<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.1 20151215//EN" "https://jats.nlm.nih.gov/publishing/1.1/JATS-journalpublishing1.dtd">
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" article-type="letter" dtd-version="1.1" specific-use="sps-1.9" xml:lang="pt">
  <front>
    <journal-meta>
      <journal-id journal-id-type="nlm-ta">Arq Bras Cardiol</journal-id>
      <journal-id journal-id-type="publisher-id">abc</journal-id>
      <journal-title-group>
        <journal-title>Arquivos Brasileiros de Cardiologia</journal-title>
        <abbrev-journal-title abbrev-type="publisher">Arq. Bras. Cardiol.</abbrev-journal-title>
      </journal-title-group>
      <issn pub-type="ppub">0066-782X</issn>
      <issn pub-type="epub">1678-4170</issn>
      <publisher>
        <publisher-name>Sociedade Brasileira de Cardiologia - SBC</publisher-name>
      </publisher>
    </journal-meta>
    <article-meta>
      <article-id specific-use="scielo-v3" pub-id-type="publisher-id">LLRXVqDMLjhtmwGjQTGJJ7J</article-id>
      <article-id specific-use="scielo-v2" pub-id-type="publisher-id">S0066-782X2023000400501</article-id>
      <article-id pub-id-type="other">00501</article-id>
      <article-id pub-id-type="doi">10.36660/abc.20220539</article-id>
      <article-categories>
        <subj-group subj-group-type="heading">
          <subject>Carta Cient&#237;fica</subject>
        </subj-group>
      </article-categories>
      <title-group>
        <article-title>Servindo a Dois Mestres: Resolvendo o Dilema entre Metan&#225;lise de Dados Individuais e Metan&#225;lise de Dados Agregados de Estudos com Estatinas</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-6003-1314</contrib-id>
          <name>
            <surname>Tramujas</surname>
            <given-names>Lucas</given-names>
          </name>
          <role>Concep&#231;&#227;o e desenho da pesquisa,Obten&#231;&#227;o de dados,An&#225;lise e interpreta&#231;&#227;o dos dados, Reda&#231;&#227;o do manuscrito e Revis&#227;o cr&#237;tica do manuscrito quanto ao conte&#250;do intelectual importante</role>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Medrado</surname>
            <given-names>Fernando Azevedo</given-names>
            <suffix>J&#250;nior</suffix>
          </name>
          <role>Obten&#231;&#227;o de dados,An&#225;lise e interpreta&#231;&#227;o dos dados, Reda&#231;&#227;o do manuscrito e Revis&#227;o cr&#237;tica do manuscrito quanto ao conte&#250;do intelectual importante</role>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Cavalcanti</surname>
            <given-names>Alexandre Biasi</given-names>
          </name>
          <role>Concep&#231;&#227;o e desenho da pesquisa,An&#225;lise e interpreta&#231;&#227;o dos dados, Reda&#231;&#227;o do manuscrito e Revis&#227;o cr&#237;tica do manuscrito quanto ao conte&#250;do intelectual importante</role>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0003-2559-8689</contrib-id>
          <name>
            <surname>Pompilio</surname>
            <given-names>Carlos Eduardo</given-names>
          </name>
          <role>Concep&#231;&#227;o e desenho da pesquisa,An&#225;lise e interpreta&#231;&#227;o dos dados, Reda&#231;&#227;o do manuscrito e Revis&#227;o cr&#237;tica do manuscrito quanto ao conte&#250;do intelectual importante</role>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
      </contrib-group>
      <aff id="aff1">
        <label>1</label>
        <institution content-type="orgdiv1">Instituto de Pesquisas</institution>
        <institution content-type="orgname">Hcor</institution>
        <addr-line>
          <named-content content-type="city">S&#227;o Paulo</named-content>
          <named-content content-type="state">SP</named-content>
        </addr-line>
        <country country="BR">Brasil</country>
        <institution content-type="original">Instituto de Pesquisas, Hcor, S&#227;o Paulo, SP &#8211; Brasil</institution>
      </aff>
      <aff id="aff2">
        <label>2</label>
        <institution content-type="orgdiv2">Hospital das Cl&#237;nicas</institution>
        <institution content-type="orgdiv1">Faculdade de Medicina</institution>
        <institution content-type="orgname">Universidade de S&#227;o Paulo</institution>
        <addr-line>
          <named-content content-type="city">S&#227;o Paulo</named-content>
          <named-content content-type="state">SP</named-content>
        </addr-line>
        <country country="BR">Brasil</country>
        <institution content-type="original">Cl&#237;nica M&#233;dica, Hospital das Cl&#237;nicas, Faculdade de Medicina da Universidade de S&#227;o Paulo (HCFMUSP), S&#227;o Paulo, SP &#8211; Brasil</institution>
      </aff>
      <author-notes>
        <corresp id="c01"> Correspond&#234;ncia: Lucas Tramujas &#8226; Instituto de Pesquisa do Hcor &#8211; Rua Des. Eliseu Guilherme, 200. CEP 04004-030, Vila Mariana, S&#227;o Paulo, SP &#8211; Brasil E-mail: <email>ltramujas@hcor.com.br</email> </corresp>
        <fn fn-type="conflict">
          <p>Potencial conflito de interesse</p>
          <p>N&#227;o h&#225; conflito com o presente artigo.</p>
        </fn>
      </author-notes>
      <pub-date date-type="pub" publication-format="electronic">
        <day>21</day>
        <month>04</month>
        <year>2023</year>
      </pub-date>
      <pub-date date-type="collection" publication-format="electronic">
        <month>03</month>
        <year>2023</year>
      </pub-date>
      <volume>120</volume>
      <issue>4</issue>
      <elocation-id>e20220539</elocation-id>
      <history>
        <date date-type="received">
          <day>28</day>
          <month>07</month>
          <year>2022</year>
        </date>
        <date date-type="rev-recd">
          <day>27</day>
          <month>10</month>
          <year>2022</year>
        </date>
        <date date-type="accepted">
          <day>14</day>
          <month>12</month>
          <year>2022</year>
        </date>
      </history>
      <permissions>
