Missing Essay on "What to say" in the most difficult situations

 Some in 2024, I think in mid-2024, I saw an essay by a physician in one of the major journals (I think.  I think, JAMA or NEJM or Ann Intern Med, etc).

The essay was about what to say in difficult situations.  In might have been triggered by a study that found that AI dialog in difficult situations (e.g. terminal diagnosis) was as emphathic as spontaneous human remarks and statements.

As I recall, the author made the point, that he had learned in his training, that in the most difficult situations, there were actually a fixed library of statements, that were generally "better" to use than simply making comments off the top of his head and hoping for the best via spontaneity.

I resonated with the physician's essay, but have not been able to find the citation again.

###

I have not reviewed the material below, but I fed the above paragraphs straight into Chat GPT 4, click on the "web" search button, and got this result.

####

Chat GPT 4 (Web search)

You might be referring to the study "Comparing Physician and Artificial Intelligence Chatbot Responses to Patient Questions Posted to a Public Social Media Forum," published in JAMA Internal Medicine in April 2023.

JAMA Network

This research found that chatbot-generated responses were preferred over physician responses and rated significantly higher for both quality and empathy. The findings suggest that AI assistants could aid in drafting responses to patient questions, potentially alleviating some of the communication burdens faced by physicians.

Following this study, several commentaries and responses were published discussing the implications of AI-generated empathy in medical communication. For instance, the article "Machine-Made Empathy? Why Medicine Still Needs Humans"

JAMA Network

and its subsequent reply

JAMA Network

delve into the nuances of AI-generated empathetic language and its role in healthcare. These discussions emphasize the importance of maintaining human oversight and the irreplaceable value of genuine human empathy in patient care.

Additionally, the viewpoint "Compassionate Machines: The Ethics of 'Artificial Empathy' in Cancer Care," published in JAMA Oncology in May 2024, explores the ethical considerations of integrating AI-generated empathy into oncology practice.

JAMA Network

This piece highlights the potential benefits and challenges of using AI to simulate compassionate communication in sensitive medical contexts.

These publications collectively address the evolving role of AI in medical communication, particularly concerning the use of standardized empathetic responses in difficult situations.

###

None of the above were exactly what I was looking for.   Adjacent to the above, JAMA suggested, "physicians and AI chatbot responses...Chen et al., JAMA Oncol 2024.

https://jamanetwork.com/journals/jamaoncology/fullarticle/2818765

###

Notes - resources found; none are the first-person essay that whether AI or memorized, there were certain time-tested based statements to say in a difficult situation.

###

Comparing physician and artificial intelligence chatbot responses to patient questions...

Ayers JAMA Intern Med 183:589

Machine-made empathy.  Why medicine still needs humans.

Cadiente JAMA Intern Med 183:1278

Machine-made empathy.  Reply.

Ayers JAMA Intern Med 183:1279.

Compassionate machines: Ethics of artificial empathy in cancer care.

Koranteng  JAMA Oncol 10:847

Physicians and artificial intelligence respones to cancer questions from social media.

Chen  JAMA Oncol 10:956

Red Tape Roadblocks - Two Physicians Can Barely Get P.A. Drug for Very Sick Child

In an essay in Annals of INternal Medicine, dated 1/21/2025, two physicians can barely, after weeks of effort, get a key drug for their very sick child.  The denials were repetitive and had to do with the exact verbiage on the FDA label.

https://www.acpjournals.org/doi/10.7326/ANNALS-24-02781

One connection I made would be when a specialty drug requires "XYZ" diagnostic test on the labeling, but an innovator lab has an LDT that provides superior performance, or a new generation of technology, but is not on-label.

CARIS - Comparison to other FDA tests

Chat GPT provided this comparison of new CARIS FDA documentation, with prior FDA genomic tests.   I have no idea how accurate these tables are.   But they were auto generated by Chat GPT.

###

Comparison of the Caris MI Cancer Seek Workflow with Other FDA-Approved NGS Panels

Below is a detailed comparison of Caris MI Cancer Seek with notable FDA-approved NGS panels such as FoundationOne CDx, Guardant360 CDx, and Oncomine Dx Target Test. This comparison focuses on bioinformatics workflows, biomarker validation, and clinical applications.

