Methods: The study employs a two phase design. Phase one involves a process analysis of the South African National Tuberculosis Programme and an electronic medical records (EMR) solution and the generation of a detailed process model grounded in the fit between individual task and technology (FITT) theoretical framework using ‘business process modelling notation’. Phase two involves a two round Delphi study in the clinical management of tuberculosis and implementers of EMR solutions. Expert opinion is analysed according to emergent thematic content. Analyses and graphical model representation are performed using Microsoft Excel^® and Visio^® software.

Results: A detailed process model is constructed which reveals 54 break points, 12 gaps, 3 risks, 5 wastes. Five participants are included in the Delphi study which support the findings of the process analysis. Thematic analysis identifies five themes: the individual, the process, technology, capacity, and collaboration. The opportunity to include synergistic relations across programmes emerges as a strong theme.

Conclusions: Overall, the findings highlight inefficiencies, risk and gaps in the current process and the need for an operational excellence intervention. The study demonstrated the value of process engineering with iterative expert consultation toward developing a meaningful EMR solution consultation in a resource constrained, developing world context.

• data collection (history and physical examination)
• data recording (clinical record)
• data processing (clinical decision making)
• information transmission (clinical orders and referrals etc.) (Tierney et al. 2010).

The electronic medical record is fast becoming an important tool to ensure accurate and timely data collection that may be used for effective health care. There is a critical collaborative relationship between technology, people and the tasks that they perform (Elske, Carola & Mahler 2006; Tsiknakis & Kouroubali 2009a). This relationship has received limited research attention in the healthcare domain. This is important because merely implementing a Health Information Systems (HIS), without revisiting the business process and its context, cannot solve challenges and improve inefficiencies (Berg & Toussaint 2003; Ludwicka & Doucettea 2009).

Some 130 years since Robert Koch’s discovery of Mycobacterium Tuberculosis, as the infectious agent of tuberculosis, the disease remains a near insurmountable global health concern, which claimed an estimated 1.3 million lives in 2012 alone, over 30% arising in sub-Saharan Africa (World Health Organisation 2013). As early as 1994, drug resistance was reported in nearly every country surveyed by the World Health Organisation (WHO) surveillance projects (Blondal 2007). Nearly two decades later, ‘the total number of MDR-TB cases estimated to have occurred worldwide was about 450 000’, comprising 3.6% of all new and 20% of all previously treated cases (WHO 2013). South Africa has the second highest burden of MDR-TB in the world (15419 notified cases in 2012), and is the only country with a growing incidence of TB (currently estimated at 1 new case per 100 persons) (WHO 2013). The total cost for treating MDR-TB is approximately 30 times more than that of drug-sensitive TB, and diverts resources away from managing a national TB programme (Tupasi et al. 2006; Resch et al. 2006; Uplekar & Lonnroth 2007). The South African National TB Control Programme provides local guidelines for the management of drug resistant tuberculosis, which is based on WHO recommendations. The current cure rate for MDR-TB in most developing countries is between 30–50%, and the second line drugs used to treat MDR-TB are poorly understood, difficult to administer, and have poor side effect profiles (Pooran et al. 2013; Resch et al. 2006; Tupasi et al. 2006; Uplekar & Lonnroth 2007). Whilst the initial response to the MDR-TB epidemic in South Africa mirrored the World Health Organisation guidelines in the provision of centralised inpatient care, the high burden of disease in this country rapidly made centralised care unsustainable. As an alternative to centralised care, KwaZulu-Natal has moved toward a decentralised model of care, with comparable outcomes (Loveday et al. 2012). Whilst tuberculosis is a curable infectious disease, successful treatment outcomes require both patient adherence and a functional health system. Health system factors have been demonstrated to significantly impact treatment outcomes, and may contribute to avoidable negative clinical outcomes (Loveday et al. 2008). Therefore, there exists the potential to improve how multidrug-resistant tuberculosis (MDR-TB) is diagnosed and treated, as a result of employing a process engineering intervention built into an electronic medical record system could be significant (Fraser et al. 2006). As a key component of the global public health response to MDR-TB, the WHO has recommended a complete migration to electronic data collection by 2015 (WHO 2013). In light of decentralised care, this would require a comprehensive electronic medical record system that is able to satisfy the data recording purposes of public health authorities as well as the clinical and operational needs of patients and their health care providers. The complexity of such an electronic medical record (EMR) system will require the collaboration of a number of key stakeholders, and specifically an iterative relationship between designers of the system and the end-users (Allen et al. 2007; Ammenwerth, Iller & Mahler 2006; Blaya, Holt & Fraser 2008; Clifford et al. 2008; Elske, Carola & Mahler 2006; Fraser et al. 2006; Gerntholtz, Van Heerden & Vine 2007).

