communication, andynchronisation between systems in clinical andaboratory settings. In the system quality con-ext, error incidents are contributed by systemerformance such as system development incom-atibility, misfit of system function with taskequirement, and poor log system and communica-ion [7,8,27,28].

Incompatible system platform and software yieldo disrupted system interaction. As a result, infor-ation and image are inaccessible and unreadable,

ausing difficulty for doctor to make decisionn patient diagnosis or treatment. Poor systemunction is attributed to various factors, includ-ng management disputes and unclear/missing userequirements. A poor log system hindered monitor-ng user activity such as unauthorised system use orser negligence to log out. An automated telephoneog is available in LIS to track and monitor labora-ory test results that may have been discussed overhe telephone [27].

High system quality is associated with its easef use; for example, instant reference for sys-em functions through tooltips. System training alsoelps user to become competent. System flexibilityefers to the ability of a system to adapt to a worketting and integrate with other systems [23]. Forxample, a patient history, treatment plan or planrepared by medical specialist are also made acces-ible to other specialists involved with the sameatient case [29].

Information quality is measured from systemsisplays in various forms such as patient record,eport, image, and prescription. Information qual-ty is subjected to user perspective on information

ccuracy, completeness, consistency, and legibility23]. Poor information quality can originated fromsers with limited information literacy, education,nd communication skills.

tsro

evaluation 771

Service quality is measured from responsiveness,ssurance, empathy, and follow up service [12].ost users think that management unit responses

responsiveness) to the requirements of systemse and function are based on their own perspec-ives instead of those of user. A service provider isesponsible for adhering to agreed system featuresnd functions and assurance to support the require-ent of user tasks. Personal and individual interests

an foster in management lack of empathy in userequirements that are critical in supporting themo perform their daily tasks smoothly.

uman

he potential of system impact and overall userxperience in using systems is defined as user satis-action [23]. User satisfaction influences the levelf system use; it also affects patient treatmentnd organisational performance in the long run.educed error due to system use could increaseser satisfaction. Users who receive good qualityf service and information show higher satisfactionevels through increased system use. [30] Systemse could be viewed as a benchmark to assesservice, system, and information quality [31—33].easures of system use include frequency of use,utput information and volunteered or mandatoryse [12,23]. Our main evaluation focus includesrequency, type, and number of error incidentttributed to system function, relevant module,nd frequency of system use.

rganization structure

linical process is one of the measures in organ-sation structure in the HOT-fit framework andatches with TTP-LIS, renamed as laboratory testrocess. Human related error can occur intention-lly or accidentally. The error usually needs to beorrected or mitigated through rigorous evaluationo avoid recurring incidents and adverse effectshat demand time, cost, and manpower.

Error evaluation can be identified and cat-gorised according to knowledge or procedure.nowledge is not only limited to health and med-cal discipline, it is also related to knowledgen information systems and technology, effectiveommunication, and process flow. Any process,ncluding a laboratory test, should follow specificrocedures that can be categorised into good andad states. A smooth process indicates the prac-

icality of its procedure while a disrupted processhows the need for procedure improvement. A dis-upted process is frequently associated with lackf user training or exposure. Knowledge related

lssr

ibapeidilucTcaso(

F

N

C

N

E

N

A

WfSJsseo

772

error generally occurs when user or medical practi-tioner faces a rare situation that requires an urgentsolution. As a consequence, the problem is solvedthrough reasoning or assumption and estimationthat are prone to error risk due to limited knowl-edge sources, reliance on the current situation, andthe use of individual intuition or hypothesis [34].

Net benefits

LIS is seen as a system that eases laboratory tasksand facilitates communication between laboratoryand clinical units to enable faster delivery of lab-oratory test orders and reports. Comprehensivesystem use in laboratory and clinical units lead tocontinuous system improvement as system weak-ness can be identified earlier. This would enablesystem developers to analyse problems that triggerthe error occurrence.

Error evaluation that involves process flow out-side and within laboratories shows the importanceof cooperation between laboratory and clinicalunits. System impacts on the initial and final phaseof laboratory testing process can be evaluatedin their instruction/procedure compliance, taskperformance, efficiency, effectiveness, accuracy,synchronisation (related to system development interms of platform, software and tool), informationaccess, decision quality, and time.

