Update app.py

app.py

@@ -45,36 +45,39 @@ def generate_systematic_review(pdf_files, review_question, include_tables=True):
         The first step in writing a systematic review paper is to identify a research field. This involves selecting a specific area of study that you are interested in and want to explore further.
         
         Step 2: Generate a Research Question
-        Once you have identified your research field, the next step is to generate a research question. This question should be specific, measurable, achievable, relevant, and time-bound (SMART).
+        Once you have identified your research field, the next step is to generate a research question. This question should be specific, measurable, achievable, relevant, and time-bound (SMART). Consider using the PICO framework (Population, Intervention, Comparison, Outcome) to structure clinical questions.
         
         Step 3: Create a Protocol
-        After generating your research question, the next step is to create a protocol. A detailed plan of how you will conduct your research, including the methods you will use, the data you will collect, and the analysis you will perform.
+        After generating your research question, create a detailed protocol. This is a comprehensive plan outlining your research methodology, including search strategies, databases to be used, and analysis techniques. The protocol should be registered in appropriate databases (e.g., PROSPERO) when applicable and follow PRISMA guidelines.
         
-        Step 4: Evaluate Relevant Literature
-        The fourth step is to evaluate relevant literature. This involves searching for and reviewing existing studies related to your research question. You should critically evaluate the quality of these studies and identify any gaps or limitations in the current literature.
+        Step 4: Define Inclusion and Exclusion Criteria
+        Clearly articulate the criteria for including or excluding studies in your review. These criteria should be directly tied to your research question and may include: publication date range, study types, population characteristics, intervention specifications, outcome measures, and language restrictions.
         
-        Step 5: Investigate Sources for Answers
-        The fifth step is to investigate sources for answers. This involves searching for and accessing relevant data and information that will help you answer your research question.
+        Step 5: Evaluate Relevant Literature
+        Conduct a comprehensive literature search using multiple databases (e.g., PubMed, Scopus, Web of Science, CINAHL) with clearly defined search terms and Boolean operators. Document your search strategy in detail to ensure reproducibility. Consider both published and unpublished (gray) literature to minimize publication bias.
         
-        Step 6: Collect Data as per Protocol
-        The sixth step is to collect data as per protocol. This involves implementing the methods outlined in your protocol and collecting the data specified. You should ensure that your data collection methods are rigorous and reliable.
+        Step 6: Quality Assessment of Studies
+        Apply established quality assessment tools appropriate for your included study designs (e.g., Cochrane Risk of Bias Tool for randomized trials, ROBINS-I for non-randomized studies, CASP checklists). Document quality assessments for all included studies.
         
         Step 7: Data Extraction
-        The seventh step is to extract the data. This involves organizing and analyzing the data you have collected, and extracting the relevant information that will help you answer your research question.
+        Create and use a standardized data extraction form to systematically collect relevant information from each study. This should include: study characteristics, participant demographics, intervention details, comparison groups, outcome measures, and results. Have multiple reviewers extract data independently when possible.
         
-        Step 8: Critical Analysis of Results
-        The eighth step is to conduct a critical analysis of your results. This involves interpreting your findings, identifying patterns and trends, and drawing conclusions based on your data.
+        Step 8: Data Synthesis and Analysis
+        Synthesize the extracted data using appropriate methods. If statistical pooling is appropriate, conduct meta-analysis with suitable models (fixed or random effects). If heterogeneity precludes meta-analysis, provide a narrative synthesis with clear explanations.
         
-        Step 9: Interpreting Derivations
-        The ninth step is to interpret the derivations. This involves taking the conclusions you have drawn from your data and interpreting them in the context of your research question.
+        Step 9: Critical Analysis of Results
+        Analyze your findings critically, examining patterns, inconsistencies, and relationships across studies. Address heterogeneity, publication bias (using funnel plots when applicable), and methodological limitations. Consider using the GRADE approach to evaluate certainty of evidence.
         
-        Step 10: Concluding Statements
-        The final step is to make concluding statements. This involves summarizing your findings and drawing conclusions based on your research. You should also provide recommendations for future research and implications for practice.
+        Step 10: Interpreting Findings
+        Interpret your findings in the context of the original research question. Discuss implications for practice, policy, and future research. Address limitations of your systematic review process and any potential biases.
         
-        Step-11:
-        Please include references in the form of citation and also link to the reference papers.
+        Step 11: Concluding Statements
+        Provide clear, substantiated conclusions based on your review findings. Ensure conclusions are proportionate to the evidence presented and acknowledge uncertainty where appropriate. Offer specific recommendations for future research.
         
-        Your response should be formatted in HTML (but avoid showing these tags ```html ```) but generate the content to look like a professional academic paper. Include proper section headers, abstracts, methodology sections, etc. Number all sections like an academic paper.
+        Step 12: References and Documentation
+        Include a comprehensive reference list following a specific citation style (APA, Vancouver, etc.). Provide links to source papers when available. Create appropriate visualizations including a PRISMA flow diagram documenting the selection process and forest plots for meta-analyses when applicable.
+        
+        Your response should be formatted in HTML (but avoid showing these tags ```html ```) but generate the content to look like a professional academic paper. Include proper section headers, abstracts, methodology sections, etc. Number all sections like an academic paper. Follow academic journal standards with double spacing, appropriate margins, and consistent formatting throughout.
         """
         
         # Extract text from each PDF