Professor David Olds, Ph.D., Professor of Pediatrics and Director, Prevention Research Center for Family and Child Health, University of Colorado Department of Pediatrics ‘Using Randomized Clinical Trials of the Family-Nurse Partnership to Inform Policy, Practice, and Developmental Science’.
Recorded on 8 March 2019 at the Emanuel Miller Memorial Lecture and National Conference focused on ‘Attachment & Early Intervention: Improving emotional wellbeing and relationships in the family, and at school’
ACAMH members can now receive a CPD certificate for watching this recorded lecture. Simply email firstname.lastname@example.org with the day and time you watch it, so we can check the analytics, and we’ll email you your certificate.
Also listen to the podcast with David.
So I just want to, first of all, thank you Stephen, for those kind words. Marinus*, thank you for those kind words also. I really value the work you’re doing, and it’s really an honour to have a chance to talk with you this morning about the work that my colleagues and I have been conducting for the last 40 plus years. And I need to just mention a couple of quick things about this. The work that I’m describing is really the result of many passionate, caring, disciplined people contributing to what I think is an approach to early intervention that shows considerable promise. And I have to say that this work also has roots here in England, and in London in particular. As I’ll mention in a moment, there are roots to John Bowlby and an ethological approach to understanding early development and our cross species drives to protect our children. I think that attachment theory has played a seminal role in the design of the intervention and my own personal thinking about how we might approach the prevention of bad things happening to families and children.
I actually personally began getting involved in this work when I was a very young person finishing up undergraduate school in Baltimore. I had actually worked with Mary Ainsworth as an undergraduate and done coding on her attachment studies in Baltimore. And I was a product of the 60s. I thought if we could just help poor preschoolers get off to a good start, that so much would be prevented in terms of their being able to function in elementary school and later on in society. So I went to work in an inner city day-care centre and it soon became clear to me that a lot was happening to children that had preceded their experiences in my classroom. One little boy couldn’t speak. He could only gesture. His mother was a drug addict and an alcoholic, and he had been abandoned to friends in the drug culture, and was being cared for by a heroic grandmother. Another little boy in my classroom couldn’t sleep at naptime, and we discovered that he couldn’t sleep at naptime because when he slept, he wet himself and if he wet himself his mum beat him.
And so it was these kinds of experiences that led me to realise that we needed to start earlier and we needed to do something with parents. But I also realised that what was happening to parents was not simply a function of their own personal experience when they were growing up. It had to do with the kinds of resources and conditions that they were having to contend with in their immediate environment. This slide looks very much… this little corner store looks very much like the corner store that was right across the street from my day-care centre in Baltimore. We talked to the day-care centre about the importance of eating healthy diets, but there were no grocery stores with fresh fruits and vegetables for miles around this neighbourhood. The housing stock was terrible, the park where I’d take children for play there were alcoholics, drug addicts, needles on the ground. I had my car stoned on the way to work by gangs of kids. It was just unsafe. So to try to protect oneself and one’s child in those kinds of environments poses considerable challenges.
And so I realised also at that time in my life that I knew not enough to really make a difference, but those experiences in Baltimore taught me that, yes, early experiences do matter and that there is an opportunity to build on this instinctual drive that we as human beings have to protect our children, that may hold a clue for how we might make a difference. But also we needed to be very thoroughly attuned to the kinds of conditions that families have to contend with in crafting any kind of an effective intervention.
So after going to graduate school and studying human ecology with Urie Bronfenbrenner, I, with my colleagues, developed what is now known as the Nurse Family Partnership, or here in the UK, the Family Nurse Partnership. I love the fact that you in the UK have put families first. It’s a programme of prenatal and infant and toddler home visiting by nurses that focuses on low-income families in which the mothers are bearing first children. And we did this… We focussed on this transition to parenthood originally because there was… we were thinking that shifting roles would create vulnerabilities in the part of mothers that would make them more receptive to offers of help. But today we also realised that there are neuroendocrine changes, massive changes that are going on in mothers themselves that also are involved in restructuring maternal brains to fulfil their evolutionarily driven mandate to protect their children. So there is an underlying neurobiology that Marinus was alluding to earlier that is critical, involved in this particular point in human development.
