For the video version, click here.
How nice would it be, as a first-time parent, to have a nurse and an educator come to your house about once a week for the first three years of your child’s life? They’ll check in, help troubleshoot problems, give advice about breastfeeding, answer questions, provide support. Especially for first time parents, such a resource seems invaluable. But let’s try to put a dollar amount on it anyway.
We’re talking about something called the “First Born Program”, which is the subject of this randomized trial appearing in the journal Pediatrics.
The First Born Program is a community-sponsored agency that provides free and frequent home visits for first-time parents. It focuses on family education, identifying family challenges (including post-partum depression), and referral to appropriate services. It seems like a pretty great idea, but until now we had no hard data to evaluate it.
Enter one FBP site in Santa Fe, New Mexico. Due to high demand, the site had to engage in a lottery system to allocate its resources. Thinking quickly, researchers used this lottery system as a de facto clinical trial to evaluate just how much these services might help.
This is a single site – not a huge study – 138 families got the intervention and 106 didn’t. Why are those numbers so different? Unfortunately, many families dropped out of the study AFTER the lottery – 128 in the intervention group and 211 in the control group. To make matters worse, the researchers didn’t have data on the people who dropped out after randomization, only those who completed a survey roughly a year later. Though they describe their analysis as “intention to treat”, it is not, at least by the classical definition.
Sure – the groups look pretty similar:
But once you lose track of people after randomization the power of randomization is quickly lost. We may just be looking at a study of people who wanted to get this intervention vs people who didn’t really care- and that is a recipe for bias.
The results were fairly compelling though.42% of control group kids had an ER visit in their first year of life, compared to just 28% of treatment group kids. That said, there was no significant difference in rates of hospitalization or injuries requiring medical attention.
If you’re a pragmatist like me, you’re probably asking – how much does this program cost? It’s nowhere to be found in the article, but the First Born Program’s annual report from 2015 states it’s around $3400 per family.
That makes the total cost of the intervention in the treatment group just under 500,000 dollars. That money saves you about 10 ED visits, so you spend around $50,000 dollars to keep one kid out of the ED.
Now, this analysis is not really fair. There are a lot more outcomes that might be helped by these programs beyond ED visits. But unfortunately, this study doesn’t give us the right tools to figure it out. I believe it’s time for communities, government agencies, and grant-funders to embrace the idea of performing robust clinical trials in this space – trials where data is collected at the time of randomization – in order to give us the ammunition we need to support early-life interventions that may be truly beneficial.