NMES

Calculating How Many ICU Patients It Takes to Prove Electrical Stimulation Prevents Muscle Loss

Biostatistics consulting — sample-size and power analysiswith Elizabeth Seewer, PT, DPT, PhD — Le Bonheur Children's HospitalConsulting analysis / presentation deliveredPower analysis, G*Power software, two-sample t-test sample-size formula

138 total participants powered (65 per arm)25% mean quadriceps CSA loss observed in pilot data (p = 0.02)~730 eligible PICU admissions per year at the partner hospitalConsulting analysis for Le Bonheur Children's Hospital

Context

What the system looked like before.

Critically ill children in the ICU can lose significant muscle mass in a short time. Dr. Elizabeth Seewer's team had already run a pilot observational study in the PICU at Le Bonheur Children's Hospital and found a mean 25% reduction in quadriceps muscle cross-sectional area, present in 57% of patients — roughly 2.7% muscle loss per day. The open question was whether Neuromuscular Electrical Stimulation (NMES) could reduce or prevent that atrophy, and before running that trial, Dr. Seewer needed to know how many patients it would actually require.

Question

Does the use of Neuromuscular Electrical Stimulation (NMES) reduce and/or prevent atrophy in critically ill pediatric patients, measured by quadriceps muscle cross-sectional area?

Worth flagging as a framing problem before getting to the statistics: as originally worded, this question will always be answered yes, since it already assumes NMES improves atrophy.

Approach & rationale

A sample-size calculation for a randomized controlled trial depends on three levers: power (the probability of detecting a real effect, commonly 0.80–0.90), alpha (the false-positive rate, usually 0.05), and the Minimal Clinically Important Difference, or MCID — the smallest change clinicians actually care about. The MCID converts into Cohen's d, a standardized effect size, so it can plug into the sample-size formula. Set the MCID too large and the trial may miss modest but real benefits; too small and the trial needs many more participants to detect an effect that may not even be clinically meaningful.

G*Power software interface for the sample-size calculation — The G*Power interface used for the power analysis.

Methodology & iteration

Including the paths that did not hold.

Both are valid — the tradeoff: a broad question needs a large N and a long timeline (δ ≈ 10–15%); a narrow question needs a smaller N and a clearer focus (≥75% threshold).

Approach A — power the study for the observed effect size

TriedUsing the clinician's own pilot data (25% mean CSA loss vs. a target of ≤10% loss in the treatment arm), set the between-group difference to power for δ = 15 percentage points. Plugging σ = 0.28, δ = 0.15, α = 0.05, and 80% power into the two-sample t-test formula gives n ≈ 55 per group.

Held upAfter accounting for the ~15–21% attrition rate seen in the pilot, that scales to 65 per arm, 138 total — matching the clinician's own original estimate, and feasible given the hospital's admission volume.

Approach B — reframe the question to need fewer patients

TriedInstead of a continuous percentage change, ask a binary, threshold-based question: in PICU patients receiving NMES, what proportion achieve ≥75% CSA preservation over 15 days compared to standard care?

Held upA binary endpoint is simpler statistically and often needs a smaller sample, and answers a clinically actionable threshold directly rather than an abstract percent-change average.

In short

Results

Numbers first, not buried in prose.

The current research question, powered at α = 0.05 and 80% power for a 15-percentage-point difference in CSA loss, requires 138 total patients (65 per arm) — which fits comfortably within Le Bonheur's patient volume. At roughly 730 eligible patients per year, that's achievable in well under a year of active recruitment, not the originally assumed two years.

Worked two-sample t-test sample-size formula — The full worked formula, arriving at n ≈ 55 per group before attrition.

Significance

Where this went.

Whether this fed into an IRB submission or trial protocol for Dr. Seewer's team isn't confirmed yet.

Reflection

This project was different from my other case studies — advising someone else's study design rather than running my own experiment. A fuller reflection on what that shift taught me is still being added.

BackAll work NextLamins →