Causal inference methods for small non-randomized studies: Methods and an application in COVID-19.
Identifieur interne : 000839 ( Main/Corpus ); précédent : 000838; suivant : 000840Causal inference methods for small non-randomized studies: Methods and an application in COVID-19.
Auteurs : Sarah Friedrich ; Tim FriedeSource :
- Contemporary clinical trials [ 1559-2030 ] ; 2020.
English descriptors
- KwdEn :
- Antiviral Agents (administration & dosage), Antiviral Agents (adverse effects), Antiviral Agents (therapeutic use), COVID-19 (drug therapy), COVID-19 (epidemiology), Causality (MeSH), Clinical Trials as Topic (organization & administration), Clinical Trials as Topic (standards), Humans (MeSH), Hydroxychloroquine (administration & dosage), Hydroxychloroquine (adverse effects), Hydroxychloroquine (therapeutic use), Pandemics (MeSH), Propensity Score (MeSH), SARS-CoV-2 (MeSH), Sample Size (MeSH).
- MESH :
- chemical , administration & dosage : Antiviral Agents, Hydroxychloroquine.
- chemical , adverse effects : Antiviral Agents, Hydroxychloroquine.
- chemical , therapeutic use : Antiviral Agents, Hydroxychloroquine.
- drug therapy : COVID-19.
- epidemiology : COVID-19.
- organization & administration : Clinical Trials as Topic.
- standards : Clinical Trials as Topic.
- Causality, Humans, Pandemics, Propensity Score, SARS-CoV-2, Sample Size.
Abstract
The usual development cycles are too slow for the development of vaccines, diagnostics and treatments in pandemics such as the ongoing SARS-CoV-2 pandemic. Given the pressure in such a situation, there is a risk that findings of early clinical trials are overinterpreted despite their limitations in terms of size and design. Motivated by a non-randomized open-label study investigating the efficacy of hydroxychloroquine in patients with COVID-19, we describe in a unified fashion various alternative approaches to the analysis of non-randomized studies. A widely used tool to reduce the impact of treatment-selection bias are so-called propensity score (PS) methods. Conditioning on the propensity score allows one to replicate the design of a randomized controlled trial, conditional on observed covariates. Extensions include the g-computation approach, which is less frequently applied, in particular in clinical studies. Moreover, doubly robust estimators provide additional advantages. Here, we investigate the properties of propensity score based methods including three variations of doubly robust estimators in small sample settings, typical for early trials, in a simulation study. R code for the simulations is provided.
DOI: 10.1016/j.cct.2020.106213
PubMed: 33188930
PubMed Central: PMC7834813
Links to Exploration step
pubmed:33188930Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en">Causal inference methods for small non-randomized studies: Methods and an application in COVID-19.</title><author><name sortKey="Friedrich, Sarah" sort="Friedrich, Sarah" uniqKey="Friedrich S" first="Sarah" last="Friedrich">Sarah Friedrich</name><affiliation><nlm:affiliation>Department of Medical Statistics, University Medical Center Göttingen, Humboldtallee 32, 37073 Göttingen, Germany. Electronic address: sarah.friedrich@med.uni-goettingen.de.</nlm:affiliation></affiliation></author><author><name sortKey="Friede, Tim" sort="Friede, Tim" uniqKey="Friede T" first="Tim" last="Friede">Tim Friede</name><affiliation><nlm:affiliation>Department of Medical Statistics, University Medical Center Göttingen, Humboldtallee 32, 37073 Göttingen, Germany. Electronic address: tim.friede@med.uni-goettingen.de.</nlm:affiliation></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">PubMed</idno><date when="2020">2020</date><idno type="RBID">pubmed:33188930</idno><idno type="pmid">33188930</idno><idno type="doi">10.1016/j.cct.2020.106213</idno><idno type="pmc">PMC7834813</idno><idno type="wicri:Area/Main/Corpus">000839</idno><idno type="wicri:explorRef" wicri:stream="Main" wicri:step="Corpus" wicri:corpus="PubMed">000839</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en">Causal inference methods for small non-randomized studies: Methods and an application in COVID-19.