Suspecting Child Abuse: Challenges and Guidance

-So my name is Jonathan Beever. I’m one of the postdoctoral
Fellows, postdoctoral Scholars, with the Rock Ethics Institute. I’m really happy
to be here, and I want to thank my
colleague Nicolae Morar, and our staff members,
Rob Peeler, who’s at the back, Deborah
Trialonas, and Carolyn Umbricht for their help
in organizing the series. I want to also
thank Doug and Julie Rock for their ongoing
support of our work here at the Rock
Ethics Institute. It’s my pleasure this evening to
welcome to University Park, all the way from Hershey, Dr.
Benjamin Levi, physician and professor of Pediatrics and
Humanities at the Penn State University College of Medicine
and the Penn State Children’s Hospital. Dr. Levi received his
MA, his PhD, and his MD from the University
of Illinois at Urbana. He’s a board
certified pediatrician and Fellow of the American
Academy of Pediatrics. Besides regular and
ongoing supervision of clinical work and his
own patient care activities, Dr. Levi also teaches
bioethics regularly to medical students,
residents, and nurses. He leads at least three
research projects, including an interactive
online decision aid related to advance
care planning called Make Your Wishes Known, an
online resource for mandated reporting of suspected child
abuse called Look Out for Child Abuse, and Penn State Suspicion
Project, an ongoing research project on reasonable suspicion
in the context of child abuse reporting. This evening, we’re very
excited to have Dr. Levi here for a talk titled
Suspecting Child Abuse– Challenges and Guidance. Dr. Levi will give us a
45- or 50-minute talk, and we’ll have a great period
of questions and answers at the end. So without further
ado, please help me in welcoming Dr. Benjamin Levi. [APPLAUSE] -Well, first of all, thank
you all for your interest in this topic, and
I hope to raise some interesting
questions and challenges and perhaps provide
some guidance as well. So uncovering child abuse is
not just a diagnostic dilemma. It’s a dilemma in terms
of values and ethics. And what I’m going to try to
do in the next 45 minutes, or so, is first provide a bit
of the landscape of the issue, talk about some of the
ethical challenges that arise, provide you some data from some
empirical work that we’ve done, talk to you a little bit
about both the analysis of these data, but also how we
think they ought to be framed. And lastly, to hopefully
provide a bit of an action plan about what we can
do to move forward. So on the one hand,
what could be simpler than the ethics of child abuse? If you do it, stop. If you see it, stop it. But actually, what may seem
clear on its surface belies much deeper problems in
terms of definitions, in terms of accuracy, in
terms of judgment calls. And I’m going to talk about
each of those in some detail. So when it comes to
underlying questions in terms of definition, the very
notion of what counts as abuse is itself disputed. The former editor of
JAMA, Catherine DeAngelis, tells a story about how when she
was in Baltimore, at Hopkins, she used to teach the police
officers about child abuse. And she would begin
her lecture by asking, how many people
spank their children? And every hand was raised. And then she said, all
right, now, how many of you beat your children? And she said, no
hand was ever raised. And then she asked how
to tell the difference, and the fun began, as she said. And that’s both in
terms of physical abuse, but it’s also what
counts as neglect. There’s an author, an essayist,
named Lenore Skenazy, who a few years ago ignited
this national fire about how old your child should be before
you allow to leave them home alone. In her case, she was
in New York City, and she allowed
her nine-year-old to go home on his
own via the subway. And some people applauded her
for not being a helicopter parent. And other people were ready
to report her for neglect. In the realm of
emotional abuse, it can be very difficult to draw a
line between harsh parenting– if you were in the car with me
and my daughter driving up– actually my whole family–
she would give you a different answer about what
counts as emotional abuse than certainly I and
her mother would. Imminent risk, if you
don’t already know, it’s when something would
have counted as abuse neglect, but for bad luck. So I take a hatchet and I
throw it at my son and I miss. The fact that I had bad aim
doesn’t make it non-abusive. It just means I had bad aim. Sexual abuse in
terms of definitions is a very different animal. Sexual abuse compared to
the other kinds of abuse is typically much
more premeditated. It is really a whole
other can of worms, and so it’s something
to be distinguished. Another kind of uncertainty
that we deal with is accuracy, both in terms
of mechanisms of injury, how did an injury occur,
people’s intentionality. And intentionality matters. If a child comes in and
they’ve got a huge bruise on their forearm, and Dad did it
when he grabbed his arm really hard, it matters whether he
was trying to pull his son out of the way of oncoming
traffic, or something else. And figuring out what people’s
intentions were or are can be challenging. So too, it’s challenging
to figure out whether a story is credible. And I’m going to tell you
several clinical stories in just a little bit. But how you believe and what you
believe is certainly in flux. And then lastly, a
question that people often don’t want to discuss
openly is the value or utility of actually
reporting a child and whether the intervention
of Child Protective Services is actually going to
help this child or not. So too, we have
issues of judgment. What are acceptable
thresholds of risk? We had an ethics consult
come to the hospital. I used to direct the Ethics
Consult Service at Penn State Hershey Medical Center. And it was a physician who
had a 10-year-old child with an immune deficiency
whose parents were planning to take the child on a
mission trip for several years to literally the jungles
of Central Africa. And the concern was, that puts
this child at greater risk, given their underlying
immune problems. Does that count as
an unacceptable risk? Reasonable protections. How far must we go in
order to protect a child? If you talked to my son Dewey
when he was in grade school, he would tell you
at great length how they weren’t allowed to
do anything on the playground. And when I asked
him why that was, he says, because
they’re all moms. All the playground
monitors were moms. He said, if there
was a dad out there, a dad would let them
do all sorts of things. And so gauging what counts
as a reasonable protection for a child is in dispute. And I should say,
before I go on, this is particularly
acute when we have a child who’s
in the hospital, and we think that they
need a particular safety plan before they go home. And Children and Youth
Services says, no, no, we don’t need that. This is sufficient. Whose judgment prevails
in terms of what is an adequate safety plan? A lot of work that I’ve
done, empirical work, is about what should be
the threshold for reporting suspected abuse. And I’ll go into that. And then some more recent
empirical work I’ve done is about if reasonable
