Strength in Numbers, the Power of Data

Health data services · public health

SignalPoint turns laboratory drug-checking results into a jurisdiction's own picture of what is circulating now: by substance, by county, by month, with the number of samples printed beside every figure. The people responsible for public health read this month's supply instead of last year's outcomes.

Shared information is the one resource that grows as more people contribute it. Every result a program contributes sharpens the picture for every county on the map, and every participant receives their own result back, anonymously.

What the platform is: a standardized, append-only record of laboratory results and the screens, maps, documents and alerts computed from it. What it is not: a laboratory, a collection program, or a holder of anyone's identity.

The source

Laboratory-confirmed

Results reported in the laboratory's own words, never reinterpreted; any laboratory, one intake contract.

The grain

County × month

Comparable to the month before and to the county next door, across laboratories.

The standard

Every figure traceable

Each number opens to the samples behind it and cites the question, version and dataset reading it was computed from.

The person

No identity exists

No field in the data model for a name, address, phone number or point on a map. A fact of the design, open to inspection.

Why — the picture New Jersey is working from

the public record on the crisis, and where the supply itself goes unmeasured

Opioid abuse and overdose remain a critical public health crisis in New Jersey, heavily driven by the proliferation of synthetic opioids like fentanyl. The state experienced a tragic surge in drug-related fatalities that peaked during the disruptions of the COVID-19 pandemic, cementing substance use disorder as one of the most pressing challenges for local families and communities. While public health interventions and harm reduction strategies have worked tirelessly to curb fatal outcomes, thousands of residents still suffer from addiction each year, stretching hospital emergency rooms and community treatment facilities to their limits.

Treatment and response

To combat this epidemic, New Jersey has expanded access to life-saving initiatives, including the widespread distribution of naloxone (Narcan) and increased funding for peer-led recovery programs. The state continues to focus on lowering barriers to substance use disorder treatment, though local admission rates reflect a persistent demand for comprehensive healthcare resources. Addressing this complex crisis requires an ongoing commitment from policymakers, healthcare providers, and community organizations to shift the focus toward compassion, early intervention, and long-term support systems.

Nationally, overdose deaths have fallen for three consecutive years, and synthetic opioids, primarily fentanyl, were involved in about seven of every ten overdose deaths in 2023. Deaths are counted months after the fact; what was in the supply when they happened is rarely recorded at all. SignalPoint exists to describe that supply, by county and by month, from the laboratory's own results.

U.S. drug overdose deaths · thousands · by year

Source: CDC National Center for Health Statistics; 2019–2023 final, 2024–2025 provisional (May 2026 release). Fentanyl share: CDC, 2023.

How it works — from a sample to a county's picture

five stations, one direction; the platform's own boundaries are drawn beneath

A sample is registered

Under a random identifier, by the program that collected it. A card with two codes goes to the participant.

registry · anonymous id · status walk

A laboratory analyzes it

Any contracted laboratory. The platform reads the laboratory's report as written and keeps the original file as received.

intake · one adapter per laboratory · raw file preserved

The result is standardized

Every substance name is resolved against one reviewed dictionary; an unknown name waits for a scientist, never a guess.

dictionary · versioned · quarantine for the unknown

The record is kept

Append-only. A correction adds a record; nothing is overwritten. Each figure cites the dataset reading it was computed from.

warehouse · dataset reading · reproducible figures

The picture is delivered

Screens, maps, monthly documents and alerts for agencies; a plain-words result for the person who submitted.

reporting · maps · documents · alerts · results

Never touchedThe platform records results; it never holds, ships or handles a physical sample.

Never reinterpretedA result is reported as the laboratory reported it. The platform does no analysis.

Never overwrittenEvery record, access and change is kept and logged. A quoted figure reproduces years later.

Never namedNo identity field exists. The participant is reachable only by the codes on their own card.

The platform — ten screens, one record

what it is, in general terms; the screens follow

What the platform is

A standardized, append-only record of laboratory drug-checking results, and the surfaces computed from it. Results arrive from any contracted laboratory through one intake contract and are resolved against one reviewed substance dictionary, so results from different laboratories and different counties can be compared. Nothing is overwritten: a correction adds a record, and every figure cites the dataset reading it was computed from, so the identical request reproduces it exactly.