</title><author><name sortKey="Friedrich, Sarah" sort="Friedrich, Sarah" uniqKey="Friedrich S" first="Sarah" last="Friedrich">Sarah Friedrich</name><affiliation><nlm:affiliation>Department of Medical Statistics, University Medical Center Göttingen, Humboldtallee 32, 37073 Göttingen, Germany. Electronic address: sarah.friedrich@med.uni-goettingen.de.</nlm:affiliation></affiliation></author><author><name sortKey="Friede, Tim" sort="Friede, Tim" uniqKey="Friede T" first="Tim" last="Friede">Tim Friede</name><affiliation><nlm:affiliation>Department of Medical Statistics, University Medical Center Göttingen, Humboldtallee 32, 37073 Göttingen, Germany. Electronic address: tim.friede@med.uni-goettingen.de.</nlm:affiliation></affiliation></author></analytic><series><title level="j">Contemporary clinical trials</title><idno type="eISSN">1559-2030</idno><imprint><date when="2020" type="published">2020</date></imprint></series></biblStruct></sourceDesc></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Antiviral Agents (administration & dosage)</term><term>Antiviral Agents (adverse effects)</term><term>Antiviral Agents (therapeutic use)</term><term>COVID-19 (drug therapy)</term><term>COVID-19 (epidemiology)</term><term>Causality (MeSH)</term><term>Clinical Trials as Topic (organization & administration)</term><term>Clinical Trials as Topic (standards)</term><term>Humans (MeSH)</term><term>Hydroxychloroquine (administration & dosage)</term><term>Hydroxychloroquine (adverse effects)</term><term>Hydroxychloroquine (therapeutic use)</term><term>Pandemics (MeSH)</term><term>Propensity Score (MeSH)</term><term>SARS-CoV-2 (MeSH)</term><term>Sample Size (MeSH)</term></keywords><keywords scheme="MESH" type="chemical" qualifier="administration & dosage" xml:lang="en"><term>Antiviral Agents</term><term>Hydroxychloroquine</term></keywords><keywords scheme="MESH" type="chemical" qualifier="adverse effects" xml:lang="en"><term>Antiviral Agents</term><term>Hydroxychloroquine</term></keywords><keywords scheme="MESH" type="chemical" qualifier="therapeutic use" xml:lang="en"><term>Antiviral Agents</term><term>Hydroxychloroquine</term></keywords><keywords scheme="MESH" qualifier="drug therapy" xml:lang="en"><term>COVID-19</term></keywords><keywords scheme="MESH" qualifier="epidemiology" xml:lang="en"><term>COVID-19</term></keywords><keywords scheme="MESH" qualifier="organization & administration" xml:lang="en"><term>Clinical Trials as Topic</term></keywords><keywords scheme="MESH" qualifier="standards" xml:lang="en"><term>Clinical Trials as Topic</term></keywords><keywords scheme="MESH" xml:lang="en"><term>Causality</term><term>Humans</term><term>Pandemics</term><term>Propensity Score</term><term>SARS-CoV-2</term><term>Sample Size</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">The usual development cycles are too slow for the development of vaccines, diagnostics and treatments in pandemics such as the ongoing SARS-CoV-2 pandemic. Given the pressure in such a situation, there is a risk that findings of early clinical trials are overinterpreted despite their limitations in terms of size and design. Motivated by a non-randomized open-label study investigating the efficacy of hydroxychloroquine in patients with COVID-19, we describe in a unified fashion various alternative approaches to the analysis of non-randomized studies. A widely used tool to reduce the impact of treatment-selection bias are so-called propensity score (PS) methods. Conditioning on the propensity score allows one to replicate the design of a randomized controlled trial, conditional on observed covariates. Extensions include the g-computation approach, which is less frequently applied, in particular in clinical studies. Moreover, doubly robust estimators provide additional advantages. Here, we investigate the properties of propensity score based methods including three variations of doubly robust estimators in small sample settings, typical for early trials, in a simulation study. R code for the simulations is provided.