suspicion is the front end, the back end is reasonable
medical certainty. When a physician
or other clinician is asked to go into court
and testify that there’s reasonable medical
certainty that the injury or condition was due to
abuse, what does that mean? So sometimes there
can be resolution. If we have better data,
we have better correlates. There’s a woman, Mary
Clyde Pierce, in Chicago. She’s an ED physician. She’s developing an algorithm
for what kinds of bruises are more likely to be abuse. It turns out the most specific
is circumferential bruising of the ankle. Almost nothing causes it, other
than an abusive type of grab. Similarly, if we have experts,
board certified pediatricians who do their work
well, we’re more likely to resolve some
of these problems. Improve communication–
this is the bugaboo. People don’t always listen. They don’t always
ask good questions. And if you develop
relationships with people, it can make things go better. It can improve understanding. You ask a child how
things went, and he says, my dad beats me every morning. You ought to find out
whether it’s in chess or whether it’s with a stick. So we can’t assume
on its face that what we think we understand is
what is actually going on. To that end, if we do a
better job analyzing things, we can often move beyond
the assumptions and biases that we begin with before
we do that kind of analysis. If we have better definitions
of what counts as abuse, we’re going to do better. And unfortunately, if
we had better consensus, because we often don’t,
there can be resolution. Where the ethical
challenges often come in is when we don’t have certainty. We don’t know what to believe. So imagine that a dad comes in. You see a child, and
the child has a rope burn halfway across their neck. And the dad says, you know
what, what happened was is I was in the front yard. I was raking leaves, and I heard
this scream from the backyard. I went in the backyard, and
my daughter was playing horse with the next door neighbor. There was a rope tied
around my daughter’s neck, because she was the horse,
and they were going around. And she fell, and
this is what happened. Do you believe that? Funny story. That’s what happened to me. My daughter, Rosie, was playing
horse with our next door neighbor, Cara. And as soon as I
made sure she was OK– this is almost 10 years
ago– I thought, oh great. She’s going to go
to school, and I’m going to get reported
for suspected abuse. We often are left
with uncertainty. What’s the actual risk of abuse? A few months ago, I
saw these two boys that I’ve been taking care
of since they were babies, 10 years old and 8 years old. Typical well-child checks. They’re fine. And I do the 8-year-old first. I do my physical. I’ll do my history,
everything else. And when I do my physical, I’ll
check out their private parts. And for both boys and girls,
as a general pediatrician for a general exam, it’s
pretty straightforward. I make sure his
testicles are both down. I push a finger up to
see if he’s got a hernia. I say, pretend like
you’re taking a poop, but don’t, and push down. And so I did it with
the 8-year-old boy, and he was fine. And then I went to
the 10-year-old, and I was doing his exam. And at the very end– and
I uncover their genitals for maybe eight seconds. And I said, now, we’re going to
check out your private parts. And we do that
because I’m a doctor, because your mom’s here. And I remind kids about a
thing called a “bad touch.” And a bad touch is if
someone touches you and it makes you feel bad. And it may be because
they’re hitting you, or they’re touching you where
you don’t want to be touched. But if it feels bad,
you tell your mom. That’s as far as I go. And before I could even finish
that, this boy started crying, I mean, inconsolable crying. We couldn’t get him to stop. It took four or five minutes. He just did not want me to
examine his private parts. And this is a big
red flag, right? And I talked to his mom
right there in the room. And I asked a
bunch of questions. And there were no risk factors,
but this was really unusual. So does that rise to the
level of needing to report? What’s the actual risk of abuse? I mean, it could be
he’s just really shy. It could be that maybe he’s
less endowed than his brother and he doesn’t want
the comparison. It could be that
he’s been hearing about what happened with
Sandusky, and he’s fearful. It could be that he
has an anxiety disorder or that he’s
internalized something, and he’s struggling with it. Or it could be abuse. I have no idea what
the likelihood of abuse is in that kind of case. So these are the kinds of
situations where we don’t have anything resembling certainty. We also don’t know how to
often interpret the guidelines. And I’m going to tell you about
a case that got me started. This is what got me started
almost a dozen years ago. I was covering
acute clinic, which means I was supervising
residents and medical students. And I was the attending, so I
needed to oversee everything. And I went into a room, and
it was a Friday afternoon, because it’s always
a Friday afternoon. And there was a dad
with his two daughters, and they were 8 and 6. And they were there for a cold,
which always makes you go, hmm, why is an experienced
parent here for a kid with a cold? And it was Dad. Dads usually don’t
bring in their kids. It’s usually moms. And so a few things
that– you’re late. He’s a friend of mine. [LAUGHTER] And when I came in the room,
there were these two girls, both with colds and their dad. And I happened to notice
that this girl had a really bruised, swollen nasal bridge. And I’m a curious guy, so I
always ask, so what happened? And if Dad had
said, well, you know what, we were playing
ball in the front yard, and threw the ball. And right as she was
about to catch it, her mom opened the screen
door, and she turned to look and it hit her right in the
nose– I might have been done. But that’s not
what the dad said. What the dad said
is, we don’t know. All right. We don’t know what happened. She just came into
our room at night. Her nose was bleeding. And she sleepwalks, so we
figured she hurt herself. And I’ll look through the chart,
no unusual emergency department visits, no documented concerns. But doctors are often terrible
about documenting things. And even if they do, you
can’t read their writing. And the dad was impatient. And the more I
asked, the angrier he got, at one point saying,
we don’t beat our kids, if that’s what you’re asking. I said, well, that’s
not what I’m asking, but it’s part of my job to
make sure kids are safe. And I noticed that Dad
had his hand bandaged. And I’m a curious guy, so
I asked him what happened. And he had 78 stitches
about 10 days ago because he’d almost cut his
hand off in a band saw accident. And I know that I’m less
patient with my kids when I’m in discomfort. And so it all sort of
figured into my thinking. So there’s the case. And I went to my colleague, who
was their regular pediatrician. I’d never met them before. And my colleague,
who’s an otherwise very good pediatrician, said,
they’re a good family. They wouldn’t hurt their kids. And that is just ignorance. Almost nobody wakes up
