A decision support system for telemedicine needs assessments in South Africa

Date
2012-03
Authors
Treurnicht, Maria Jacoba
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: The various applications of Information and Communication Technologies (ICTs) in healthcare are increasingly effective to improve the cost-effectiveness and quality of healthcare service delivery. Telemedicine is such an application, using ICTs to provide health services over a distance. Since 1997, the South African Department of Health has invested large amounts of capital to implement telemedicine systems in South Africa. Unfortunately, telemedicine programs have had little success since, leading to many workstations standing dormant. Telemedicine implementation decision making that is based on insufficient evidence is identified as one of the underlying problems that cause telemedicine programs to fail. It is proposed that implementation decisions should be based on quantifiable evidence regarding the potential benefits of telemedicine. A decision support system is developed that can be used to quantify potential benefits and plan telemedicine implementation programs accordingly. The decision support system is modelled and demonstrated using data from the Eastern Cape public health sector. The first phase of the system guides decision makers to identify potential telemedicine benefits as well as data sources that can be used to measure these benefits. The system is scoped to focus on the application of telemedicine to support patient referrals between hospitals. Data sources are considered accordingly, with electronic health record (EHR) data proving to be a feasible primary source for needs assessments, however limiting the benefits that can be quantified. The analysis of the needs assessment is included in the second phase of the decision support system. Data are extracted, transformed and loaded into a data warehouse from where it can be analysed. The system includes three analysis steps to: map referral patterns, analyse potential benefits of telemedicine programs and determine cost-effective telemedicine solutions by allocating equipment at different hospitals. Analysis techniques used in the system include Pareto analysis, economic analysis, linear programming and the use of a genetic algorithm. It is proposed that the potential benefit results and equipment allocation algorithm be used to plan telemedicine programs for continuous evaluation. The final phase of the system therefore guides decision makers to use the results for implementation planning as well as evaluability assessments, for future management and evaluation of telemedicine programs. The decision support system is validated using patient referral data from the Western Cape public health sector. The case study proved that the system is applicable to the real-world and could be a valuable tool for decision makers to base telemedicine implementation planning on quantifiable evidence. The limitation on size and quality of both the Eastern Cape and Western Cape data sets, caused that the full potential of the system could not be demonstrated and validated. It is recommended that the quality standards of EHR referral reports be improved, to ensure more accurate benefit results. Future work is recommended to include qualitative needs assessments in the scope of the decision support system, hereby increasing the amount of benefits to be assessed. Although it is expected that the developed system is capable to support even better resolution decisions with more detailed data sets, the system developed in this study proved already adequate for improved implementation decision making. This could lead to higher success rates of telemedicine programs and ultimately better quality healthcare for all.
AFRIKAANSE OPSOMMING: Die verskillende toepassings van Informasie en Kommunikasie Tegnologie (IKT) in gesondheidsorg, speel ʼn rol in toenemende doeltreffendheid om die koste-effektiwiteit en kwaliteit van gesondheidsorg dienslewering te verbeter. Tele-geneeskunde is een van hierdie toepassings, wat IKT gebruik om gesondheidsdienste oor ʼn afstand te kan voorsien. Die Suid-Afrikaanse Departement van Gesondheid belê sedert 1997, groot bedrae kapitaal in die implementering van tele-geneeskunde stelsels, in Suid-Afrika. Ongelukkig het tele-geneeskunde programme min sukses behaal sedertdien, wat veroorsaak dat vele werkstasies dormant is. Die basering van implementeringsbesluite op onvoldoende getuienis, is geïdentifiseer as een van die onderliggende probleme wat veroorsaak dat tele-geneeskunde programme misluk. Daar word voorgestel dat implementeringsbesluite gebaseer moet word op kwantifiseerbare getuienis ten opsigte van die potensiële voordele van telemedisyne. ʼn Besluitnemingsondersteuning stelsel is ontwikkel wat gebruik kan word om die potensiële voordele te kwantifiseer en dienooreenkomstig implementering van tele-geneeskunde programme te beplan. Die stelsel is gemodelleer en gedemonstreer aan die hand van data uit die Oos-Kaap publieke gesondheidsektor. Die eerste fase van die stelsel begelei besluitnemers om potensiële voordele van telegeneeskunde, sowel as data-bronne wat gebruik kan word om hierdie voordele te meet, te identifiseer. Die stelsel is beperk tot ʼn fokus op die ondersteuning wat tele-geneeskunde aan hospitaal pasiënt verwysingstelsels, kan bied. Data bronne is dienooreenkomstig oorweeg: elektroniese mediese rekords (EMR) word erken as ʼn gunstige primêre databron, maar veroorsaak egter beperkings op die aantal voordele wat gekwantifiseer kan word. Die behoeftebepaling word uitgevoer in die tweede fase van die besluitnemingsondersteuning stelsel. Data is onttrek, getransformeer is en gelaai in 'n data stoor, vanwaar dit ontleed kan word. Die stelsel sluit drie analisestappe in: verwysingspatroon analise, berekening van potensiële voordele vir tele-geneeskunde programme en die bepaling van koste-effektiewe oplossings deur toekenning van toerusting by verskillende hospitale. Die analise tegnieke wat in die stelsel gebruik word, sluit die volgende in: Pareto analise, ekonomiese analise, lineêre programmering en 'n genetiese algoritme. Die gebruik van potensiële voordeel resultate en die toerusting toekenning algoritme word voorgestel vir die beplanning vir deurlopende evaluering in tele-geneeskunde programme. Die finale fase van die stelsel is gestruktureer, om besluitnemers te begelei in die gebruik van analise resultate, vir implementering beplanning sowel as evalueerbaarheid studies, wat sodoende deurlopende evaluering en bestuur van tele-geneeskunde programme sal verbeter.Die besluitnemingsondersteuning stelsel is gevalideer deur pasiënt verwysings data van die Wes-Kaap publieke gesondheidsektor, te gebruik. Die gevallestudie het bewys dat die stelsel toepaslik is in die werklike wêreld en kan as ʼn waardevolle hulpmiddel vir besluitnemers dien om tele-geneeskunde implementering beplanning op kwantifiseerbare bewyse te baseer. Die beperkings op die grootte en gehalte van beide die Oos-Kaap en Wes-Kaap datastelle het veroorsaak dat die stelsel nie tot sy volle reg gedemonstreer en gevalideer kon word nie. Verbeterings in kwaliteit standaarde van EMR verwysing data word aanbeveel om meer akkurate resultate te bekom. Verdere studies wat die byvoeg van kwalitatiewe meetings in die stelsel ondersoek, sal die omvang van potensiële voordele verbeter en dus die algehele waarde van die stelsel verbeter. Alhoewel die ontwikkelde stelsel in staat is om beter resolusie besluite te kan ondersteun met meer gedetailleerde data, is dit bewys dat die huidige stelsel reeds voldoende is om besluitneming te verbeter. Beter besluitneming gevolglik lei tot hoër sukseskoerse van tele-geneeskunde programme en uiteindelik verbeterde gehalte gesondheidsorg vir almal.
Description
Thesis (MScEng)--Stellenbosch University, 2012.
Keywords
Telemedicine, Electronic health records, Health systems, Cost optimization, Medical needs assessment, Health informatics, Decision support, Dissertations -- Industrial engineering, Theses -- Industrial engineering
Citation