        <license license-type="open-access" xlink:href="https://creativecommons.org/licenses/by/4.0/" xml:lang="en">
          <license-p> This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. </license-p>
        </license>
      </permissions>
      <kwd-group xml:lang="pt">
        <kwd>Metan&#225;lise</kwd>
        <kwd>Colesterol</kwd>
        <kwd>Doen&#231;a cardiovascular</kwd>
      </kwd-group>
      <counts>
        <fig-count count="0"/>
        <table-count count="2"/>
        <equation-count count="0"/>
        <ref-count count="7"/>
      </counts>
    </article-meta>
  </front>
  <body>
    <sec sec-type="intro">
      <title>Introdu&#231;&#227;o</title>
      <p>N&#237;veis elevados de colesterol de lipoprote&#237;na de baixa densidade (LDL-C) foram estabelecidos como um importante fator de risco para o desenvolvimento de aterosclerose.<sup><xref ref-type="bibr" rid="B1">1</xref></sup> Como corol&#225;rio desse racioc&#237;nio, surgiu uma abordagem centrada no LDL com base em estatinas e outros medicamentos capazes de reduzir o LDL-C e o risco cardiovascular.<sup><xref ref-type="bibr" rid="B1">1</xref></sup></p>
      <p>A <italic>Cholesterol Treatment Trialists&#8217; Collaboration</italic> (Colabora&#231;&#227;o de Investigadores Cl&#237;nicos no Tratamento de Colesterol), uma metan&#225;lise de dados individuais do paciente, encontrou uma rela&#231;&#227;o logar&#237;tmica entre o grau de redu&#231;&#227;o de LDL-C e eventos cardiovasculares.<sup><xref ref-type="bibr" rid="B2">2</xref></sup> De acordo com isso, diretrizes especializadas para prevenir doen&#231;as cardiovasculares recomendam o uso de estatinas para preven&#231;&#227;o prim&#225;ria e secund&#225;ria.<sup><xref ref-type="bibr" rid="B2">2</xref></sup></p>
      <p>Por outro lado, uma metan&#225;lise de dados agregados publicada por Byrne et al. sobre a associa&#231;&#227;o entre as redu&#231;&#245;es induzidas por estatinas nos n&#237;veis de LDL-C e as redu&#231;&#245;es absolutas e relativas em desfechos cl&#237;nicos individuais mostrou resultados discrepantes.<sup><xref ref-type="bibr" rid="B3">3</xref></sup> Os autores conclu&#237;ram que as redu&#231;&#245;es de risco absoluto do tratamento com estatinas na mortalidade por todas as causas, infarto do mioc&#225;rdio, e acidente vascular cerebral s&#227;o modestos. A metan&#225;lise tamb&#233;m sugere uma associa&#231;&#227;o inconclusiva entre as redu&#231;&#245;es induzidas por estatina nos n&#237;veis de LDL-C e os resultados cl&#237;nicos.<sup><xref ref-type="bibr" rid="B3">3</xref></sup></p>
      <p>Aqui, descreveremos as principais diferen&#231;as entre metan&#225;lise de dados individuais de pacientes e metan&#225;lise de dados agregados e como entender esses resultados para evitar vi&#233;s na tomada de decis&#245;es m&#233;dicas, considerando o exemplo de metan&#225;lises de ensaios com estatinas.</p>
    </sec>
    <sec>
      <title>Defini&#231;&#245;es de metan&#225;lise de dados agregados e metan&#225;lise de dados individuais de pacientes</title>
      <p>A metan&#225;lise de estudos de interven&#231;&#227;o &#233; um m&#233;todo estat&#237;stico para combinar todas as pesquisas publicadas sobre uma quest&#227;o de pesquisa espec&#237;fica e estimar os efeitos gerais do tratamento agrupados.<sup><xref ref-type="bibr" rid="B4">4</xref></sup></p>
      <p>Revis&#245;es sistem&#225;ticas tradicionais e metan&#225;lises s&#227;o baseadas em dados publicados agregados, o que significa que n&#227;o s&#227;o necess&#225;rios bancos de dados de estudos originais com dados individuais dos pacientes.<sup><xref ref-type="bibr" rid="B4">4</xref></sup> As medidas de efeito mais comuns usadas como uma estat&#237;stica resumida s&#227;o odds ratio (raz&#227;o de chances), risco relativo ou diferen&#231;a de risco para comparar resultados dicot&#244;micos (por exemplo, doen&#231;a versus aus&#234;ncia de doen&#231;a) ou diferen&#231;a m&#233;dia ou diferen&#231;a m&#233;dia padronizada para comparar resultados cont&#237;nuos (por exemplo, medi&#231;&#227;o dos n&#237;veis de LDL-C) entre os grupos de tratamento.<sup><xref ref-type="bibr" rid="B4">4</xref></sup></p>
      <p>Por outro lado, a metan&#225;lise de dados individuais de participantes envolve obter, combinar e analisar bancos de dados de estudos originais.<sup><xref ref-type="bibr" rid="B4">4</xref></sup> O conceito de dados individuais de paciente &#233; diferente de dados agregados, pois o primeiro se refere aos dados coletados para cada participante do estudo, enquanto o segundo se refere &#224;s informa&#231;&#245;es m&#233;dias ou estimadas de todos os indiv&#237;duos em um estudo.<sup><xref ref-type="bibr" rid="B5">5</xref></sup> Por exemplo, em um ensaio cl&#237;nico de colesterol, os dados individuais do paciente podem ser os n&#237;veis de colesterol pr&#233;-tratamento e p&#243;s-tratamento ou caracter&#237;sticas cl&#237;nicas basais. Isso permitiria estimar a diferen&#231;a no LDL-C para cada paciente e avaliar o tamanho da associa&#231;&#227;o dessas diferen&#231;as no LDL-C e os desfechos cl&#237;nicos. A <xref ref-type="table" rid="t1">Tabela 1</xref> lista as vantagens e desvantagens dos dados agregados e da metan&#225;lise de dados individuais de pacientes.</p>
      <p>
        <table-wrap id="t1">
          <label>Tabela 1</label>
          <caption>
            <title>&#8211; Resumo das vantagens e desvantagens da metan&#225;lise de dados agregados e da metan&#225;lise de dados individuais de pacientes</title>
          </caption>
          <table frame="hsides" rules="groups">
            <colgroup>
              <col/>
              <col/>
              <col/>
            </colgroup>
            <thead>
              <tr>
                <th align="left" style="font-weight:normal"> </th>
                <th align="left">Vantagens</th>
                <th align="left">Desvantagens</th>
              </tr>
            </thead>
            <tbody>
              <tr>
                <td>
                  <bold>Metan&#225;lise de dados agregados</bold>