</div></front></TEI><pubmed><MedlineCitation Status="MEDLINE" Owner="NLM"><PMID Version="1">33188930</PMID><DateCompleted><Year>2021</Year><Month>01</Month><Day>08</Day></DateCompleted><DateRevised><Year>2021</Year><Month>01</Month><Day>29</Day></DateRevised><Article PubModel="Print-Electronic"><Journal><ISSN IssnType="Electronic">1559-2030</ISSN><JournalIssue CitedMedium="Internet"><Volume>99</Volume><PubDate><Year>2020</Year><Month>12</Month></PubDate></JournalIssue><Title>Contemporary clinical trials</Title><ISOAbbreviation>Contemp Clin Trials</ISOAbbreviation></Journal><ArticleTitle>Causal inference methods for small non-randomized studies: Methods and an application in COVID-19.</ArticleTitle><Pagination><MedlinePgn>106213</MedlinePgn></Pagination><ELocationID EIdType="pii" ValidYN="Y">S1551-7144(20)30291-3</ELocationID><ELocationID EIdType="doi" ValidYN="Y">10.1016/j.cct.2020.106213</ELocationID><Abstract><AbstractText>The usual development cycles are too slow for the development of vaccines, diagnostics and treatments in pandemics such as the ongoing SARS-CoV-2 pandemic. Given the pressure in such a situation, there is a risk that findings of early clinical trials are overinterpreted despite their limitations in terms of size and design. Motivated by a non-randomized open-label study investigating the efficacy of hydroxychloroquine in patients with COVID-19, we describe in a unified fashion various alternative approaches to the analysis of non-randomized studies. A widely used tool to reduce the impact of treatment-selection bias are so-called propensity score (PS) methods. Conditioning on the propensity score allows one to replicate the design of a randomized controlled trial, conditional on observed covariates. Extensions include the g-computation approach, which is less frequently applied, in particular in clinical studies. Moreover, doubly robust estimators provide additional advantages. Here, we investigate the properties of propensity score based methods including three variations of doubly robust estimators in small sample settings, typical for early trials, in a simulation study. R code for the simulations is provided.</AbstractText><CopyrightInformation>Copyright © 2020 Elsevier Inc. All rights reserved.</CopyrightInformation></Abstract><AuthorList CompleteYN="Y"><Author ValidYN="Y"><LastName>Friedrich</LastName><ForeName>Sarah</ForeName><Initials>S</Initials><AffiliationInfo><Affiliation>Department of Medical Statistics, University Medical Center Göttingen, Humboldtallee 32, 37073 Göttingen, Germany. Electronic address: sarah.friedrich@med.uni-goettingen.de.</Affiliation></AffiliationInfo></Author><Author ValidYN="Y"><LastName>Friede</LastName><ForeName>Tim</ForeName><Initials>T</Initials><AffiliationInfo><Affiliation>Department of Medical Statistics, University Medical Center Göttingen, Humboldtallee 32, 37073 Göttingen, Germany. Electronic address: tim.friede@med.uni-goettingen.de.</Affiliation></AffiliationInfo></Author></AuthorList><Language>eng</Language><PublicationTypeList><PublicationType UI="D016428">Journal Article</PublicationType><PublicationType UI="D013485">Research Support, Non-U.S. Gov't</PublicationType></PublicationTypeList><ArticleDate DateType="Electronic"><Year>2020</Year><Month>11</Month><Day>11</Day></ArticleDate></Article><MedlineJournalInfo><Country>United States</Country><MedlineTA>Contemp Clin Trials</MedlineTA><NlmUniqueID>101242342</NlmUniqueID><ISSNLinking>1551-7144</ISSNLinking></MedlineJournalInfo><ChemicalList><Chemical><RegistryNumber>0</RegistryNumber><NameOfSubstance