Little research is available for healthcare process management both in South Africa and other developing countries. South African health care processes have been described as ‘fundamentally broken’ and, thus, research in this area is much needed (Gerntholtz, Van Heerden & Vine 2007). South Africa’s boldest attempt to implement an EMR solution across all government hospitals in Limpopo failed in 1998. Healthcare workers were inadequately prepared and a lack of attention to the intent of processes and their unique application in South Africa appear to have played a role in this failure (Littlejohns, Wyatt & Garvican 2003).

Technology is a mechanism to enhance delivery; and one such technology is OpenMRS which has a specific module for management of MDR-TB programmes (Choi & Fraser n.d.; Seebregts et al. 2006). OpenMRS is one of the most widely used open source EMR solutions in Africa (Seebregts et al. 2006; Tierney et al. 2010). OpenMRS is designed using international standards (HL7, DICOM, and LOINC) for interfacing with other technologies and is designed for universal deployment. The OpenMRS MDR-TB module that is discussed in this study was developed to provide an intuitive ‘front end’ to support the treatment of MDR-TB for WHO sponsored projects. The module can be customised with some medium to high level computer skills for specific geographical or treatment requirements (Choi & Fraser n.d.). To date OpenMRS has been implemented in over 25 countries, these being mostly low income, and supports HIV and TB programmes. The OpenMRS MDR-TB module may be used as an electronic medical record solution, but may, in addition, provide the electronic framework for providing process engineering support to the critical MDR-TB clinical programme. The combination of a grounded process analysis tool, together with expert consultation, is a novel method for designing and optimising an EMR solution. This study aims to describe the role of process engineering and iterative consultation in shaping an EMR solution (OpenMRS) in the South African MDR-TB programme.

FIGURE 1: Contributions of process analytics by process analysis and by Delphi technique.

In the Delphi study five participants responded to both rounds in the study, and profiled themselves as spending their time doing research, implementing EMR solutions and performing clinical activities. All three clinicians in the group had knowledge of the South African clinical guidelines. Only one participant refrained from indicating their level of experience with EMR solutions, whilst two of the participants expressed ‘some experience’ and another two expressed a ‘great deal of experience’. With the exception of the one participant who practices process analysis on a daily basis, all other participants had limited exposure to process analysis. One of the clinicians expressed an interest to learn process analysis. The majority (four out of five) of the participants regarded the alignment of process, technology and individual as ‘important’. The participants’ responses affirm the underlying principle of the FITT framework that has been used as a theoretical framework for the study (Chan & Kaufman 2010; Elske, Carola & Mahler 2006; Tsiknakis & Kouroubali 2009a; Tsiknakis & Kouroubali 2009b). On reviewing the participants’ statements, a pattern emerged, that 80% of statements were related to either ‘Process’, ‘Individual’, or ‘Technology’. The participants tended to use the terms from the FITT framework that were used in the questions. The remaining statements were then reviewed to identify common themes, and two additional themes, namely ‘collaboration’ and ‘capacity’, were added. This illustrated that participants contributed new ideas to the process analysis as opposed to simply validating what they were presented with. Overall, the Delphi study demonstrated support for and affirmation of the process analysis findings (Table 2).

A summary of gaps identified in the process analysis is shown in Table 3. Delphi participants identified five unique gaps after final coding which were not identified in the process analysis phase (displayed in bold). This finding demonstrates the synergistic potential of process analysis, with an iterative consultation process, with stakeholders.

Wastes were stated as improvements, gaps or risks in the second round of the Delphi study and tested for agreement with the participants. In agreement with existing research, their responses show that two of the greatest inefficiencies in current non-EMR settings are the redundant capture of information and laborious data analysis (Clifford et al. 2008; Blaya, Holt & Fraser 2008; Gerntholtz et al. 2007; Vine 2007). As noted by a participant, it is important for an EMR solution to be customisable, ‘The reality of the operational set up is that the care process is highly fluid’. Understanding the gap between the technology and the process is a critical exercise that must be conducted, to ensure that there are limited work-a-rounds once the hospital information system (HIS) implementation is completed. The gap analysis indicates there is a high concentration of gaps between the process and the technology.

