ICML 9

9º Congresso Mundial de Informação em Saúde e Bibliotecas

Salvador, Bahia - Brasil, 20 a 23 de setembro de 2005

BVS4

4ª Reunião de Coordenação Regional da BVS

19 e 20 de setembro de 2005

Artificial intelligence in medical decision support: the role of Virtual Health Libraries in systematic review

Participantes:
  • Professora, Departamento de Epidemiologia e Bioestatística, Instituto de Saúde da Comunidade, Universidade Federal Fluminense  - Brasil
  • Pesquisadora, Núcleo de Investigação em Doenças Infecciosas Emergentes do Instituto de Bioquímica Médica do Centro de Ciências da Saúde da Universidade Federal do Rio de Janeiro  - Brasil
  • Chefe, Departamento de Epidemiologia e Antropologia do Instituto de Pesquisa Clínica Evandro Chagas da Fiocruz  - Brasil
  • Professor, Departamento de Sistemas de Informação, Instituto de Computação da Universidade Estadual de Campinas e do Departamento de Informática em Saúde da Universidade Federal de São Paulo  - País:
  • Professor, Departamento de Saúde Materno-Infantil e Coordenador da área de Informática em Saúde da Faculdade de Medicina da Universidade Federal Fluminense  - Brasil
  • Chefe, Serviço de Hemoterapia e Coordenador do Setor de Saúde do Trabalhador do Instituto de Pesquisa Clínica Evandro Chagas da Fiocruz  - Brasil
  • Médico infectologista, Instituto de Pesquisa Clínica Evandro Chagas da Fiocruz  - Brasil
  • Médica Infectologista, Ambulatório de Doenças Febris Agudas do Instituto de Pesquisa Clínica Evandro Chagas da Fiocruz  - Brasil
  • Chefe, Serviço de Vigilância em Saúde do Departamento de Epidemiologia e Antropologia do Instituto de Pesquisa Clínica Evandro Chagas da Fiocruz  - Brasil

Artificial intelligence in medical decision support: the role of Virtual Health Libraries in systematic review

Computerized medical decision support systems based on artificial intelligence (aimed at diagnostic or therapeutic guidelines) can improve safeness, quality and cost- efficiency of health care. In underdeveloped countries, this is a new research area, where impact could be greater than in developed ones. There is a need for advances in increasing capacity, accessing facilities and flexibility for such systems. We performed a systematic review on the use of artificial intelligence in medical decision support, based on database search in Cochrane Library and Health Sciences Latin American and Caribbean Library (Lilacs), both available at Virtual Health Library website. The keywords “artificial intelligence” and “telemedicine” were used. Cochrane Library search retrieved 71 references, and 10 were obtained in Lilacs. The most relevant findings of this review were: (a) systems aimed a t health professionals can make the choice of best evidence easier, since they present information in a more accessible or immediate format, at a specially relevant moments of care; (b) studies performed on emergency applications showed positive results regarding injuries such as eye lesions and maxillary fractures, but the only immediate improvement was consultation time reduction. In underdeveloped countries, specially in Latin America, health informatics implementation initiatives were more sparse. Efficacy studies were performed on teledermatology services in remote islands in Chile and on visual acuity examinations in Cuba, with satisfactory results. Nevertheless, virtually all studies are related to pilot projects in central capitalist countries, and outcome measures are mainly related to process variables or short-term results. Randomized clinical trials must be performed, with priority in developing countries, where possibly those technologies would be most beneficial. Long-term studies are needed in order to estimate remote impact of such interventions. The virtual libraries researched were important sources of references about this theme.