PyStatsClinical

Clinical-trial and clinical-research statistical computing for Python.

Status: 0.1.0 — first module shipped. The treatment-effect module is available now; more clinical-research methods are on the roadmap below.

PyStatsClinical is part of the open-core PyStatistics family:

Like its siblings, it builds on pystatistics for the general statistical layer and adds methods specific to clinical research.

What's available now (0.1.0)

Treatment-effect measures from a two-arm 2×2 — the canonical readout for a binary outcome:

from pystatsclinical import effect

r = effect.risk_measures(
    treated_events=15, treated_n=100,
    control_events=30, control_n=100,
)
print(r.summary())

risk_measures reports the control and experimental event rates (CER, EER), the absolute risk reduction (ARR) with a Newcombe or Wald confidence interval, the risk ratio (RR) with the Katz log-method interval, the relative risk reduction (RRR), and the number needed to treat / harm (NNT/NNH) with the Altman (1998) interval. It fails loud on invalid counts and on an undefined risk ratio (CER = 0).

Roadmap (candidates, not commitments)

• Responder analysis: responder rate with a confidence interval at a clinical threshold.
• Diagnostic accuracy: sensitivity, specificity, predictive values, likelihood ratios.
• Non-inferiority / equivalence testing and stratified (Mantel–Haenszel) effects.

Installation

pip install pystatsclinical

License

MIT © Hai-Shuo. Part of the SGCX open-core ecosystem.