1. Why Ethics Committees Scrutinize Sample Size First
Institutional Ethics Committees (IEC / IRB) and funding agencies reject protocols more frequently for inadequate sample size justifications than almost any other methodological flaw.
An underpowered study (sample too small) is ethically unacceptable because it subjects patients to invasive procedures, phlebotomy, or experimental drugs without having enough statistical power to detect a true clinical effect. Conversely, an overpowered study (sample unnecessarily large) wastes institutional resources and needlessly exposes excess participants to inferior treatment arms.
2. The 4 Fundamental Pillars of Power Calculation
Every sample size calculation requires four quantitative parameters derived from prior peer-reviewed pilot studies or literature reviews:
1. Type I Error (Alpha, α) & Significance Level
The probability of falsely rejecting the null hypothesis (false positive). In medical science, α is standardly set at \(0.05\) (5%), corresponding to a 95% Confidence Level (\(Z = 1.96\)).
2. Type II Error (Beta, β) & Statistical Power (1 - β)
The probability of failing to detect a true difference (false negative). Power is the probability of correctly identifying a true clinical effect. Standard medical studies require Power = 80% (\(Z_eta = 0.84\)) or 90% (\(Z_eta = 1.28\)).
3. Clinically Meaningful Effect Size (Δ or δ)
The minimum difference between two groups that is clinically relevant (Minimal Clinically Important Difference - MCID), such as a 5 mmHg drop in SBP or a 15% increase in survival.
4. Population Variance (σ) or Baseline Event Rate (P)
The expected standard deviation or baseline incidence rate derived from published literature.
3. Clinical Formulas with Worked Examples
Scenario A: Estimating Disease Prevalence (Observational Study)
Suppose anticipated prevalence of diabetic retinopathy from literature is \(p = 20\%\ (0.20)\), with absolute precision \(d = 5\%\ (0.05)\) at 95% confidence (\(Z = 1.96\)):
Scenario B: Comparing Two Independent Means (Clinical Trial)
Where \((\mu_1 - \mu_2)\) is the target effect difference between arms, \(\sigma\) is the pooled standard deviation, \(Z_{lpha/2} = 1.96\) (5% alpha), and \(Z_eta = 0.84\) (80% power).
4. Adjusting for Anticipated Loss to Follow-up (Attrition)
In prospective clinical trials, patients discontinue therapy or are lost to follow-up. The calculated sample size must be adjusted upwards before ethics submission:
N_final = N_calculated / (1 - Loss Rate)
If \(N_{ ext{calculated}} = 246\) and anticipated dropout rate is \(15\%\ (0.15)\):
\(N_{ ext{final}} = 246 / (1 - 0.15) = 246 / 0.85 = 289.4
ightarrow\) Final Enrollment Target = 290 patients
Dr. Manjinder Singh Sidhu
MBBS, MD (Radiotherapy)
Senior Consultant in Radiation Oncology at DMCH Ludhiana with 18+ years experience, international fellowship at UCSD/Scripps USA, and 18 peer-reviewed publications.