Document(s) 19 de 31

Introduction

Telemedicine is the process of providing medical expertise remotely with the help of telecommunication. It is particularly valuable in remote areas, and is therefore useful in Nepal, where the few specialists are separated from most of the population. Nepal has one of the lowest gross national products (GNPs) (US$300) per capita¹ and one of the lowest literacy rates (50%)² of the South Asia Region. These factors have contributed to the prevalence of communicable, respiratory and nutrition deficiency diseases, which are among the most common disorders seen in hospital outpatient departments. Telemedicine is therefore an attractive potential means of improving health services. In industrialized countries, real-time telemedicine is commonly used. However, this modality is not applicable in a developing country, like Nepal, because of the high cost of bandwidth and the poor telecommunication infrastructure.

The alternative modality, more suitable for a developing country, is store-and-forward telemedicine. In a store-and-forward interaction, the referring doctor usually enters the clinical information and digital images in a computer. The information can then be transmitted via a dial-up Internet connection as an email attachment to a central computer. The expert doctor can access the data independently at a convenient time. The size of the images attached to email messages is a potential problem with this technique, which may necessitate some degree of image compression.

HealthNet Nepal

HealthNet Nepal is a non-governmental organization (NGO) that provides affordable Internet services to the Nepalese health community, access to health information and technical support for various regional information-sharing initiatives. It has developed its own software, called Hnet telemedicine.

Development of the software

When we examined the available telemedicine software, we found that open source software was not suitable for our purpose³ and that commercial software could not be easily modified. It was therefore decided to develop software suitable for telemedicine in the context of Nepal.

The local system collects a clinical history and images, which are then transmitted to a specialist for diagnosis. For the clinical history, both general information and discipline-specific information are collected. The general information includes the following:

•  patient information
  
  
• basic information
  
  
• personal history
  
  
• past medical history
  
  
• family history.

There is also discipline-specific information for the following specialties:

• pathology
  
  
• radiology
  
  
• dermatology
  
  
• cardiology.

In acquiring digital images, the following components were considered.

Pathology

Images were obtained using a digital camera (Nikon CoolPix 4500). This type of camera is suitable for pathology. It has a resolution of 4 megapixels (Mp), which is sufficient for telemedicine according to the recommendations of the American College of Radiology.⁴ An additional advantage of this camera is that it can be easily mounted on a microscope with a 30 mm eyepiece adapter.

Radiology and dermatology

It was found that using the same camera for pathology was not practical. The camera used for radiology and dermatology was a Nikon CoolPix P4, which has a higher resolution (8 Mp). An additional feature of this camera is its vibration-reduction ability, which makes it suitable for taking pictures in shaky conditions.

Cardiology

For cardiology, an ECG machine was used (ECG 9620L, Cardiofax) that directly transfers an ECG image to a computer. Thus, there is no need to use a digital camera to photograph ECG recordings.

Software for data transfer and security

The software ensured data security during transfer in the following ways.

Image processing

Once images have been stored in the computer, they need to be cropped to reduce the file size and annotation has to be provided for better description of the clinical problem. A simple graphics package was developed for this purpose and incorporated into the client software of the Hnet telemedicine system. This graphics package allows cropping, annotation, rotation, colour balance and conversion to black and white. In addition, the free image-viewing software IrfanView was used.⁵

Compression

Image compression was used to reduce the quantity of memory required to store an image. For example, an image that has a resolution of 640 × 480 pixels with 24-bit colour requires 900 kbyte of storage. If this image can be compressed at a compression ratio of 20 : 1, the quantity of storage required is reduced to about 45 kbyte. Hnet has used the camera’s built-in JPEG image compression at a ratio of 16 : 1. It has been found that image compression of 60–80% can be achieved without significant loss of diagnostic quality.

Security and confidentiality

Data stored in a computer must not be read or compromised by unauthorized users. The Hnet software uses password-level encryption. The passwords are stored in MD5-encrypted format for security. Without an authorized login, no one can run the application.

Receiving and reviewing cases

After logging on to the Hnet system, the two physicians concerned, typically a remote physician or health worker and a medical specialist in a hospital, communicate by email. After uploading a case, the server selects a doctor on a simple round-robin basis from the names listed in the roster. The server regularly watches for a reply from the doctor. If a doctor does not reply to a case within 12 hours, then the case is automatically assigned to another doctor listed in the roster. Doctors are categorized into groups, such as dermatology, radiology, cardiology and pathology. These specialists can view only their own cases.

