quarto_meta.dash.qmd

---
title: "Interactive Meta-Analysis Dashboard"
format: html
editor: visual
---

```{r setup, include=FALSE}
library(tidyverse)  # Data manipulation
library(meta)       # Meta-analysis
library(plotly)     # Interactive plots
```

## **Introduction**

This dashboard allows users to explore a living meta-analysis interactively. Users can visualize meta-analysis plots, filter studies, and explore how study inclusion affects overall effect sizes and heterogeneity.

---

## **Data Import and Preparation**

```{r}
# Simulated example of meta-analysis data
meta_data <- tibble(
  study_id = paste("Study", 1:10),
  effect_size = rnorm(10, mean = 0.5, sd = 0.2),
  lower_ci = effect_size - runif(10, 0.1, 0.3),
  upper_ci = effect_size + runif(10, 0.1, 0.3),
  sample_size = sample(50:200, 10),
  analysis_type = sample(c("Change-from-baseline", "End-point", "Self-report", "Clinician-administered"), 10, replace = TRUE),
  include = TRUE  # Column to toggle study inclusion in analysis
)

# Display the data table for context
meta_data
```

---

## **Analysis Type Selection**

```{r}
# Create a dropdown menu to select analysis type
selected_analysis <- "All Studies"  # Default selection
analysis_options <- c("All Studies", "Change-from-baseline", "End-point", "Self-report", "Clinician-administered")

cat("**Select Analysis Type:**\n")
cat(paste0("- [", analysis_options, "](javascript:void(0))\n"), sep="")
```

---

## **Interactive Filtering of Studies**

```{r}
# Filter studies based on selected analysis type
filtered_data <- if (selected_analysis == "All Studies") {
  meta_data
} else {
  meta_data %>% filter(analysis_type == selected_analysis)
}

# Interactive table with checkboxes to select/deselect studies
plotly_table <- plotly::plot_ly(
  type = 'table',
  header = list(values = names(filtered_data)),
  cells = list(values = t(filtered_data))
)

plotly_table
```

> **Instructions:** Use the table above to review the list of studies included in the meta-analysis. While you can't toggle studies on this static dashboard, the table allows you to view each study's information.

---

## **Meta-Analysis Calculation**

```{r}
# Filter only included studies
filtered_data <- filtered_data %>% filter(include == TRUE)

# Perform a random-effects meta-analysis
meta_result <- meta::metagen(
  TE = effect_size, 
  seTE = (upper_ci - lower_ci) / 3.92,  # Approximation of SE from CI
  studlab = study_id, 
  data = filtered_data, 
  sm = "SMD" # Standardized Mean Difference
)

# Display the meta-analysis results
summary(meta_result)
```

---

## **Interactive Forest Plot**

```{r}
# Create the forest plot with meta-analysis results
meta::forest(
  meta_result, 
  studlab = TRUE, 
  xlab = "Effect Size (SMD)", 
  col.square = "darkblue"
)
```

> **Instructions:** Use the forest plot above to visualize the effect sizes of the studies. You can hover over points to see the study name, effect size, and confidence intervals.

---

## **Heterogeneity Assessment**

```{r}
# Extract heterogeneity measures
heterogeneity <- tibble(
  I2 = meta_result$I2, 
  Q = meta_result$Q, 
  pval_Q = meta_result$pval.Q
)

heterogeneity
```

> **Interpretation:** The I2 value represents the percentage of total variation across studies due to heterogeneity rather than chance. Use this measure to assess the consistency of study results.

---

## **Conclusion**
This dashboard provides an interactive view of meta-analysis results. While static on GitHub Pages, a more dynamic version could allow users to toggle inclusion of specific studies or adjust filters for sensitivity analyses. For an interactive version, consider hosting on shinyapps.io, Streamlit Cloud, or similar platforms.

---

## **Future Enhancements**
- Add a sidebar filter to toggle study inclusion/exclusion.
- Deploy an interactive version via Shiny or Streamlit for real-time updates.
- Allow users to upload their own meta-analysis data for analysis.