We think the programme is… Marinus has indicated… has had replicated effects with different populations, living in different contexts, to give us confidence that this is something that is really worth pursuing. And one of the reasons we think the programme is showing consistency is that we’ve tried to be really clear about what it is that nurses are trying to accomplish and how they might go about doing this. And I think that it’s critical… It’s… yes, nurses. If you talk to people around the world about this programme, they’ll say nurses are heroes, and there’s no question in my mind that nurses are heroes. But if you talk to nurses, they’ll say that the real heroes in this are the parents, the mothers who’ve had to contend with often unimaginable, unspeakable sometimes, adversity, torture. And it’s the parents whose lives… and who are the heroes in this. And we need to, in some ways, what we hope to do at this particular point in human development is to activate and support this powerful drive to protect one’s child. The intervention is also strengths based. I need to mention that because especially, throughout the world, as we’ve developed the programme now, people will say to us that they really love the fact that the programme is looking for strengths on the part of mothers and not going in with a critical stance. I think that that is a critical feature, and I want to emphasise how nurses have played a systematic and deep, deep role in shaping the intervention that I’m going to be talking about here.
The nurses have three major goals. The first is to help women improve the outcomes of pregnancy by helping women improve their prenatal health. The focus is on things like prenatal tobacco use, alcohol, illegal drugs, identification of emerging obstetric complications, to have those problems treated more promptly and reliably so that they won’t affect the developing foetal brain. The second major goal is to help parents improve their children’s subsequent health and development by helping them provide sensitive, responsive, competent care for the child in the early years of life to set in motion a process of secure attachment and growth promotion on the part of parents in supporting their children’s cognitive and language development, helping parents gain satisfaction out of providing… seeing the rewards that they can observe in their children’s affect and developmental accomplishments, and the mother’s own feelings of competence and being able to manage increasing adversities in their lives. The third major goal is to help parents become… to improve their own health but also to create a more stable economic environment for their families. And critical in all of this is planning the timing of subsequent pregnancies, so the nurses focus on helping women complete their educations, find work, and again critically, planning the timing of subsequent pregnancies. A critical theory in all of this is self-efficacy theory, helping women identify small, achievable objectives that can be accomplished between visits that will reinforce their growing sense of mastery in coping with adversities in their lives.
The nurses also systematically involve fathers and grandmothers and other family members and friends in the family network who can play a role in supporting the goals and objectives of the programme and the mother’s sense of security. And also nurses link families up with other needed health and human services in the communities. Nurses are critical to all of this, of course, but they can’t do it alone. And so they work with primary care providers, mental health treatment services, substance abuse treatment services. They assist families in obtaining housing, emergency food assistance, all of those kinds of basic survival needs that families have to contend with nurses address through the intervention.
Now, we’ve tested the programme first in three separate randomised clinical trials in the US over the last 42 years, first with a sample of primarily low-income whites living in a semirural community in upstate New York. At the time we began the study we wanted to focus on those who were more vulnerable, but we didn’t have a clear idea about what vulnerability actually might look like. So we registered… We actively recruited women who were poor, who were unmarried and who were teenaged. But we allowed anyone in the community bearing a first child to register in the programme, in the trial, because we didn’t want to create a programme that was stigmatised as being only for the poor or for people with problems. So 15 percent of the 400 families enrolled in this study had no risks, none of the socio-demographic risks that we’ve identified, and that Marinus just listed in his presentation a while ago. Ninety percent of the sample was white. And many of the findings from our first trial were very promising. And many people in the US said, gee, you’ve got a programme that works; you need to make it more widely available. We took the position that we ought not to do that, that instead we needed to hold off. We needed to see whether the programme effects would replicate with minorities living in a major urban area first.
So we took four years and raised nine… from nine funding sources to raise the money to conduct a second trial in Memphis, Tennessee, with a sample that was essentially 90 percent African-American. But this time, one of the things you’ll see in a moment, is that the benefits in the Elmira trial were truly concentrated where there were overlapping socio-demographic risks. So we actively recruited in our second trial families where there were overlapping socio-demographic risks, and we registered this time 1,100 or so women for the prenatal phase of the trial. And because we were paying for this intervention and the research, all out of research dollars, we followed up on only a predetermined smaller sample, 742 turns out, followed postnatally, in this design. And then more recently, those findings were also replicating many of the beneficial effects that we found in the first trial, and so it gave us greater confidence.
