Accuracy of store-and-forward diagnosis in leprosy

Introduction

Leprosy is a chronic, incapacitating and infectious disease. Despite having been identified as a mycobacteriosis even before tuberculosis, the source and the modes of transmission are still undefined. ¹ Leprosy is still a public health problem in countries such as Brazil, ² and is a major cause of physical disability since it is frequently only diagnosed at an advanced stage. ^3,4

The World Health Organization (WHO) includes leprosy in the multi-agency special programme for research and training in tropical diseases, promoting treatment and advocating global goals for its elimination. Leprosy diagnosis is based on the patient's history and a clinical examination, without laboratory tests. ⁵ The poor training of non-specialist doctors contributes to late leprosy diagnosis. ⁶

Telemedicine allows a second opinion to be obtained rapidly, which may be important in cases in which diagnosis is difficult. Telemedicine may also contribute to the ongoing education of clinicians in public health clinics. ^7,8 Store-and-forward telemedicine is used in teledermatology, in particular for tumour-like skin lesions, thereby decreasing the cost of services by releasing personnel from diagnosis and the performance of laboratory tests; and it has received a high level of patient acceptance and satisfaction. ⁹ Such a service for leprosy could reduce the time delay for diagnosis and avoid the development of disabilities. To this end, a project was implemented at the School of Medicine of the University São Paulo (FMUSP). This work had the collaboration of PAHO/WHO, the Brazilian Ministry of Health and the Department of Health of both the State and the City of São Paulo. ¹⁰ The aim of the study was to compare the accuracy of diagnosis made by telemedicine with that made by in-person examination.

Methods

We selected eight Basic (first or second resolution) Public Health Unity clinics in outlying areas of the city of São Paulo. All the clinics had large numbers of cases of leprosy, a doctor prepared to participate, a director who agreed to participate and Internet access. All patients gave written consent to participate. The project was approved by the appropriate committees.

The doctors were instructed to forward cases in which leprosy was suspected either by the doctor him/herself or by a referring physician, without laboratory tests. For the purposes of analysis, opinions regarding diagnosis were classified either as leprosy or not leprosy. The diagnosis of leprosy used the WHO method, based on the history and clinical examination.

All of the doctors received a 5.1 megapixel digital camera (Cyber-Shot, Sony), as well as basic training in digital photography ^8,11 and in forwarding data to the website. Clinical images were saved as 1 MByte images (960 × 1280 pixels). The case history and clinical examination data, and at least two clinical images for each patient, were stored in an Internet-based system developed for teledermatology. The case records were stored in a database. Diagnosis and care of patients was the responsibility of the doctors who forwarded their cases.

After the examination in the public clinic, patients then attended a teaching hospital for an in-person examination. The benchmark was the clinical examination of one dermatologist involved as a researcher at the university hospital. Laboratory examinations were conducted only to confirm leprosy or no leprosy. The dermatologist that made the telemedicine diagnosis was not involved as a researcher in the project, and had no access to laboratory test results. The telemedicine diagnosis was based on the case file data stored in the database.

To analyse agreement, the accuracy, sensitivity and specificity were calculated. ^12,13 The data were processed using the statistical package SPSS (version 11.0).

Results

From August 2005 to April 2006, 142 suspected cases of leprosy were forwarded to the website by the doctors at the eight clinics. Of these, 36 cases were excluded: 17 because the patients did not attend the teaching hospital for examination and six because they were sent with laboratory examination results. In nine cases a telemedicine diagnosis was not possible based on the data sent to the website because the images were inadequate. In four cases a diagnosis could not be made in-person: two neurological examinations were necessary (paresthesis in fingers, foot or hands) and two required laboratory or clinical examination at another time (hypochromic lesions).

In 49 of the 106 cases laboratory examinations were conducted for leprosy classification: skin biopsy with baciloscopy in 47 cases and mycological examination at the teaching hospital in two cases.

The telemedicine diagnosis for 106 cases was compared with the in-person diagnosis and there was agreement in 74%. The sensitivity was 78% and the specificity was 31% (Table 1).

Discussion

Although leprosy has been described since antiquity and the aetiological agent was discovered even before that of tuberculosis, diagnosis is still based on clinical examination. This is largely due to a lack of resources, since the public health clinics where the majority of cases are diagnosed and treated do not have access to laboratory methods for diagnosis and lack specialists. ¹⁴

Many dermatology studies suggest that telemedicine can make diagnosis quicker and more effective, and thereby improve the effectiveness of the health services in diagnosing and caring for users. ^15,16 A few studies, mainly in teledermatology, have reached phase III and IV (randomized studies). ^17,18 We are not aware of any published telemedicine studies in infectious diseases that represent public health problems.

In the majority of cases that we studied, two medium resolution images, together with the case history, were sufficient for a rapid and effective second opinion, based on the fact that the clinical history is fundamental when it comes to suspecting leprosy, even when the patient attends for examination in person. However, 25% of the cases in the present study were excluded due to the poor quality of the images or for other reasons.

The accuracy of the telemedicine diagnosis of leprosy was similar to that reported in many general dermatology studies in the literature (about 70%). ^15–19 The specificity was very low. This is probably because leprosy is a polymorphic disease and hence difficult to diagnose in the initial stages. ¹⁴ A second opinion by store-and-forward methodology generally results in a less certain diagnosis than one based on in-person examination, in many cases because palpation is not possible. ²⁰ However, the results of the present study suggest that telemedicine may be useful for obtaining second opinions in programmes to control leprosy. Because store-and-forward telemedicine allows a case discussion to be generated, ^21,22 it may also contribute to greater involvement of primary care physicians and improved collaboration between medical colleagues.

The results of the present study suggest that the implementation a telemedicine service can assist in eradicating leprosy in areas where it is a public health problem. A mechanism can be established for referring suspected cases of leprosy in outlying regions of a major city with poor living conditions for a second opinion at a tertiary hospital. This is the principle behind the National Telehealth Programme (Programa Nacional de Telesaúde) currently being implanted in Brazil to improve the quality of basic health-care services and provide resources for education and epidemiology. ²³