---

1. Bioinformatics Workflows

Data Processing

Feature	MI Cancer Seek	FoundationOne CDx	Guardant360 CDx	Oncomine Dx Target Test
Sequencing Platform	Illumina NovaSeq 6000	Illumina HiSeq 4000	Illumina NextSeq 550	Ion Torrent PGM
Target Regions	Whole exome with boosted coverage of 228 genes	Comprehensive genomic profiling of 324 genes	74 genes in cell-free DNA	~50 genes for hotspot mutations
Input Material	FFPE tumor tissue	FFPE tumor tissue	Plasma (liquid biopsy)	FFPE tumor tissue
Data Alignment	Sentieon BWA to hg38	BWA-MEM to hg19	BWA to hg19	Torrent Suite alignment to hg19
Variant Calling	Custom algorithms for SNVs, indels, TMB, MSI, CNA	Custom algorithms for SNVs, indels, fusions	Custom algorithms optimized for cfDNA	Proprietary mutation detection pipeline
Reporting Levels	Companion Diagnostic (CDx), Clinically Significant, Emerging Evidence	Companion Diagnostic and Clinical Evidence Levels	Companion Diagnostic and Select Variants	Companion Diagnostic Only

Strengths of MI Cancer Seek:

• Broader Scope: The inclusion of whole exome sequencing (WES) allows for higher flexibility in adding emerging biomarkers compared to hotspot-based assays like Oncomine Dx Target Test.
• Advanced Bioinformatics: Use of synthetic RNA/DNA differentiation during library preparation enables RNA and DNA processing from the same sample, although RNA is not reported under the current approval【10†source】.

Limitations:

• FFPE Limitation: While comprehensive, MI Cancer Seek does not yet support liquid biopsies, which Guardant360 CDx offers for less invasive sampling.

---

2. Biomarker Validation Criteria

Criteria	MI Cancer Seek	FoundationOne CDx	Guardant360 CDx	Oncomine Dx Target Test
Analytical Validation	≥100x read depth; ≥16 billion reads per run; ≤1% error rate	≥500x for hotspot regions; stringent quality metrics	≥5000x (cfDNA); highly sensitive for low VAFs	≥500x for hotspot regions
Clinical Validation	CDx biomarkers tied to therapies; evidence-based profiling	CDx biomarkers and investigational markers	CDx biomarkers and select investigational markers	CDx biomarkers only
Biomarker Types	SNVs, indels, CNAs (ERBB2), TMB, MSI	SNVs, indels, CNAs, MSI, TMB	SNVs, indels, CNAs, MSI	SNVs, CNAs, fusions
Companion Diagnostic Ties	Therapies for breast cancer, colorectal cancer, NSCLC, melanoma	Similar breadth with additional rare cancers	Primarily NSCLC, breast cancer, and urothelial carcinoma	Limited; specific to NSCLC

Strengths of MI Cancer Seek:

• High Biomarker Sensitivity: The inclusion of MSI, TMB, and specific CNAs (e.g., ERBB2 in breast cancer) reflects advanced tumor profiling capabilities【10†source】【11†source】.
• Emerging Evidence Layer: Reporting Level 3 biomarkers supports exploratory research and clinical trials.

Limitations:

• Narrow Clinical Implementation: While broadly validated, MI Cancer Seek's clinical utility depends on centralized testing at the Caris lab, unlike decentralized kits like Oncomine Dx Target Test.

---

3. Clinical Applications

Feature	MI Cancer Seek	FoundationOne CDx	Guardant360 CDx	Oncomine Dx Target Test
Therapies Supported	15+ targeted therapies	30+ targeted therapies	10+ targeted therapies	Limited set for NSCLC
Turnaround Time	10-14 days	~2 weeks	7 days	~5 days
Specimen Type	FFPE only	FFPE only	Plasma (cfDNA)	FFPE only
MSI and TMB Assessment	Comprehensive	Comprehensive	MSI only	Not included

Strengths of MI Cancer Seek:

• Companion Diagnostic Integration: Direct linkage to therapies like PIQRAY® (alpelisib) for PIK3CA mutations and KEYTRUDA® (pembrolizumab) for MSI-H tumors【8†source】【10†source】.
• Tumor Profiling Depth: Enhanced by whole exome sequencing and comprehensive TMB calculations.

Limitations:

• Restricted Access: Centralized testing limits availability, contrasting with the Guardant360 CDx liquid biopsy model, which offers decentralized and patient-friendly access.

