The Profiling System in Every State You Never Signed Up For
Black Mirror’s “Nosedive” is already happening — the problem lies in murky consent.
Last year, I found out that I had a rare, incurable disorder.
I also found out my government was paying a privately owned company to profile and assign me a “score”.
I don’t know which news was more upsetting.
I wasn’t allowed to know my score or even how it was derived.
What is this “score”?
It is an addiction/overdose risk assessment score, assigned by an algorithm developed by Appriss, Inc., in Louisville, Kentucky.
It assessed a few values entered into data fields, then ran them through its weighted algorithm, and PRESTO!
It spits out a score just for me.
I have a science background.
I am reasonably intelligent.
I didn’t like this.
I began to dig.
Saving Us From Ourselves
Appriss, Inc., began their quest for data mining domination in the middle of the 1990s by the development of the VINE system.
VINE is a victim notification system that alerts the victims of a crime when the perp is released from jail/prison.
Reviews of this system by its users have been mixed at best. The VINE app on iPhone, for example, which is a preferred notification system for most victims, currently sits at a 1.9-star rating on the iTunes Store, with review after review of horrified, scared victim stating they were not informed.
A software and data company has 3 apps on iTunes — 1 with no reviews, 2 with 1.9-star reviews and very angry customers because the developers appear to have abandoned any updates for over a year.
Appriss, Inc., has spent the better part of the last twenty years developing and perfecting their algorithms around schedule prescribed medications, PDMPs (prescription drug monitoring programs), and a specialized PDMP in house profiling project they ultimately named NarxCare.
This is far more lucrative to current management than the altruistic vision of the founder.
A spectacular failed attempt at slowing methamphetamine abuse by the government.
Failed because between the years 2011–2016 methamphetamine overdose deaths in the United States have more than tripled in number, according to the CDC.
A failure doesn’t get illustrated quite so resoundingly very often.
Yet, they still continue their course of stubbornly throttling prescribed opioid medication to the patients who aren’t abusing it in the first place.
Of course, this is all framed “for the best interest of public health”.
Does anyone else cringe when a public official mouths the words “best interest of public health”?
Every time those words exit a politician’s mouth, you can be certain a civil rights violation will be on its heels.
To wit, “the public” is left in the dark regarding these plots, never asked their input, and, most importantly, never asked for informed consent.
Little is known about the data fields Appriss actually inputs.
The algorithms are proprietary.
Furthermore, the state statutes written around PMP/PDMPs have this little nugget hidden inside:
“Program to contract with vendors/3rd parties for the purpose of “obtaining technical assistance in the design, implementation, operation, and maintenance” of the system. “
Questions addressed to the state agencies /reps regarding them are ignored.
What I have been able to ascertain independently is troubling.
Appriss has introduced a number of “add-on” packages to the basic PDMP that each state has in place- thanks to Brandeis University.
The Brandeis proprietary PDMPassist has been the author of the state PDMP metrics and algorithms for each state PDMP program.
Congressman Hal Rogers, of Kentucky, is the PDMP ‘face man’ at a national level.
The ‘Hal Rogers’ PDMP grant program has funded many states’ programs and gotten them up and running.
The Congressman’s moniker is attached to all things PDMP for no obvious reason.
The only thing Rep. Rogers appears known for besides the PDMP programs in his name is being dubbed “the Prince of Pork” by government spending watchdog groups.
He also served as the chairman of the House Appropriations Committee for a short time.
Appriss, Inc., HQ is located in Kentucky as previously mentioned, same as Rep. Rogers.
One thing I have learned in my research throughout the bizarre turns of the opioid crisis is that there are no coincidences.
One of the add-on programs that almost all states have implemented to their PDMPs is called “PDMP AwaRxe”.
This program is nothing more than Appriss, Inc.’s, insidious NarxCare under a different name.
This is also where the “scoring” comes into play.
With NarxCare, a patient gets what is referred to as a NarxScore.
Herein the problem lies.
The PDMP, NarxCare, PDMP AwaRxe, et al., are all rudimentary algorithm programs.
They do not allow for any inputs by a physician into data fields.
This is a problem.
It’s a problem because of many factors, firstly, that patients are not told that they are being high scored or “red flagged” for innocent behavior.
Here is an example of how a patient can be pressured into rehab for a computer algorithm decided diagnosis of addiction when it is, in fact, normal behavior.
Betty is a pastor’s wife who suffers chronic pancreatitis. On Sept 1, she went to an appointment with a new pain management physician.
Her new physician wanted her to be committed for treatment for SUD based on what the computer told him and the flag it gave him regarding Betty.
When she refused to go, he called her husband and enlisted his help to try to convince Betty to go.
How did this happen?
The computer algorithm told her new physician that Betty was an extremely high-risk patient.
It said she was a doctor shopper who went to multiple providers and pharmacies over the past 9 months.
It recommended that she seek treatment immediately.
The physician would not listen to Betty when she attempted to explain what happened in the past 9 months.
Addicts are liars.
What did actually happen?