UI="D000998">Antiviral Agents</NameOfSubstance></Chemical><Chemical><RegistryNumber>4QWG6N8QKH</RegistryNumber><NameOfSubstance UI="D006886">Hydroxychloroquine</NameOfSubstance></Chemical></ChemicalList><CitationSubset>IM</CitationSubset><MeshHeadingList><MeshHeading><DescriptorName UI="D000998" MajorTopicYN="N">Antiviral Agents</DescriptorName><QualifierName UI="Q000008" MajorTopicYN="N">administration & dosage</QualifierName><QualifierName UI="Q000009" MajorTopicYN="N">adverse effects</QualifierName><QualifierName UI="Q000627" MajorTopicYN="Y">therapeutic use</QualifierName></MeshHeading><MeshHeading><DescriptorName UI="D000086382" MajorTopicYN="N">COVID-19</DescriptorName><QualifierName UI="Q000188" MajorTopicYN="Y">drug therapy</QualifierName><QualifierName UI="Q000453" MajorTopicYN="Y">epidemiology</QualifierName></MeshHeading><MeshHeading><DescriptorName UI="D015984" MajorTopicYN="N">Causality</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D002986" MajorTopicYN="N">Clinical Trials as Topic</DescriptorName><QualifierName UI="Q000458" MajorTopicYN="Y">organization & administration</QualifierName><QualifierName UI="Q000592" MajorTopicYN="N">standards</QualifierName></MeshHeading><MeshHeading><DescriptorName UI="D006801" MajorTopicYN="N">Humans</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D006886" MajorTopicYN="N">Hydroxychloroquine</DescriptorName><QualifierName UI="Q000008" MajorTopicYN="N">administration & dosage</QualifierName><QualifierName UI="Q000009" MajorTopicYN="N">adverse effects</QualifierName><QualifierName UI="Q000627" MajorTopicYN="Y">therapeutic use</QualifierName></MeshHeading><MeshHeading><DescriptorName UI="D058873" MajorTopicYN="N">Pandemics</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D057216" MajorTopicYN="N">Propensity Score</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D000086402" MajorTopicYN="N">SARS-CoV-2</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D018401" MajorTopicYN="N">Sample Size</DescriptorName></MeshHeading></MeshHeadingList><KeywordList Owner="NOTNLM"><Keyword MajorTopicYN="Y">COVID-19</Keyword><Keyword MajorTopicYN="Y">Causal inference</Keyword><Keyword MajorTopicYN="Y">Propensity score</Keyword><Keyword MajorTopicYN="Y">Small samples</Keyword></KeywordList></MedlineCitation><PubmedData><History><PubMedPubDate PubStatus="received"><Year>2020</Year><Month>07</Month><Day>31</Day></PubMedPubDate><PubMedPubDate PubStatus="revised"><Year>2020</Year><Month>10</Month><Day>09</Day></PubMedPubDate><PubMedPubDate PubStatus="accepted"><Year>2020</Year><Month>11</Month><Day>06</Day></PubMedPubDate><PubMedPubDate PubStatus="pubmed"><Year>2020</Year><Month>11</Month><Day>15</Day><Hour>6</Hour><Minute>0</Minute></PubMedPubDate><PubMedPubDate PubStatus="medline"><Year>2021</Year><Month>1</Month><Day>9</Day><Hour>6</Hour><Minute>0</Minute></PubMedPubDate><PubMedPubDate PubStatus="entrez"><Year>2020</Year><Month>11</Month><Day>14</Day><Hour>20</Hour><Minute>8</Minute></PubMedPubDate></History><PublicationStatus>ppublish</PublicationStatus><ArticleIdList><ArticleId IdType="pubmed">33188930</ArticleId><ArticleId IdType="pii">S1551-7144(20)30291-3</ArticleId><ArticleId IdType="doi">10.1016/j.cct.2020.106213</ArticleId><ArticleId IdType="pmc">PMC7834813</ArticleId></ArticleIdList><ReferenceList><Reference><Citation>Expert Rev Clin Immunol. 2020 Jul;16(7):659-666</Citation><ArticleIdList><ArticleId IdType="pubmed">32620062</ArticleId></ArticleIdList></Reference><Reference><Citation>Am J Epidemiol. 2006 Jun 15;163(12):1149-56</Citation><ArticleIdList><ArticleId IdType="pubmed">16624967</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2020 Sep 23;:</Citation><ArticleIdList><ArticleId IdType="pubmed">32964526</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Methods Med Res. 2019 Aug;28(8):2455-2474</Citation><ArticleIdList><ArticleId IdType="pubmed">29966490</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2007 Jul 20;26(16):3078-94</Citation><ArticleIdList><ArticleId IdType="pubmed">17187347</ArticleId></ArticleIdList></Reference><Reference><Citation>Lancet. 