in the morning and says, I think I’m going
to hurt a kid today. They get frustrated. They don’t know what to
do, and they lash out. And the things that
predispose to abuse, if you cut across everything
else, it’s stress. It may be financial stress. It may be social stress. It may be that the kid
has a chronic illness. It may be the kid is
being a pain that day. It may be that parents
have untreated ADHD, or that they have alcohol,
or drug abuse problems. There are any number of
things, but stress is what cuts across everything else. And I went to the
nurse and the nurse said to me, “This is a
family that’s on the brink. The slightest push could blow
this family out of the water.” That was a direct quote. I remember. And so the question
is, does this rise to the level of
reasonable cause to suspect? So let me ask you
these questions, if you’d raise
your hand for them. How many people
think it’s possible that this was due to abuse? Go ahead and raise your hands. All right. Come on. Keep your hands up. So how many people
have some suspicion that this is due to abuse? No hands have gone down. How many people would say that
you have reasonable suspicion? That’s the threshold
for reporting. At least half, 2/3
of the hands go down. How many of you believe
that this was due to abuse? Oh, almost all hands– all
but a couple of hands go down. And that is an exercise
in the power of language. Language matters. It matters because how you
conceptualize something determines whether
you may report it or you’re not going to, right? And this is a big deal. It shows that words matter. So the laws around the country
use very different kinds of words. This is arguably,
I think, the best. This is from the
California Penal Code. it says, “It is objectively
reasonable for a person to have reasonable suspicion,
based upon facts that could cause a reasonable
person in a like position, drawing, when appropriate, upon
his or her training experience, to suspect abuse.” This is a normative standard. And it’s what most
people would do. Now, you saw the difference
between suspect and belief. And if you look at laws
in the United States, the yellow are those that
use some variant of belief, and the blue use some
variant of suspicion. And there are 11 different
statutory wordings. And guess what, it actually
makes a difference which wording you use. So a student and
I did an analysis of these different
statutory wordings. And it turns out that you get
different reporting rates. So the total is
the number of kids who are reported per 1,000. All right, so Reason to
Suspect or Cause to Suspect, both of those, 41.9
children per 1,000. And so you see some variation. I’ll make it a little
bit more straightforward. If we divide them into
Suspicion versus Belief, different categories, you
see there’s a difference. Those states that
use suspicion as part of their statutory wording
have 50.4 kids per 1,000 that are reported, versus 41.9. And of those 50.4,
29.3 of those kids, they say, yeah, we’re
going to investigate this. And 21.1 say, no, they’re not. And you again see this is
not just general public. This also includes
health care personnel. And this matters. I mean, we’re talking about,
between these two thresholds, a difference of over 600,000
referrals and 324,000 screened-in calls. And an investigation
for child abuse can be about as
gentle as a tax audit. And what happens depends
upon what county you’re in, at least in Pennsylvania. In some counties,
they’ll whip a kid out of a home in a heartbeat. In other counties, your kid has
to have a spear sticking out of their head before
they’ll do anything. So it matters, and it
matters to families. And so, we’re often
left with uncertainty. And in a survey that we did
of Pennsylvania pediatricians, we asked people, given the
practical implications, will intervention help? They don’t know. What people say is,
well– one person said, I have to be sure
enough to warrant the disruption,
the loss of trust, the decrease in the
relationship, and the turmoil that I put a family through. Because these things
actually make a difference. And the problem is that the
system is deeply flawed. The child protection
system, if not broken, is enormously bent out of shape. And it’s the only
system we have. So it’s the only system
we can rely upon. And to not rely upon the system
is to throw a child’s safety to the winds. And it matters. It seems to me–
and this is why I began to take this
question seriously– that I need to be prepared
for this girl coming back to my office in 10
years and saying to me, why didn’t you report me? You were the only person who
ever saw physical evidence of what my father did to me. It took me 10 years to
escape his clutches. Why didn’t you report me? I need to have an
answer for that. I also need to have an answer
if she comes back in 10 years and says, why did you
do that to my family? It blew us apart. I haven’t seen my
sister for four years. Why did you do that to us? And it seems to
me, conceptually, all of us who are
mandated reporters of suspected abuse,
which is a lot of people, we owe it to the children. We owe it to the families. We owe it to ourselves
to have a conceptual and an applied understanding
of what this is. Now, the problem is, of
course, is that lots of things put kids at risk. And there’s a certain role that
luck plays about who gets hurt and who doesn’t. Kids will get on your last
nerve and jump on it just to see what happens. One time, when my kids
were a little bit younger– they’re 14 and almost 11
now, and my son Dewey, he was about four and
a half or five then– and we were running late. Because we always run late,
and it drives me crazy. And I had kids in the
car, and I need to go. And Dewey was on the floor
searching for something, Fruit Loops, or something. I don’t know what. And I said, Dewey, you’ve
got to get up in the seat. And he said, I’ll
just be a minute. I said, Dewey, we’ve got to go. He said, I’ll just be–
I’ve just got to find this. And finally, I said,
Dewey, I have a meeting. I’ve got to get going. And he sort of threw it back
over his shoulder, that’s OK, you can wait. That’s what kids will do. Another time, we
were running late. We were going to
someone’s house, and we were all dressed up. I remember exactly when it was. We had just gotten
what for us was a brand new car, a nice Honda
Pilot, the first car I ever I had that had a
CD player in it. I was excited. And we were running
late, and Dewey was looking for something,
and he didn’t want to go. So he’s– he was
four and a half, and so I do what
any good dad does. I picked him up, put
him under my arm. We’re walking out the
door, and he grabs a hold of the door knob with one hand. He grabs a hold of my
shirt with the other. And as I go, he yanks
this button off. And there was no time to change. I mean, we were running late. These people were expecting us. I was furious. We went to the car, and
I tossed him in the car. I mean, I didn’t
throw him overhand. I just tossed him in the
car, and I slammed that door. And as soon as I slammed that
door, I thought to myself, oh my goodness, what
if he had reached out? If he had reached out, it
would have broken his arm in four places. It would have been an