Digital photography guides

A technical manual was developed as a basic guide for non-medical professionals and general practitioners on how to take clinical photographs in radiology, pathology and dermatology using a digital camera.

Pilot project

The pilot telemedicine network was implemented from July 2004 to December 2006. The project was financed by the PAN Asia programme of the IDRC, Canada (see Chapter 7). The following hospitals referred cases for telemedicine:

• AMDA Hospital, Damak (eastern region)
  
  
• Siddhartha Children’s and Women Hospital, Butwal (western region)
  
  
• Siddhi Memorial Hospital, Bhaktapur (central region).

Specialist expertise was provided from the following central-level hospitals:

• Teaching Hospital, Tribhuvan University (Department of Radiology and Dermatology)
  
  
• Kathmandu Medical College (Department of Pathology and Department of Dermatology)
  
  
• Sahid Ganga Lal National Heart Centre.

The medical experts located at the medical colleges and special hospitals of the central level (capital city) were chosen because the doctors there had better knowledge of IT and were willing to participate in the project. The local hospitals were chosen based on the size and location of the local population.

It was found that the availability of telemedicine increased the volume of teleconsultation and provided exposure to a rural community for the medical interns of medical colleges. It also provided continuing education and reduced the professional isolation of health professionals working in rural areas. All these factors improved the quality and efficiency of the health service. With the use of telemedicine, access to the literature through the HINARI system of the World Health Organization increased. The observations below are based on informal communication with the referring physicians.

Study

A study was conducted in Nepal to assess whether telemedicine based on store-and-forward technology would be satisfactory for the diagnosis of cases sent from remote rural areas in the specialties of dermatology, radiology, pathology and cardiology. There were two components to the work: a basic component and a clinical component.⁶ The basic component was to define the minimum technical specifications to ensure that diagnostically important information could be captured through a clinical history and digital image. This component also included an investigation of the medical expert’s ability to make an accurate diagnosis based on digital images and clinical history. The second component was to evaluate whether the technology would improve the process of health care delivery by increasing information flow and reducing professional isolation.

For inclusion of cases, data were collected from the field sites. These cases were entered into the computer of the client site and uploaded to the server. On the server, the specialist doctors completed their perception of the image exposure and diagnostic quality using a binary scale (0 = poor or 1 = adequate). At the second stage, the specialist doctors recorded their confidence in the presence of each diagnostic indicator on a five-point scale (1 = definitely absent to 5 = definitely present). The ratings from these scales were used for receiver operating characteristic (ROC) analysis. The ROC analysis was used to determine whether the image quality was sufficient for diagnosis.

A two-step rating method was used to plot the ROC data. First, a dichotomous variable indicating quality of image (0 = negative or 1 = positive) was used. Second, the dichotomous variable based on the quality of the image was further rated based on a five-point scale (1 = definitely negative to 5 = definitely positive. The use of five categories is a reasonable compromise between the needs of ROC analysis and the precision with which the observers can be expected to reproduce their ratings. The five-point rating scale was used to calculate predicted probability. This predicted probability along with the dichotomous variable can be used to plot an ROC curve in order to provide a visual impression of the reliability of the points. The statistical package SPSS 11.5 was used for calculating predicted probability using a logistic regression model. Based on the dichotomous variable, a predicted probability ROC curve was plotted.

Similarly, provisional diagnosis by primary care physicians and diagnosis by clinical experts was used for comparing baseline medical knowledge of primary care physicians with clinical experts. These data were used only in dermatology.

Study findings

A total of 218 cases were observed:

• dermatology : 75 cases
  
  
• radiology : 90 cases
  
  
• histology : 34 cases
  
  
• cytologu : 19 cases.

The gold standard for any specialist referred remains the traditional method of consultation, i.e. in the case of radiology, it is the viewing of an X-ray film; in the case of pathology, it is viewing through a microscope; and in the case of dermatology, it is face-to-face consultation.