But many people in the US came to us. They were… Home visiting was starting to really catch on… and they came to us and said, you know, you’ve got a very good programme here, but you know something? You could do an even better job if you were to hire people from the community to do this because they are closer to the families. There’s a reduced social distance. They’ll be able to do an even better job. And there were many, what we call paraprofessional or community health delivered home visiting programmes in the US at the time. So we said, well, you may be right. Let’s test it; let’s conduct a trial where families are randomly assigned to receive nurse home visiting based on the NFP, model paraprofessional home visiting based on a version of the NFP model or randomly assigned to usual care.
And so I’m going to share with you some of the major findings that have come out of these trials, but let me also just let you know that in all of the trials our usual care groups received more than what was usual care in the community. For example, in the Elmira trial, all of the mothers received free transportation for prenatal and well childcare. The children were screened regularly for sensory and developmental vulnerabilities and referred for further evaluation and treatment. A similar kind of approach was used in our Memphis trial and the same in Denver. And of course, you know a little bit, I think, about the quality of our health and human services system in the US and we are deeply envious of the system here in the UK, in spite of the fact that I know that it’s under assault right now with reduced funding.
These are the findings, some of the key findings where we see replicated effects on at least two of the three trials. We see significant improvements in prenatal health, replicated effects on reductions in tobacco use during pregnancy, biochemically validated reductions, significant reductions in hypertensive disorders of pregnancy. And there are other benefits that, in some cases, were only measured in one of our trials. For example, we found significant improvements in maternal diet in our first trial, but we did not choose to measure diet in our second trial. So things were going on in alignment with the model in all of the programmes that gave us confidence that nurses were accomplishing or achieving those objectives that might lead to better pregnancy outcomes. We found significant reductions in children’s injuries across trials, and I should say that the effects were most pronounced where families were more vulnerable. And I think that in all of this work, we need to be careful about distinguishing health care utilisation from health. That’s a huge issue for us. And I think that… We could pick up on this. I’ll illustrate this in just a moment.
We also see significant, consistent cross trial effects on children’s early language, cognitive development, school achievement, that these effects are limited to those children born to mothers who are more psychologically vulnerable, and by that we mean the mothers have lower cognitive functioning, limited sense of control over the life circumstances and higher rates of symptoms of depression and anxiety at baseline. So it’s the accumulation of those conditions in the control group that lead to… that are associated with poor developmental accomplishments on the part of the child that are mitigated in the presence of the nurse. We see replicated effects on children’s behavioural problems at school entry, based on both parent and teacher report. We see significant reductions across trials in children’s reports of depression, self-reports of depression and anxiety early in adolescence, significant reductions early in adolescence in the use of substances. And cross trial reductions in maternal behavioural impairment due to their own use of substances.
And consistently across all three trials, we see significant reductions in the rates of closely spaced subsequent pregnancies. That’s critically important because closely spaced subsequent pregnancies are associated with poor pregnancy outcomes in subsequent births, but they also mean closely spaced pregnancies [inaudible 00:20:36] the parenting [inaudible 00:20:38] to make it more difficult for parents to care well for the first born child. So it also means that if mother is working at McDonald’s at the counter and she gets pregnant quickly after the birth of her first child, she has to drop out of the workforce. But if she is able to postpone the birth of the second child, it means that she’s able to move on to becoming an assistant manager. So all of these kinds of elements of the intervention are designed to reinforce one another and to create pathways for improvements in these domains of functioning. And of course, if families are functioning better economically and have fewer pregnancies, then it’s easier for parents to provide nurturing, supportive care to the first-born child.