In January, 9 months previously, Betty had her normal pain management appointment.
Three weeks later she was in a terrible car accident (not her fault) and taken by ambulance to a local hospital.
When she was discharged, four days later, it was a Saturday afternoon, around 5 PM.
The orthopedic physician who treated her broken ankle in the hospital wrote her a prescription for pain medication after he spoke to her pain management physician and they came up with a plan.
Her husband filled the prescription at a Walmart pharmacy because their normal pharmacy was closed.
Betty had her normal pain management appointment in March. She was hospitalized at the end of March with appendicitis that had perforated her bowel.
The GI doc discharged her with a few days of pain medication on top of her pain management after consulting her pain management physician.
The prescription was filled at the hospital pharmacy before they left the hospital.
Betty had her normal pain management appointment in May.
She had an infected root canal removed in June and required extensive dental surgery along with a bone graft in her jaw.
Her oral surgeon prescribed pain medicine after consulting with her pain management physician.
Her husband filled the script for her at the Rite Aid near the dentist office because she was in terrible pain and he wanted her to have the medication for the trip home (a 2-hour drive).
Betty had her normal pain management appointment in July.
Her physician told her he was retiring.
He gave her a few names of physicians in the area.
Betty made an appointment with the new physician for September. She had all her records forwarded there.
Betty had a flare-up of acute pancreatitis at the beginning of August.
She was hospitalized for three days, then discharged with a 4 day prescription of extra pain medicine from the hospitalist.
Her husband filled the script at the Rite Aid by the hospital because it was the only pharmacy open on Sunday.
Because Betty had five different physicians and five different pharmacies recorded within the past 9 months in her records dispensing prescribed opioids, the algorithm automatically flagged her and recommended rehab — all per the NarxCare programming.
The literature says to “have a discussion”.
No physician in practice utilizing the NarxCare system is following any of the “recommendations” provided in any of these “guides” past the words; “cut off” or “rehab” that I have heard of, anecdotally.
After all, the entire premise of this is based on the CDC guidelines for opioid prescribing for chronic pain, a voluntary guide for new patients, not existing ones, which have never once been utilized in the way they were written — to include the premise of Narxcare in the first place!
The new physician had a prejudice against her based on NarxScore before he ever laid eyes on her.
There was no way for any of her former physicians to input a note into her “NarxReport” or PDMP AwaRxe report.
No way to clarify extenuating circumstances or about their consultations.
EHRs are incomplete.
Physicians won’t believe the patients when they try to explain.
What are patients supposed to do?
They are fighting a system that 99% of them aren’t even aware is in place.
They certainly have not been told the “rules”.
I have found most physicians don’t even know the “rules” themselves or where to find them.
Indeed, what exactly are “the rules”? They vary state to state.
Innocent, non-pathological, patient appropriate use of prescribed opioid medication as directed by a physician gets a patient flagged as high risk.
Not only that, it guarantees the patient will not be considered for pain management by any physician.
The patient will be forced to undergo extreme pressure to go to rehab for a disorder that doesn’t exist.
Patients are “allowed” to have erroneous information removed from their PDMP/ Narxcare.
As with many “safeguards”, this benevolent statute is rarely practically applied.
This operates under the premise that a patient:
- Has easy access to and regularly reviews their report(s), can afford fees
- Has a physician willing to believe them when they report an error
- Has a physician willing to sign off verifying the error
- The pharmacy will remove the error from the database (physician offices punt on this)
I challenge any doubting Thomas to attempt to have an error stricken from their PDMP.
NO PROFILING BEFORE NOON, PLEASE
As an advocate for the incurable painful disease patient and physician community, I am commonly asked the following:
How did this get to this point? Why weren’t we told?
We have to wonder… why hasn't anyone ever asked for a signature of consent, saying to the effect of:
”I, (patient), have had PDMP and Appriss Inc. NarxCare fully explained to me. I understand my rights and responsibilities as a citizen and patient as I am enrolled in this database.”
The pain contracts given to patients do not address this issue, outside of vague mentions in some about “monitoring”. These mentions are buried in the contract, not discussed, and of course, a patient cannot object.
Perhaps the practitioners themselves cannot explain these systems?
Or maybe because the entire sham is a civil rights violation of such magnitude there is no way to ask consent.
Therefore we all just play pretend and hope no one ever notices.
The amount of people who have access to the sensitive patient PDMP information who are not bound by HIPAA is stomach churning to contemplate.
There are so many.
Few truly understand this.
It has been well hidden.
A running joke around Appriss HQ is they use the PDMP/NarxCare system to pull up people’s addresses for the company Christmas party invites.
Far from humorous, this illustrates what a joke online security is around our protected, sensitive patient information.
Are the employees of Appriss, Inc. bound by HIPAA? Brandeis University? Yale? The students there?
How many stakeholders, virtually all people in the employ of a veterinary office, billing, coding, law enforcement… I honestly get sick to my stomach when the scope of this is truly considered.
What about HIPAA? Isn’t this a direct violation of that law?
