09/29/2026
When Systems Fail Families, Why Do We Evaluate the Parents?
Several times this week alone, I have watched the same pattern play out in different forms.
A service is withdrawn, and the explanation shifts toward a parent’s supposed lack of investment or dedication.
A family does not engage with an offered intervention, and the file begins to describe them as disengaged.
A parent repeatedly says, “This isn’t working for my child,” and somehow the question becomes whether the parent is willing enough, motivated enough, consistent enough, or compliant enough.
What strikes me is how rarely we seem to entertain the most neutral explanation:
Maybe the service was a poor fit.
That possibility should be unremarkable. Not every intervention will work for every child. Not every service model will be accessible to every family. Timing matters. Relationships matter. Neurotype matters. Culture matters. Disability matters. Family capacity matters. The assumptions built into the intervention matter.
And sometimes, quite simply, what we offered was not what this particular family needed.
Yet our systems have an extraordinary capacity to transform information about fit into information about parents.
“Disengaged” is an interpretation, not an observation
Consider the word disengaged.
It sounds objective. Clinical, even.
But what did we actually observe?
Perhaps a parent stopped attending appointments.
Perhaps they declined a particular service.
Perhaps homework was not completed.
Perhaps they did not respond to several phone calls.
Perhaps they attended inconsistently.
Those are observations.
“Disengaged” is the meaning we assign to them.
And there are dozens of other possible meanings.
The appointment time may have been impossible. The intervention may have increased distress. The child may have refused to participate. The parent may have already tried similar approaches repeatedly without success. The service may have required more executive functioning than the family had available. The clinician-family relationship may not have felt safe. The treatment goals may not have reflected the family's priorities.
Or the parent may have made a thoughtful decision that the service simply wasn't helping.
That is very different from not caring.
We turn system outcomes into parent characteristics
This matters because language doesn't stay neutral once it enters a record.
“Family declined service.”
“Parent did not follow recommendations.”
“Limited follow-through.”
“Poor engagement.”
“Parent appears insufficiently invested.”
One interpretation begins to accumulate alongside another until a story about the family emerges.
And then the story becomes data.
The next clinician reads the previous clinician's notes. The next agency receives the referral. A case conference occurs. Decisions are made about eligibility, readiness, risk, compliance, capacity, or whether additional resources should be offered.
Suddenly, a service that did not fit has become evidence about the people it failed to serve.
This is particularly concerning because service systems hold considerably more power over the narrative than families do. Professionals write the charts. Agencies define successful participation. Programs establish the expectations. Systems determine what counts as engagement.
When the service provider is also the person interpreting why the service did not work, there is an obvious blind spot:
We may be evaluating the family's response to our intervention without adequately evaluating the intervention itself.
What if refusal is data?
There is another way to look at this.
Instead of asking:
Why won't this family engage?
We could ask:
What is their lack of engagement telling us about the service we offered?
Instead of:
Why won't this parent follow through?
We could ask:
What makes this recommendation difficult, undesirable, inaccessible, or inappropriate in the context of their actual life?
Instead of:
Why aren't they invested in treatment?
We could ask:
Are we working toward something they actually identify as important?
That shift seems small, but it changes where we locate the problem.
It also makes room for curiosity.
A family declining an intervention is information.
A child refusing to enter the therapy room is information.
A parent repeatedly missing an appointment scheduled during school pickup is information.
A neurodivergent family struggling with a highly structured, compliance-heavy program is information.
None of those observations automatically tell us what the information means.
That requires curiosity.
The danger of the single lens
This is where I think systems can unintentionally become remarkably self-protective.
If our underlying assumption is that the service is appropriate, then poor outcomes have to be explained somewhere else.
The family wasn't ready.
The parent didn't follow through.
The child wasn't motivated.
There wasn't enough consistency at home.
They disengaged.
But if we allow ourselves to hold multiple explanations at once, another set of questions becomes possible.
Was the intervention appropriate?
Was it accessible?
Did the family experience it as useful?
Did we understand what they were asking for?
Did our model account for this child's neurotype, developmental needs, trauma history, autonomy, communication style, and relationships?
Did the service require capacities from the parent that we never actually supported them in developing?
And perhaps most importantly:
What evidence would convince us that our service was the problem?
If the answer is nothing, then we are no longer evaluating the situation neutrally.
We have built a framework in which the system cannot fail.
Only families can.
Parents are not infinitely resourced implementation systems
There is also a broader assumption embedded in many child-serving systems: that parents exist primarily as the implementation arm of professional recommendations.
We recommend the strategy.
We provide the worksheet.
We establish the routine.
We suggest the intervention.
And then we measure whether the parent carries it into the home.
But parents are not blank delivery mechanisms for treatment plans.
They are people embedded within complicated relational systems. They may be parenting multiple children with competing needs. They may be disabled or neurodivergent themselves. They may be working, caregiving, advocating, coordinating appointments, managing crises, navigating financial strain, or recovering from their own histories.
None of this means we should assume parents are always right, or that every difficulty is attributable to systems. Parents can struggle. Parents can make choices that don't help. Clinicians sometimes need to have difficult conversations.
But difficult conversations should be grounded in evidence and curiosity—not in a default assumption that when our intervention meets resistance, the resistance tells us something deficient about the parent.
We need to separate observation from interpretation
Perhaps one practical place to begin is with our documentation.
Instead of the family is disengaged, we can document what actually happened.
Instead of the parent lacks motivation, we can describe what the parent said, did, declined, or found difficult.
Instead of poor compliance, we can identify which recommendations were not implemented and ask why.
Then we can hold multiple hypotheses about what those observations might mean.
Because the distinction matters.
“The parent attended two of six sessions” is data.
“The parent does not prioritize their child's treatment” is a conclusion.
And there is an enormous amount of unexplored territory between those two sentences.
Curiosity is not the absence of clinical judgment
None of this requires abandoning professional expertise.
It requires using it more carefully.
Good clinical judgment should allow us to examine the child, the parent, the therapeutic relationship, the intervention, the organization, and the broader environment simultaneously.
It should permit the possibility that several things can be true.
A parent can be overwhelmed and a service can be poorly designed.
A child can need support and reject the particular support we offered.
A clinician can be skilled and still be the wrong fit for a family.
A program can be evidence-based and inaccessible to the person sitting in front of us.
A family can struggle to engage and care desperately about their child.
When we collapse all of that complexity into a single explanation—the parents aren't engaged—we don't just risk unfairly judging families.
We lose information.
And when we lose information, we become less capable of helping the children our systems were built to support.
Perhaps the question shouldn't be, “Why won't this family engage with what we're offering?”
Perhaps it should be:
“What is this family's response teaching us about what we're offering?”
That question leaves room for accountability on all sides.
More importantly, it leaves room for us to learn.