open comminuted fracture. He would need at least
two or three surgeries. And every person would have
been right to report me for suspected child
abuse, injuring my son in a fit of anger. Why didn’t I get reported? Because of the luck that
my son did not reach out. We had a young, newly minted
child abuse pediatrician at our institution
some time back, who was telling a
courageous story about how there was a
four-year-old boy who had gotten a broken arm. And the way he’d gotten
a broken arm was he was riding a scooter,
one of these scooters. And he and his family, this
is part of their culture. And they go riding every
weekend at these rallies. And he was going
40 miles an hour. He was wearing his helmet. And he’s four and a half years
old, and he broke his arm. And the parents were
totally non-apologetic. I mean, this was just
something that happens. It’s part of the sport. And the Child Protection Team
was increasingly frustrated. And eventually, they
threatened the family that if they didn’t
sell this motorcycle, they would report them
for abuse and neglect. And so the family did. But who did they sell it to? They sold it to somebody else
with a four and a half year old. There are rallies where four
and a half year olds race. So does that mean we
should be justified in going to these rallies
and charging everyone whose kids have not
gotten injured yet for abuse and neglect? They haven’t had the bad luck
that their kid’s been injured. I mean, how do we figure
out how luck makes its way into our understanding? Now, not everybody views this. Some people, their world
is much more simple. The world is a world
of black and white. So when we ask people what
reasonable suspicion means, we had one person
who wrote– I’m projecting this
was their voice– that “reasonable suspicion
constitutes any suspicion. It’s ludicrous to quantify
any percentage before you report any incident.” And their cousin then also
wrote, “Reasonable suspicion is just that. If there’s one iota of
suspicion, I report.” Wow. It seems to me you’d be
reporting a lot of people. I think a better answer, and
this was somebody else’s. My colleague, Cheryl Dellasega
and I wrote this paper a couple of years ago. “That it is a reasonable,
not far-fetched or outlandish possibility that the
injury occurred by abuse. Not just that it’s in a
long line of possibilities.” This person, this physician,
wrote, “Cancer can always be on the differential or on
the rank order list for a kid with fever. But most of the time, it’s
not reasonable to think it is a likely cause.” And I think these things require
careful thought if you’re being intellectually honest. Because the first
grader who says that she fell from the monkey
bars may have been abused. So, too, the
happy-go-lucky preschooler who’s suddenly always sad. Or the seventh grader who
is inexplicably withdrawn. I mean, I once saw a 15-year-old
boy– Friday afternoon, always– who came in for
a 10-minute appointment for a sore throat. And I went to swab his throat,
and he was really anxious. And well, you’re getting
to sense I’m a curious guy. So I turned to Mom, and I said,
is he always this anxious? And Mom said, well, he has
been for the last few months whenever he’s
getting ready to go visit his dad for the weekend. And over the next 10 minutes,
to my shock and mom’s horror, we learned that, at least
according to the young man, that Dad had not only been
physically abusing him. Dad had been
sexually abusing him. And the last time he
was at Dad’s house, they were fishing in the
river, and Dad held his head underneath the
water for a minute. That was from a kid
who’s unusually anxious. So if you’re not looking, you’re
not asking the right questions, you’re not going to
find these things. All right, so when there’s
one iota of doubt, report, you’re going to be reporting
a lot of your practice, a lot of the kids you see
in your class in school. You’re going to be
reporting a lot of people. There’s enormous
variability in terms of individual interpretation. And this creates
real ethical problems in terms of how we
allocate scarce resources, and how we treat
parents, and if we do so in a fair, just manner. And frankly, how we
protect children. If you don’t have
standards– gesundheit– it’s a hard business to protect
kids in an inappropriate way. So here’s some
empirical data about how do professionals understand and
interpret reasonable suspicion? And this was my question. This is what I really
wanted to understand. And I didn’t want to
know what are the signs and symptoms of abuse. Or what do they think the
risk factors or red flags are. And I, along with
various colleagues, did a variety of studies
asking these questions. And these are the various
populations of people. RR stands for Response Rate. And you can see
quite readily when I discovered that a $5 gift
certificate to Starbucks will get you a
better response rate. And I wanted to
abstract away from all the clinical relevant issues. I’ll give you the
physician version of this, which was, imagine
you’ve encountered– actually, no, this is everybody. Imagine you’ve
encountered a child who has an injury or condition
that could have been caused by abuse, and you’ve
gathered as much information as you feel is
reasonable to do, given the sources readily available. On a list of possible
explanations, how likely would
abuse have to be? So if I come in, and you
say, how you doing, Benny? I say, I’ve got a headache. My headache could
be due to stress. It could be due
to lack of sleep. It could be a migraine. It could be I got
hit by somebody. It could be I have increased
intracranial pressure. I could have a brain tumor. I could have any
number of things. And one of those things,
you’re going to rank. If you’re smart, it will be
stress and lack of sleep, as those are the two highest
ones on this rank order list. And so what I want to know is,
how high on this rank order list does abuse have
to be before you feel you have reasonable suspicions? Would it have to be most likely? Could it be the
least likely thing? And we just arbitrarily cut
this off at 10, or somewhere in the middle. All right, and so this is what
we found– all over the board. Now, it’s sort of a busy slide,
but you see the first group that I surveyed were
Pediatricians, as the hot pink, Nursing Students, Nurses,
Day Care Providers, Community Mandated Reporters, Abuse
Experts, and Undergrads. All over the map. As many people said
that abuse would have to be second most likely
as seventh most likely. So there’s no standard at all. And the reason this matters
is that sometimes people say, well, if you think
it’s a possibility, report. Well, how likely a
possibility must it be? And you notice, by the way,
that for child abuse experts, these are clinicians
and researchers in the field of child
abuse, the help for society, they’re no different than
really anybody else on this. So this is I don’t think a great
way to conceptualize the issue. I’d rather conceptualize
it in this way. I use the same stem. You’ve encountered
a child, yada, yada. How likely would
abuse have to be? How great a likelihood
would it have to be before you felt you