A participant, who is an international OpenMRS implementer, highlighted that nine of the thirteen gaps between task and technology could be addressed by customisations. OpenMRS has been customised and integrated with other applications such as Chasqui in Peru, FrontlineSMS in Ghana and AMPATH in clinics in sub-Saharan Africa, and Google maps in Pakistan (Tierney et al. 2010; Staccini et al. 2000; Seerbregts et al. 2009; Seebregts et al. 2006; Frasier, May & Wanchoo 2008; Choi & Fraser n.d.; Blaya et al. 2007; Allen et al. 2007). Of the nine improvements proposed by participants, five of these related to technology improvements (Figure 2). The figure further illustrates that the majority of the respondents agreed with the need for technology related improvements, with the exception of increased access to GeneXpert diagnostic technology. This was possibly because the South African Department of Health had announced its plans for a national roll-out of the technology just prior to this study, making such an improvement unnecessary. The technology has since been widely rolled out and enjoys the growing support of the medical and scientific community (Theron et al. 2013).

Given the growing use of technologies to support clinical decision making, it was not surprising that participants, particularly clinicians, made recommendations with regard to the need to integrate specialised technologies (Andersson, Hallberg & Timpka 2003; Isern & Moreno 2008; Terazzi et al. 1998). This is reinforced by the outcome in which the majority of participants responded positively to the technology improvement in OpenMRS regarding clinical decision making support to aid healthcare workers.

One of the greatest opportunities to enhance the OpenMRS system is to ensure that the processes that are supported by the system adequately provide relevant communication to stakeholders involved in the process. Quality improvement research highlights the need for effective communication amongst healthcare workers during the clinical care process, to support care co-ordination (Boston-Fleischhauer 2008; Taneva et al. 2010). Not all participants responded to statements relating to the need for the system to support communication, and responses varied widely. An individual’s perspective on this issue may be dependent on the environment in which the participant operates.

The participants with OpenMRS implementation experience also contributed to the synergies identified, indicating that electronic monitoring systems could support patient treatment adherence. This is not surprising, as OpenMRS implementations in other African countries have already encountered such challenges and have worked on solutions, such as SMS reminders (Allen et al. 2007; Choi & Fraser n.d.).

Whilst there are numerous gaps highlighted in the out-of-the-box installation of the OpenMRS MDR-TB module, it does help to mitigate various identified risks and eliminate waste, making the diagnosis and treatment process of MDR-TB more efficient (see Table 1). However, it would be better if the gaps identified in the study were closed before an implementation is carried out. This might have been possible if the implementation was preceded by a process analysis with iterative consultation, as described in this study.

Five improvement opportunities were identified by participants, four of which were from a clinical perspective. One of the suggested process improvements (Simplify the process by always insisting on 3 sputum samples) to address the perceived waste in the system was not supported by the experts, possibly because new modalities of diagnosis, such as GeneXpert, no longer require multiple first contact sputum specimens.

According to the second round of the Delphi study the most frequently experienced risks were:

• Paper records that are more likely to be lost or corrupted than electronic records
• Long laboratory turn-around times that result in delayed treatment
• Unclear timing of integration of antiretroviral therapy
• Dependency on the patient providing the correct information
• The limited linkage of health records between health facilities
• Suboptimal patient adherence.

All of these risks may have a negative impact on clinical decisions made by providers.

Ten synergies were identified by four participants, three of whom are MDR-TB clinicians. One of the main themes that were stressed by participants is the collaboration of treatment facilities for HIV positive and MDR-TB patients. Some of the synergies that participants disagreed on were: electronic monitoring systems that do not require specialised data capturers, and separate clinic notes and registers and provision of isoniazid prophylaxis for all immune-compromised individuals, especially post-TB treatment. In contrast to the fact that most participants disagreed with the synergy to provide isoniazid prophylaxis to all immune-compromised individuals, especially post-TB treatment, all participants agreed with the comprehensive treatment of other opportunistic infections, including the provision of co-trimoxazole prophylaxis. The synergies that received the most agreement from participants were:

• Identification of delays, and the reasons for the delay, in initiating HIV treatment of MDR-TB patients. This is especially significant as recent data suggests that delayed initiation of treatment is a major challenge to the health system and is a significant contributor to morbidity and mortality in patients with MDR-TB (O’Donnell et al. 2009; Padayatchi et al. 2014).
• The provision of support to HIV and MDR-TB patients to adhere to their treatment programmes.
• One service provider focusing on treatment of both conditions.
• Ensuring that all TB facility attendees are offered an HIV test.