We adopted an affordable system that was appropriate to our needs.⁷ Compared with real-time systems, store-and-forward systems are more practical and less expensive. Store-and-forward methods allow the use of low-cost equipment, low-bandwidth connections and asynchronous consultation. This approach has been used successfully in industrialized countries where high-quality images are transmitted via tele-medicine networks for consultation.⁸

Digital cameras are an efficient method of obtaining digital images. Krupinski et al^9,10 evaluated the effectiveness of digital photography for dermatology and bone trauma diagnoses, and found that digital camera images were of satisfactory quality. Piccolo et al¹¹ found high diagnostic concordance between telepathology and histopathology diagnosis. Lim et al¹² and High et al¹³ evaluated a digital camera for teledermatology, and also found high concordance with face-to-face diagnoses.

For determining the quality of images sufficient for diagnosis, a ROC analysis was performed. It has been suggested that qualitative conclusions can be drawn from ROC experiments performed with at least 100 observations.¹⁴ In our example, all four tests (dermatology, radiology, histology and cytology) had low p values (p < 0.5).¹⁵ Figure 17.1 shows the ROC curve for dermatology. It can be concluded that the images had the ability to detect disease.

Figure 17.1 ROC curve for dermatology (the broken line represents no discrimination, i.e. a random guess)

Willingness to participate in the telemedicine project

In Nepal, there are few doctors. There is only one doctor for 6000 people (the doctor-to-population ratio is 1 : 6000). Our study showed that practising doctors usually did not use IT for knowledge updating. The reasons given for not using IT were that most of the doctors were busy practitioners and they perceived that they could manage communicable, respiratory, tropical and nutritional deficiency diseases without recent information. Another reason was the heavy patient workload.

However, the circumstances are different in medical colleges, since doctors are required to update their knowledge in order to teach students. IT has become a useful medium for updating their knowledge. So, doctors working in medical colleges and special research institutes are chosen as experts for the diagnosis of certain cases. Another reason is that doctors working in medical colleges and research institutes feel comfortable in using IT.

Age of users

All the users were aged 31–50 years. This included participants from both central and local hospitals. The reason may be that the younger generation grew up with computers, witnessing developments in computers and the Internet first hand.

Gender

Only one (10%) female was involved in a telemedicine project, and she was a health professional from a central hospital. In Nepal, medicine has been and still is a male-dominated profession. Nurses were not involved in telemedicine in the local hospital. The reason is that consultations are like a second opinion, which exclude nurses’ participation because they do not have the doctors’ level of competence.

Organizational position

With reference to the organizational position of participants, it was found that staff involved in the telemedicine project were doctors or technicians responsible for the technical part of the project.

Users’ attitudes to telemedicine

The technical components of telemedicine were related to the computerization process of central and remote sites. The problems related to the computerization process were related to low computer literacy and the absence of reliable communication infrastructure. In this context, the users’ attitudes depended on their general knowledge and familiarization with computer technology. All participants in the project from local hospitals and central hospitals felt comfortable with IT as a whole and telemedicine in particular. Telemedicine gave them an opportunity to enhance their efficiency and distance learning and to minimize professional isolation. With the help of telemedicine, they were able to create databases for various diseases, their forms and methods of treatment. Doctors perceived that the databases helped to improve their efficiency.

Consultation areas

Telemedicine was used for the purpose of general guidance, assessment of diagnosis, treatment and examinations. It was also found that confidence between local and remote hospitals was crucial to achieve and maintain a fruitful dialogue. Often it is difficult to state the diagnosis from the first information transferred. For this, two-way communications in the form of chat is provided to make a precise diagnosis.

Collaboration between doctors at local and central level

The users in the telemedicine project noted that their collaboration with professionals changed with the help of telemedicine. Doctors become more open to advice from their colleagues and specialists. Doctors came to believe that the process of patient care could be enhanced through telemedicine, and their trust in information technology increased.

Exposure to rural areas

In Nepal, most doctors are reluctant to work in rural areas owing to professional isolation, lack of facilities to update their knowledge and lack of continuing education. In medical colleges, as a part of an internship programme, a doctor has to work in rural areas. The telemedicine process minimized the gap of professional isolation and gave the opportunity for continuing education in rural areas. This first-hand experience in rural areas with telemedicine may motivate doctors to work in rural areas when their medical courses are completed.