had reasonable suspicion? Would it have to
be only 1% likely? 99% likely, or
somewhere in between? Mark an X on this line. And I’d encourage you
all to take five seconds. Just with your finger,
write on your palm where would you put this? Would you put 50%? 75%? 1%? I’m telling you,
you can’t be wrong. This is what we found–
all over the board. As many people put
it at 30% as 90%. So this means that
whether I’m going to get reported for
my daughter playing horse with our next door
neighbor, Cara, on a given day, depends entirely who I run into. And this is a problem. It’s a problem in terms
of protecting children. It’s a problem in terms of
the ethics of people being treated fairly and equally. It’s a problem for us who
are mandated reporters, because we’re told
we have to report if we have reasonable suspicion,
and there’s no standard. So that reasonable
person standard is gone. There is no reasonable person. So that’s the problem
that I was engaged in. So I did this conceptual work. You see, again,
abuse experts are no different than anybody else. And then we did this
paired relationship. So if you start with
that rank order list, the number one thing
on that rank order list might be 50% likely, or
it might be 1% likely. And number 2 might
be 0.9% likely. So you have no idea
where you start. But if you start this way, you
can make some correlations. Because the question of
whether something is abuse is a binary issue. Either it’s abuse
or it’s not abuse. There’s nothing else. It’s like arson. A house burns down
either because of arson or it’s not arson. They may have faulty wiring. It may be old. It may be cluttered. It may be more prone to be put
on fire, but it’s either arson or it’s not. So the list of things that
are possible explanations, all together, have to add
up to no more than 100%. Because it’s either
abuse or everything else. So if someone says that to
be reasonable suspicion it has to be greater than 50%,
then all the other possibilities have to be less than 50%. So if I’m being
logically consistent, if I put it has to
be greater than 50% estimated probability to
be reasonable suspicion, I should put it as number
one on my rank order list. And if I say it’s greater
than 33%, at least second. And you can do the math. And I’ve done this math many
times to make sure it’s right. And here’s what we found. The blue are those that
are logically consistent. So let’s start in the
middle, 326, right bang in the middle of the circle. So those 326 people
said that abuse had to be 50% likely before
it was a reasonable suspicion. But it could be the fifth most
likely thing of causing it. Down towards the corner,
there are six people who said it had
to be 90% likely, but it could be as low as
the ninth most likely thing. So you could say people are
just illogical and inconsistent. Or another way to
interpret this is to say– and this was across all
these different populations– and there was no difference. It was almost– for
every population, it was roughly 13%
to 15% of people were logically consistent. And what this tells
me is that how you conceptualize
reasonable suspicion, again makes a difference. Because you’re not going
to have the same answer. If someone says, well, if it’s
in the top 10 possibilities, that’s reasonable
suspicion, you’re going to have a very
different answer than if someone says, well, it
should be more likely than not. Or there should be at
least a one in three chance that this is abuse. So that’s the top end. That’s reasonable suspicion. So all mandated
reporting is triggered by someone having
reasonable suspicion, so it’s a big funnel. The end of the funnel is
reasonable medical certainty. When a case goes to court,
and someone has to say, yes, I think there’s
reasonable medical certainty that this is caused by abuse. And our hypothesis was
adapted from the views of a famous existentialist,
Humpty Dumpty. “when I use a word,
it means just what I choose it to mean–
neither more nor less.” And we got this because this
was the advice, or the response given to one of my
colleagues, Mark Dias, a pediatric neurosurgeon, when
he asked one of the lawyers in the court where he
was about to testify. They said, but the definition
of reasonable medical certainty is just what you define
it to be, Doctor. So we asked people, how do they
interpret this, presumably, diagnosis? And we did a listserv
survey of various listservs that deal with– sorry about
this– with child abuse. And what we found
among these groups, we had just under
300 respondents. Go back there. Almost half were child
abuse pediatricians. We had forensic pathologists,
neurosurgeons, and so forth. And these were people
who, 95% of them, overall, had testified in court. Almost 2/3 of them had
testified in more than 25 cases in the last five years. So these are people who
actually are going into court and testifying. Almost half of them had
been directly involved in the care of the child
for whom they had testified. Others were
independent witnesses. And, interestingly,
fewer than half said that they had ever been
trained about what constituted reasonable medical certainty. So one of the first
questions we asked was, how comfortable are you
with this terminology? Because doctors don’t talk about
reasonable medical certainty. They’re asked to talk
that way in court. And what we found was
that only less than 40% had any level of
comfort with this term. Over 40% said that they were
either somewhat uncomfortable or very uncomfortable
with the term. We asked how many had ever
received formal training on reasonable medical suspicion. Interestingly, pathologists,
more so than others, said that they had, which
is sort of interesting. We asked them what
was “your” threshold. If you’re going to set the
mark, where is that mark? And you see it’s again
all over the board, not near as much as the
reasonable suspicion. But again, it’s interesting. Actually, I don’t think I’ve
included these data here, but childhood pediatricians
set the threshold the highest. But one of the things this
means is that if I’m on a jury and someone says reasonable
medical certainty, they may be meaning,
well, if I flip a coin, it’s more likely
than not child abuse. Or they may be meaning, oh,
there’s a 99% likelihood that it’s abuse. And I can’t but imagine
that a judge or a jury would be influenced
by whether or not– which threshold they chose. Because it’s framed
almost like a diagnosis. And interestingly,
there was no correlation for where people
set this threshold with any demographic factor–
their age, their training, their field, nothing. Now, interestingly,
though, it was influenced by what type of court. So in those cases where there
was a criminal case versus it was a civil, a family
case, a quarter of people, of these experts,
changed the threshold for where they would set
reasonable medical certainty, which is sort of odd. Because again, if
it’s a diagnosis, that shouldn’t
make a difference. And I would argue that
these findings tell us more about these
people than they tell us about what this actually means. And the absence of any
significant correlation suggests that training
and experience also have little bearing on how
