Thirty-nine resources were included in the review (Fig 1). There were 27 journal articles, two book chapters, one letter to the editor, one special report to PAHO, one systematic review protocol, one systematic review, one poster abstract and five institutional documents. The results show that substantive ehealth initiatives exist in Botswana, which are first summarised below, before focussing on telemedicine activity. For the literature summary, retrieved papers were categorised into five foci: technology, health information technology (HIT), training, strategy, and clinical services (telemedicine).
The technology focus included applications involving Skype, SMS, and robotics. Skype has been used to share patient information and deliver educational services between Botswana and Canada . Short Message Service (SMS) applications have included transmission of CD4 test results from the laboratory to the clinic , patient appointment reminders , and invoicing and bill payment . Telerobotics for dermatopathology was reported which was replaced with telepathology using WhatsApp .
HIT activities within Botswana have covered surveillance or monitoring for TB [17-19], and access to information by healthcare workers . In addition Coppock et al.  reported on a pharmacy mobile application used to track antiviral medication refill data. Nine discrete EMRs have been identified and remain active within the public health institutions [22-24]. Between 2012 and 2015 the government reduced the number of HIT solutions from 37 to 12  and now focusses on the Integrated Patient Management System (IPMS), District Health Information System 2 (DHIS2), Central Stores Drug Management, and Patient Information Management System (PIMS) .
Several applications have been applied to training or education. Okrainec et al.  used Skype for simulation to teach Laparoscopic Surgery in Botswana, whilst Armstrong et al.  described SMS for clinical practice guidelines. Witt et al.  examined the use of tablets in mlearning, and the University of Botswana School of Medicine (UBSOM)/UPenn partnership developed a mobile learning programme for training medical students . Finally, Oladokun  examined the information needs and information seeking behaviour of the populace, and described the responsibility of the government (and others) and associated policy in establishing an information rich setting for the country.
No study described the process of ehealth strategy development within the developing world. However, strategy related to ICT and ehealth overall has been developed in Botswana. In 2004, the Government developed an ICT strategy to guide the development of the country’s ICT infrastructure branded Maitlamo . The policy assisted the country in developing an almost nationwide ICT infrastructure, including a fibre backbone network (BOFINET), making viable other ICT related initiatives such as egovernment and ehealth. A first draft of an ehealth strategy was embedded in the ICT Policy of 2004, but only as an appendix . In 2012 the Botswana Parliament released its ICT Master Plan in which the intent to pursue an ehealth Botswana programme was identified . This document also recognised and differentiated telemedicine (described as “curative services provided through telecommunication technologies, including telephone, videoconferencing and Internet”) and telehealth (described as “wider healthcare advice provided through telecommunications”). In 2015, Botswana released version 4 of a specific ehealth strategy . This draft document provided a detailed single framework to help all the key stakeholders in the health sector to implement the required ehealth initiatives.
Clinical services (telemedicine) have also taken place within Botswana. Most studies (18) have been performed by a single research group, the BUP Partnership [13, 18, 20, 21, 26-28, 33-43]. The available evidence suggests this partnership is no longer functional. These studies have largely been on mobile telemedicine (mhealth), concentrated around the capital city Gaborone, and report on the same four mhealth pilot projects. The projects addressed cervical cancer screening, teledermatology, teleradiology and oral medicine telemedicine services. Only the PEEK project, the Portable Eye Examination Kit, was confirmed to still be active.
Telemedicine services have addressed cervical cancer screening, teledermatology, teleradiology, oral medicine and eye screening. Quinley et al.  studied the diagnostic agreement between remote diagnosis by a gynaecologist using photographs of the cervix and in-person visual inspection with acetic acid by midwives. The images for remote evaluation were taken using a mobile phone camera and transmitted through multimedia messaging service (MMS).
Azfar et al.  examined the reliability and accuracy of outcomes derived from the store and forward images. Subsequently, Azfar et al.  investigated mobile dermatology for taking and transmitting images of skin lesions in HIV positive patients. Schwartz et al.  compared chest X-rays with digital photos taken from a mobile phone, opening the possibility of access to the three radiologists practising in Botswana’s public health system, all located in Francistown and Gaborone. Mobile phones have been used by local health practitioners to collect and upload patient information for oral medicine purposes, sending it to a remote server for second opinion . PEEK, has been adopted in several countries and uses a smartphone app and a low-cost adapter for retinal imaging to screen for core vision problems . The BUP Partnership, in collaboration with the Standard Chartered Bank Botswana, MOH, Ministry of Education and Botswana Optometrists Association, commissioned PEEK Botswana, a national eye screening and treatment management project for school children . In addition, Joubert et al.  explored the use of screening to address hypertension in an urban population from Gaborone, and proposed a telemedicine-supported model of care as a potential solution.
The BUP official was interviewed solely to confirm the status of their telemedicine projects. Most were non-functional. All but one project is no longer active because donor funding has ceased, and the Government could not take on the projects once the pilot phases were completed. Only the PEEK project is active and is being scaled up with the intervention of the Government.
Botswana ehealth strategy review
The 2015 document ‘Botswana eHealth Strategy v4’  was critically reviewed to understand the strengths and weakness of the Strategy, in particular in relation to telemedicine. The document adopted the WHO definition for ehealth. However, there were omissions, oversights, and biases identified that impair the utility of the current draft strategy. For example, although the terms telehealth, telemedicine, and mhealth are mentioned there are no clear definitions provided of these and other key terms. Further, the general tone of the document is very ‘informatics’ focussed, with the term ‘telemedicine’ used just 5 times but ‘health information’ 17 times.
Connected to these issues, there is no clear description of the depth and breadth of interventions possible through non-informatics ehealth approaches. The draft eHealth Strategy also failed to emphasise the over-riding need for interventions to address specific and identifiable health needs. Although some health status data for Botswana is presented, and improvement in Millenium Development Goals (MDGs) has been emphasised, there is little consideration of the broad identification of health issues. Of further note was the lack of any consideration of identifying non-eHealth solutions to identified health needs. In addition, although mention was made of ‘high-priority eHealth services and applications’ there was no description of how priority was determined, nor if or how any priority was given to identified health needs before consideration of eHealth solutions.
The reviewed literature identified infrastructure issues as the main barrier to implementation of telemedicine in developing countries, including Botswana. According to Oluoch et al.  issues included “unstable electrical power, loss of Internet connectivity, and access to mobile phones.” Low levels of computer literacy, lack of end user training, bandwidth cost, lack of clinical and technical expertise, poor user acceptance, increased workload of medical practitioners and poor network coverage were documented as additional barriers to mobile telemedicine [22, 48]. Other barriers were weak organisational capacities, lack of technological training, lack of evidence based decision making, lack of financing, lack of strong political leadership (free of corruption), negative attitudes towards use of ICT, lack of technological awareness, resistance to new processes, poor data quality, and a lack of development of data standards .