We see short term at least, cross trial effects on maternal employment. And in the US, in our first two trials, we see significant reductions in cash assistance, welfare, food stamps, and use of public health insurance for the poor. That effect doesn’t replicate in our third trial after more restrictive welfare rules and regulations went into effect during the Clinton administration. I’m going to illustrate this pattern of results with a few findings. Let me see if I can help you see what’s going on here. We found in our first trial, for example, that among… 45 percent of the women, by the way, smoked five or more cigarettes per day at registration, and in that group, there was both a reduction in prenatal tobacco use and a corresponding reduction in the rates of preterm delivery that you see on this slide. We also see, in the first two years of life, that there was a trend for there to be fewer cases of state verified reports of child abuse or neglect in that segment of the sample where there were overlapping socio-demographic risks, where the mothers were poor, unmarried and teenaged. These were the criteria used for sample recruitment. And while this is not a statistically significant difference… It’s a trend…this is a clinically highly meaningful effect. And so it’s… and you can see how much more pronounced the effect is in the control group without support. This group, by the way, was visited by nurses during pregnancy alone and not followed up postnatally. And we see that in that group of poor, unmarried teens, the effect of the intervention was further pronounced among women who at baseline had limited sense of control over their life circumstances.
These are… I’ll try not to do too much of this… These are separately fitted regressions of maternal… of child abuse and neglect, whether or not the child was abused or neglected, on maternal sense of control, fitted separately for the comparison group versus those who received the nurse. And you can see that that large difference in the preceding slide, this difference here, was really limited to those, or let me say this, more pronounced among mothers… children born to mothers with limited sense of control over their life circumstances. And remember, building self-efficacy is part of the model. So this led us to hypothesise that the benefits would be more pronounced in subsequent trials, not only on the basis of women’s limited sense of control over their life circumstances, but by corresponding limited intellectual functioning and higher rates of depression and anxiety. We’re going to pick up on that piece in a moment.
We also see 15 years after the birth of the first child that there was a treatment main effect. By that we mean irrespective of socio- demographic characteristics, there was roughly a 50 percent reduction in the rates of state verified reports of child abuse and neglect, but this effect was again more pronounced where they were overlapping socio-demographic risk, where the mothers at baseline were poor and unmarried, and poor. They were poor and unmarried. In our… And we also found in that same 15-year period that mothers were less likely to be arrested over the first 15 years of life in that group that was poor and unmarried at baseline. And that over the 19 year period following the birth of the first child, there was a significant reduction in the rates of self reported arrests on the part of the children. This effect, by the way, was there at 19, but the effect… or at 15, but at 19, in the second half of adolescence, the beneficial effect was really limited to females. And we see that, if anything, the intervention control difference wasn’t statistically significant in the wrong direction, but there were higher rates of arrests in the second half of adolescence in the recidivity group compared to their counterparts in the control group, among males. And exactly why that is, is not entirely clear. There is some thinking that adolescence leads to normative eruptions of antisocial behaviour, that in many… in families that are functioning well are limited to adolescents. We have not been able to do the kind of follow up that I would like to really see whether there are lifetime differences in arrests and convictions. So the story is still out on the life course history of criminal involvement.
These findings led us to our Memphis trial where we concentrated our sample recruitment on a population that was at very high risk, 92 percent African-American, 98 percent unmarried. Eighty five percent of these families enrolled in the study were living below the US federal poverty guidelines and in the US that is really, really poor. It means that you have virtually no discretionary income. As a matter of fact, the families in the control group had less than what… On average, they had insufficient income to cover even the basics. This was a largely teen sample, and the neighbourhoods in which families were living were among the very worst in the US. Two and a half standard… 2.4 standard deviations above the national mean in terms of neighbourhood adversity. So we are dealing with a population and a sample that is at very, very high risk. This sample… this slide shows in the Memphis trial how those 1,100 cases were randomised.
We originally created a… had these three groups: a group that would receive prenatal transportation and screening for the children, a group that would receive the transportation compared to prenatal and a single postpartum visit, and a group that received transportation, screening, prenatal and postpartum home visits. Within the first six months of the trial, we realised that the rates of recruitment into the trial were higher than we could accommodate with the nurses we had hired, so we created another group that was randomised after that point to receive just the transportation for prenatal care. For the test of prenatal effects, we combined these two groups and compared them to the combination of these two groups and for the analysis of postnatal effects we compared group two to group four.