reasonable medical certainty is understood, and should leave
us wondering what it means– reasonable medical certainty,
that is– as a threshold, other than personal opinion. So what we see depends
really on who we are. Now, on one level, this is
not necessarily surprising when you have competing
values at stake. And when there is
not certainty, there are competing values in place. So we asked people,
which is worse, mistakenly leaving a kid in a
home where abuse is occurring? Or is it worse to
remove a kid from a home where there is no abuse? And we found that seven out of
eight people– sorry, six out of seven people said
it’s worse to leave a kid in an abusive home. When we asked, which is worse,
convicting an innocent person by mistake or letting a
guilty person go free? We found that seven
out of eight said convicting an innocent
person is worse than letting a guilty person go free. But when we really posed
the dilemma, which is worse, putting an innocent
person in jail or leaving a child
in an abusive home? It was pretty
evenly split, 50/50. Now what’s challenging is
that what we’re working with is an enormously
overwhelming problem. And it’s something
that’s chameleon-like. You may have seen these. There’s this guy in
China who paints himself like a background. He actually paints his
body and his clothes, and he stands in front of
things and gets photographed. And child abuse is a
little bit like that. It’s everywhere. So when I get asked
in interviews, as in my capacity as director
for the Center of Protection of Children, what do I think
about the scandal at Penn State, and so forth, I say,
for people in this field, we know that this kind of
behavior, what did happen and what didn’t happen,
happens in every community all throughout the United
States and throughout the world. What was different is that
this got people’s attention. People in the field of child
abuse and child protection have been screaming about
this issue for years. And now there’s a platform on
which we can focus our efforts to make a difference. So this is everywhere. And if you’re not
looking for it, you’re not going to find it. And part of the problem is
we have inadequate data. We don’t have
sufficient resources. We have poor predictive value. And we often lack the time,
the personnel, and the tools to actively find the truth. And the problem is that
this is a big deal. Because if you miss it, then
a child remains in harm’s way. And if you’re not looking,
you won’t find it. If you don’t know what to
look for, you won’t find it. And here’s the man, right there. You can see him now. Because people blend in
with their surroundings, and certainly, child abuse does. Now the challenge, of
course, is how do you define where these
thresholds ought to be? And I think it’s
a social calculus. It depends upon
what we as a society believe, how much we’re willing
to spend to protect children, what kinds of harm we’re
willing to tolerate, how much harm we’re
willing to tolerate. And there is no objectively
correct answer to this. It’s also the case that
improving this system can be very challenging. It’s a behemoth. The things that
we have to change, everything is connected
to something else. Now especially here
in Pennsylvania, this is a big deal because the
laws that are on the horizon have looming problems. When it becomes the case
next year, as it will, that if you’re a professional
who’s a mandated reporter, and you don’t report
when you should have had reasonable cause to
suspect, your license is going to be taken
away, a lot of people are going to report
a lot of instances that– well, I don’t know
this kid or this family. If it’s between fulfilling my
professional responsibility to losing my license,
well, sure, why not? I’ll report them. It’s not clear that it’s
going to help anybody if the reporting rate
triples or quadruples and we have a
signal-to-noise ratio that results in the
system being overwhelmed. And it’s calling upon
the wisdom of Dumbledore that sometimes we must face
a choice between what is easy and what is right. And there’s a price with doing
the right thing sometimes. And this is where
hard cases make bad law, as is famously said. So we’re facing conceptual
and practical dilemmas that don’t have
necessarily a ready answer. So one question is,
well, what can we do? A colleague and I who’s down
at University of North Carolina Charlotte, Sharon Portwood
and I wrote a conceptual paper about both raising this
issue from policy standpoint, but asking how we should
ourselves operationalize this. And I’ll break this
down a little bit. So I think what we
can begin with is just the clear-cut question
of is, are you certain that abuse occurred? And if the answer
is yes, then report. If the answer is no,
then the question is, well do I have
reasonable suspicion? And reasonable suspicion
is the umbrella terminology for all those 11
statutory thresholds. And if the answer to
that question is yes, then go ahead and report. If the answer is no, I
don’t have reasonable cause to suspect in Pennsylvania,
then don’t report. The $64 question is, how
do we figure out when I have reasonable suspicion? And what I and
Sharon came up with was that there
are some questions that we can ask ourselves. And this is a partial–
this is one part of it. I think that the key
part is to recognize that suspicion is a feeling. And so one of the things
you want to ask yourself is whether you have
confidence in the soundness of your observations. Have you spent enough
time observing? Are you familiar with
what you’re looking at? You see all these round scars
or blisters on a child’s back. Do you know that those
blisters are not– what’s the– cupping, which is a common
tradition in Asian medicine, to remove evil humors, so
to speak, from the body. I mean, are you
familiar with this? Do you know that that
bruise on the back is definitely not
a Mongolian spot? We had a couple of residents
at our medical center. We had a little baby that
got taken away from them for three days because
someone was ignorant and did not know that that
mark on their lower back was a Mongolian spot
and not a bruise. Are we clear about
what we think we heard? The call that my
father beats me. Beats me at what? Did you observe it firsthand? Or was this one of those
games of telephone? By the time it got
to you, it’s morphed. Do you have confidence
in your judgment because of your own experience and
the fit of the explanation? And I’ve argued
elsewhere that I think that suspicion is a feeling. Now oftentimes it’s thought
of as a weak form of belief. I think that’s wrong. I think it’s a feeling that’s,
as I say, imbued with meaning. And what that means to me is
that it’s a sensation you have. It’s a feeling in
your gut, in a sense. And the meaning comes into
play if you walk into a room and there’s this horrible smell. And you say, what is that? And someone says, oh, it’s this
food, that it turns out you actually quite like. You smell it again,
and the smell actually changes because of the
meaning that the smell had. So the sensation itself is
varied based on meaning. And I think that reasonable
suspicion is exactly that thing, or suspicion is. It’s based part upon meaning. So how do we approach this? Well, that’s a little
