Please enable JavaScript.

Coggle requires JavaScript to display documents.

L5 Systematic Review2 :princess::skin-tone-3: “STEPS to Systematic review”

- - - - Length ~ 1-1.5 pages
      - Show reason why review&study this, Explain the importance of problem (clinical q), อ้างอิงแหล่งข้อมูล
      - Description of condition — short description abt important problem (eg. Info abt ds, Dx, prognosis, economic importance<include prevalence or incidence>)
      - Description of intervention (should compare ≥ 2 things)
      - How the intervention might work (mechanism of action)
      - Why it’s important to do this review?
    - - Should be one sentence noting Main objective of study
        (ไม่ต้องมีหลายข้อ*)
      - Should note Interventions, Suffered fr health problems population & outcomes
      - Should cover other objective รองอื่นๆ
      - Clinical outcomes** should cover both Benefit parts & Harmful parts
      - Eg. Study effeciency & Adverse effects of drug A in treating pt with suicidal risk
    - - Criteria for considering studies for this review
        
        Types of studies — RCTs, Quasi-RCT, etc
        (Quasi main objective is no bias)
        
        Types of participants
        
        Ds or Conditions of interest should be defined using Explicit criteria
        
        Board population & Setting of interest should be defined
        
        Inclusion criteria
        
        Exclusion criteria
        
        Types of interventions
        
        Intervention vs Placebo
        Intervention A vs Int. B or Int. C
        
        Dose, Freq, Duration, Mode of administration
        
        Types of Outcome measures
        
        Primary outcome
        
        Should normally :!: reflect ≥ 1 potential benefit & ≥ 1 potential area of harm
        
        Should be as few as possible
        (as many will need to prove a lot -> inc. sample size)
        
        Secondary outcomes
        (non-primary outcomes)
        
        Eg. Mortality rate, Relapse rate, Success rate, Quality of life issues, Adverse events, Cost-effectiveness
      - Search methods for identification of studies
        
        Electronic searches
        
        Bibliographic databases searched
        
        Dates & Period searched
        
        Any constraints eg language
        
        Searching other resources
        
        Unpublished literature
        
        Grey literature sources such as internal report, conferences proceedings
        
        Etc
      - Data collection & analysis
        
        :pineapple:Selection of studies
        
        Studies (not reports) as Unit of interest
        
        Identifying multiple reports fr the same study
        
        A typical process of selecting studies
        
        Implementation of selection process
        eg ทำเรื่อง DM ก็ควรมีspecialist for DM
        
        Selection ‘excluded studies’ :!?:
        
        Title & Abstract evaluation form
        
        Title & Abstract not related with Intervention
        
        Not RCT or Quasi-RCT
        
        Not outcomes between Intervention & Comparison
        
        No relevant Health problem
        
        Children or Pregnancy
        
        Studied in Animal only
        (if ja try drugs in human, it should hv been test in Animal first)
        
        :pineapple:Data extraction & Mx
        
        Extract data fr published reports (use data collection form)
        
        Data are extracted independently by > 1 author should be stated + how disagreement are solved
        
        If relevant, methods for processing data in preparation for analysis should be described
        
        :pineapple:Quality assessment in included studies
        
        :star::star: Bias
        
        :pineapple:Measures of treatment effect
        
        Dichotomous data
        (Yes/No question)
        
        Risk ratio (RR)
        
        Odds ratio (OR)
        
        Risk difference (RD)
        
        Continuous data
        
        Difference in means (MD)
        
        Standardized difference in means (SMD)
        
        Time-to-event data
        
        :pineapple:Dealing with missing data
        
        Strategies for dealing with missing data should be described
        
        Include missing participants due to drop-out
        Whether an Intention-to-treat analysis will be conducted?
        
        :pineapple:Assessment of heterogeneity
        
        :pineapple:Assessment of reporting bias
        
        Should describe how publication bias & other reporting biases are addressed
        
        :pineapple:Data synthesis
        
        :pineapple:Subgroup analysis & Inx of heterogeneity
        
        :pineapple:Sensitivity analysis