The synergies that received the most agreement from participants are supported by the growing support for the integration of TB and HIV programmes (Loveday & Zweigenthal 2011; Perumal, Padayatchi & Stiefvater 2009; Van Rie et al. 2013). Participants went so far as to suggest one service provider to support co-treatment, in keeping with the integration of TB and HIV care as a current major health systems priority (Loveday & Zweigenthal 2011; O’Donnell et al. 2009; Padayatchi et al. 2014; Perumal et al. 2009).

In the Delphi study five participants responded to both rounds in the study, and profiled themselves as spending their time doing research, implementing EMR solutions and performing clinical activities. All three clinicians in the group had knowledge of the South African clinical guidelines. Only one participant refrained from indicating their level of experience with EMR solutions, whilst two of the participants expressed ‘some experience’ and another two expressed a ‘great deal of experience’. With the exception of the one participant who practices process analysis on a daily basis, all other participants had limited exposure to process analysis. One of the clinicians expressed an interest to learn process analysis. The majority (four out of five) of the participants regarded the alignment of process, technology and individual as ‘important’. The participants’ responses affirm the underlying principle of the FITT framework that has been used as a theoretical framework for the study (Chan & Kaufman 2010; Elske, Carola & Mahler 2006; Tsiknakis & Kouroubali 2009a; Tsiknakis & Kouroubali 2009b). On reviewing the participants’ statements, a pattern emerged, that 80% of statements were related to either ‘Process’, ‘Individual’, or ‘Technology’. The participants tended to use the terms from the FITT framework that were used in the questions. The remaining statements were then reviewed to identify common themes, and two additional themes, namely ‘collaboration’ and ‘capacity’, were added. This illustrated that participants contributed new ideas to the process analysis as opposed to simply validating what they were presented with. Overall, the Delphi study demonstrated support for and affirmation of the process analysis findings (Table 2).

A summary of gaps identified in the process analysis is shown in Table 3. Delphi participants identified five unique gaps after final coding which were not identified in the process analysis phase (displayed in bold). This finding demonstrates the synergistic potential of process analysis, with an iterative consultation process, with stakeholders.

Wastes were stated as improvements, gaps or risks in the second round of the Delphi study and tested for agreement with the participants. In agreement with existing research, their responses show that two of the greatest inefficiencies in current non-EMR settings are the redundant capture of information and laborious data analysis (Clifford et al. 2008; Blaya, Holt & Fraser 2008; Gerntholtz et al. 2007; Vine 2007). As noted by a participant, it is important for an EMR solution to be customisable, ‘The reality of the operational set up is that the care process is highly fluid’. Understanding the gap between the technology and the process is a critical exercise that must be conducted, to ensure that there are limited work-a-rounds once the hospital information system (HIS) implementation is completed. The gap analysis indicates there is a high concentration of gaps between the process and the technology.

A participant, who is an international OpenMRS implementer, highlighted that nine of the thirteen gaps between task and technology could be addressed by customisations. OpenMRS has been customised and integrated with other applications such as Chasqui in Peru, FrontlineSMS in Ghana and AMPATH in clinics in sub-Saharan Africa, and Google maps in Pakistan (Tierney et al. 2010; Staccini et al. 2000; Seerbregts et al. 2009; Seebregts et al. 2006; Frasier, May & Wanchoo 2008; Choi & Fraser n.d.; Blaya et al. 2007; Allen et al. 2007). Of the nine improvements proposed by participants, five of these related to technology improvements (Figure 2). The figure further illustrates that the majority of the respondents agreed with the need for technology related improvements, with the exception of increased access to GeneXpert diagnostic technology. This was possibly because the South African Department of Health had announced its plans for a national roll-out of the technology just prior to this study, making such an improvement unnecessary. The technology has since been widely rolled out and enjoys the growing support of the medical and scientific community (Theron et al. 2013).

FIGURE 2: Improvements identified and rated by participants.

Given the growing use of technologies to support clinical decision making, it was not surprising that participants, particularly clinicians, made recommendations with regard to the need to integrate specialised technologies (Andersson, Hallberg & Timpka 2003; Isern & Moreno 2008; Terazzi et al. 1998). This is reinforced by the outcome in which the majority of participants responded positively to the technology improvement in OpenMRS regarding clinical decision making support to aid healthcare workers.