bit of an algorithm. I think part of it is there’s
a certain kind of a stance we take, an attitude, being
aware, a sensitivity that this is an issue, and that
we need to understand what needs to be understood. That we approach these
situations with humility. I consternate the residents
who work with me periodically there, reminding them
that how little we know. 95% of the universe
is dark matter. We don’t know what it is at all. So let’s be humble about
what we assume we know and what we don’t know. In that regard, we need
to be open to possibility, open to possibility that
a wonderful family we like and trust might
act out of anger. Or that the person
we think is great with children has been sexually
abusing them his entire adult life. And that requires a certain
intellectual honesty. And by that– and I put
a lot of stock in this– is that intellectual
honesty demands that we acknowledge where
our reasoning is weak, where we don’t have good
evidence for what we believe or what we’re concluding, and
that we acknowledge our biases. In my house, there’s
a room– there’s a Yiddish word called mishigas. And mishigas means– it’s
something you believe and you sort of hold to
it, but you don’t really have any good reasons. So an example is, if you
believe– any of you– that it’s better
for the toilet paper to come from the top or the
bottom, that’s mishigas, OK? And in my house, if you
are willing to acknowledge that what you want– where
the forks are supposed to be placed, or whatever– if
you acknowledge it’s mishigas, you get your way. But you have to
acknowledge it’s mishigas. And so part of what this
demands, in a sense, is that you acknowledge
where your biases are. I think there’s also
a process of discovery that we need to pay attention
to, how we respond to people. it’s easy to jump on
people and judge them for what they’ve done. What about my son and
the door that I slammed? How would I be judged now
versus if he had actually reached his hand out? We need to think about
how we ask questions. When I’m in clinic, I don’t
say, so do you understand? It’s an intelligence test. If I say, does that make sense? Because I don’t always
explain things real well. I take it on me if I
didn’t communicate well. I don’t put it on them. And learning to
ask good questions is important, and relying
on your resources, taking a systematic
approach to things. Talk to your colleagues. Talk to people who
are responsible. Be prepared for resistance. People say, that’s
not a problem. They’re fine. This is a good family. They wouldn’t hurt
their children. And be transparent. Acknowledge people
what you’re thinking. When I report people
for suspected abuse. I tell them that I’m going to. And I tell them, I don’t
know what happened. I wasn’t there, and I’m worried. And we need to invoke
the system that is set up to protect
children, even though it’s a flawed system. And that they need to
be prepared for people accusing them and judging them. And that the more honest they
are, and the less defensive they are, the better the
process will go for them. And I come back, as
always, to this business of being intellectually
honest about that. And people will challenge you. You’re about to put me
through this horrible ordeal, and my family. Why are you doing that? And you can hide behind
terminology and all sorts of things. Or you can be
intellectually honest and acknowledge
what you’re thinking and what your uncertainties are. And it requires a
certain, I think, strength of mind and of
purpose to do that well. And if what we’re after
is to protect children, there are a whole range of
these conceptual and practical matters that are unresolved. And yet, I think we
need to take a stand. These are the Moeraki Boulders. They’re on the South
Island of New Zealand. And they only exist on this
one beach and one beach in South America. And no one knows
exactly how they form, but it’s clear that some
are more stable than others. So where you take your stand,
you want to be careful. Because some foundations
are better equipped to support your
weight than others. And thinking through
these conceptual issues, and thinking through their
practical implications is, I think, very important. And with that, I’ll stop and
ask if people have questions. [APPLAUSE] -I wonder if you could speak a
little bit about that broader conception of
preventing child abuse and how people in
your field think about that in terms
of if– just if you have any thoughts
on what it might mean to take some of the steps. And whether you
think it actually might be better to start
talking about those broader preventative issues, as opposed
to the responding issues, not to the exclusion
of, but in addition to. -I mean, your point is just
perfectly well-timed here. Everything you say is
exactly on the money. And the challenge is who
addresses these issues and how? There’s a wonderful paper in
the Journal of Pediatrics, the flagship journal for the
American Academy of Pediatrics, about all the things that as a
physician, as a pediatrician, you’re supposed to do in
terms of anticipatory guidance to help people at any given age. And if you did them
all, a well-child check would last three and a
half hours for every child. So the question
is, what do you do? And I think that there are
some practical answers. And I will tell every parent of
kids up through at least school age, I’ll ask them about
are they getting time to themselves? Where someone else takes
care of their child, and they get to get out and
just talk, go for a walk, go to Applebee’s for dinner. Because they’re going to have
different ideas about how the child should be raised. And if they want to
know the right way, I’ll tell them it’s my way. And most of us think that. And so they need time
to sort out differences. They need time to invest
in their own relationship, because presumably, they
didn’t come together just to have children. So I think there are a
whole bunch of those things and none of them are sexy. None of them will
grab a front page. I can make them humorous. I’ll write people a prescription
to go out at least once every three weeks. And usually there’s
one partner who needs that more than the other
person to show the one and say, the doctor said,
we need to go out. So those kinds of preventive
things are really important. And I think that
the challenge is to link what is good parenting,
being a good family member, being a supportive friend,
being a supportive community. How do you take those
things and you link them with a problem,
which is child abuse? And I think that you’ve
got to find the lever. Nowadays, the lever is
Penn State and Sandusky to move things forward. My center, the Center for
the Protection of Children, was a direct
outgrowth of something that has been happening in
communities all over America. So why? Well, because it got
someone’s attention. One of the ways you get
people’s attention in America is by talking about money. And you’re looking at the costs. And the direct costs
alone from child abuse– annual direct costs–
are $70 billion. The indirect costs add
about another– they estimate at $200 billion. And when they do studies
