Skip to main content

Table 4 Power calculations

From: Protocol for a randomized controlled trial evaluating the impact of the Nurse-Family Partnership’s home visiting program in South Carolina on maternal and child health outcomes

Primary outcomes Assumptions regarding control group mean Full sample (unadjusted) Full sample (Bonferroni) Vulnerable subgroup (unadjusted) Vulnerable subgroup (Bonferroni)
Low attrition High attrition Low attrition High attrition Low attrition High attrition Low attrition High attrition
Adverse birth outcome 24% − 3.5 pp. (14%) − 3.6 pp. (15%) − 4.3 pp. (18%) − 4.4 pp. (18%) − 5.2 pp. (21%) − 5.4 pp. (22%) − 6.3 pp. (26%) − 6.6 pp. (27%)
Birth spacing 13% − 2.7 pp. (20%) − 2.8 pp. (21%) − 3.3 pp. (25%) − 3.4 pp. (26%) − 3.9 pp. (30%) − 4.1 pp. (31%) − 4.8 pp. (36%) − 5.0 pp. (38%)
Acute injury, abuse or neglect 18% − 3.1 pp. (17%) − 3.2 pp. (18%) − 3.8 pp. (21%) − 3.9 pp. (22%) − 4.6 pp. (26%) − 4.7 pp. (26%) − 5.6 pp. (31%) − 5.8 pp. (32%)
  1. Note(s):
  2. Power calculations show the minimum detectable effect between treatment and control in percentage points (pp) with percentages in parentheses. Percentages are rounded to the nearest whole percent. We assume a significance level (alpha) of 5% and a statistical power level of 80%, and that 96% of study participants randomized into the treatment group will receive NFP services, based on current enrollment trends. Additionally, we assume that none of the control group receives NFP services. Low attrition calculations assume 7.5% attrition, while High attrition estimates assume 15% attrition across both treatment and control groups through miscarriages, migration, imperfect identifier for matching, or withdrawing consent
\