Skip to contents

Using jsurvival in jamovi: Complete Workflow

ClinicoPath Development Team

2025-06-30

Source: vignettes/general-26-jamovi-workflow-legacy.Rmd
general-26-jamovi-workflow-legacy.Rmd

Overview

This vignette demonstrates how to use the jsurvival module within jamovi for comprehensive survival analysis. The jsurvival module is part of the ClinicoPath suite and provides a user-friendly interface for conducting survival analyses without programming.

Installation in jamovi

  1. Open jamovi
  2. Click on the + button in the top-right corner
  3. Select jamovi library
  4. Search for “ClinicoPath” or “jsurvival”
  5. Click Install

Option 2: Manual Installation

  1. Download the latest .jmo file from releases
  2. Open jamovi
  3. Click + → Sideload → Select the .jmo file

Data Preparation

Required Variables

For survival analysis, you need at minimum:

  • Time variable: Time from start of study to event or censoring
  • Event indicator: Binary variable indicating whether event occurred (1 = event, 0 = censored)

Optional Variables

  • Grouping factors: Categorical variables for group comparisons
  • Continuous predictors: Variables for cut-point analysis
  • Multiple explanatory variables: For multivariable analysis

Example Data Structure 

Patient_ID | Time_months | Death | Treatment | Age | Grade
1          | 24.5        | 1     | A         | 65  | High
2          | 36.2        | 0     | B         | 58  | Low
3          | 12.1        | 1     | A         | 72  | High

Analysis Workflows

1. Single Arm Survival Analysis

Use case: Overall survival characteristics of your entire study population

Steps: 1. Navigate to Survival → ClinicoPath Survival → Single Arm Survival 2. Assign variables: - Time Elapsed: Your time variable - Outcome: Your event indicator - Outcome Level: Select the level that indicates an event (usually “1”) 3. Configure options: - Cut Points: Time points for survival estimates (e.g., “12, 36, 60” for 1, 3, 5 years) - Time Type Output: Choose months or years - Confidence Intervals: Enable 95% CI - Risk Table: Show numbers at risk 4. Click Run

Output includes: - Overall survival curve - Median survival time with 95% CI - Survival rates at specified time points - Natural language summary

2. Survival Analysis (Group Comparisons)

Use case: Compare survival between different groups

Steps: 1. Navigate to Survival → ClinicoPath Survival → Survival Analysis 2. Assign variables: - Time Elapsed: Your time variable - Explanatory: Grouping variable (e.g., treatment, stage) - Outcome: Your event indicator - Outcome Level: Event level 3. Configure analysis: - Analysis Type: Choose overall, pairwise, or combination - P-value Adjustment: Method for multiple comparisons - Proportional Hazards: Enable Cox regression 4. Plotting options: - Risk Table: Show numbers at risk - Censored Points: Mark censored observations - Confidence Intervals: Display CI bands - End Plot: Set maximum time for plot

Output includes: - Kaplan-Meier curves by group - Log-rank test results - Cox regression hazard ratios - Pairwise comparisons (if selected) - Survival tables by group

3. Continuous Variable Survival Analysis

Use case: Find optimal cut-point for a continuous biomarker

Steps: 1. Navigate to Survival → ClinicoPath Survival → Survival Analysis for Continuous Variable 2. Assign variables: - Time Elapsed: Your time variable - Continuous Explanatory: Your continuous predictor - Outcome: Your event indicator 3. Cut-point options: - Find Cut-point: Let algorithm find optimal threshold - Manual Cut-point: Specify your own threshold 4. Configure analysis similar to group comparisons

Output includes: - Optimal cut-point determination - Survival curves for high/low groups - Hazard ratio for dichotomized variable - ROC analysis for cut-point validation

4. Multivariable Survival Analysis

Use case: Adjust for multiple risk factors simultaneously

Steps: 1. Navigate to Survival → ClinicoPath Survival → Multivariable Survival Analysis 2. Assign variables: - Time Elapsed: Your time variable - Explanatory: Multiple explanatory variables - Outcome: Your event indicator 3. Model options: - Model Type: Choose Cox proportional hazards - Variable Selection: Manual or automated selection - Interaction Terms: Include interactions if needed

Output includes: - Multivariable Cox regression table - Adjusted hazard ratios with 95% CI - Model fit statistics - Adjusted survival curves

5. Odds Ratio Analysis

Use case: Binary outcome analysis (case-control studies)

Steps: 1. Navigate to Survival → ClinicoPath Survival → Odds Ratio Table and Plot 2. Assign variables: - Outcome: Binary outcome variable - Explanatory: Risk factors 3. Configure: - Reference Level: Choose reference category - Confidence Level: Usually 95%

Output includes: - Odds ratio table - Forest plot - Chi-square test results

6. Time Interval Calculator

Use case: Calculate time differences from dates

Steps: 1. Navigate to Survival → Data Preparation → Time Interval Calculator 2. Assign variables: - Start Date: Beginning date - End Date: End date or follow-up date 3. Options: - Output Unit: Days, months, or years - Handle Missing: How to treat missing dates

Interpretation Guidelines

Kaplan-Meier Curves

  • Steep drops: High event rate at specific times
  • Flat portions: Low event rate (good prognosis periods)
  • Wide confidence intervals: High uncertainty due to small sample size
  • Crossing curves: Proportional hazards assumption may be violated

Statistical Tests

  • Log-rank test: Compares overall survival distributions
  • p < 0.05: Statistically significant difference between groups
  • Hazard Ratio > 1: Increased risk of event
  • Hazard Ratio < 1: Decreased risk of event (protective factor)

Clinical Interpretation

  • Median survival: Time when 50% of subjects have experienced the event
  • 5-year survival rate: Percentage alive at 5 years
  • Confidence intervals: Uncertainty range around estimates
  • Numbers at risk: Sample size remaining at each time point

Best Practices

Data Quality

  1. Check for data completeness: Missing time or event data
  2. Validate event coding: Ensure consistent coding (0/1 or No/Yes)
  3. Review follow-up times: Check for unrealistic or negative times
  4. Assess censoring pattern: High censoring rates may bias results

Analysis Considerations

  1. Sample size: Minimum 10 events per variable in Cox regression
  2. Proportional hazards: Check assumption using Schoenfeld residuals
  3. Multiple comparisons: Adjust p-values when testing multiple groups
  4. Clinical relevance: Statistical significance vs. clinical importance

Reporting

  1. CONSORT guidelines: Follow reporting standards for survival studies
  2. Number at risk tables: Always include in survival plots
  3. Confidence intervals: Report alongside point estimates
  4. Effect sizes: Focus on hazard ratios and median survival differences

Troubleshooting

Common Issues

  • “No events observed”: Check event coding and follow-up time
  • “Convergence failed”: May indicate separation or small sample size
  • “Proportional hazards violated”: Consider stratified Cox model
  • Missing survival curves: Check variable assignments and data types

Getting Help

Further Reading