2020 Jul 18;396(10245):e2-e3</Citation><ArticleIdList><ArticleId IdType="pubmed">32653079</ArticleId></ArticleIdList></Reference><Reference><Citation>Pharmacoepidemiol Drug Saf. 2005 Apr;14(4):227-38</Citation><ArticleIdList><ArticleId IdType="pubmed">15386700</ArticleId></ArticleIdList></Reference><Reference><Citation>Int J Biostat. 2012 Sep 18;8(1):</Citation><ArticleIdList><ArticleId IdType="pubmed">22992289</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Methods Med Res. 2019 Aug;28(8):2439-2454</Citation><ArticleIdList><ArticleId IdType="pubmed">29921162</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2007 Feb 20;26(4):754-68</Citation><ArticleIdList><ArticleId IdType="pubmed">16783757</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2020 Jul 10;39(15):2017-2034</Citation><ArticleIdList><ArticleId IdType="pubmed">32185801</ArticleId></ArticleIdList></Reference><Reference><Citation>Acad Emerg Med. 2002 Dec;9(12):1430-4</Citation><ArticleIdList><ArticleId IdType="pubmed">12460851</ArticleId></ArticleIdList></Reference><Reference><Citation>Biometrics. 2005 Dec;61(4):962-73</Citation><ArticleIdList><ArticleId IdType="pubmed">16401269</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2014 Mar 15;33(6):1057-69</Citation><ArticleIdList><ArticleId IdType="pubmed">24123228</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Sci. 2007;22(4):569-573</Citation><ArticleIdList><ArticleId IdType="pubmed">18516239</ArticleId></ArticleIdList></Reference><Reference><Citation>N Engl J Med. 2020 Aug 6;383(6):517-525</Citation><ArticleIdList><ArticleId IdType="pubmed">32492293</ArticleId></ArticleIdList></Reference><Reference><Citation>Am J Med. 2020 Sep 18;:</Citation><ArticleIdList><ArticleId IdType="pubmed">32950502</ArticleId></ArticleIdList></Reference><Reference><Citation>BMC Med Res Methodol. 2016 Nov 24;16(1):163</Citation><ArticleIdList><ArticleId IdType="pubmed">27881078</ArticleId></ArticleIdList></Reference><Reference><Citation>JAMA. 2018 Nov 20;320(19):2041</Citation><ArticleIdList><ArticleId IdType="pubmed">30458484</ArticleId></ArticleIdList></Reference><Reference><Citation>BMJ. 2000 May 27;320(7247):1468</Citation><ArticleIdList><ArticleId IdType="pubmed">10827061</ArticleId></ArticleIdList></Reference><Reference><Citation>Biometrics. 2018 Sep;74(3):977-985</Citation><ArticleIdList><ArticleId IdType="pubmed">29451947</ArticleId></ArticleIdList></Reference><Reference><Citation>Eur Heart J. 2020 Jun 7;41(22):2109-2117</Citation><ArticleIdList><ArticleId IdType="pubmed">32498081</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2013 Aug 30;32(19):3388-414</Citation><ArticleIdList><ArticleId IdType="pubmed">23508673</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2017 Jun 30;36(14):2302-2317</Citation><ArticleIdList><ArticleId IdType="pubmed">28295456</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2000 Feb 29;19(4):453-73</Citation><ArticleIdList><ArticleId IdType="pubmed">10694730</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 1991 Dec;10(12):1897-913</Citation><ArticleIdList><ArticleId IdType="pubmed">1805317</ArticleId></ArticleIdList></Reference><Reference><Citation>Contemp Clin Trials. 2020 Oct;97:106145</Citation><ArticleIdList><ArticleId IdType="pubmed">32927092</ArticleId></ArticleIdList></Reference><Reference><Citation>Am J Epidemiol. 2011 Apr 1;173(7):731-8</Citation><ArticleIdList><ArticleId IdType="pubmed">21415029</ArticleId></ArticleIdList></Reference><Reference><Citation>Multivariate Behav Res. 2011 May;46(3):399-424</Citation><ArticleIdList><ArticleId IdType="pubmed">21818162</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Methods Med Res. 2020 Mar;29(3):644-658</Citation><ArticleIdList><ArticleId IdType="pubmed">32186264</ArticleId></ArticleIdList></Reference><Reference><Citation>N Engl J Med. 2020 Jun 11;382(24):2282-2284</Citation><ArticleIdList><ArticleId IdType="pubmed">32289216</ArticleId></ArticleIdList></Reference><Reference><Citation>Am J Epidemiol. 