                </td>
                <td>Mais f&#225;cil de executar</td>
                <td>Depende da qualidade dos dados originais e de como os autores os relatam</td>
              </tr>
              <tr>
                <td> </td>
                <td>Requer menos tempo</td>
                <td>Os desfechos prim&#225;rios e secund&#225;rios do estudo podem diferir daqueles examinados na metan&#225;lise</td>
              </tr>
              <tr>
                <td> </td>
                <td>Menos caro</td>
                <td>Achados negativos s&#227;o mais propensos a serem negligenciados devido ao vi&#233;s de publica&#231;&#227;o</td>
              </tr>
              <tr>
                <td> </td>
                <td> </td>
                <td>Poder limitado para analisar os efeitos do tratamento em subgrupos espec&#237;ficos de interesse que diferem daqueles avaliados em estudos prim&#225;rios</td>
              </tr>
              <tr>
                <td>
                  <bold>Metan&#225;lise de dados individuais de pacientes</bold>
                </td>
                <td>Dados preliminares e n&#227;o publicados podem ser inclu&#237;dos, reduzindo o risco de vi&#233;s de publica&#231;&#227;o</td>
                <td>Demorado devido &#224; aquisi&#231;&#227;o de dados, extra&#231;&#227;o, an&#225;lise e necessidade de colabora&#231;&#227;o</td>
              </tr>
              <tr>
                <td> </td>
                <td>Avalia&#231;&#227;o abrangente do protocolo, m&#233;todos e qualidade geral do estudo</td>
                <td>A coleta de dados de certos estudos, particularmente os mais antigos, pode ser dif&#237;cil</td>
              </tr>
              <tr>
                <td> </td>
                <td>Permite a an&#225;lise de subgrupo para entender melhor o impacto de uma interven&#231;&#227;o no n&#237;vel do paciente</td>
                <td>Mais caro</td>
              </tr>
              <tr>
                <td> </td>
                <td>Se aplic&#225;vel, a an&#225;lise estat&#237;stica pode ser ajustada &#224;s diferen&#231;as da linha de base</td>
                <td>Mais experi&#234;ncia necess&#225;ria</td>
              </tr>
            </tbody>
          </table>
        </table-wrap>
      </p>
    </sec>
    <sec>
      <title>Quais s&#227;o as diferen&#231;as entre a metan&#225;lise de dados individuais do paciente e a metan&#225;lise de dados agregados dos ensaios com estatinas em termos de resultados?</title>
      <p>A <italic>Cholesterol Treatment Trialists&#8217; Collaboration</italic> , uma metan&#225;lise de dados individuais de pacientes com 170.000 pacientes e 26 estudos cl&#237;nicos, teve como objetivo avaliar a seguran&#231;a e efic&#225;cia de redu&#231;&#245;es mais intensivas de LDL-C.<sup><xref ref-type="bibr" rid="B2">2</xref></sup> Dados individuais de participantes de ensaios cl&#237;nicos randomizados comparando tratamentos mais intensivos versus regimes de estatina menos intensivos (5 ensaios; 39.612 indiv&#237;duos) e estatinas versus placebo foram inclu&#237;dos (21 ensaios; 129.526 indiv&#237;duos). Eles mostraram uma redu&#231;&#227;o nos principais desfechos cardiovasculares para cada redu&#231;&#227;o de 1 mmol/L no LDL-C (raz&#227;o de taxas, 0,78; intervalo de confian&#231;a [IC] de 95%, 0,76 a 0,80; p&lt;0,0001).<sup><xref ref-type="bibr" rid="B2">2</xref></sup></p>
      <p>A metan&#225;lise de dados agregados publicada por Byrne et al. para avaliar uma quest&#227;o semelhante obteve resultados inconclusivos.<sup><xref ref-type="bibr" rid="B4">4</xref></sup> A associa&#231;&#227;o entre redu&#231;&#227;o do LDL-C e infarto do mioc&#225;rdio n&#227;o foi estatisticamente significativa, a associa&#231;&#227;o entre redu&#231;&#227;o do LDL-C e &#243;bito foi significativa apenas na escala relativa, mas n&#227;o absoluta, e a associa&#231;&#227;o entre a redu&#231;&#227;o do LDL-C e acidente vascular cerebral foi significativa com uma redu&#231;&#227;o nos riscos absolutos e relativos.<sup><xref ref-type="bibr" rid="B3">3</xref></sup> Este estudo usou meta-regress&#227;o, que analisa a redu&#231;&#227;o agregada no LDL-C e o tamanho do efeito (agregado) de cada ensaio. Por outro lado, nos dados individuais dos participantes, a metan&#225;lise considera a varia&#231;&#227;o no LDL-C e o resultado de cada paciente, com uma granularidade muito maior de informa&#231;&#245;es.<sup><xref ref-type="bibr" rid="B3">3</xref></sup></p>
    </sec>
    <sec>
      <title>Como entender esses resultados discrepantes? Al&#233;m disso, em que situa&#231;&#245;es se espera que a metan&#225;lise de dados agregados e a metan&#225;lise de um paciente individual gerem respostas id&#234;nticas para uma pergunta espec&#237;fica?</title>
      <p>Quando a quest&#227;o diz respeito ao efeito comparativo de dois tratamentos alternativos em um resultado cl&#237;nico em uma determinada popula&#231;&#227;o, a estimativa combinada do efeito dos dados agregados ou da metan&#225;lise de dados de pacientes individuais deve ser semelhante.<sup><xref ref-type="bibr" rid="B5">5</xref></sup></p>
      <p>No entanto, discrep&#226;ncias podem surgir por v&#225;rios motivos. Em primeiro lugar, pode haver diferen&#231;as nos estudos originais inscritos: regimes de estatinas de alta intensidade versus regimes de baixa intensidade foram considerados apenas na metan&#225;lise de dados de pacientes individuais. Como resultado, os pacientes inscritos na metan&#225;lise de dados agregados apresentavam menor risco cardiovascular. Essa caracter&#237;stica da metan&#225;lise de dados agregados de LDL-C ajuda a explicar tamanhos de efeito absolutos menores, que dependem do risco basal (assumindo redu&#231;&#227;o constante do risco relativo em diferentes riscos basais).</p>