---

Summary of Key Differentiators

• Bioinformatics Sophistication: MI Cancer Seek leads in data handling with whole exome capabilities and RNA/DNA integration.
• Comprehensive Biomarker Profile: It incorporates advanced profiling (TMB, MSI, CNAs) unmatched by some competitors like Oncomine Dx Target Test.
• Centralized Workflow: While this ensures control and consistency, it limits patient access compared to models like Guardant360 CDx.

---

Would you like further detail on clinical validation examples for specific biomarkers (e.g., MSI or PIK3CA mutations) or an analysis of the implications of centralized versus decentralized testing models?

MolDxOlogy: Two Tables for TA Listings

 I was puzzled by different entry points for TA documents?

(1)
If I start at Palmetto MolDx home page

   https://www.palmettogba.com/moldx

I click on Technology Assessments and go here:

   https://www.palmettogba.com/palmetto/moldxv2.nsf/DID/9W2LPH1487

Then I click on Technology Assessments Forms Updated 7 17 2024 and get here

   https://www.palmettogba.com/palmetto/moldxv2.nsf/DIDC/TJ4XC2M5IX~Technical%20Assessment

For example the first form GEN-CQM-003 Checklist, seems to be V1 2023.

I count 13 TA entries.

(2)
But if I use Google I get here:

https://www.dexzcodes.com/palmetto/dex.nsf/T/Technical%20Assessment

https://www.dexzcodes.com/palmetto/dex.nsf/DID/UMPUJNFSV0

And the first form is GEN CQM 003 Checklist, but now, it is updated 10 17 2024 and it is V4 October 2024.

I tally 17 TA entires (plus "gapfill").

Extra ones here include venous thromboembolism screening and prenatal carrer screening and prenatal cell free dna screening

###

I could understand if the Dexzcodes.com included public facing forms like prenatal, not on Palmetto Medicare, but why would things like Checklist have different versions?  

CMS Updates Certain Medical Audit Rules (PIM Chapter 3)

 https://www.cms.gov/files/document/r13008pi.pdf

Unrelated:
Certain very tedious audits for NCDs (re: Comment on Med Adv using "only NCDs."

https://www.cms.gov/files/document/r12990cp.pdf

Moldx and "patients without cancer"

The MRD TA form at MOLDX inquires about presurgical patients and post surgical patients.  Post surgical patients are divided into those with cancer and those without cancer.   How does it define the latter?

Not on the TA form.   

Go to the LCD itself, as MolDx explains below.

• As per L38779, a patient without cancer is a patient with “no clinical, radiographical, or other biological evidence that tumor cells remain post treatment and subsequently the patient is no longer being subjected to therapeutic interventions for cancer.”
• If treatment is ongoing, then the patient is considered to still have cancer. So, a patient who had a lumpectomy is considered to be “without cancer” IF the above statement is true - for example, she is also not undergoing radiation or other treatments and there is no evidence that cancer cells exist (i.e. margins and scans are negative, etc.).

MolDx defines e.g. postsurgical adjuvant radiation or chemo as "treatment for cancer" but a 5 or 10 year treatment with e.g. tamoxifen is NOT considered a treatment for cancer.

If the patient qualifies as treatment for cancer period, then MRD testing would be either one time testing, or, in some cases, a series of several assays defined as one test.   If the patient qualifies as POST cancer, NO cancer, as above, then MRD tests can be billed individually as it is a "patient without cancer."   (This paragraph is inference to the one-test-per-patient clause in NCD 90.2)

#

tagNCD90_2  tagpatientswithoutcancer

Medicare Advantage and Out of Network Rates

 tagmedicareadvantage, tagoutofnetowork

Labs Often Get Paid CMS Rate for Out of Network Lab Claims

Why?

###

Rule 422.214 on Out of Network M.A. Payments

Where this comes from is 42 CFR 422.214.   This rule does not directly state that Medicare Advantage must pay the FFS rate to non contracted (out of network ) providers.  However, it says that out of network providers must accept the FFS rate.   That is, the provider can appeal any rate paid that is less that the FFS rate.   Also, there is a clause that when an out of network provider submits a claim to Med Adv, it is "deemed" to be for the Medicare FFS payment rate. (422.214(c).)

https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-B/part-422/subpart-E/section-422.214

####

There is also a Medicare Advantage policy manual (Internet only manual 16) with about 20 chapters on different aspects of Medicare Advantage.  Chatper 4 section 100 manualizes a reference to 422.214.

https://www.cms.gov/regulations-and-guidance/guidance/manuals/downloads/mc86c06.pdf   see section 100 for reference to 422.214.