Access is by role and by organization. An agency sees its own scope; the operator's rooms (console, accounts, audit) are separate; the participant's result page asks for two codes from a card and nothing else.

The ten screens

Reporting	the county's figures, findings, strength, confidence; the drill to one sample's record
Maps	county heat maps by month: share, strength, change; profile match for a combination
Reports	composed, printable, reproducible documents for county and state audiences
Alerts	scientist-confirmed detections published in the audience's own format
Results	the participant's anonymous result page
Console	the scientist's room: quarantine, the dictionary, review, every decision attributed
Accounts	organizations, seats and terms, as versioned records
Audit	every privileged read, export and change, traced; the access review document
Data API	the same governed answers for research teams, machine-readable, under agreement
Sign in	one front door; one session honors every screen

Reporting — the county's picture

headline figures · findings by substance · strength · level · confidence score · the drill from a figure to one sample's laboratory recordDemonstration data

Reporting screen: headline figures and the findings table for a county and month range

Every number on the screen opens: the figure, the exact samples behind it, one sample's complete laboratory record. The dial beside the findings is a display choice, never a data restriction; the served figures do not change.

Maps — share of samples

Demonstration data

Maps screen: county heat map of a substance's share of samples

Real Census county shapes, animated month by month. A substance enters a region and moves county to county; the map shows it.

Maps — strength and change

Demonstration data

The same map switched between how often a substance appears, how strong it is, and how it changed against the period before or the year before.

Documents — composed, reproducible

Demonstration data

$A composed report document with headline figures and a findings table$

Monthly documents composed from the same figures as the screen, printable, and reproducible exactly, forever. Each carries its sources, its method and its dataset reading.

Alerts — confirmed, then published

Demonstration data

Alerts desk: a published alert with its evidence tier, state and audience

A newly detected substance is confirmed by a scientist, never by the machine alone, then published to the agencies it concerns in their own format. The evidence is frozen with the alert; a later retraction changes the banner and never a figure.

Results — the participant's page

The one public-facing screen. It asks for two codes from a card and nothing else: no account, no name, no phone number. Four steps, Registered · Received · Testing · Ready, then the result in plain words with the laboratory's wording beneath it.

Who it serves — one record, read at the grain each reader needs

access is by role and by organization; a reader sees its own scope and nothing beyond it

The record
One standardized, append-only record of laboratory results
every figure computed from it · every access logged · nothing overwritten
- Public health
  Agencies
  - County health departments
    This month's supply
    The county's dashboard, a monthly document for the commissioners, an alert the health alert network can send as written.
  - State agencies
    Every county on one map
    A statewide picture comparable county to county and month to month; an alert template in the state's own format.
  - Coroners · fatality review
    Supply context beside the case
    The month a compound first appeared in the county, on the record before toxicology.
- Clinical
  Care
  - Hospitals · health systems
    Warning before the first admission
    The month a new sedative or opioid appears in the region, on the record before it appears in the emergency department.
  - Treatment providers
    Why a protocol stopped working
    The county trend shows a change in the supply before the next admission does.
- Supply side
  Contributors
  - Community programs
    Their evidence, returned
    Programs that collect the samples see their own contribution as a document and hold a seat where presentation rules are written.
  - Laboratories
    Results, credited and read
    A laboratory's work reaches every agency in its region, in its own words, with its name on the record.
  - Researchers
    A dataset a method can cite
    Every figure pinned to a question, a version and a dataset reading, so a reviewer can reproduce it.

Participants sit outside the chart on purpose: they hold no account and are read by no one. They receive their own result by code, and the platform holds nothing that could name them.

Built to be trusted — what the platform cannot do is the point

No identity exists

There is no field in the data model for a name, an address, a phone number or a point on a map. Not a policy: a fact of the design, and the schema is open to inspection.

Never touches a sample

The platform records results. Collection, shipping and analysis belong to the programs and laboratories, and no feature may add a physical step.

Nothing is overwritten

Every record, every access and every change is kept and logged. A correction adds a record. A figure quoted today reproduces years from now.

A written firewall

Law enforcement may receive only an aggregate county report with small counts withheld, and never sample, site or submitter data. The same statement rides every agreement.

Contact

For county and state agencies, health systems, laboratories and programs who want to see the platform with their own region on the screen.

info@signalpointhds.com

Write to info@signalpointhds.com with "SignalPoint demonstration" in the subject line.