One of the greatest opportunities to enhance the OpenMRS system is to ensure that the processes that are supported by the system adequately provide relevant communication to stakeholders involved in the process. Quality improvement research highlights the need for effective communication amongst healthcare workers during the clinical care process, to support care co-ordination (Boston-Fleischhauer 2008; Taneva et al. 2010). Not all participants responded to statements relating to the need for the system to support communication, and responses varied widely. An individual’s perspective on this issue may be dependent on the environment in which the participant operates.

The participants with OpenMRS implementation experience also contributed to the synergies identified, indicating that electronic monitoring systems could support patient treatment adherence. This is not surprising, as OpenMRS implementations in other African countries have already encountered such challenges and have worked on solutions, such as SMS reminders (Allen et al. 2007; Choi & Fraser n.d.).

Whilst there are numerous gaps highlighted in the out-of-the-box installation of the OpenMRS MDR-TB module, it does help to mitigate various identified risks and eliminate waste, making the diagnosis and treatment process of MDR-TB more efficient (see Table 1). However, it would be better if the gaps identified in the study were closed before an implementation is carried out. This might have been possible if the implementation was preceded by a process analysis with iterative consultation, as described in this study.

Five improvement opportunities were identified by participants, four of which were from a clinical perspective. One of the suggested process improvements (Simplify the process by always insisting on 3 sputum samples) to address the perceived waste in the system was not supported by the experts, possibly because new modalities of diagnosis, such as GeneXpert, no longer require multiple first contact sputum specimens.

According to the second round of the Delphi study the most frequently experienced risks were:

• Paper records that are more likely to be lost or corrupted than electronic records
• Long laboratory turn-around times that result in delayed treatment
• Unclear timing of integration of antiretroviral therapy
• Dependency on the patient providing the correct information
• The limited linkage of health records between health facilities
• Suboptimal patient adherence.

All of these risks may have a negative impact on clinical decisions made by providers.

Ten synergies were identified by four participants, three of whom are MDR-TB clinicians. One of the main themes that were stressed by participants is the collaboration of treatment facilities for HIV positive and MDR-TB patients. Some of the synergies that participants disagreed on were: electronic monitoring systems that do not require specialised data capturers, and separate clinic notes and registers and provision of isoniazid prophylaxis for all immune-compromised individuals, especially post-TB treatment. In contrast to the fact that most participants disagreed with the synergy to provide isoniazid prophylaxis to all immune-compromised individuals, especially post-TB treatment, all participants agreed with the comprehensive treatment of other opportunistic infections, including the provision of co-trimoxazole prophylaxis. The synergies that received the most agreement from participants were:

• Identification of delays, and the reasons for the delay, in initiating HIV treatment of MDR-TB patients. This is especially significant as recent data suggests that delayed initiation of treatment is a major challenge to the health system and is a significant contributor to morbidity and mortality in patients with MDR-TB (O’Donnell et al. 2009; Padayatchi et al. 2014).
• The provision of support to HIV and MDR-TB patients to adhere to their treatment programmes.
• One service provider focusing on treatment of both conditions.
• Ensuring that all TB facility attendees are offered an HIV test.

The synergies that received the most agreement from participants are supported by the growing support for the integration of TB and HIV programmes (Loveday & Zweigenthal 2011; Perumal, Padayatchi & Stiefvater 2009; Van Rie et al. 2013). Participants went so far as to suggest one service provider to support co-treatment, in keeping with the integration of TB and HIV care as a current major health systems priority (Loveday & Zweigenthal 2011; O’Donnell et al. 2009; Padayatchi et al. 2014; Perumal et al. 2009).

In addition, new evidence which has demonstrated the substantial survival benefit, of integrating antiretroviral therapy early within MDR-TB treatment, would need rapid incorporation into an MDR-TB EMR solution that contributes to the ‘process’ of MDR-TB management (Loveday et al. 2012; O’Donnell et al. 2009; Padayatchi et al. 2014). The benefit of using an open-source EMR solution such as OpenMRS lies in the ability of a wide community of developers to be called upon to deal with gaps and the need for revising the existing version, as has been performed with the MDR-TB module in this setting. Whilst EMR implementers might attempt to remain abreast of clinical developments, regular iterative input from clinical experts may serve as a more pragmatic response to keeping an EMR both useful and relevant. The move from centralised care for MDR-TB patients to growing support for a decentralised model of care will further challenge an EMR solution for MDR-TB. Central services are well documented to be easier for the implementation of EMR solutions, whilst decentralised care, especially in rural, resource-constrained settings, presents significant challenges to the implementation of an EMR solution (Heeks 2006; Lapao et al. 2009; Littlejohns, Wyatt & Garvican 2003; Seebregts et al. 2006; Seerbregts et al. 2009; Tierney et al. 2010; Tsiknakis & Kouroubali 2009a). The responsiveness of a health informatics solution to such programmatic changes will be crucial for its sustainability, and will make iterative feedback (through methods such as the Delphi technique) essential to understand unique challenges that may emerge only once a programme is shifted to more rural and outlying settings. Clinical support for any EMR solution will only be possible if clinicians are in agreement that the EMR adds value to the process of MDR-TB management, and that the value added, in terms of existing advantages, matches its accuracy and relevance when placed in the context of prevailing clinical guidelines. This is the major potential benefit of including a process analysis approach to EMR design and development. Including stakeholders, particularly healthcare workers in an EMR system selection and design, may improve their openness to the technology and reduce resistance to change (Tierney et al. 2010).