Here you see one of the effects that I mentioned earlier. This pattern of results was replicated in our Elmira trial. We see significant reductions in pregnancy-induced hypertension. We see significant reductions in the numbers of days that children were hospitalised with injuries, and I think this is particularly important because hospitalisation for injury is not simply a reflection of the worried well. This is a reflection of children who are seriously injured and have to be hospitalised because of injury or ingestions. This slide shows the three children who are hospitalised with an injury or an ingestion in the first two years of life in the nurse visitor condition, group four. Notice that these children are all 12 months of age or older. And two of these children picked up things on the ground and ingested them while crawling around. They were mobile; they were creating this for themselves. This child was crawling on a bed where his grandmother had been ironing and picked up an iron and put it on his face.
This slide shows the corresponding rates or the diagnoses associated with hospitalisations in the first two years of life in the control group. Now notice that the denominator here is twice as large as what we see for the nurse visiting group. But notice that these children… 40 percent of these children are hospitalised before six months of age. These children are not mobile. They are not creating risks for themselves because of their mobility. Notice the nature of the diagnoses associated with these hospitalisations: head trauma, fractured long bones, bilateral subdural haematomas, bleeding under the skull, fractured skulls, another… a rehospitalisation of the child who had been hospitalised earlier with bilateral subdural haematomas, fractured skulls, child abuse and neglect suspected. These children were hospitalised for dramatically different reasons and for longer time periods and at much younger ages, and all of these effects, virtually all of them, were limited to that segment of the sample where the mothers at registration fell into this category of having limited psychological resources to cope with adversity. These are the mothers with higher rates or lower… They had lower cognitive functioning, higher rates of depression and anxiety, and limited sense of control over their life circumstances. And this index… What we did is we created an index that standardised each of those dimensions that I just described for you, averaged them, standardised them to a mean of 100 with a standard deviation of 10, and you can see that all of the treatment control difference is limited to the sample in the lower half of the distribution.
Now in our Memphis trial the rates of state verified reports of child abuse and neglect were three percent in the general population over the same time period. We conducted pre-trial work in this community and found that the rates of… official rates of maltreatment were far too low for this to serve as a viable outcome in this trial. So we hypothesised this pattern of results, and it just reminds me once again that use of state verified reports of child abuse and neglect is a very unstable and not very meaningful indicator of whether the children have actually been abused and neglected. I think it’s a general problem that we have for all the preventive intervention work that we are doing. I say we, I mean, collectively, all of us because, again, we need to distinguish between health care utilisation patterns, social service utilisation patterns, and real health. We need generally to come up with much better indicators of neuro-biologic functioning early in life that may reflect children’s experiences.
In the first trial… Let me just circle back a little bit… In the very first trial that we conducted we found that the… that was conducted in Chemung County, New York… They had the highest rates of state verified reports of child abuse and neglect in the entire state, in the entire… in all of New York State. The economic conditions were very poor. And I should point out that our nurses worked with Child Protective Service workers to make sure that suspected maltreatment was referred to local Child Protective Services, to make sure that those cases where maltreatment would occur would not be lost. Nurses are required, just like they are here in the UK, to report suspected maltreatment. It became clear to us in the conduct of that trial that we have a kind of surveillance bias operating in this programme as well. And by that I mean nurses are required to report and ethically they’re going to make sure that children who need these services or need to be protected are going to be protected. I think that in general, as we think about maltreatment, we have to keep that issue in mind when we’re trying to evaluate interventions. So we were never able to… We never even sought the records in our Memphis trial to look at maltreatment and the same… We never got the records in our third trial as well.