It's cited to SSA 1852(a)(2) and to 1852(k)(1), and to a preamble that appeared June 29, 2000.   The June 29, 2000, rulemaking is presented at 65 FR 40325 (and alter amended; see 422.214 webpage, bottom).

https://www.federalregister.gov/documents/2000/06/29/00-15648/medicare-program-medicarechoice-program#page-40325

Discusson at:

https://www.federalregister.gov/d/00-15648/p-1066

The preamable text which helps interpret the intenton of the regulations, is here:

Special Rules for Services Furnished by Noncontract Providers (§ 422.214) 

Consistent with sections 1852(k)(1) and 1866(a)(1)(O) of the Act, § 422.214 requires that any health care provider that does not have in effect a contract establishing payment amounts for services furnished to a beneficiary enrolled in an M+C coordinated care plan must accept, as payment in full, the amounts that they could collect if the beneficiary were enrolled in original Medicare (less the amounts specified in §§ 412.105(g) and 413.86(d) of the regulations on hospital graduate medical education payments, when applicable). Any statutory provisions (including penalty provisions) that apply to payment for services furnished to a beneficiary not enrolled in an M+C plan also apply to the payment described in § 422.214(a)(1) of our regulations. We received three comments regarding this section. 

Comment:

Several commenters suggested that we revise § 422.214 to provide that payment to a noncontracting provider must equal the amount that provider would be allowed to collect under original Medicare. These commenters believe that M+C organizations should only be permitted to pay the billed amount when this is the same amount that Medicare would pay under original Medicare. 

Response:

Section 422.214 implements section 1866(a)(1)(O) of the Act, with respect to services furnished by a “provider of services” as defined in section 1861(u), and section 1852(k)(1), with respect to other services. Neither of these provisions requires an M+C organization to pay a provider more than the amount of the provider's bill, or even impose obligations on M+C organizations at all. Rather, these provisions serve as a limit on the amount the provider can collect from the M+C organization. Specifically, each of these provisions states that a provider “shall accept as payment in full” the amount (less the amounts specified in §§ 412.105(g) and 413.86(d) of the regulations) that it would receive under original Medicare, including cost sharing and permitted balance billing (“the Medicare payment amount”). While this means that under these provisions the provider cannot collect 

more than the Medicare payment amount if its billed amount is 

higher, this obligation to “accept” the Medicare amount as payment in full does not obligate the M+C organization to pay this amount if the provider's bill is 

lower. Thus, in the case of emergency services and certain other services referred to in section 1852(d)(1)(C) of the Act furnished to an enrollee in a coordinated care plan, the provider or providers must accept the Medicare payment amount for the services if their billed amount is higher, but would have no right under sections 1852(k)(1) or 1866(a)(1)(O) to be paid more than the amount of their bill if the billed amount is lower than the Medicare payment amount. 

We note, however, that a provision in the BBA does give providers furnishing services to coordinated care plan enrollees the right to be paid the Medicare payment amount under certain circumstances. Section 1852(a)(2) provides that where an M+C organization chooses to furnish services through providers that do not have contracts with the organization in order to meet its obligation under section 1852(a)(1) to make Medicare services available, it must provide for payment “equal to at 

least

” the Medicare payment amount. (Emphasis added.) This new provision, unlike section 1866(a)(1)(O) or section 1852(k)(1), establishes a “floor” for payment when it applies. This “floor,” combined with the “ceilings” under sections 1866(a)(1)(O) and 1852(k)(1), essentially requires that the Medicare payment amount be paid where section 1852(a)(2) applies. Because section 1852(a)(2) refers to an M+C organization's furnishing services in fulfillment of its obligations under section 1852(a)(1), we are interpreting section 1852(a)(2), in the coordinated care plan context, as providing M+C organizations with the opportunity to arrange to provide nonemergency services through noncontracting providers. Under this interpretation, the “minimum payment” requirement in section 1852(a)(2) would only apply where the M+C organization has arranged for the services in question to be provided by a noncontracting provider...