In keeping with evidence from other settings, participants in this study identified the use of process analysis, in the development of clinical protocols, as the highest ranked advantage (Taneva et al. 2010). The second and third ranked advantages (The ability of healthcare workers to personally identify problems in the healthcare system, and the ability to identify operational health system factors which may negatively impact on clinical outcomes.) attested to the use of process engineering as a quality improvement tool. There is a growing trend in healthcare quality improvement programmes and research to promote healthcare workers to initiate improvement identification opportunities (Chassin et al. 2010; Martikainen, Korpela & Tiihonen 2014).

Ammenwerth, E., Iller, C. & Mahler, C., 2006, ‘IT-adoption and the interaction of task, technology and individuals: A fit framework and a case study’, BMC Medical Informatics Decision Making 6, 3. http://dx.doi.org/10.1186/1472-6947-6-3

Andersson, A., Hallberg, N. & Timpka, T., 2003, ‘A model for interpreting work and information management in process-oriented healthcare organisations’, International Journal of Medical Informatics 72, 47–56. http://dx.doi.org/10.1016/j.ijmedinf.2003.09.001

Berg, M. & Toussaint, P., 2003, ‘The mantra of modeling and the forgotten powers of paper: A sociotechnical view on the development of process-oriented ICT in health care’, International Journal of Medical Informatics 69, 223–234. http://dx.doi.org/10.1016/S1386-5056(02)00178-8

Blaya, J., Holt, B. & Fraser, H.S., 2008, ‘Evaluations of the Impact of eHealth Technologies in Developing Countries: A Systematic Review’, Working paper for Rockefeller eHealth Meeting, Harvard-MIT Division of Health Sciences and Technology, Partners In Health, Division of Social Medicine and Health Inequalities, Cambridge, MA.

Blaya, J.A., Shin, S.S., Yagui, M.J., Yale, G., Suarez, C.Z., Asencios, L.L. et al., 2007, ‘A web-based laboratory information system to improve quality of care of tuberculosis patients in Peru: functional requirements, implementation and usage statistics’, BMC Medical Informatics Decision Making 7, 33. http://dx.doi.org/10.1186/1472-6947-7-33

Blondal, K., 2007, ‘Barriers to reaching the targets for tuberculosis control: multidrug-resistant tuberculosis’, Bull World Health Organzation 85, 387–394.

Boston-Fleischhauer, C., 2008, ‘Enhancing healthcare process design with human factors engineering and reliability science, part 2: Applying the knowledge to clinical documentation systems’, Journal of Nursing Administration 38, 84–89. http://dx.doi.org/10.1097/01.NNA.0000295632.80345.3d

Chan, C.V. & Kaufman, D.R., 2010, ‘A technology selection framework for supporting delivery of patient-oriented health interventions in developing countries’, Journal of Biomedical Informatics 43, 300–306. http://dx.doi.org/10.1016/j.jbi.2009.09.006

Chassin, M.R., Loeb, J.M., Schmaltz, S.P. & Wachter, R.M., 2010, ‘Accountability measures – using measurement to promote quality improvement’, New England Journal of Medicine 363, 683–688. http://dx.doi.org/10.1056/NEJMsb1002320

Choi, S. & Fraser, H., n.d., Developing Multidrug-resistant TB Systems Using OpenMRS. Partner in Health, Harvard Medical school.

Clifford, G.D., Blaya, J.A., Hall-Clifford, R. & Fraser, H.S., 2008, ‘Medical information systems: A foundation for healthcare technologies in developing countries’, BioMedical Engineering OnLine 7.