So we’ve had to rely on other indicators that this child is dysregulated and there are challenges with caregiving in both our Elmira and our Memphis trial. I don’t have data here to show you, but there were interventions, differences in observed qualities of dyadic interaction. And they’re more pronounced in this segment of the samples. And we see in our Memphis trial that there are significant intervention control differences in children’s dysregulated aggression revealed in their response to the MacArthur Story Stem Batteries. I don’t know whether you even know about this measure, but it’s an approach that gives children little opening narratives about stories and then children are asked to complete the stories. Well, what we see is that in the children born to mothers with the lowest psychological resources, the rates of dysregulated aggression in their stories is significantly higher than in the nurse visited course… in the group visited with that same high risk group visited by nurses. And we see a similar kind of pattern when we look at the degree to which children are incoherent in their narrative responses to these story stems, and again, the rates of incoherence are substantially higher in children born to mothers with low psychological resources and the nurse visited group has offset that risk. And you can see that the rates are substantially lower for both the intervention and control group in the higher resource sample and these differences are not statistically significant. So that’s where the action is.
In the Memphis trial, at the age 12 we see significant intervention control differences in children’s reading and maths achievement directly assessed. But it’s limited to those children born to mothers with low psychological resources living in these neighbourhoods and environments where levels of adversity are off the charts. And when we look at the emergent use of substances at age 12, intervention control differences, irrespective of the mother’s psychological resources overall in children’s emergent use of substances at age 12, corresponding reductions in internalising disorders at age 12, again, irrespective of the mother’s psychological abuse, [inaudible 00:40:16] as a baseline. And over the first two decades following the birth of the first child, there was a significant reduction in mortality among the children for preventable causes, and that includes sudden infant death syndrome, which is reflected right here, injury, and homicide, by the time the children reach adolescence. And because of the programme’s effect on helping women become more economically self-sufficient, we see significant reductions in children’s use… government expenditures for food stamps, Medicaid, which is the US health insurance programme for the poor, and cash assistance welfare.
In our Denver trial, what we see is that in spite of the claim that paraprofessional or community health workers would be able to do a better job than nurses, in fact, nurses produce effects that are roughly twice as large as paraprofessional visitors. And it’s really… This is not the whole story, but part of it… is that families open the doors more for nurses. And even if you control statistically for differences in the dosage of the programme received, nurses produce greater effects because of their greater clinical sophistication. This programme requires nursing to be operating at the highest levels of functioning. And it’s that when we see it’s not just a matter of dosage, it’s a reflection of the clinical sophistication in being able to weave all of these components of the programme together. The programme is organised; it’s operationalised; it is detailed. There’s a lot of guidance that’s provided to nurses about how to do this. But ultimately, it’s up to nurses to make decisions about how they’re going to adapt or adjust the content and dosage of the programme on a moment to moment basis for the particular mother, father, child they’re interacting with at that time. So it’s this individualisation that is critical to really the success of the programme.
In Elmira or in the Denver trial, we see, for example, significant reductions in urine cotinine. Cotinine is the major nicotine metabolite. So we see significant reductions for the nurse visited group, smaller, non- significant reductions for the paraprofessional group, almost no reductions for the control group. We see at 21 months significant differences in children’s language delays at 21 months, but they are limited to children born to mothers with low psychological resources. You can see that the rates of language delay are really very high in the control group, children born to mothers with low psychological resources, and they’re cut substantially for the group visited by nurses. And we see at 21 months or 20… or not 20 months, four years, there was a significant overall difference in children’s language development, but it’s limited to, again, children born to mothers with low psychological resources. It’s the group… in the control group that is born to mothers with low psychological resources that is the most vulnerable and nurse visited group, the nurse visited group, is performing significantly better than their counterparts in the control group.
And we’ve also created in the Denver trial a synthesis of different components of executive functioning and created an executive functioning composite, and what we see for the children is that there is a significant improvement in the nurse visited children born to low psychological resource mothers executive functioning compared to their counterparts in the control group. So this gives us a lot of hope for the future. I’m going to show you one other slide that I think is important. And these are survival analyses of timing to the first subsequent pregnancy. Here is the group visited in the control group. Here is the paraprofessional visited group, and here is the nurse visited group. Again the paraprofessional group is following right in between the control and nurse visited group. And its nurses are, for a whole host of reasons, able to really provide women with good guidance and access to contraceptives to plan the timing of subsequent pregnancies.