Discussion and conclusion

Aligning IS use with clinical workflow to fulfil actualwork reality in healthcare is a challenging task. Therelationship between and combination of TTP andLIS formed our proposed framework, known as TTP-LIS that aims to facilitate the evaluation of errorincident for laboratory setting. We analysed theexisting frameworks to identify their suitability inaddressing error incidents related to TTP and LISthrough their strengths and limitations and subse-quently extend TTP to construct a new framework.Its comprehensive measures that encompass theoverall socio-technical HOT-fit framework enableda rigorous evaluation. The combination of fac-tor and dimension in the HOT-fit and TTP modelsresulted in a comprehensive laboratory test processflow and HIS evaluation dimensions.

LIS plays an important role in managing labo-ratory test process in clinical unit and laboratory.

However, the misfit of technology with healthorganisation structure and clinical practice in lab-oratory testing process resulted in various errors.Therefore, continuous evaluation in the overall

R

M.M. Yusof, A. Arifin

aboratory testing process is crucial in addressingystem problems and creating user awareness ofystem potentials and advantages to overall labo-atory and clinical units.

To validate its usefulness, TTP-LIS can be testedn clinical and laboratory settings, preferably withaseline data. Access to relevant documents suchs incident report, error management report, androcedure process flow could be very useful invaluating error incident factor. However, obtain-ng access to these documents may be challengingue to their sensitivity, privacy and confidentialityssues. Synergy and cooperation between clinical,aboratory, and IT, and support from managementnits, is required to improve laboratory testing pro-ess and LIS usefulness. Evaluation measures inTP-LIS could be extended to evaluate factors thatontributes to error in laboratory testing processesnd LIS that are caused by external factors such asystem incompatibility that affect LIS capabilitiesr other factor related to organisation managementlatent failure).

unding

o funding sources.

ompeting interests

one declared.

thical approval

ot required.

cknowledgements

e gratefully acknowledge the funding receivedrom the Malaysia Exploratory Research Grantcheme ERGS/1/2011/STG/UKM/02/46 and theapan Sumitomo Foundation Grant that helpedponsor this study. The project was scientificallyupported by King Saud University, Deanship of Sci-ntific research, research chairs and Research Chairf Health Informatics and Promotion.

eferences

[1] Fernald D, Hamer M, James K, Tutt B, West D. Launch-ing a laboratory testing process quality improvement

T stem

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

1996;72:201—14.

owards a framework for Laboratory Information Sy

toolkit: from the Shared Networks of Colorado Ambula-tory Practices and Partners (SNOCAP). J Am Board Fam Med2015;28:576—83.

[2] Carraro P, Plebani M. Errors in a stat laboratory: typesand frequencies 10 years later. Clin Chem: Lab Manag2007;53:1338—42.

[3] Hawkins R. Managing the pre-and post-analytical phases ofthe total testing process. Ann Lab Med 2012;32:5—16.

[4] Plebani M, Piva E. Medical errors: pre-analytical issue inpatient safety. J Med Biochem 2010;29:310—4.

[5] Plebani M. Errors in clinical laboratories or errors inlaboratory medicine? Clin Chem Lab Med: CCLM/FESCC2006;44:750—9.

[6] Yusof MM, Papazafeiropoulou A, Paul RJ, Stergioulas LK.Investigating evaluation frameworks for health informationsystems. Int J Med Inform 2008;77:377—85.

[7] Lichenstein R, O’Connell K, Funai T, Blumberg S, Shaw K,Ruddy R, et al. Laboratory errors in a Pediatric EmergencyDepartment Network—–an analysis of incident reports. Pedi-atr Emerg Care Appl Res Netw 2015;00:1—5.

[8] Blaya JA, Shin SS, Yagui M, Contreras C, Cegielski P, Yale G,et al. Reducing communication delays and improving qualityof care with a tuberculosis laboratory information systemin resource poor environments: a cluster randomized con-trolled trial. Open Access J: PLoS One 2014;9:1—7.

[9] Kristensen GB, Moberg Aakre K, Kristoffersen AH, SandbergS. How to conduct external quality assessment schemes forthe pre-analytical phase? Biochem Med 2014;24:114—22.

10] Takian A, Sheikh A, Barber N. We are bitter, but we arebetter off: case study of the implementation of an elec-tronic health record system into a mental health hospitalin England. BMC Health Serv Res 2012;12.

