Scientific methodology: study design

Metodologia científica: o desenho da pesquisa

Abstracts

Esse trabalho se propõe a mostrar como elaborar um projeto de pesquisa, iniciando pelo acesso à literatura e a sua análise crítica, como escolher os tipos de estudos e classificá-los em níveis, segundo as Diretrizes e Bases feitas pela AMB (Associação Médica Brasileira) e CFM (Conselho Federal de Medicina), aprovadas e adotadas pelo Ministério da Educação; e também analisa e comenta cada um desses tipos de estudo. Tece comentários de como definir a pergunta que deve ser respondida pelo próprio estudo, e a sua importância na elaboração de todos os tópicos do projeto. Finalmente fala da necessidade de termos em nosso país, trabalhos publicados seguindo a metodologia correta, para atingirmos um nível desejado na literatura internacional.

Metodologia; Pesquisa; Saúde

ORIGINAL ARTICLE

Scientific methodology: study design

Maria Luiza L. Amatuzzi^I; Marco Martins Amatuzzi^II; Luiz Eugênio Garcez Leme^III

^IBibliotecária e Pós-Graduanda em Ciências

^IIChairman at FMUSP and Head of the Department of Orthopedics and Traumatology. Post Graduation Department Coordinator.

^IIIAssociate PhD Professor, Geriatrician, Post Graduation Instructor

^{Correspondence}

SUMMARY

This paper aims to explain how to prepare a study project starting from literature evaluation and its critical analysis, how to choose study type as well as rating in levels according to Guidelines by AMB (Associação Médica Brasileira - Brazilian Medical Association) and CFM (Conselho Federal de Medicina – Brazilian Medical Council) approved and adopted by Education Ministry; also evaluates and comments each of these study types. Comments how to establish the question the study should answer and its importance in preparing all project items. Finally comments about the need of more studies performed with correct methodology being published in Brazil, in order to reach an international level.

Key words: Methodology; Research; Health.

The large number of technical health literature makes very difficult for a professional to update, in the best possibility making him/her to get too many knowledge with no relevancy criteria.

In fact, from the 100 scientific journals in 1900, there were above 100,000 in 1990. One single visit to Medline, only one among other data banks of medical literature as SciElo, Lilacs, etc, informs that there above 11 million articles are registered since the sixties. On the other hand, it can be observed that 80% of all references in the Science Citation Index are from less than 1,000 journals, and more importantly, 90% of the most important scientific achievements are published in only 150 among these journals.

In order to avoid such a publication accumulation, many low or fair in quality, some new rules appeared that, if followed by editorial boards would limit this number making published papers to be really profitable for most of researchers.

Scientific Methodology is a rigorous analysis of the methods used for obtaining scientific information, and when properly applied draws the way for the search or the truth necessarily in the correct context and statistic perspective.

Scientific Methodology is in the basis of Study Project preparation, in the way of structuring the truth search which is reproducible at any time by the investigators themselves (intra-observers) or by other groups (inter-observers), and their results may be accepted without any restriction.

In Scientific Research all starts with the formulation of the Question, originated from a doubt when seeing a patient, a previous reading or a suggestion for a project. The objectivity and simplicity of this question are quality factors in a research planning. The temptation of trying to answer other questions during a project may be dangerous, frequently jeopardizing the quality of the job. In any case, methodology should be driven towards finding the answer.

The Question drives the all research process and indicate the best answer design. Once the question is asked, the work, still in a project phase, should be rated according to its type: etiology, diagnosis, treatment, prognosis, prophylaxis or cost/benefit rate.

So, we have that for Treatment, the best design is Randomized Clinical Trials; for Prognosis, cohorte studies and finally, for Prophylaxis, Randomized Clinical Trials (Table 1).

The studies can be prospective or retrospective, observational or interventional.

Prospective studies: are programmed to evaluate future events, thus can more easily be structured under rigid scientific quality criteria. On the other hand are based on questions from the daily practice, from the perception of an eventual problem or from previous retrospective studies. Generally involve pilot studies protocols.

A Prospective Study may be:

Randomized when data are took by chance; it is impossible to know the next one (for example heads or tails).

Quasi Randomized: when it is possible to previously know which will be the next one (example: even or uneven days).