      <p>Em segundo lugar, a metan&#225;lise de dados agregados tem um poder menor para detectar rela&#231;&#245;es entre caracter&#237;sticas no n&#237;vel do paciente e tamanhos de efeito. Um estudo meta-epidemiol&#243;gico mostra que as estimativas de meta-regress&#227;o de dados agregados s&#227;o menos precisas do que aquelas das metan&#225;lises de dados individuais de pacientes.<sup><xref ref-type="bibr" rid="B6">6</xref></sup> Erros aleat&#243;rios em estimativas de meta-regress&#227;o frequentemente far&#227;o com que difiram substancialmente daqueles obtidos em metan&#225;lise de dados individuais.<sup><xref ref-type="bibr" rid="B7">7</xref></sup> Essa fal&#225;cia &#233; conhecida como vi&#233;s de agrega&#231;&#227;o, que &#233; definido pela suposi&#231;&#227;o de que uma associa&#231;&#227;o entre duas vari&#225;veis em n&#237;vel de grupo &#233; igual &#224; associa&#231;&#227;o entre a vari&#225;vel correspondente em n&#237;vel individual.<sup><xref ref-type="bibr" rid="B7">7</xref></sup></p>
      <p>Em terceiro lugar, nem todas as caracter&#237;sticas que podem impactar significativamente o funcionamento de uma interven&#231;&#227;o podem ser adequadamente avaliadas em uma metan&#225;lise de dados agregados, mais comumente porque os efeitos de acordo com as caracter&#237;sticas de linha de base de interesse n&#227;o est&#227;o dispon&#237;veis nos artigos relatados.<sup><xref ref-type="bibr" rid="B7">7</xref></sup></p>
      <p>Em quarto lugar, no caso das metan&#225;lises do LDL-C, a metan&#225;lise dos dados individuais do paciente avaliou a associa&#231;&#227;o entre a varia&#231;&#227;o do LDL-C e um desfecho composto (morte coronariana ou infarto do mioc&#225;rdio n&#227;o fatal). Em contraste, a metan&#225;lise de dados agregados avaliou a associa&#231;&#227;o entre LDL-C e morte, acidente vascular cerebral ou infarto do mioc&#225;rdio separadamente.<sup><xref ref-type="bibr" rid="B2">2</xref> , <xref ref-type="bibr" rid="B3">3</xref></sup></p>
      <p>&#192; medida que o n&#250;mero de eventos para um resultado composto &#233; maior, a precis&#227;o das estimativas de efeito aumenta. Quando os estudos originais n&#227;o relatam os efeitos do tratamento para o resultado combinado de interesse, a metan&#225;lise usando dados agregados n&#227;o &#233; poss&#237;vel. Por outro lado, &#233; poss&#237;vel gerar e analisar resultados combinados avaliando bancos de dados de ensaios originais.</p>
      <p>Portanto, em compara&#231;&#227;o com metan&#225;lises de dados agregados, metan&#225;lises de dados individuais s&#227;o superiores na investiga&#231;&#227;o de rela&#231;&#245;es entre efeitos em vari&#225;veis intermedi&#225;rias e desfechos cl&#237;nicos e na an&#225;lise de poss&#237;veis modificadores de tratamento. Por outro lado, uma metan&#225;lise de dados agregados ou individuais de pacientes deve oferecer respostas semelhantes para quest&#245;es de efeito do tratamento, desde que use as mesmas defini&#231;&#245;es para popula&#231;&#227;o, tratamento, comparador e desfechos e inclua os mesmos estudos originais.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <title>Refer&#234;ncias</title>
      <ref id="B1">
        <label>1</label>
        <mixed-citation>1. Faludi AA, Izar MCO, Saraiva JFK, Chacra APM, Bianco HT, Afiune A Neto, et al. Atualiza&#231;&#227;o da Diretriz Brasileira de Dislipidemias e Preven&#231;&#227;o da Aterosclerose &#8211; 2017. Arq Bras Cardiol. 2017;109(2 Supl 1):1-76. doi: 10.5935/abc.20170121.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Faludi</surname>
              <given-names>AA</given-names>
            </name>
            <name>
              <surname>Izar</surname>
              <given-names>MCO</given-names>
            </name>
            <name>
              <surname>Saraiva</surname>
              <given-names>JFK</given-names>
            </name>
            <name>
              <surname>Chacra</surname>
              <given-names>APM</given-names>
            </name>
            <name>
              <surname>Bianco</surname>
              <given-names>HT</given-names>
            </name>
            <name>
              <surname>Afiune</surname>
              <given-names>A</given-names>
              <suffix>Neto</suffix>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Atualiza&#231;&#227;o da Diretriz Brasileira de Dislipidemias e Preven&#231;&#227;o da Aterosclerose &#8211; 2017</article-title>
          <source>Arq Bras Cardiol</source>
          <year>2017</year>
          <volume>109</volume>
          <issue>2</issue>
          <supplement>Supl 1</supplement>
          <fpage>1</fpage>
          <lpage>76</lpage>
          <issue-id pub-id-type="doi">10.5935/abc.20170121</issue-id>
        </element-citation>
      </ref>
      <ref id="B2">
        <label>2</label>
        <mixed-citation>2. Cholesterol Treatment Trialists&#8217; (CTT) Collaboration; Baigent C, Blackwell L, Emberson J, Holland LE, Reith C, et al. Efficacy and Safety of More Intensive Lowering of LDL Cholesterol: A Meta-Analysis of Data from 170,000 Participants in 26 Randomised Trials. Lancet. 2010;376(9753):1670-81. doi: 10.1016/S0140-6736(10)61350-5.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <collab>Cholesterol Treatment Trialists&#8217; (CTT) Collaboration</collab>
            <name>
              <surname>Baigent</surname>
              <given-names>C</given-names>
            </name>
            <name>
              <surname>Blackwell</surname>
              <given-names>L</given-names>
            </name>
            <name>
              <surname>Emberson</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Holland</surname>
              <given-names>LE</given-names>
            </name>
            <name>
              <surname>Reith</surname>
              <given-names>C</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Efficacy and Safety of More Intensive Lowering of LDL Cholesterol: A Meta-Analysis of Data from 170,000 Participants in 26 Randomised Trials</article-title>
          <source>Lancet</source>
          <year>2010</year>
          <volume>376</volume>
          <issue>9753</issue>
          <fpage>1670</fpage>
          <lpage>1681</lpage>