So it was with these findings under our belt… We held off for 20 years from offering up the programme for public investment before… We wanted to make sure that findings would replicate, that findings would replicate with different populations, living in different contexts, the different points in US social and economic history, under different policy contexts, and we wanted to know the findings would endure before offering the programme up for public investment. But in 1996, we were invited by the US Justice Department to set up a programme in high crime neighbourhoods in the US, and it was with a lot of apprehension that I endorsed that work and we began a process of careful community replication of the programme in the US, in which we paid a lot of attention to making sure that organisations and communities were well-prepared to deliver the programme. We developed detailed educational and consultation models for nurses. We operationalised the programme in visit-by-visit guidelines meant to be adapted to families’ individual needs. We created a consistent information system that monitored programme implementation in maternal and child health, and that information was then used to assess programme performance, and that information was used to guide continuous quality improvement.
And so today the programme is operating in the US in over 300 communities, serving 50,000 families a year in the US, and then ten or 12 years ago, actually maybe more than that, we were invited first by the Dutch and then the English to replicate the programme here, and we developed a model for international exportation of the replication of the programme. And we took the position that we didn’t know whether the programme would work in new contexts. The cultures are different. Families may have different attitudes about caring for their children. Would this drive to protect children hold up in Aboriginal contexts or Roma families in Bulg… in Eastern Europe? Say nothing about even more dramatically different cultural contexts. So we took the position that we don’t know, and that if we’re going to work with other societies, we needed to carefully adapt the programme to their populations, their context, and then to conduct a small pre-test, a small scale trial of the programme to see whether it was really feasible. And then if it was that, if it looked like it was really going to work, then we would really urge our collaborators, government… It’s almost always government… to invest in a randomised clinical trial of the programme in those new settings, so that we would have some estimate about what the added value of the programme would be in those other contexts. We felt that that was the responsible thing to do because in many societies, whether it’s Norway or here in the UK, you’ve got much, much better well-developed health and human service systems. What’s the added value of this programme in your context? We needed to know.
And so we’ve been working for the last… If the programme works… And you saw the evidence in from Holland; they give us a lot of reassurance that it can work in other contexts. We’ve been now working here in the UK, in England, Scotland, Northern Ireland. I know that there are some nurses here from those countries and I am thrilled that they’re here this morning. We’ve been working in Australia to serve Aboriginal families, now serving 13 Aboriginal communities in Australia, working in Ontario, and British Columbia, and Canada. There’s a separate randomised clinical trial going on there. We’re working in Norway to serve really vulnerable segments of their population, in Bulgaria, serving Roma families, and throughout the US serving American Indians and Alaskan natives. In all of this, our approach has been that the programme itself, and our capacity to deliver it well, will always be a work in progress that we need to pay attention to where things are not going well. And so we’ve been doing work now, translational kind of research, looking at the predictors of participant retention, and completed home visits, and devising interventions to improve that.
We’ve devised and supported the conduct of a cluster based randomised clinical trial to address intimate partner violence, to support hormonal contraception delivered by nurses in the home. We’ve developed a new method to support nurses’ observations of qualities of dyadic interaction in the home, and using that information to guide nurses even more effective, supportive parent dyadic interaction. We’ve developed an intervention for nurses to deliver in addressing moderate maternal depression and anxiety in the home. We’ve developed a tool for nurses to use that aligns with the underlying theory in developmental epidemiology, the programme to, and the clinical operations of the programme, to identify proximal and more distal risks, and to use as a tool in guiding the delivery of the programme called the STAR Framework.
We have been working to modernise the NFP with new tele-health methodologies and to promote even, we think, better collaboration between the NFP and child welfare and primary care providers. And now in the US, we’re also developing and testing a version of the programme for women who’ve had previous live births and those who are involved with substance abuse, and in the US we are particularly concerned about the terrible, terrible opioid epidemic there. Opioid overdoses are now the highest cause of death in the USA, exceed automobile accidents as the cause of death in the US. So in either case, we’re trying to prevent bad things from happening to mothers and their offspring with this. And in all of this, we tried to honour these principles in the design and implementation of the programme. And with that, I will take some questions. So thank you very much.
Thank you, David, for an inspirational talk, and gosh, how you’re developing it wider and wider. Terrific. Questions? Jonathan Green?