Elske, A., Carola, I. & Mahler, C., 2006, ‘IT - Adoption and the interaction of task, technology and individuals: a fit framework and a case study’, BMC Medical Informatics Decision Making 6, 1472–6947.

Fraser, H.S., Blaya, J., Choi, S.S., Bonilla, C. & Jazayeri, D., 2006, ‘Evaluating the impact and costs of deploying an electronic medical record system to support TB treatment in Peru’, AMIA Annual Symposium Proceedings, 264–268.

Frasier, H., May, M.A. & Wanchoo, R., 2008, ‘e-Health Rwanda Case Study’, in American Medical Informatics Association , viewed 12 September 2014, from http://ehealth-connection.org/files/resources/Rwanda%20+%20Appendices.pdf

Gerntholtz, T., Van Heerden, M.V. & Vine, D.G., 2007, ‘Electronic Medical Records – Why should you consider implementing an EMR?’, Continuing Medical Education 25, 24–28.

Heeks, R., 2006, ‘Health information systems: Failure, success and improvisation’, International Journal of Medical Informatics 75, 125–137. http://dx.doi.org/10.1016/j.ijmedinf.2005.07.024

Isern, D. & Moreno, A., 2008, ‘Computer-based execution of clinical guidelines: a review’, International Journal of Medical Informatics 77, 787–808. http://dx.doi.org/10.1016/j.ijmedinf.2008.05.010

Lapao, L.V., Rebuge, A., Silva, M.M. & Gomes, R., 2009, ‘ITIL Assessment in a healthcare environment: the role of IT governance at Hospital Sao Sebastiao’, Studies in Health Technology and Informatics 150, 76–80.

Littlejohns, P., Wyatt, J.C. & Garvican, L., 2003, ‘Evaluating computerised health information systems: Hard lessons still to be learn’t’, Journal of Biomedical Informatics 326, 860–865.

Loveday, M., Thomson, L., Chopra, M. & Ndlela, Z., 2008, ‘A health systems assessment of the KwaZulu-Natal tuberculosis programme in the context of increasing drug resistance’, International Journal of Tuberculosis and Lung Disease 12, 1042–1047.

Loveday, M., Wallengren, K., Voce, A., Margot, B., Reddy, T., Master, I., et al., 2012, ‘Comparing early treatment outcomes of MDR-TB in decentralised and centralised settings in KwaZulu-Natal, South Africa’, International Journal of Tuberculosis and Lung Disease 16, 209–215. http://dx.doi.org/10.5588/ijtld.11.0401

Loveday, M. & Zweigenthal, V., 2011, ‘TB and HIV integration: obstacles and possible solutions to implementation in South Africa’, Tropical Medicine & International Health 16, 431–438. http://dx.doi.org/10.1111/j.1365-3156.2010.02721.x

Ludwicka, D.A. & Doucettea, J., 2009, ‘Adopting electronic medical records in primary care: Lessons learned from health information systems implementation experience in seven countries’, International Journal of Medical Informatics 78, 22–31. http://dx.doi.org/10.1016/j.ijmedinf.2008.06.005

Martikainen, S., Korpela, M. & Tiihonen, T., 2014, ‘User participation in healthcare IT development: A developers’ viewpoint in Finland’, International Journal of Medical Informatics 83, 189–200. http://dx.doi.org/10.1016/j.ijmedinf.2013.12.003

National Department of Health, 2009, ‘South African National Tuberculosis Guidelines’, in HEALTH, Pretoria.

O’Donnell, M.R., Padayatchi, N., Master, I., Osburn, G., Robert, C. & Horsburgh, C.R., 2009, ‘Improved Early Results for Patients with Extensively Drug Resistant Tuberculosis and HIV in South Africa’, International Journal Tuberculsosis Lung Disease 13, 855–861.

Padayatchi, N., Abdool Karim, S.S., Naidoo, K., Grobler, A. & Friedland, G., 2014, ‘Improved survival in multidrug-resistant tuberculosis patients receiving integrated tuberculosis and antiretroviral treatment in the SAPiT Trial’, International Journal of Tuberculosis and Lung Disease 18, 147–154. http://dx.doi.org/10.5588/ijtld.13.0627

Perumal, R., Padayatchi, N. & Stiefvater, E., 2009, ‘The whole is greater than the sum of the parts: recognising missed opportunities for an optimal response to the rapidly maturing TB-HIV co-epidemic in South Africa’, BMC Public Health 9, 243. http://dx.doi.org/10.1186/1471-2458-9-243