          <issue-id pub-id-type="doi">10.1016/S0140-6736(10)61350-5</issue-id>
        </element-citation>
      </ref>
      <ref id="B3">
        <label>3</label>
        <mixed-citation>3. Byrne P, Demasi M, Jones M, Smith SM, O&#8217;Brien KK, DuBroff R. Evaluating the Association between Low-Density Lipoprotein Cholesterol Reduction and Relative and Absolute Effects of Statin Treatment: A Systematic Review and Meta-analysis. JAMA Intern Med. 2022;182(5):474-81. doi: 10.1001/jamainternmed.2022.0134.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Byrne</surname>
              <given-names>P</given-names>
            </name>
            <name>
              <surname>Demasi</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Jones</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Smith</surname>
              <given-names>SM</given-names>
            </name>
            <name>
              <surname>O&#8217;Brien</surname>
              <given-names>KK</given-names>
            </name>
            <name>
              <surname>DuBroff</surname>
              <given-names>R</given-names>
            </name>
          </person-group>
          <article-title>Evaluating the Association between Low-Density Lipoprotein Cholesterol Reduction and Relative and Absolute Effects of Statin Treatment: A Systematic Review and Meta-analysis</article-title>
          <source>JAMA Intern Med</source>
          <year>2022</year>
          <volume>182</volume>
          <issue>5</issue>
          <fpage>474</fpage>
          <lpage>481</lpage>
          <issue-id pub-id-type="doi">10.1001/jamainternmed.2022.0134</issue-id>
        </element-citation>
      </ref>
      <ref id="B4">
        <label>4</label>
        <mixed-citation>4. Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, et al. Cochrane Handbook for Systematic Reviews of Interventions version 6.3 (updated February 2022) [Internet]. Chichester: Cochrane; 2022 [cited 2022 07 Jul]. Available from: <ext-link ext-link-type="uri" xlink:href="http://www.training.cochrane.org/handbook">www.training.cochrane.org/handbook</ext-link>.</mixed-citation>
        <element-citation publication-type="webpage">
          <person-group person-group-type="author">
            <name>
              <surname>Higgins</surname>
              <given-names>JPT</given-names>
            </name>
            <name>
              <surname>Thomas</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Chandler</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Cumpston</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Li</surname>
              <given-names>T</given-names>
            </name>
            <name>
              <surname>Page</surname>
              <given-names>MJ</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <source>Cochrane Handbook for Systematic Reviews of Interventions version 6.3 (updated February 2022)</source>
          <comment>Internet</comment>
          <publisher-loc>Chichester</publisher-loc>
          <publisher-name>Cochrane</publisher-name>
          <year>2022</year>
          <date-in-citation content-type="cited-date">cited 2022 07 Jul</date-in-citation>
          <ext-link ext-link-type="uri" xlink:href="http://www.training.cochrane.org/handbook">www.training.cochrane.org/handbook</ext-link>
        </element-citation>
      </ref>
      <ref id="B5">
        <label>5</label>
        <mixed-citation>5. Smith CT, Marcucci M, Nolan SJ, Iorio A, Sudell M, Riley R, et al. Individual Participant Data Meta-Analyses Compared with Meta-Analyses Based on Aggregate Data. Cochrane Database Syst Rev. 2016;9(9):MR000007. doi: 10.1002/14651858.MR000007.pub3.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Smith</surname>
              <given-names>CT</given-names>
            </name>
            <name>
              <surname>Marcucci</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Nolan</surname>
              <given-names>SJ</given-names>
            </name>
            <name>
              <surname>Iorio</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Sudell</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Riley</surname>
              <given-names>R</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Individual Participant Data Meta-Analyses Compared with Meta-Analyses Based on Aggregate Data</article-title>
          <source>Cochrane Database Syst Rev</source>
          <year>2016</year>
          <volume>9</volume>
          <issue>9</issue>
          <size units="pages">MR000007</size>
          <issue-id pub-id-type="doi">10.1002/14651858.MR000007.pub3</issue-id>
        </element-citation>
      </ref>
      <ref id="B6">
        <label>6</label>
        <mixed-citation>6. Lambert PC, Sutton AJ, Abrams KR, Jones DR. A Comparison of Summary Patient-Level Covariates in Meta-Regression with Individual Patient Data Meta-Analysis. J Clin Epidemiol. 2002;55(1):86-94. doi: 10.1016/s0895-4356(01)00414-0.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Lambert</surname>
              <given-names>PC</given-names>
            </name>
            <name>
              <surname>Sutton</surname>
              <given-names>AJ</given-names>
            </name>
            <name>
              <surname>Abrams</surname>
              <given-names>KR</given-names>
            </name>
            <name>
              <surname>Jones</surname>
              <given-names>DR</given-names>
            </name>
          </person-group>
          <article-title>A Comparison of Summary Patient-Level Covariates in Meta-Regression with Individual Patient Data Meta-Analysis</article-title>
          <source>J Clin Epidemiol</source>
          <year>2002</year>
          <volume>55</volume>
          <issue>1</issue>
          <fpage>86</fpage>
          <lpage>94</lpage>
          <issue-id pub-id-type="doi">10.1016/s0895-4356(01)00414-0</issue-id>
        </element-citation>
      </ref>
      <ref id="B7">
        <label>7</label>