of people who are in jail, it turns out that about
75% to 80% of them will admit that they
were abused as children. How do you get people to
invest in those mundane types of processes? That’s the big question. But I think your
point is well taken. And when a mom comes in four
times in the first three weeks of their child’s birth for
things that are nothing, and some people think,
why is she here? And other people should
be thinking– I mean, many people do–
what’s going on here? Is there a point of stress? Because if you can relieve
that, if you can help someone recognize that children cry. The average six week old
baby cries three hours a day. And if you can turn
the noise down, it’s a lot easier to take. And those are called earplugs. I’ve looked for earplugs that
get rid of all the noise. They don’t exist. Or that you teach a parent
that if their child cries for an extra hour and a
half on a given night, they’ll be just fine. But they’re better off
walking away for 20 minutes to gather themselves
than shaking a child and injuring it forever. Those kind of preventive
things are really important. And building them into
the daily practice of all the other
things we do, it’s a challenge to get
people to recognize the antecedents of a crisis
and how important they are. But your point is
perfectly taken. We actually didn’t
ask about reporting. We asked about how
they understood the concept, because the
concept drives reporting, but only partially. There are plenty of studies
that have demonstrated– there’s actually a guy up here
named King– I’m forgetting his first name– who’s done a
bunch of this work in the 90s– I think he’s in psychology–
who showed that fully a third of pediatricians and
almost half of social workers, when surveyed, said that
there are times when they had reasonable cause to suspect
and they decided not to report. And the reason they
decided not to report was because they didn’t think it
was in the kid’s best interest. Or they didn’t think
it would do any good. Or they didn’t want to
ruin their relationship with the family. Or they thought they could
do it better themselves. So I’m trying to–
as a philosopher, I’m interested in the concept. As a pediatrician and
a member of society, I’m interested in
the impact it has. But the relationship between
the concept and how people act, that’s a separate
question– related, but definitely separate. -So thank you for
the terrific talk. And it brings to mind
an [INAUDIBLE] debate in another sphere,
but it’s related. So in criminal law,
the standard of proof for a criminal conviction is
beyond a reasonable doubt. And judges have many times
tried to explain to juries what on earth beyond a
reasonable doubt means and what a reasonable doubt is. And many of them have
truly botched it. Is there some empirical
data in that field that would be relevant to and
of help to you in your work? -Yes, [? Kathleen ?]
[? Kagehiro ?]. She has written– she used to
be at the Temple University School of Law. She is out in private practice
now in Virginia somewhere. And she took college students,
and she gave them cases and gave them instructions,
either in verbal probability– likely, probable, et cetera–
or in numerical probability. And, as it turns out, there’s
an enormous literature among cognitive
psychologists about this. And these people are either
the most generous people in the world or the loneliest
people in the world. Because when you
email them, they send you back pages and
pages of emails in response. And there are some people at
the University of Illinois, [INAUDIBLE] and others that
I’ve corresponded with. And I actually tried to
track down [? Kathleen ?] [? Kagehiro. ?] I
could not find her, and I called a
co-author on her paper. And I called him at his
law office, and I says, you have any idea where she is? And he said, I think
she’s next door today. So I actually spoke
with her about this. And what she found, and
what cognitive psychologists have found, is that if you
give people information, they prefer to have it,
and they do a better job if you give them that kind
of information in terms of numerical probability. You say, what’s the
lowest standard? The one where it’s
more likely than not? -[INAUDIBLE]. -Pardon? -[INAUDIBLE] -No, the American language–
the US language is– I’ll think of it. I just present– I
have it in my data for the reasonable
medical certainty. But it’s more likely than not,
so if you say greater than 50% likelihood, people
do a better job in the court of law
adhering to that standard. Similarly, if you say
you have greater than 70% or 75% for substantial evidence. And if you say greater than 95%
for beyond a reasonable doubt, people do better. They don’t like to
give the information. So most people like
to receive information in terms of numerical
probabilities. They prefer to give
information in terms of verbal probabilities. So I think the way to go is to
use numerical probabilities. And people will
usually say, well, people are terrible at numeracy. But the examples I’ve
given in the past is about if someone says
there’s a 5% chance of rain, you don’t bring your umbrella. If there’s a 95% chance of rain,
everyone brings their umbrella. And it turns out, if you
look in the literature, weather forecasters
have actually tested this about whether
people who hear weather forecasts– this paper came
out in 2010– whether do they understand and can they
make appropriate daily activity judgments based on
numerical probabilities for weather forecasts? And the answer is yes. So if you ask people
to manipulate numbers, they’re not very good. But if you ask
them to understand numerical probability,
they know that when Barry Bonds has a
330 batting average, that he’s getting a hit
every one in three at-bats. -Can I just [INAUDIBLE]? -Sure. -If numbers are that
[? diverse, ?] do you ever [INAUDIBLE]
what kind of numbers you should give
people [INAUDIBLE] in the reasonable
suspicion field? -I’m often pressed for that. I’m hesitant, because I
think there’s your calculus. What we need is we need a set of
well-done, broadly disseminated experiments that
look to see what’s the effect on the system, if
you set the threshold at 75%, versus 50%, versus 25%. We know– we have some evidence
that it makes a difference. There are a couple of
graduate students, Black and [? Pfleger, ?] who in the 90s
did their PhD dissertations on this with school
teachers, where they either used verbal probability, or
they said 25% likelihood. Well, if they used
25% likelihood, they found that school teachers
were 2.3 times more likely to report physical abuse
in the same scenario, and over 3 times more likely
to report sexual abuse. So we know it
makes a difference, but should it be set at 25%? 50%? 55%? 70%? I don’t know. And I think part of it depends
upon which population you’re talking about. Because if you ask
a pediatrician, you may get a different answer
than if you ask a school teacher or someone else. So I don’t give an answer. I mean, I have my
own internal answer. But I do a lot of work on this,
and I think about it a lot. Well, thank you very
much for your interest. [APPLAUSE]

Leave a Reply

(*) Required, Your email will not be published