Pooran, A., Pieterson, E., Davids, M., Theron, G. & Dheda, K., 2013, ‘What is the cost of diagnosis and management of drug resistant tuberculosis in South Africa?’, PLoS One 8, e54587. http://dx.doi.org/10.1371/journal.pone.0054587

Resch, S.C., Salomon, J.A., Murray, M. & Weinstein, M.C., 2006, ‘Cost-effectiveness of treating multidrug-resistant tuberculosis’, PLoS Medicine 3, e241. http://dx.doi.org/10.1371/journal.pmed.0030241

Seebregts, C., Mars, M., Fourie, C., Singh, Y. & Weyer, K., 2006, ‘Inexpensive Open Source TB and HIV electronic medical record system (OpenMRS) in South Africa Collaborating Toward an EMR for Developing Countries’, Proceedings of the AMIA Symposium. Washington DC. November 11–15.

Seebregts, C., Mamlin, B., Biondich, P., Fraser, H., Wolfe, B., Jazayeri, D., et al., 2009, ‘The OpenMRS implementers network’, International Journal of Medical Informatics 78, 711–720. http://dx.doi.org/10.1016/j.ijmedinf.2008.09.005

Staccini, P., Joubert, M., Quaranta, J.F., Fieschi, D. & Fieschi, M., 2000, ‘Integration of health care process analysis in the design of a clinical information system: applying to the blood transfusion process’, Proceedings of the AMIA Symposium 824–828, California, November 4-8.

Taneva, S., Grote, G., Easty, A. & Plattner, B., 2010, ‘Decoding the perioperative process breakdowns: a theoretical model and implications for system design’, International Journal of Medical Informatics 79, 14–30. http://dx.doi.org/10.1016/j.ijmedinf.2009.10.001

Terazzi, A., Giordano, A. & Minuco, G., 1998, ‘How can usability measurement affect the re-engineering process of clinical software procedures?’, International Journal of Medical Informatics 52, 229–234. http://dx.doi.org/10.1016/S1386-5056(98)00141-5

Theron, G., Zijenah, L., Chanda, D., Clowes, P., Rachow, A., Lesosky, M. et al., 2013, ‘Feasibility, accuracy, and clinical effect of point-of-care Xpert MTB/RIF testing for tuberculosis in primary-care settings in Africa: a multicentre, randomised, controlled trial’, Lancet 383, 424-435. http://dx.doi.org/10.1016/S0140-6736(13)62073-5

Tierney, W.M., Achieng, M., Baker, E., Bell, A., Biondich, P., Braitstein, P. et al., 2010, ‘Experience implementing electronic health records in three East African countries’, Studies in Health Technology and Informatics 160, 371–375.

Tsiknakis, M. & Kouroubali, A., 2009a, ‘Organizational factors affecting successful adoption of innovative eHealth services: a case study employing the FITT framework’, International Journal of Medical Informatics 78, 39–52. http://dx.doi.org/10.1016/j.ijmedinf.2008.07.001

Tsiknakis, M. & Kouroubali, A., 2009b, ‘Organizational factors affecting successful adoption of innovative eHealth services: A case study employing the FITT framework’, International Journal of Medical Informatics 78, 39–52. http://dx.doi.org/10.1016/j.ijmedinf.2008.07.001

Tupasi, T.E., Gupta, R., Quelapio, M.I., Orillaza, R.B., Mira, N.R., Mangubat, N.V. et al., 2006, ‘Feasibility and cost-effectiveness of treating multidrug-resistant tuberculosis: a cohort study in the Philippines’, PLoS Medicine 3, e352. http://dx.doi.org/10.1371/journal.pmed.0030352

Uplekar, M. & Lonnroth, K., 2007, ‘MDR and XDR - the price of delaying engagement with all care providers for control of TB and TB/HIV’, Tropical Medicine & International Health 12, 473–474. http://dx.doi.org/10.1097/01.qai.0000434954.65620.f3

Van Rie, A., Patel, M.R., Nana, M., Driessche, K.V., ‘Tabala, M., Yotebieng, M. et al., 2013, Integration and task-shifting for TB/HIV care and treatment in highly resource-scarce settings: one size may not fit all’, Journal of Acquired Immune Deficiancy Syndromes 65, e110–117. http://dx.doi.org/10.1097/01.qai.0000434954.65620.f3

Vine, D.G., 2007, ‘Communicating between colleagues - pitfalls and practical solutions’, Continuing Medical Education 25, 14–16.

World Health Organisation, 2013, ‘Global Tuberculosis Report 2013’, Geneva.