Grantee: University of Colorado-Denver. The full study report is linked here.
Description of the Intervention: This was a study of the long-term impact of the Nurse Family Partnership (NFP) on mortality among the mothers and children served by the program. NFP is an evidence-based nurse home visitation program for first-time mothers – mostly low-income and unmarried – during their pregnancy and children’s infancy. NFP has been evaluated in three well-conducted RCTs in the United States (in Elmira, N.Y., Memphis, Tenn., and Denver, Colo.). These studies have found a pattern of sizable, sustained effects on important child and maternal outcomes, including reductions in child abuse and neglect and mothers’ number of subsequent births in their late teens and early twenties.
Study Design: This study used National Death Index (NDI) data to measure mortality outcomes for all sample members in the three aforementioned RCTs. The sample members included (i) 1,138 mothers and 1,089 children who were born alive (i.e., not miscarried or stillborn) in Memphis; (ii) 400 mothers and 388 live-born children in Elmira; and (iii) 735 mothers and 709 live-born children in Denver. Consistent with the study’s pre-registered analysis plan for estimating mortality effects, the Elmira and Denver trials were analyzed together, while the Memphis trial was analyzed separately because Memphis’ study sample was substantially more disadvantaged than the Elmira and Denver samples. The Elmira-Denver sample was tracked for 21.1 years and the Memphis sample was tracked for 24.9 years after random assignment. Mothers in the Elmira-Denver sample averaged 41 years of age at the time of follow-up, while mothers in the Memphis sample averaged 43 years of age.
Impact on the Primary Outcomes: The study found that NFP had no statistically significant effects on any of the four primary outcomes: (i) all-cause mortality for mothers in the Memphis sample; (ii) all-cause mortality for mothers in the combined Elmira and Denver samples; (iii) child mortality from preventable causes in the Memphis sample; (iv) child mortality from preventable causes in the combined Elmira and Denver samples. In each case, the mortality rate was marginally lower for the NFP group than the control group, but the differences were not statistically significant and could well be the result of chance. Overall mortality rates were very low in all groups (between 0.0% and 5.1%), limiting the study’s ability to detect effects on mortality.1
Study Quality: Based on a careful review, we believe this RCT evaluation of NFP’s long-term impacts on mortality was well-conducted and produced valid findings.2
Based on the sample sizes, the study had the ability to detect differences in mortality rates between the NFP and control groups of 2-3 percentage points or greater.↩︎
For example, random assignment was successful in each of the three RCTs (as evidenced by highly similar treatment and control groups); and this evaluation of mortality impacts had virtually no sample attrition (fewer than 1% of sample members could not be matched to NDI data), and used valid analyses that were publicly pre-registered.↩︎