According to the evaluation they can be:

Blind when only the study subjects do not know the treatment

Double-blind when not only the subjects but also the investigator do not know the intervention.

Triple-blind when additionally to both above, the data analyzer also ignores the treatment of that individual subject.

Observational Studies are those where the observer only observes the evolution or the occurrence of a given event, as in epidemiological studies.

In Interventional Studies the investigator interfere in study subjects evolution, as in therapeutic studies.

Retrospective Studies: are based in alredy existent facts. Frequently present a higher number of bias than prospective studies, since structuring depends on data with quantity and quality sometimes irregular.

Once the project is ready, and its type is defined, next step is literature. In its review data are extracted with quality evaluation and later presentation and results interpretation.

Literature Review can be performed through Internet, by consulting data bases as MEDLINE, LILACS, EMBASE, etc. The search strategy for each kind of research is previously defined: for clinical and intervention situations, the research starts from the question words, their synonymous and associated key words.

Inside the chapter Literature Review it is mandatory a rigorous selection of the referenced articles, which will be considered by their importance an applicability to the theme. This is called Critical Evaluation of the Literature.

Finding a Systematic Review of the Literature on the subject could represent an enormous progress due to quality and quantity of information regarding the project subject.

In Treatment projects, patients allocation should be randomized, control groups clearly identified in regard to important characteristics of the outcome and patients included in the study adequately validated at their finish in regard to outcome and destination.

In Prognosis projects the sample should be representative of patients, in a well defined point of disease evolution, and the follow-up should be long enough and complete, with an independent and "blind" comparison to a well pre-defined reference standard.

In any circumstance, the Results should be clear, valid and contribute to patient care.

Based on its quality, studies are ranked in recommendation and evidence levels (^{Board 1}):

Level A – Systematic Literature Reviews;

Level B – Randomized Clinical Trials;

Level C – Prospective controlled non-randomized (cohorte);

Level D – Retrospective Studies;

Level E – Expert reports and consensual decisions.

In Brazil, in a range of 4,600 journals, the Conselho Federal de Medicina and Associação Médica Brasileira already structured standardization criteria for a adequately performing scientific studies, being its quality mandatory from ethical point of view.

Certainly works performed this way will then be published in impacting journals and referenced, being thus important in scholarship granting and post-graduation evaluations. CAPES (Fundação Coordenação de Aperfeiçoamento de Pessoal de Nível Superior) uses publication impact for evaluation of teachers, courses and the institutions offering post-graduation courses.

Trabalho recebido em 19/08/2002

Aprovado em 20/11/2002

Este arigo não tem vínculo financeiro

• 1. Castro AA. Avaliação crítica da Literatura: roteiro para a sessão de artigos de revista. Disponível em: <D:\My \documents\ ald_livro\Lv1_BEM\avaliação critica ald-incluir.doc.>. Acesso em maio 2002.
• 2. Castro AA, Clark OAC. Localizando informações para a prática clínica. Disponível em: <http://www.metodologia.com >. Acesso em maio 2002.
• 3. Castro AA, Clark, OAC. Medicina baseada em evidências. Disponível em: <http://www.evidencias.com>. Acesso em maio 2002.
• 4. Secaf, V. Artigo científico: do desafio à conquista. São Paulo: Editorial, 2000.
• 5. Vieira S, Hossne WS. Metodologia científica para área de saúde. Rio de Janeiro: Campus, 2001.
• 6. Vilhena V, Crestana MF. Produção Científica: critérios de avaliação de impacto. Rev Assoc Med Bras 48:1-25, 2002.
• ⁷
  7. PROJETO DIRETRIZES. Associação Médica Brasileira/ Conselho Federal de Medicina. 2001. Disponível em: <http://www.amb.org.br>. Acesso em maio 2002.

Correspondence to

Work performed at Instituto de Ortopedia e Traumatologia do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo

Rua Dr. Ovídio Pires de Campos, 333, Hall 3º andar, Cerqueira César

CEP 05403-010, São Paulo, Sp

Publication Dates

• Publication in this collection
  16 Apr 2003
• Date of issue
  Jan 2003

History

• Accepted
  20 Nov 2002
• Received
  19 Aug 2002