Drawbacks

Although telemedicine clearly has a number of potential benefits, it also has some disadvantages. The main drawbacks are a potential breakdown in professional relationships and a number of organizational problems.

Breakdown in the relationship between health professionals

Skilled staff at the remote site sometimes perceive that their autonomy is threatened by the use of telemedicine, or, worse still, think that they will become no more than technicians, acting solely at the guidance of the remote specialist.¹⁶

Organizational problems

The following organizational problems were identified in Nepal:

• Fear of the unknown. Nepal remained poorly networked because of inertia, fear and technical inexperience of higher management personnel in government and NGOs.
  
  
• Lack of trained manpower. In spite of widespread enthusiasm about IT, there are very few local facilities for higher-level technical training.
  
  
• Possessiveness. In a country with the lowest GNP in the region and having few opportunities, people, including higher-level management, tend to hoard everything. They do not hesitate to break contracts, even for minor financial benefits. This attitude runs counter to networking and information sharing.

Economic evaluation

An analysis was carried out to compare the cost of telemedicine with the cost of traditional methods of providing health care services (Table 17.1). Only variable costs were considered. The cost charged by the remote site was approximately 50% less than the charge to be paid in city areas. However, the patient need not bear the cost of transportation and lodging. The telemedicine service charge would be sufficient to cover the fixed costs and to pay the fees of the expert. Thus, only the cost of the expert’s Internet connection would have to be borne by the remote centre. It is too early to calculate all the costs (fixed and variable) in order to estimate whether tele-medicine could become financially sustainable.

Research findings

We also compared the baseline medical knowledge of primary care physicians with clinical experts. Here, we also found that there was no difference in the diagnosis among primary care physicians and expert doctors. This supports the view that textual information supplemented by images is sufficient for diagnosis.

The telemedicine system helped primary care physicians to interact with medical experts. This, in turn, helped to reduce the professional isolation of health care staff working in rural areas.

At the start of the project, we found technophobia among health professionals working in rural areas. To reduce this, we provided IT training courses to all the health professionals working in rural areas. These training courses helped rural health workers to become familiar with IT. These health workers also found that IT could enhance their efficiency. The technical knowledge encouraged staff to use IT in other areas such as hospital information systems, to improve the efficiency of the hospital and to provide access to the health literature to update their knowledge.

Future developments

In our research project, we found that there was potential for telemedicine in Nepal, especially in rural areas, but also in urban areas. It was also found that textual information with images was sufficient for diagnosis in many cases. Telemedicine helped medical education, medical care and collegial support. However, continuation of the work after the project has finished will depend on how the local health professionals and medical experts support the services. To support and continue the telemedicine work, there should be other computerized health services, such as distance education, digital libraries and computerized hospital information systems. That is, telemedicine cannot develop in isolation. These types of activities will encourage rural health workers and medical specialists to continue the telemedicine work. As users become more comfortable with the technology, they are inclined to experiment with new applications. These activities help to enhance their efficiency, update their knowledge and reduce professional isolation.

Low-cost telemedicine based on store-and-forward systems has also been carried out in other developing countries. For example, the Swinfen Charitable Trust established a telemedicine link to support the Centre for the Rehabilitation of the Paralysed (CRP) in Savar, near Dhaka in Bangladesh, in July 1999¹⁷ (see Chapter 19). Based on the success of the Bangladesh project, the Swinfen Charitable Trust supplied digital cameras and tripods to more hospitals in other developing countries. These included the Patan Hospital in Nepal (March 2000), Gizo Hospital in the Solomon Islands (March 2000), Helena Goldie Hospital in the Solomon Islands (September 2000) and LAMB Hospital in Bangladesh (September 2000). In South Africa, the technique has been used for referrals from remote clinics. The US Armed Forces have made use of store-and-forward telemedicine for telepathology (see Chapter 15), and a link has been established between Italian hospitals worldwide (see Chapter 16).¹⁸ Store-and-forward teledermatology has been used in linking remote parts of Africa to teaching hospitals in the USA and Europe.

Suggestions were invited from the users in Nepal for improving the telemedicine service. The following were the main suggestions:

• Include separate funding for telemedicine in hospitals with telemedicine units and in central hospitals providing expert advice.
  
  
• Improve the technology for telemedicine from still images to videoconferencing.
  
