Guidance on literature review

Gowtham A. Rao

2024-01-06

Source: vignettes/GuidanceOnLiteratureReview.Rmd

PhenotypeLibrary is part of HADES

OHDSI Phenotype Development and Evaluation: A Guide to Systematic Literature Search for Published Phenotype Algorithms

This guide outlines a systematic approach to conducting literature searches for published phenotype algorithms for any phenotype of interest.

Step 1: Define Your Phenotype of Interest

  • Clearly define the phenotype you are interested in. Use the OHDSI Phenotype Library Clinical Description template.
  • Identify key characteristics and features associated with the phenotype.

Step 2: Keyword Development

  • Generate a list of keywords related to your phenotype of interest.
  • Include synonyms, related terms, and both broad and narrow terms.
  • Utilize Medical Subject Headings (MeSH) terms where possible for these terms.

Step 3: Develop a Search Strategy

  • Utilize Boolean operators (AND, OR, NOT) to combine keywords.
  • Group keywords and phrases in parentheses to ensure the search is carried out in the desired order.
  • Make use of wildcards and truncation to capture variations in spelling and terminology.

Step 4: Select Databases

  • Use multiple databases to ensure comprehensive coverage: We mostly use
    • PubMed/MEDLINE
    • EMBASE
    • Cochrane Library
    • Web of Science
  • Input your search strategy into the databases.
  • Keep a record of the number of results returned and any modifications made to the search strategy.

Step 6: Review the Results

  • Screen titles and abstracts to identify relevant articles.
  • Retrieve full texts of potentially relevant articles for further assessment.
  • Check the references of included articles to identify additional relevant papers.

Step 7: Extract Relevant Information

  • Extract information on the phenotype definition and algorithm used.
  • Note the validation methods, databases used, and performance metrics reported.

Step 8: Summarize and Analyze

  • Summarize the findings from the literature.
  • Consider using a spreadsheet to document your findings.
  • Compare different algorithms and their performance.
  • Identify gaps in the literature and areas for future research.
  • Consider using a spreadsheet to document your findings with one row per reviewed publication.

Step 9: Document Your Process

  • Keep a detailed record of the search strategy, databases used, and articles included/excluded.
  • This will help in replicating the search in the future or adjusting the strategy as necessary.

Step 10: Seek Expert Guidance

  • Consult with experts in phenotyping and OHDSI for guidance and feedback on your search strategy and findings.

Example of Parasomnia

Parasomnias are a category of sleep disorders that involve abnormal movements, behaviors, emotions, perceptions, and dreams that occur while falling asleep, sleeping, between sleep stages, or during arousal from sleep.

Keyword Development

  • Main Keywords: Parasomnias, Sleep Disorders
  • Related Terms and Synonyms: Sleep Arousal Disorders, REM Sleep Parasomnias, Sleep Bruxism, Nocturnal Myoclonus Syndrome, etc.
  • MeSH Terms: “Parasomnias”[Mesh], “Sleep Arousal Disorders”[Mesh], “REM Sleep Parasomnias”[Mesh], etc.

Develop a Search Strategy

Construct a Boolean query to capture articles related to parasomnia phenotype algorithms.

Example PubMed Search Query: (Thank you Gayle Murray for your help is this search criteria creation.)

(((((("Parasomnias"[Mesh:NoExp]) OR "Nocturnal Myoclonus Syndrome"[Mesh]) OR "Nocturnal Paroxysmal Dystonia"[Mesh]) OR "REM Sleep Parasomnias"[Mesh]) OR "Sleep Arousal Disorders"[Mesh]) OR "Sleep Bruxism"[Mesh]) OR "Sleep-Wake Transition Disorders"[Mesh]]

AND 

(((((((((((((((retrospective cohort) OR (epidemiology [MeSH Terms])) OR (Epidemiologic Methods[MeSH Terms])) OR (phenotype[Text Word])) OR (Validation Study[Publication Type])) OR (positive predictive value[Text Word])) OR (Validation Studies as Topic[MeSH Terms])) OR (Sensitivity and Specificity[MeSH Terms])) OR algorithm))) ) OR (Diseases Category/epidemiology[MeSH Terms])) OR (Predictive Value of Tests[MeSH Terms])) OR (Reproducibility of Results[MeSH Terms]) OR (Predictive Value[Text Word]))

AND

 (((((((((((((((((((((Outcome Assessment) OR (Data Warehouse)) OR (Database Management Systems[MeSH Terms])) OR (Medical Records Systems, Computerized[MeSH Terms])) OR (Current procedural terminology)) OR ("drug surveillance")) OR (Databases, Factual[MeSH Terms])) OR (Databases as topic[MeSH Terms])) OR (Medical Record Linkage[MeSH Terms])) OR ("registered persons database")) OR ("Medical Records Systems")) OR ( "Population Surveillance")) OR ("Data Collection")) OR ("Automatic Data Processing")) OR (Incidence[MeSH Terms])) OR ("Medical Records"[MeSH Terms])) OR ("Patient Discharge"[MeSH Terms])) OR ("Hospital Records"[MeSH Terms])) OR (((((((((Medicaid) OR (Medicare)) OR (Truven)) OR (Optum)) OR (Medstat)) OR (Nationwide Inpatient Sample)) OR (National Inpatient Sample)) OR (PharMetrics)) OR ("National Health Insurance Research"))) OR ("claims administrative"[Title/Abstract:~1])) OR ("administrative data"[Title/Abstract:~1]))
NOT
(((((((("Clinical Trial"[pt] OR "Editorial"[pt] OR "Letter"[pt] OR "Randomized Controlled Trial"[pt] OR "Clinical Trial, Phase I"[pt] OR "Clinical Trial, Phase II"[pt] OR "Clinical Trial, Phase III"[pt] OR "Clinical Trial, Phase IV"[pt] OR "Comment"[pt] OR "Controlled Clinical Trial"[pt] OR "Case Reports"[pt] OR "Clinical Trials as Topic"[Mesh] OR "double-blind"[All] OR "placebo-controlled"[All] OR "pilot study"[All] OR "pilot projects"[Mesh] OR "Prospective Studies"[Mesh])) OR ("Genetics"[Mesh]) OR ("Genotype"[Mesh]) ) OR (biomarker[Title/Abstract]))))) OR ("review"[Publication Type])) OR (Meta-Analysis[Publication Type])