Kids are freaking fragile creatures. There is simply nothing more heartbreaking than when we hear of cases of child abuse and neglect. And now, when good healthcare is defined as more than just the absence of disease, and there is a greater focus on interconnected healthcare systems to create multiple healthcare access points, the promise of technology to address this horrific condition should never be more promising.
The underlying causes of child abuse may never be fully understood but it is clear that the rising incidence of abuse correlates to disturbing societal trends, many of which are quite obvious. Undeniably this country is experiencing unprecedented class segregation along residential, political and educational lines – all exacerbated by profoundly skewed income distribution. Ironically, demographers report that the U.S. is experiencing less segregation along religious and racial lines (inter-racial marriages were 0.7% of all married couples in 1970 and was 3.9% in 2008, per United States Census Bureau). A significant contributor to rising rates of abuse is the impact of poverty on certain populations and the associated desperation and pressures on family structure.
Earlier this year the Boston Globe reported that Massachusetts had the highest rate of abused and neglected children in the country, which quite frankly staggered me. In 2014 there were 31,867 “victimized children” in the Commonwealth which equated to 22.9 cases per 1,000 children. Of those children, 6,587 had to be removed from their homes to ensure their personal safety. Unconscionable. Nationally, there were 702,208 cases of abuse or 9.4 victims for every 1,000 children in 2014. The Crimes Against Children Research Center determined that abuse is directly tied to illegal drug use and that both poverty and a lack of affordable housing were also significant contributors to this tragedy.
Separate but related, my brother, Dr. Christopher Greeley, is a leading national expert in child abuse and has dedicated much of his career to developing community-based approaches to addressing this type of abuse. Over the years, and having heard about many of the heart wrenching cases he has confronted, I was shocked to see the incidence data above, which overshadow the data of many of the most feared diseases that we commonly discuss. My brother and other clinicians in the field are frustrated by the lack of effective diagnostic and intervention technologies that might determine more precise care pathways – all of these buzzwords are used when discussing other diseases like cancer, heart disease and even obesity.
Again, separate but related, gang violence has profoundly touched my family – twice. I am utterly baffled by the phenomenon of kids shooting kids. Inner city youth violence is one the greatest tragedies of our day and is yet another form of child abuse that we witness time and painful time again. And this too is directly tied to poverty and the associated break downs in social order.
Then there is the explosion of reported sexual misconduct cases at boarding schools. My prep school – Phillips Exeter Academy – was recently caught up in this with a series of revelations of inappropriate behavior by revered faculty. A dozen years ago the U.S. Department of Education released a study suggesting that nearly 10% of all children are the victims of “unwanted sexual attention” from educators and school employees over the course of their academic careers. Child abuse in yet another form.
Two months ago a group of neuroscientists, geneticists and social scientists convened in New York City for the “Poverty: the Brain and Mental Health” meeting. This group described the concept of “social concussions” for children raised in poverty given the multiple and chronic stress conditions that they are subjected to like parental discord, maternal depression, crime, intermittent hunger and poor nutrition. Notably, these conclusions are consistent with the pivotal Adverse Childhood Experiences study conducted in the mid-1980’s in California which studied 17,421 adults to understand how similar stress circumstances when they were children led to mental health issues later in their lives. These epigenetics studies suggested that damage from early abuse can be partially reversed through compelling community-based support systems that allow children to rebuild resilience via caring, consistent relationships with other adults, but it is an on-going struggle for most.
So how bad is it? in a meeting with Professor Robert Putnam at Harvard recently, he shared a wide range of social data which were very disheartening. Much of his research cuts across educational levels and compares those cohorts along a number of dimensions often considered predictive of childhood quality. For instance, as of 2010, for households with less than a high school education, 65% of those are single parent households; this was less than 20% in 1950. The statistics below describe children in each of those households.
And on and on. Clearly children born into less educated, poorer households are placed into more stressful, less supportive environments – and it is only getting worse with increased income disparity. The LENA (Language ENvironment Analysis) Foundation has developed a small wearable digital recording device (LENA System) to measure language patterns in children to assess impact of poverty on development. This “talk pedometer” has exposed the dramatic fact that poor four year olds have heard as many as 30 million fewer words than affluent children by that same age.
Earlier this year a nearly ten-year-old class action lawsuit was settled in Florida which sought to improve access to quality healthcare for children in low-income Medicaid families. It was shown that state reimbursement rates were so low that literally hundreds of thousands of children in that state had never had a check-up and that 80% of them had never seen a dentist. The primary remedy mandated in the settlement required that state contracts ensure that providers maintain adequate medical and dental provider networks so as to eliminate issues around access. While obviously not directly defined as child abuse, this structural disadvantage resulted in many cases that mirrored the impact often found in classic child abuse. The Florida legislature determined that steps taken to raise the general condition of a population ought to certainly help with overall well-being of its children.
Just a few months ago the Commission to Eliminate Child Abuse and Neglect Fatalities issued its final report after a two-year intensive study. A subtext throughout the report is a sense of frustration that technologies simply do not exist to provide an early warning system for healthcare providers. Shockingly but not surprising, the Commission identified that the best predictor of abuse is a call to a child protection hotline. Really?!? The most effective approach continues to be “home visiting” programs which unfortunately requires competent coordination across numerous public agencies, which at times can be a tall order. Sadly, buried in the data in the report, one learns that African American children die from abuse a rate 2.5x greater than the general population – shameful.
Which brings me back to my brother. Considerable VC investments have been made in disease specific diagnostics and therapeutics but when it comes to child abuse, it is very hard to precisely identify those children most at risk. Obvious cases of abuse are obvious – usually after the abuse has occurred. And while there are clear “markers” for likely conditions that would lead to abuse, such as poor maternal mental health or domestic partner violence, too often healthcare systems simply are not very good at predicting individual cases of abuse. Existing screening technologies still have far too high false positive rates; estimates for sensitivity are around 70%, which would not be commercially viable in the molecular diagnostics marketplace. “Meaningful use” was a significant step forward but too often there are stranded “data pools” that do not clearly correlate parental EMR data with related children.
In the general population the incidence of child abuse is thought to be about 1% but in “at risk” populations, the incidence spikes up to 2 – 3%. The field of child abuse prevention is moving to be more focused on population management; that is to roll out technologies to engage and educate broader “at risk” populations under the belief that a rising tide will help all children in a given community. For instance, there has been a dramatic increase in the number of “parenting apps” introduced to certain Medicaid populations; some of these require weekly recordings of parents reading to their children which both reinforces that reading is good for children and allows providers to witness those interactions for troublesome telltale signs.
Geo-coding of “at risk” populations also facilitates more sophisticated cluster analysis to see incidence patterns. Such an exercise allows providers and local governments and community leaders to better allocate resources such as community groups and other intervention programs.
And while the technologies to predict, assess and intervene in child abuse cases are developing, the clinical field at times struggles with the basic definition of many cases of abuse and neglect. Broken bones are easy to diagnose but what about latchkey children who are not eating regularly at home? Or cases of intermittent abuse – “my dad only hit me a few times” – or cases where an older child beats on a younger sibling? When does that behavior become a crime?