Run your first analysis
This guide walks through the full EasyHTA workflow, from creating a project and picking an analysis template through compute and reviewing the results in the app. Estimated time: 10 minutes.
Prerequisites
- An EasyHTA account: sign in at
app.easyhta.com(the 14-day trial gives full access) - For PSM / PSM-cure templates: digitized KM coordinates and an at-risk table from the published trial, typically extracted with a tool such as WebPlotDigitizer. See Preparing KM coordinates for how to get these right.
- For the BIA template: patient-population estimates, drug and administration costs, and (if comparing against current treatments) market-share assumptions. No curves required.
Create a project and choose a template
Sign in at the app (app.easyhta.com). Then open the Projects page, click + New Project, give it a name, and pick a template:
- Partitioned survival (PSM): three-state PSM (progression-free, progressed, dead) with state occupancy derived from PFS/OS curves.
- Partitioned survival with cure model (PSM-cure): three-state PSM with cure-model extrapolation for therapies with long-term responders.
- Budget Impact Analysis (BIA): estimate the budget impact of introducing a new treatment versus continuing current standard of care.
Survival templates (PSM / PSM-cure)
Upload your digitized curve
From the project, drop your CSV onto the upload zone, or click to browse.
The CSV needs just two columns: time and survival (survival as a probability between 0 and 1). Headers are auto-detected, so column order and extra columns don’t matter. The at-risk table is entered separately in the next step; it does not belong in the CSV. For how to digitize and format this file well, see Preparing KM coordinates.
Enter the at-risk table and reconstruct IPD
Enter the at-risk numbers and the time step reported in the publication, then run reconstruction. EasyHTA uses the Guyot algorithm (via the IPDfromKM R package) to reconstruct individual patient-level data from the digitized KM summary.
Once reconstruction completes the project view shows the reconstructed IPD alongside an overlay comparing the digitized and reconstructed KM curves, so you can visually check the fidelity of the reconstruction.
Fit parametric models
Move on to the model step. Tick the distributions you want to fit. On a first pass, run all six standard distributions so you can compare them, then click Fit. EasyHTA fits each model by maximum likelihood using the flexsurv R package and returns AIC, BIC, log-likelihood, fitted parameters, the variance–covariance matrix, and RMST for each one.
Review the fits
Once the fit completes, the project view shows each fitted curve overlaid on the reconstructed KM, alongside a comparison table of median survival, RMST, AIC, BIC, log-likelihood, and parameter count per distribution. Use the overlay to sanity-check the extrapolated tail behaviour before settling on a distribution; see Parametric distributions for guidance on what to look for.
Export to Excel (optional)
When you need the full results outside the app, for sharing, archival, or downstream PSA, click Export → Excel to download a multi-sheet .xlsx workbook. It covers:
- Model fit statistics (median survival, RMST, and the standard information criteria) for each distribution
- Fitted parameter estimates per distribution
- Observed and fitted survival across the analysis time horizon
- A summary of the IPD reconstruction
- Run metadata
- The variance–covariance matrices needed for downstream PSA
Budget Impact Analysis
If you picked the BIA template, the project opens straight to an inputs form. No digitized curves required.
Fill in BIA inputs
The inputs view groups everything you need into a few sections:
- Project basics: indication label and time horizon.
- Patient population: cohort settings and a choice of methods for deriving the eligible population.
- New drug: dosing, pricing, administration, and adverse-event costs.
- Current treatments: the same cost fields for each comparator (optional if you run without a comparator).
Review results
The results view shows net budget impact and percentage-change headline metrics, with three switchable tabs:
- Aggregate: total / net spend across the horizon
- Per-year: year-by-year spend breakdown
- Per-patient: cost per patient-year and per patient-month
Per-weight-category breakdowns appear automatically when weight-based dosing is configured.
Export to Excel (optional)
For sharing or archiving the analysis outside the app, click Export to Excel on the results view to download a multi-sheet .xlsx workbook. It covers:
- The full set of inputs behind the analysis
- Per-year spend breakdowns for the new drug and for current treatments
- Aggregate budget-impact totals
- Per-weight-category breakdowns when weight-based dosing is used
- An audit trail of source citations and notes
- Captured charts from the results view
Next steps
- Two-arm comparison: repeat the upload + reconstruction for a control arm to compute RMST differences and incremental life-year gains.
- Cure-fraction models: for therapies with long-term responders, the PSM-cure template layers in SMR-adjusted background mortality using built-in national lifetables. Full walkthrough in the cure-fraction guide (coming soon).
- Parametric distributions: read the methodology reference before justifying a model choice in a submission dossier.
LinkedIn