11] Ash JS, Berg M, Coiera E. Some unintended consequences ofinformation technology in health care: the nature of patientcare information system-related errors. J Am Med InformAssoc 2004;11:104—12.

12] Yusof MM. A case study evaluation of a Critical Care Infor-mation System adoption using the socio-technical and fitapproach. Int J Med Inform 2015;84:486—99.

13] McGrowder D, Bishop R. An evaluation of laboratoryinformation systems in medical laboratories in Jamaica.In: Moumtzoglou A, Kastania A, Archondakis S, edi-tors. Laboratory management information systems: currentrequirements and future perspectives. Hershey, PA; 2015.

14] Plebani M. Harmonization in laboratory medicine: thecomplete picture. Clin Chem Lab Med: CCLM/FESCC2013;51:741—51.

15] Plebani M, Panteghini M. Promoting clinical and laboratoryinteraction by harmonization. Clin Chim Acta: Int J ClinChem 2013;432:15—21.

16] Hopp WJ, Lovejoy WS, Myers J. Diagnostic services. In: HoppWJ, Lovejoy WS, editors. Hospital operations: principles ofhigh efficiency health care. New Jersey: Pearson Education;2014. p. 339—462.

[

Available online at www

ScienceD

evaluation 773

17] Plebani M. The detection and prevention of errors in labo-ratory medicine. Ann Clin Biochem 2010;47:101—10.

18] O’Kane M. The reporting, classification and grading ofquality failures in the medical laboratory. Clin Chim Acta2009;404:28—31.

19] Lundberg GD. How clinicians should use the diagnosticlaboratory in a changing medical world. Clin Chim Acta1999:3—11.

20] Ellwood PM, Lundberg GD. Managed care: a work inprogress. JAMA 1996;276:1083—6.

21] Laposata M, Dighe A. ‘‘Pre-pre’’ and ‘‘post-post’’ analyticalerror: high-incidence patient safety hazards involving theclinical laboratory. Clin Chem Lab Med 2007;45:712—9.

22] Deetz CO, Nolan DK, Scott MG. An examination of the use-fulness of repeat testing practices in a large hospital clinicalchemistry laboratory. Am J Clin Pathol 2012;137:20—5.

23] Yusof MM, Kuljis J, Papazafeiropoulou A, Stergioulas LK.An evaluation framework for health information systems:human, organization and technology-fit factors (HOT-fit).Int J Med Inform 2008:1—14.

24] Delone WH, McLean ER. The DeLone and McLean model ofinformation systems success: a ten-year update. J ManagInf Syst 2003;19:9—30.

25] Scott Morton MS. The corporation of the 1990. New York:Oxford University Press; 1991.

26] Yusof MM. HOT-fit evaluation framework: validation usingcase studies and qualitative systematic review in healthinformation systems evaluation adoption. In: Proceedings ofthe 5rd European conference on information managementand evaluation. 2011. p. 359—65.

27] Nguyen MT, Fuhrer P, Pasquier-Rocha J. Enhancing E-healthinformation systems with agent technology. Int J TelemedAppl 2009:1—13.

28] Plebani M, Sciacovelli L, Aita A, Padoan A, Chiozza ML. Qual-ity indicators to detect pre-analytical errors in laboratorytesting. Clin Chim Acta: Int J Clin Chem 2013:1—5.

29] Becher EC, Chassin MR. Improving quality, minimizing error:making it happen. Health Aff 2001;3:68—81.

30] Markovic S, Loncaric D, Loncaric D. Service quality andcustomer satisfaction in the health care industry—–towardshealth tourism market. Tour Hosp Manag 2014;20:155—70.

31] Antreas D, Opoulos A. Modeling customer satisfactionin telecommunication: assessing the multiple transactionpoints on perceived over all performance of the provider.Prod Oper Manag 2003;12:224—45.

32] Cronin Jr JJ, Taylor SA. Measuring service quality: a reex-amination and extension. J Mark 1992;56:55—68.

33] Spreng RA, Mackoy RD. An empirical examination of amodel of perceived service quality and satisfaction. J Retail

34] Reason J. Safety in the operating theatre—–Part 2: humanerror and organisational failure. Qual Saf Health Care2005;14:56—60.

.sciencedirect.com

irect