        <mixed-citation>7. Geissb&#252;hler M, Hincapi&#233; CA, Aghlmandi S, Zwahlen M, J&#252;ni P, Costa BR. Most Published Meta-Regression Analyses Based on Aggregate Data Suffer from Methodological Pitfalls: A Meta-Epidemiological Study. BMC Med Res Methodol. 2021;21(1):123. doi: 10.1186/s12874-021-01310-0.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Geissb&#252;hler</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Hincapi&#233;</surname>
              <given-names>CA</given-names>
            </name>
            <name>
              <surname>Aghlmandi</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Zwahlen</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>J&#252;ni</surname>
              <given-names>P</given-names>
            </name>
            <name>
              <surname>Costa</surname>
              <given-names>BR</given-names>
            </name>
          </person-group>
          <article-title>Most Published Meta-Regression Analyses Based on Aggregate Data Suffer from Methodological Pitfalls: A Meta-Epidemiological Study</article-title>
          <source>BMC Med Res Methodol</source>
          <year>2021</year>
          <volume>21</volume>
          <issue>1</issue>
          <size units="pages">123</size>
          <issue-id pub-id-type="doi">10.1186/s12874-021-01310-0</issue-id>
        </element-citation>
      </ref>
    </ref-list>
    <fn-group>
      <fn fn-type="other">
        <p>Vincula&#231;&#227;o acad&#234;mica</p>
        <p>N&#227;o h&#225; vincula&#231;&#227;o deste estudo a programas de p&#243;s-gradua&#231;&#227;o.</p>
      </fn>
      <fn fn-type="other">
        <p>Aprova&#231;&#227;o &#233;tica e consentimento informado</p>
        <p>Este artigo n&#227;o cont&#233;m estudos com humanos ou animais realizados por nenhum dos autores.</p>
      </fn>
    </fn-group>
    <fn-group>
      <fn fn-type="supported-by">
        <p>Fontes de financiamento: O presente estudo n&#227;o teve fontes de financiamento externas.</p>
      </fn>
    </fn-group>
  </back>
  <sub-article article-type="translation" id="TRen" xml:lang="en">
    <front-stub>
      <article-categories>
        <subj-group subj-group-type="heading">
          <subject>Research Letter</subject>
        </subj-group>
      </article-categories>
      <title-group>
        <article-title>Serving Two Masters: Resolving the Dilemma between Individual Patient Data Meta-Analysis and Aggregate Data Meta-Analysis from Statin Trials</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-6003-1314</contrib-id>
          <name>
            <surname>Tramujas</surname>
            <given-names>Lucas</given-names>
          </name>
          <role>Conception and design of the research,Acquisition of data,Analysis and interpretation of the data, Writing of the manuscript and Critical revision of the manuscript for important intellectual content</role>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Medrado</surname>
            <given-names>Fernando Azevedo</given-names>
            <suffix>J&#250;nior</suffix>
          </name>
          <role>Acquisition of data,Analysis and interpretation of the data, Writing of the manuscript and Critical revision of the manuscript for important intellectual content</role>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Cavalcanti</surname>
            <given-names>Alexandre Biasi</given-names>
          </name>
          <role>Conception and design of the research,Analysis and interpretation of the data, Writing of the manuscript and Critical revision of the manuscript for important intellectual content</role>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0003-2559-8689</contrib-id>
          <name>
            <surname>Pompilio</surname>
            <given-names>Carlos Eduardo</given-names>
          </name>
          <role>Conception and design of the research,Analysis and interpretation of the data, Writing of the manuscript and Critical revision of the manuscript for important intellectual content</role>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
      </contrib-group>
      <aff id="aff1001">
        <label>1</label>
        <country country="BR">Brazil</country>
        <institution content-type="original">Instituto de Pesquisas, Hcor, S&#227;o Paulo, SP &#8211; Brazil</institution>
      </aff>
      <aff id="aff2001">
        <label>2</label>
        <country country="BR">Brazil</country>
        <institution content-type="original">Cl&#237;nica M&#233;dica, Hospital das Cl&#237;nicas, Faculdade de Medicina da Universidade de S&#227;o Paulo (HCFMUSP), S&#227;o Paulo, SP &#8211; Brazil</institution>
      </aff>
      <author-notes>
        <corresp id="c01001">Mailing Address: Lucas Tramujas &#8226; Instituto de Pesquisa do Hcor &#8211; Rua Des. Eliseu Guilherme, 200. Postal Code 04004-030, Vila Mariana, S&#227;o Paulo, SP &#8211; Brazil E-mail: ltramujas@hcor.com.br</corresp>
        <fn fn-type="conflict">
          <p>Potential conflict of interest</p>
          <p>No potential conflict of interest relevant to this article was reported.</p>
        </fn>
      </author-notes>
      <kwd-group xml:lang="en">
        <kwd>Meta-Analysis</kwd>
        <kwd>Cholesterol</kwd>
        <kwd>Cardiovascular Disease</kwd>
      </kwd-group>
    </front-stub>
    <body>
      <sec sec-type="intro">
        <title>Introduction</title>
        <p>Elevated levels of low-density lipoprotein-cholesterol (LDL-C) have been established as an important risk factor for the development of atherosclerosis.<sup><xref ref-type="bibr" rid="B1">1</xref></sup> As a corollary of this reasoning, an LDL-centric approach has emerged based on statins and other medications able to reduce LDL-C and cardiovascular risk.<sup><xref ref-type="bibr" rid="B1">1</xref></sup></p>
        <p>The Cholesterol Treatment Trialists&#8217; Collaboration, an individual patient data meta-analysis, found a log-linear relationship between the degree of LDL-C reduction and cardiovascular events.<sup><xref ref-type="bibr" rid="B2">2</xref></sup> According to this, expert guidelines for preventing cardiovascular diseases recommend statin use for primary and secondary prevention.<sup><xref ref-type="bibr" rid="B2">2</xref></sup></p>