2007 Dec 1;166(11):1337-44</Citation><ArticleIdList><ArticleId IdType="pubmed">18000021</ArticleId></ArticleIdList></Reference><Reference><Citation>J Crit Care. 2020 Jun;57:279-283</Citation><ArticleIdList><ArticleId IdType="pubmed">32173110</ArticleId></ArticleIdList></Reference><Reference><Citation>JAMA. 2020 Jun 9;323(22):2262-2263</Citation><ArticleIdList><ArticleId IdType="pubmed">32364561</ArticleId></ArticleIdList></Reference><Reference><Citation>Biometrics. 1996 Mar;52(1):249-64</Citation><ArticleIdList><ArticleId IdType="pubmed">8934595</ArticleId></ArticleIdList></Reference><Reference><Citation>Lifetime Data Anal. 2015 Oct;21(4):579-93</Citation><ArticleIdList><ArticleId IdType="pubmed">26100005</ArticleId></ArticleIdList></Reference><Reference><Citation>Arch Pediatr Adolesc Med. 2009 May;163(5):438-45</Citation><ArticleIdList><ArticleId IdType="pubmed">19414690</ArticleId></ArticleIdList></Reference><Reference><Citation>Lifetime Data Anal. 2013 Jul;19(3):279-96</Citation><ArticleIdList><ArticleId IdType="pubmed">23329123</ArticleId></ArticleIdList></Reference><Reference><Citation>BMC Med Res Methodol. 2012 May 30;12:70</Citation><ArticleIdList><ArticleId IdType="pubmed">22646911</ArticleId></ArticleIdList></Reference><Reference><Citation>N Engl J Med. 2020 Nov 19;383(21):2041-2052</Citation><ArticleIdList><ArticleId IdType="pubmed">32706953</ArticleId></ArticleIdList></Reference><Reference><Citation>Pharm Stat. 2021 Jan;20(1):15-24</Citation><ArticleIdList><ArticleId IdType="pubmed">32776719</ArticleId></ArticleIdList></Reference><Reference><Citation>Int J Biostat. 2010 May 17;6(1):Article 17</Citation><ArticleIdList><ArticleId IdType="pubmed">20628637</ArticleId></ArticleIdList></Reference><Reference><Citation>J Clin Epidemiol. 2020 Jul;123:120-126</Citation><ArticleIdList><ArticleId IdType="pubmed">32330521</ArticleId></ArticleIdList></Reference><Reference><Citation>Epidemiology. 2009 Nov;20(6):930-1</Citation><ArticleIdList><ArticleId IdType="pubmed">19829189</ArticleId></ArticleIdList></Reference><Reference><Citation>Epidemiology. 2016 May;27(3):356-9</Citation><ArticleIdList><ArticleId IdType="pubmed">26680297</ArticleId></ArticleIdList></Reference><Reference><Citation>Proc Mach Learn Res. 2019 Apr;89:2445-2453</Citation><ArticleIdList><ArticleId IdType="pubmed">31198908</ArticleId></ArticleIdList></Reference><Reference><Citation>Contemp Clin Trials. 2020 Nov;98:106154</Citation><ArticleIdList><ArticleId IdType="pubmed">32961361</ArticleId></ArticleIdList></Reference><Reference><Citation>Int J Antimicrob Agents. 2020 Jul;56(1):105949</Citation><ArticleIdList><ArticleId IdType="pubmed">32205204</ArticleId></ArticleIdList></Reference><Reference><Citation>Stat Med. 2010 Sep 10;29(20):2137-48</Citation><ArticleIdList><ArticleId IdType="pubmed">20108233</ArticleId></ArticleIdList></Reference></ReferenceList></PubmedData></pubmed></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Sante/explor/CovidChloroV1/Data/Main/Corpus
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000839 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Corpus/biblio.hfd -nk 000839 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Sante
|area= CovidChloroV1
|flux= Main
|étape= Corpus
|type= RBID
|clé= pubmed:33188930
|texte= Causal inference methods for small non-randomized studies: Methods and an application in COVID-19.
}}
Pour générer des pages wiki
HfdIndexSelect -h $EXPLOR_AREA/Data/Main/Corpus/RBID.i -Sk "pubmed:33188930" \
| HfdSelect -Kh $EXPLOR_AREA/Data/Main/Corpus/biblio.hfd \
| NlmPubMed2Wicri -a CovidChloroV1
| This area was generated with Dilib version V0.6.38. |  |