After the form — does how we record disclosure actually change what happens next?
This week's threads have largely asked how we
capture harm well: flagging divergence, tracking referral cascades, choosing
whose definition of violence counts. All essential. But I want to ask the
question one step downstream: once a disclosure is recorded well — flagged as
divergent, tracked through a fuller cascade — does that improve what actually
happens to the woman afterward, or does it just improve the dataset?
I ask from the service-delivery side, not the
research side. In a peer-led recovery model, disclosure isn't a discrete
measurement event — it resurfaces across an ongoing relationship, and what a
peer does with it in the moment (refer, hold, follow up personally)
often matters more to outcome than how precisely it gets coded afterward. I'd
like to hear from others doing direct service work: has better instrument
design or better recording practice measurably changed care outcomes in your
setting, or has the improvement stayed mostly on the data side while service
response lagged behind? And if there's a gap between the two, whose job is it
to close it — the researcher who designed the better instrument, or the
implementer who has to act on what it surfaces?
That's exactly the right question, and the honest
answer is: the peer holds it first, but the program owns it
structurally. The peer agent is the one who initially receives and
acts on the disclosure — that's deliberate, because the relationship is where
trust lives, not the filing system. But the peer doesn't hold it alone:
every case is logged into a shared program record with supervisory oversight,
precisely so the follow-through doesn't collapse if that one peer is
unavailable, burns out, or the relationship itself becomes complicated.
Your "record as handoff, not data
point" framing is exactly right, and it clarifies something I hadn't put
into words before: the record only works as a handoff if the person writing it
and the person accountable for the next 72 hours are the same person, or at
minimum sit inside the same supervision structure. The moment recording and
action get separated into different roles — researcher designs the field,
someone else has to act on what it surfaces — you've recreated the exact gap
you named.
As for the woman herself holding it: she always
retains the right to what's shared and how, but I wouldn't say she "holds
the record" in the accountability sense — that would quietly shift the
burden of follow-through back onto her, which defeats the purpose of a peer-led
structure in the first place. The record needs an owner who is not the
person harmed.
Charles, your point about the peer holding it first while the program owns it structurally highlights the exact breakdown point in most systems.
The gap between better datasets and better care usually opens up during the administrative handoff. When an instrument is designed primarily for M&E or research, the act of recording turns a human disclosure into an administrative object. Unless that recording tool doubles as a real-time trigger for resources (e.g., auto-flagging an urgent voucher, scheduling an immediate supervisor review, or reserving a bed), improved data quality only serves the reporting cycle.
Closing the gap shouldn't fall on either the researcher or implementer alone it requires designing actionable instruments, where a field cannot be added to a form unless there is a pre-funded, pre-mapped service response attached to every positive response."
I think the critical issue is not whether the researcher or implementer “owns” the gap, but whether the system has been designed so that disclosure automatically creates responsibility and a pathway to action.
A well-designed instrument should do more than document what happened. It should help answer three operational questions in real time: What level of risk is present? Who is responsible for the next action? By when must that action occur? Without those three elements, even an excellent dataset can become an endpoint rather than a bridge to care.
This is where I see an important distinction between data capture, case management and accountability. Data capture tells us that a problem exists. Case management determines what happens to the individual. Accountability ensures that the case does not disappear between referral, supervision and service delivery.
For example, a disclosure requiring urgent intervention should not simply generate a coded variable in an M&E database. It should trigger a defined escalation pathway, a named responsible person, a timeframe for action and documented confirmation that the action occurred. Routine cases may follow a different pathway, but the principle remains the same.
I also agree that placing the burden on the woman to navigate the system after disclosure is problematic. The person experiencing harm should retain agency over consent, confidentiality and decisions about disclosure, but institutional responsibility for appropriate follow-through should sit with the programme and its designated safeguarding/service structures.
This has implications for researchers too. If we design instruments that identify needs without mapping the available response, we may improve measurement without improving outcomes. Conversely, implementers cannot reasonably be expected to act on needs that the system has failed to translate into clear responsibilities, resources and referral pathways.
Perhaps the better design principle is therefore:
Every important data field should have an action pathway attached to it.
If a positive response does not change assessment, referral, supervision, treatment, protection or follow-up in some defined way, we should ask whether collecting that information is actually serving the person—or merely serving the reporting system.
Ultimately, the quality of an M&E system should not be judged only by completeness, accuracy or timeliness of records. We should also ask whether it produces timely action, continuity of care and demonstrable outcomes for the person behind the record.
That is where data governance, implementation science and service delivery genuinely meet.