        <p>On the other hand, an aggregate data meta-analysis published by Byrne et al. on the association between statin-induced reductions in LDL-C levels and the absolute and relative reductions in individual clinical outcomes has shown discrepant results.<sup><xref ref-type="bibr" rid="B3">3</xref></sup> The authors concluded that the absolute risk reductions of treatment with statins in all-cause mortality, myocardial infarction, and stroke are modest. The meta-analysis also suggests an inconclusive association between statin-induced reductions in LDL-C levels and clinical outcomes.<sup><xref ref-type="bibr" rid="B3">3</xref></sup></p>
        <p>Here, we will describe the main differences between individual patient data and aggregate data meta-analysis and how to make sense of these results to avoid bias in medical decision-making, considering the example of meta-analyses of statin trials.</p>
      </sec>
      <sec>
        <title>Definitions of meta-analysis of aggregate data and meta-analysis of individual patient data</title>
        <p>Meta-analysis of interventional studies is a statistical method to combine all published research on a specific research question and estimate pooled overall treatment effects.<sup><xref ref-type="bibr" rid="B4">4</xref></sup></p>
        <p>Traditional systematic reviews and meta-analyses are based on aggregate published data, meaning that databases from original trials with data from individual patients are not needed.<sup><xref ref-type="bibr" rid="B4">4</xref></sup> The most common measures of effect used as a summary statistic are odds ratio, relative risk, or risk difference to compare dichotomous outcomes (e.g., disease versus no disease) or mean difference or standardized mean difference to compare continuous outcomes (e.g., LDL-C levels measurement) between the treatment groups.<sup><xref ref-type="bibr" rid="B4">4</xref></sup></p>
        <p>Conversely, the individual participant data meta-analysis involves obtaining, combining, and analyzing databases from original studies.<sup><xref ref-type="bibr" rid="B4">4</xref></sup> The concept of individual patient data is distinct from aggregate data in that the first refers to data collected for each study participant, while the second refers to information averaged or estimated across all individuals in a study.<sup><xref ref-type="bibr" rid="B5">5</xref></sup> For example, in a cholesterol trial, the individual patient data could be the pre-treatment and post-treatment cholesterol levels or baseline clinical characteristics. This would allow estimation of the difference in LDL-C for each patient and assess the size of the association of these LDL-C differences and clinical outcomes. <xref ref-type="table" rid="t1001">Table 1</xref> lists the advantages and drawbacks of aggregate data and individual patient data meta-analysis.</p>
        <p>
          <table-wrap id="t1001">
            <label>Table 1</label>
            <caption>
              <title>&#8211; Summary of advantages and drawbacks of aggregate data meta-analysis and individual patient data meta-analysis</title>
            </caption>
            <table frame="hsides" rules="groups">
              <colgroup>
                <col/>
                <col/>
                <col/>
              </colgroup>
              <thead>
                <tr>
                  <th align="left" style="font-weight:normal"> </th>
                  <th align="left">Advantages</th>
                  <th align="left">Drawbacks</th>
                </tr>
              </thead>
              <tbody>
                <tr>
                  <td rowspan="4">
                    <bold>Aggregate data meta-analysis</bold>
                  </td>
                  <td>Easier to perform</td>
                  <td>It is dependent on the original data quality and how the authors report it</td>
                </tr>
                <tr>
                  <td>Require less time</td>
                  <td>Primary and secondary study endpoints may differ from those examined in the meta-analysis</td>
                </tr>
                <tr>
                  <td>Less expensive</td>
                  <td>Negative findings are more likely to be overlooked due to publication bias</td>
                </tr>
                <tr>
                  <td> </td>
                  <td>Limited power to analyze treatment effects in specific subgroups of interest that differ from those evaluated in primary studies</td>
                </tr>
                <tr>
                  <td rowspan="4">
                    <bold>Individual patient data meta-analysis</bold>
                  </td>
                  <td>Preliminary and unpublished data can be included, reducing the risk of publication bias</td>
                  <td>Time-consuming due to data acquisition, extraction, analysis, and the need for collaboration</td>
                </tr>
                <tr>
                  <td>Comprehensive assessment of the protocol, methods, and overall study quality</td>
                  <td>Data collection from certain studies, particularly the oldest, may be difficult</td>
                </tr>
                <tr>
                  <td>Allows for subgroup analysis better to understand an intervention's impact at the patient level</td>
                  <td>More expensive</td>
                </tr>
                <tr>
                  <td>If applicable, statistical analysis can be adjusted to baseline differences</td>
                  <td>More expertise required</td>
                </tr>
              </tbody>
            </table>
          </table-wrap>
        </p>
      </sec>
      <sec>
        <title>What are the differences between individual patient data and aggregate data meta-analysis from statins trials in terms of results?</title>