  
• Increase the frequency of consultations and organize videoconferences.
  
  
• Define a list of cases for mandatory advice apart from emergency cases.
  
  
• Computerize the hospital information systems, which will help to follow up patients treated earlier.

It is not easy to answer the question whether telemedicine in a developing country is cost-effective, since this requires evaluation in a properly controlled, scientific trial. Unless this can be done, the answer to the question whether telemedicine represents an appropriate use of resources will remain unknown.¹⁹

Further reading

Blanchet KD. Innovative programs in telemedicine. Telemed J E Health 2008; 14: 318–22.

Clarke M, Thiyagarajan CA. A systematic review of technical evaluation in telemedicine systems. Telemed J E Health 2008; 14: 170–83.

iPath Association. Telemedicine in Developing Countries. Available at: ipath.ch/site/node/19.

Latifi R, ed. Establishing Telemedicine in Developing Countries: From Inception to Implementation. Amsterdam: IOS Press, 2004.

Wurm EM, Campbell TM, Soyer HP. Teledermatology: how to start a new teaching and diagnostic era in medicine. Dermatol Clin 2008; 26: 295–300.

References

¹ United Nations Development Programme. Nepal Millennium Development Goals (MDG): Progress Report 2005. Available at: www.undp.org.np/publication/html/mdg2005/mdg2005.php.

² World Health Organization. Country Health System Profile: Nepal. Available at: www.searo.who.int/en/Section313/Section1523_6870.htm.

³ Carnall D. Medical software’s free future. BMJ 2000; 321: 976.

⁴ American College of Radiology. ACR Standard for Teleradiology, 1998. Available at: imaging.stryker.com/images/ACR_Standards-Teleradiology.pdf.

⁵ IrfanView. Available at: www.irfanview.com.

⁶ Perednia DA, Brown NA. Teledermatology: one application of telemedicine. Bull Med Libr Assoc 1995; 83: 42–7.

⁷ McGee R, Tangalos EG. Delivery of health care to the undeserved: potential contributions of telecommunications technology. Mayo Clin Proc 1994; 69: 1131–6.

⁸ Szot A, Jacobson FL, Munn S et al. Diagnostic accuracy of chest X-rays acquired using a digital camera for low-cost teleradiology. Int J Med Inform 2004; 73: 65–73.

⁹ Krupinski E, Gonzales M, Gonzales C, Weinstein RS. Evaluation of a digital camera for acquiring radio-graphic images for telemedicine applications. Telemed J E Health 2000; 6: 297–302.

¹⁰ Krupinski EA, LeSueur B, Ellsworth L et al. Diagnostic accuracy and image quality using a digital camera for teledermatology. Telemed J 1999; 5: 257–3.

¹¹ Piccolo D, Soyer HP, Burgdorf W et al. Concordance between telepathologic diagnosis and conventional histopathologic diagnosis: a multiobserver store-and-forward study on 20 skin specimens. Arch Dermatol 2002: 138: 53–8.

¹² Lim AC, Egerton IB, See A, Shumack SP. Accuracy and reliability of store-and-forward teledermatology: preliminary results from the St George Teledermatology Project. Australas J Dermatol 2001; 42: 247–51.

¹³ High WA, Houston MS, Calobrisi SD et al. Assessment of the accuracy of low-cost store-and-forward teledermatology consultation. J Am Acad Dermatol 2000; 42: 776–83.

¹⁴ Metz CE. Basic principles of ROC analysis. Semin Nucl Med 1978; 8: 283–98.

¹⁵ Zweig MH, Campbell G. Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in medicine. Clin Chem 1993; 39: 561–77.

¹⁶ Hjelm NM. Benefits and drawbacks of telemedicine. J Telemed Telecare 2005; 11: 60–70.

¹⁷ Vassallo DJ, Swinfen P, Swinfen R, Wootton R. Experience with a low-cost telemedicine system in three developing countries. J Telemed Telecare 2001; 7(Suppl 1): 56–8.

¹⁸ Della Mea V, Beltrami CA. Telepathology applications of the Internet multimedia electronic mail. Med Inform (Lond) 1998; 23: 237–44.

¹⁹ Wootton R. The possible use of telemedicine in developing countries. J Telemed Telecare 1997; 3: 23–6.

Document(s) 19 de 31