Hello, David. Thank you very much, great considered programme over a long time. The results make me even more interested in some of the more detailed fine-grain comments on actually the delivery of the programme, how much nurse time it takes, how frequent the visits are, what your attrition rates are, what your education of your nurses might be, etcetera. So some more details about how the programme actually works in practise?
Professor David Olds
Particularly interested in dosage issues, so how often are these families being visited on average and the range of that? And if you’ve got any dosage effects you could share?
Professor David Olds
So thank you for that question. We have 62 structured, 64 now, 64 structured visits that nurses have at their disposal to guide their work with families, but nurses are instructed to adapt the dosage to the needs of the families. Nurses in the US are unable to carry more than 25 families per nurse. Here in the UK, it’s probably less than that because of just differences in their workload and hours of working. But from the beginning, we’ve known that you cannot deliver 64 visits over the course of that two and a half year period to caseloads of 25 families. So we’ve guided nurses to focus on those where the needs are greatest. And the rates of attrition out of the programme are in the first… We’re dealing with rates of attrition of around 50 percent overall. These are dropout rates, and a lot of this has to do with families moving. There’s greater transience in the US. I don’t know what the corresponding data are here in England or Scotland, but it’s less. And the rates of uptake of the programme are higher here, and I think that it’s higher because you have universal services; you have health visiting; it’s part of your culture. It’s not so potentially stigmatising to have someone come to your home. So all of that makes this kind of service much more acceptable in your context.
Professor David Olds
So the nurses, just to give you some insight on this, the nurses are asked to visit once a week for the first month after families register in pregnancy. And that’s different, by the way, than the way usual prenatal care is organised, where the visits are really bunched up at the end of pregnancy. We try to get to families and have more frequent visits early in pregnancy because that’s when all of the action is happening in terms of exposures to neurotoxins, for example. And we think it’s critical that the nurse establish a relationship. Everything depends on the quality of the nurse’s relationship with the mother. So getting back in there, understanding mother’s needs, her aspirations, her fears, all of that is critical to the success of the programme. Then the nurses drop back to visit every other week, and then after the week… until the baby is born… and then once a week for the first six weeks after delivery, and then every other week, and then every three weeks, and by the end of the two-year period, the nurses are visiting once a month. Now, nurses are making adjustments to the frequency of visits, depending on their observation about the family’s needs. So this adjustment of content, adjustment of dosage, is part of the clinical judgement, sophistication, that nurses bring to the table in doing this work. We’ve been able to build upon, and I think we… The UK government has been able to build upon the wonderful existing systems that you already have in place with health visiting, for example, so… You have a wonderful workforce in health visiting, and so it gives this programme a leg up in the UK in getting up and running well. So we’ve learnt a lot. We will continue to learn a lot from UK nurses and counterparts in all of this.
Right, one more question here.
Professor David Olds
Shall I go first? Thank you. Aiden Phillips from WAVE Trust. I was very interested to see the results comparing high psychological resources with low psychological resources. Do you think this evidence provides a strong argument, not just for investing more in perinatal mental health services for the parents themselves, but actively trying to encourage parents to look into their mental health during the prenatal period to try and acknowledge possibly childhood trauma, insecure attachment, issues they may not have considered before, rather than waiting for them to come to you?
Professor David Olds
I want to make sure that I’ve understood your question. So I do think that these findings don’t apply to just… This pattern of results doesn’t apply to just the Family Nurse Partnership, and that it does tell us that this is a population that is both vulnerable and maybe susceptible to the positive effects of these types of interventions, in spite of some of these kinds of things that may be heavily influenced by genetics. There is… We may also… There may be underlying this group some fairly high concentration of susceptibility genes. There may be. So I don’t… We could dig into this in lots of different ways. I do think that these findings suggest that there is a vulnerable population there. If we spend the time engaging them in a respectful, caring way that they will be responsive to offers of help, but there is often natural resistance and natural apprehension about engaging helpers if they experience those helpers as somehow being judgemental, so that whatever is done, it has to be done in an open, caring, supportive way. And I do think that pregnancy is an opportune time to do this kind of thing because of all of the changes that I just mentioned at the beginning of my talk. I hope that answers your question. Okay, thank you.