        <p>The Cholesterol Treatment Trialists&#8217; Collaboration, a meta-analysis of individual patient data with 170.000 patients and 26 clinical studies, aimed to assess the safety and efficacy of more intensive reductions of LDL-C.<sup><xref ref-type="bibr" rid="B2">2</xref></sup> Individual participant data from randomized clinical trials comparing more intensive versus less intensive statin regimens (5 trials; 39,612 individuals) and statins versus placebo were included (21 trials; 129,526 individuals). They showed a reduction in major cardiovascular outcomes for every 1 mmol/L reduction in LDL-C (rate ratio, 0.78; 95% confidence interval [CI], 0.76 to 0.80; p&lt;0.0001).<sup><xref ref-type="bibr" rid="B2">2</xref></sup></p>
        <p>The aggregate data meta-analysis published by Byrne et al. to evaluate a similar question obtained inconclusive results.<sup><xref ref-type="bibr" rid="B4">4</xref></sup> The association between LDL-C reduction and myocardial infarction was not statistically significant, the association between LDL-C reduction and death was significant only in the relative but not in the absolute scale, and the association between LDL-C reduction and stroke was significant with a reduction in both absolute and relative risks.<sup><xref ref-type="bibr" rid="B3">3</xref></sup> This study used meta-regression, which analyzes the aggregate reduction in LDL-C and (aggregate) effect size from each trial. By contrast, in the individual participant data, meta-analysis considers the variation in LDL-C and outcome for each patient, with a much greater granularity of information.<sup><xref ref-type="bibr" rid="B3">3</xref></sup></p>
        <p>How one makes sense of these discrepant results? Furthermore, in what situations are aggregate data meta-analysis and an individual patient meta-analysis expected to generate identical answers for a specific question?</p>
        <p>When the question concerns the comparative effect of two alternative treatments on a clinical outcome in a given population, the pooled estimate of the effect of either aggregate data or individual patient data meta-analysis should be similar.<sup><xref ref-type="bibr" rid="B5">5</xref></sup></p>
        <p>However, discrepancies may emerge for several reasons. First, there may be differences in the original studies enrolled: high-intensity statins regimens versus low-intensity regimens were considered only in the individual patient data meta-analysis. As a result, patients enrolled in the aggregate data meta-analysis were at lower cardiovascular risk. This characteristic of the LDL-C aggregated data meta-analysis help explain smaller absolute effect sizes, which are dependent on baseline risk (assuming constant relative risk reduction across different baseline risk).</p>
        <p>Second, aggregate data meta-analysis has a lower power to detect relationships between patient-level characteristics and effect sizes. A meta-epidemiological study shows that estimates from meta-regression of aggregate data are less precise than those from individual patient data meta-analyses.<sup><xref ref-type="bibr" rid="B6">6</xref></sup> Random error on meta-regression estimates will often cause them to differ substantially from those obtained on individual patient data meta-analysis.<sup><xref ref-type="bibr" rid="B7">7</xref></sup> This fallacy is known as aggregation bias, which is defined by the assumption that an association between two group-level variables equals the association between the corresponding variable at the individual level.<sup><xref ref-type="bibr" rid="B7">7</xref></sup></p>
        <p>Third, not all characteristics that may significantly impact how an intervention works can be adequately assessed in an aggregate data meta-analysis, most commonly because effects according to the baseline characteristics of interest are not available in the reported papers.<sup><xref ref-type="bibr" rid="B7">7</xref></sup></p>
        <p>Fourth, in the case of the LDL-C meta-analyses, the individual patient data meta-analysis assessed the association between LDL-C variation and a composite outcome (coronary death or non-fatal myocardial infarction). In contrast, the aggregate data meta-analysis assessed the association between LDL-C and death, stroke, or myocardial infarction separately.<sup><xref ref-type="bibr" rid="B2">2</xref> , <xref ref-type="bibr" rid="B3">3</xref></sup></p>
        <p>As the number of events for a composite outcome is greater, the precision of effect estimates increases. When original studies did not report treatment effects for the combined outcome of interest, meta-analysis using aggregate data is not possible. Conversely, generating and analyzing combined outcomes is feasible by assessing original trial databases.</p>
        <p>Therefore, compared to aggregate data meta-analyses, individual patient data meta-analyses are superior in investigating relationships between effects on intermediate variables and clinical outcomes and examining potential treatment modifiers. Conversely, a meta-analysis of aggregate or individual patient data should offer similar answers for questions of treatment effect, as long as they use the same definitions for population, treatment, comparator, and outcomes and include the same original studies.</p>
      </sec>
    </body>
    <back>
      <fn-group>
        <fn fn-type="other">
          <p>Study association</p>
          <p>This study is not associated with any thesis or dissertation work.</p>
        </fn>
        <fn fn-type="other">
          <p>Ethics approval and consent to participate</p>
          <p>This article does not contain any studies with human participants or animals performed by any of the authors.</p>
        </fn>
      </fn-group>
      <fn-group>
        <fn fn-type="supported-by">
          <p>Sources of funding: There were no external funding sources for this study.</p>
        </fn>
      </fn-group>
    </back>
  </sub-article>
</article>