Since HIPAA’s adoption in 1996, we have had less than 24 documented convicted cases of medical records tampering/ abuse/ confidentiality breaches.
Out of 350 million people over 22 years.
Concern for HIPAA breaches is astounding at HHS, it appears.
This number is derived by going to the HHS website on HIPAA compliance and painstakingly pulling the data out of their “reports”.
Can I opt out?
Therein lies the rub.
Even if you decided today that you were sufficiently freaked out enough to want to try to taper off your medications to see if you could manage your life without them, you are in these databases for life.
No getting out.
How do you opt out of a system you never signed up for in the first place?
Remember, even people who have never been prescribed any schedule meds are in there with scores.
Like my autistic daughter and son, who hit the ground running with a high score solely due to their autism.
They will have one hell of a time getting any opioid therapy if something awful were to happen.
Because, as of right now, a damn few docs are treating any child’s pain.
Children apparently are immune to any pain as of the 2010s.
Moreover, as soon as a child hits “teenager”, thanks to the idiots who decided to steal prescriptions and party with Grandma’s Percocet, all are painted with the brush of “drug seeker”.
Autistic individuals are considered “high risk” for drug abuse because the algorithm says so. Along with those who have anxiety disorder, depression… if you have any mental health diagnosis.
Autism is a mental deficiency, according to the algorithm’s programming. (NOT ME. I don’t have column space to delve into the fury I feel about this.)
My friend who was raped is another person who is judged unfairly under the scoring.
Assault is another flag for the algorithm.
Her horrible assault makes her have a score that is “at risk”, although she has several years between herself and that awful day.
She worked very hard in therapy to regain her poise, confidence and self-esteem.
Rape victims are more likely to become addicts, according to the algorithm’s programming.
A young man who was molested one summer at camp- he is also tagged high risk.
Children who were abused are more likely to become addicts later in life, as per the algorithm’s programming.
There are several inherent problems with this kind of profiling algorithm.
First of which, the types of studies its engineers used to declare which populations were at risk in the first place.
No inputs, remember?
The computer has no idea if these patients were treated for their trauma.
Unresolved trauma is the key takeaway.
It is true that statistics show those who have suffered a trauma in their past have a higher incidence of addiction.
What of those who have worked hard to get past it?
Moreover, doesn’t everyone deserve appropriate pain management anyway? Or is it just too much work?
Should a patient who trusts his physician with past trauma that is resolved be reminded of it again and again, simply due to the fact an algorithm proclaims he/she may become a drug addict due to that trauma?
Should they be unfairly punished due to “pre-crime” Minority Report-type profiling and denied compassionate care?
Does that one input define them as a person?
Do we deny children of alcoholics the right to purchase or use alcohol, even though the science and statistics overwhelmingly show COA may become alcoholics too?
Why not embrace attaching EHRs (electronic health records) to the PDMPs as suggested? Wouldn’t that help?
EHRs are inherently a horrendous idea, and the reason why is security, and to a lesser extent, data integrity.
Many think the convenience of the EHR is wonderful, game-changing.
Many physicians, however, have complained about the time suck of EHR technology, as well as missing data in EHRs.
They want paper charts.
Many have flat out refused to switch.
Forcing physicians to switch to EHRs for Medicare patients, which CMS has done for 2019, probably ensured massive dumping of thousands of patients by physicians, instead of forcing physicians to switch to EHRs as they hoped.
Framing it as physicians “holding patient records hostage” in a new way to spark patient outrage is a nice touch by CMS.
Who was the marketing genius who thought that up? He deserves a raise.
Telling consumers they will be in control of their records is the best way to get them to pressure their providers into folding.
Also, in attaching EHRs to the PDMPs, government and insurers may get their hands on long coveted information about data on people and potential insurers will screen heavily before deciding to take them as clients.
As Cigna has begun to do with their own “proprietary” algorithm with 16 datasets they won’t disclose. It has begun.
In the government’s case, I shudder to think why.
This is catastrophically bad for the consumer. Imagine the insurer being able to see your entire medical record, every nook, and cranny, before deciding what rate they will charge you.
The laws have not been in line with the technology.
Or is this merely another by product of the digital era?
A google search reveals nothing.
Pages of praise for EHRs. Pages of praise for PDMPs. A few blog posts with warnings. No laws that reflect any patient protections.
As I said previously, legislators either ignore the issue when asked or reveal their ignorance by saying they don’t know anything about the topic.
FOIA requests sit in limbo.
The one and only thing America should be focused on is a complete overhaul of the insurance industry, along with their lobby.
Add on requiring consent to be obtained for all this nonsense.
Only when the power of the insurers and their lobbyists are broken and scaled back forever, only then we will see a difference.
We will see meaningful and lasting positive change in the American healthcare industry.
This is Part I in a groundbreaking series on the algorithm data science and lack of informed consent surrounding the opioid crisis in 2019 America.
Please check back to continue reading the never before coverage of this issue.
Thank you for reading! If you enjoyed reading this article you may like this one also: