Retrospective aggregate cohort study · TriNetX US Collaborative Network · Aug 19, 2026

The Missed Moment of ADT Initiation

Henry M. Blair · Departments of Urology, SUNY Downstate Health Sciences University & Maimonides Medical Center, Brooklyn, NY

When an older man starts androgen-deprivation therapy (ADT) for prostate cancer, one moment carries two guideline duties: monitor his bones (ADT causes bone loss and fracture) and, if his cancer is already metastatic, intensify treatment (add an ARPI or docetaxel, which improves survival). Examined together in one population, both duties fail at the same moment.

69.3%

never receive a central bone scan

of 35,851 men on ADT

41.1%

intensified when metastatic

58.9% get neither agent

3.6×

screening tracks prior bone disease

not triggered by ADT itself

0.90

Black vs White screening (RR)

intensification 0.96

One snapshot, sourced from this folder. Every number is read from Henry - PCa Data Master.xlsx and every figure from pca_figures/ (produced by pca_figures.py); the page regenerates with build_hub.py. All counts are the Aug 19, 2026 extraction (network 61,114,901; 72 of 74 HCOs) — nothing is rescaled.

How the cohorts were built →

Full derivation from 61,114,901 network patients to every numerator, with the query-builder logic.

The figures →

Screening, intensification, agent mix, comorbidity and disparities — the manuscript figures.

The paper — on request →

The manuscript is not posted publicly while the work is under review. Email me and I will send it.

The two care gaps, in plain language

Gap 1 — bone-health monitoring is largely absent

ADT accelerates bone loss, so guidelines say to check bone density (a DXA scan) around when it starts. In practice 69.3% of men never get a central DXA at any point, and uptake barely rises over ten years. When it happens it is 3.6× more likely in men who already had a bone-disease diagnosis before ADT — the scan follows a problem already on the chart, not the treatment that creates new risk. Screening is also delivered less often to Black and Asian men.

Gap 2 — survival-improving treatment is under-delivered

For metastatic men, adding an ARPI or docetaxel improves survival, yet only 41.1% were intensified; 58.9% received neither agent. Use rose from 25.5% (2017) to 46.7% (2023) — driven by ARPI while docetaxel stayed flat — then plateaued. The un-intensified men were only modestly frailer, and men with visceral disease were intensified no more often (RR 1.00).

The unifying thesis: both failures converge on the same clinical moment — ADT initiation — and share a pattern. Care is organized around problems that have already declared themselves (a coded metastasis, a diagnosed bone disease) rather than around the foreseeable consequences of starting the treatment.

Cohort & query derivation

Each cohort was carved by ANDing one criterion at a time onto its parent, anchored to the first incident ADT term. The Mets1 partition closes exactly (3,170 + 4,541 = 7,711).

Query builders — how each query was constructed

Plain-language build logic for every cohort and numerator: terms, codes, relationship anchors, windows, instance/exclusion rules.

Rules applied to (almost) every query

Global filter: Male patients only.
Prostate cancer base: ICD-10-CM C61 recorded on ≥2 occasions (two ANDed C61 groups).
Age ≥70 applied as Age-at-Event on the C61 term (not the platform's global current-age filter).
Every temporal relationship is anchored to the FIRST incident ADT term; other criteria are defined relative to it.
Counts of ≤10 are obfuscated by the platform (low-count floor).

Networkn = 61,114,901

TriNetX US Collaborative Network

Full de-identified network denominator at the Aug 19, 2026 extraction.

Aggregate counts computed per contributing HCO; ≤10 obfuscated.

Base cohortn = 1,048,285

Prostate cancer, men ≥70 · “PCa(2x) ≥70M”

C61 on ≥2 occasions + Age-at-Event ≥70 on the C61 term + Male global filter.

ICD-10-CM C61 (×2); Age-at-Event ≥70

Base C61 male, Aug 5 snapshot (1,048,285).

Exposuren = 61,022

+ ADT union (8 terms)

AND an 8-term ADT-union group (all ORed): GnRH agonists leuprolide/goserelin/triptorelin/histrelin, antagonists degarelix/relugolix, and orchiectomy (CPT 54520/54690).

RxNorm 42375·50610·38782·50975·2472778·475230; CPT 54520·54690

First-generation antiandrogens (bicalutamide/flutamide) deliberately excluded from the defining exposure.

Window filtern = 38,685

+ within 2017–2023 index window

Restrict the first ADT term to Jan 1 2017 – Dec 31 2023 (pre-washout ADT Union (2x)).

ADT Union (2x) in 2017–23 window.

Analytic cohortn = 35,851

Incident-ADT cohort (IADT)

AND a 12-month clean look-back CANNOT-HAVE group (no ADT of any kind in the year before the first ADT term) to enforce incident use.

PRIMARY denominator for Study 6.

Analytic cohortn = 7,711

Mets1 (metastatic ≤1 mo)

AND a distant-metastasis group (any of C79.51/C79.52/C78.7/C78.0/C79.31) related to the first ADT term from any time before to ≤1 month after it. Nodal C77.x deliberately excluded (N1, not M1).

ICD-10-CM C79.51·C79.52·C78.7·C78.0·C79.31 (nodal C77.x excluded)

PRIMARY metastatic cohort. Partition below closes: 3,170 + 4,541 = 7,711.

Analytic cohortn = 33,957

v1_SIADT (sustained-ADT base)

Within IADT, require a documented prostate-cancer (C61) encounter ≥6 months after the first ADT term (sustained exposure), reducing misclassification of transient/limited ADT.

C61 encounter ≥6 mo after ADT

Sustained-care denominator for the screened/never comparison.

Numeratorn = 2,549

ARPI-Ia4 (ARPI ≤4 mo)

Within Mets1, AND an ARPI-union group (abiraterone/enzalutamide/apalutamide/darolutamide) initiated ≤4 months on/after the first ADT term.

RxNorm 1100072·1307298·1999574·2180325; ≤4 mo of ADT

ARPI component of intensification.

Numeratorn = 782

ADT-Id4 (docetaxel ≤4 mo)

Within Mets1, AND a docetaxel group (RxNorm 72962 OR HCPCS J9171) initiated ≤4 months on/after the first ADT term. Both billing routes captured.

RxNorm 72962 · HCPCS J9171; ≤4 mo of ADT

Docetaxel component; chemo is often billed under HCPCS, not RxNorm.

Numeratorn = 3,170

Intensified = Ia4 OR Id4

Direct union query: Mets1 AND (ARPI ≤4 mo OR docetaxel ≤4 mo). Primary intensification numerator.

41.1% of Mets1.

Numeratorn = 4,541

Not intensified = Mets1 AND NOT (Ia4 OR Id4)

Direct complement query: Mets1 AND CANNOT-HAVE (ARPI ≤4 mo OR docetaxel ≤4 mo).

58.9% of Mets1. Exactly complements Intensified.

Numeratorn = 4,691

Central DXA ≤1 y

Within IADT, AND a central-DXA group (CPT 77080/77085/77086) related to the first ADT term from same day to ≤1 year after.

CPT 77080·77085·77086; ≤1 y of ADT

12-mo PRIMARY screening numerator (13.1%).

Numeratorn = 24,849

Never central DXA (any time)

Within IADT, CANNOT-HAVE any central-DXA code at any point in the record (no temporal relationship).

CANNOT HAVE CPT 77080·77085·77086

HEADLINE: 69.3% never screened. Immune to the ‘screened later’ objection.

Subsetn = 161

Triplet = Ia4 AND Id4

Mets1 AND ARPI ≤4 mo AND docetaxel ≤4 mo (ADT + ARPI + docetaxel).

Both agents within 4 months.

Subsetn = 2,078

Mets1 + visceral (lung/liver/brain)

Within Mets1, AND a visceral-metastasis group (C78.0 lung / C78.7 liver / C79.31 brain).

ICD-10-CM C78.0·C78.7·C79.31

vs bone-only (non-visceral) 5,633.

Sensitivityn = 8,279

Mets3 (metastatic ≤3 mo) — sensitivity

As Mets1 but the metastasis window is extended to ≤3 months after first ADT.

same metastasis code set

Sensitivity variant; differs <0.3 pp at every ARPI horizon.

Comparison armn = 4,633

2a: SIADT × central DXA ≤1 y (screened)

v1_SIADT AND central DXA ≤1 y of ADT. Pre-ADT bone-disease covariates attached.

Screened arm of the screened-vs-never comparison.

Comparison armn = 23,633

2b: SIADT × never central DXA (never)

v1_SIADT AND CANNOT-HAVE any central DXA. Pre-ADT bone-disease covariates attached.

Never-screened arm; exact complement within SIADT of the screened arm.

Code lists & ladders

ADT union (8 terms, ORed)

Agent	System	Code
Leuprolide	RxNorm	42375
Relugolix	RxNorm	2472778
Degarelix	RxNorm	475230
Goserelin	RxNorm	50610
Triptorelin	RxNorm	38782
Histrelin	RxNorm	50975
Orchiectomy, simple	CPT	54520
Orchiectomy, laparoscopic	CPT	54690

Excluded from ADT union

Item	Reason
CPT 55250	VASECTOMY — inflated the cohort to 79,094; excluded
bicalutamide / flutamide / nilutamide	peripheral antiandrogens; flare cover, not castrating ADT
ketoconazole	not ADT
Identifier 1014948	internal rollup, redundant with 77080

ARPI union

Agent	System	Code
Abiraterone	RxNorm	1100072
Enzalutamide	RxNorm	1307298
Apalutamide	RxNorm	1999574
Darolutamide	RxNorm	2180325

Docetaxel

Agent	System	Code
Docetaxel (ingredient)	RxNorm	72962
Injection, docetaxel 1mg	HCPCS	J9171

Metastasis (M1) ICD-10-CM

Site	Code
Bone	C79.51
Bone marrow	C79.52
Liver	C78.7
Lung	C78.0
Brain	C79.31

DXA (Central=77080/85/86; Any +77081)

Type	System	Code
Axial	CPT	77080
Axial + VFA	CPT	77085
VFA	CPT	77086
Peripheral	CPT	77081

Pre-ADT bone-disease covariates

Diagnosis	System	Code
Osteoporosis	ICD-10-CM	M81
Osteopenia / bone-density	ICD-10-CM	M85.8
Disorders of bone density	ICD-10-CM	M80–M85
Osteoporosis fracture history	ICD-10-CM	Z87.310
Osteoporosis screening encounter	ICD-10-CM	Z13.820

Data & results workbook · Aug 19, 2026

Headline results as held in the master workbook — every value read live from the raw-counts sheet.

Study 6 — bone-density screening

Window / measure	n	% of 35,851	HCOs
Central DXA ≤1 y (primary)	4,691	13.1%	49
Central DXA ≤2 y	6,486	18.1%	51
Central DXA ≤5 y	8,772	24.5%	51
Central DXA ≤10 y	9,425	26.3%	51
Never central DXA (headline)	24,849	69.3%	—

Study 6 screening (denominator = incident-ADT 35,851).

Study 7 — treatment intensification by year

Index year	Metastatic (Mets1)	ARPI n (%)	Union n (%)	Docetaxel n (%)	Triplet n (%)
2017	1,773	220 (12.4%)	452 (25.5%)	244 (13.8%)	12 (0.7%)
2018	2,165	379 (17.5%)	663 (30.6%)	303 (14.0%)	19 (0.9%)
2019	2,541	555 (21.8%)	884 (34.8%)	353 (13.9%)	24 (0.9%)
2020	2,730	729 (26.7%)	1,048 (38.4%)	346 (12.7%)	27 (1.0%)
2021	3,148	1,005 (31.9%)	1,334 (42.4%)	365 (11.6%)	36 (1.1%)
2022	3,685	1,313 (35.6%)	1,637 (44.4%)	390 (10.6%)	66 (1.8%)
2023	4,054	1,570 (38.7%)	1,893 (46.7%)	437 (10.8%)	114 (2.8%)
Pooled	7,711	2,549 (33.1%)	3,170 (41.1%)	782 (10.1%)	161 (2.1%)

Table 5. Treatment intensification by index year against the Mets1 cohort (workbook, Aug 19, 2026). Year denominators are independent single-year queries and sum above the pooled cohort.

Comorbidity prevalence ratios

Comorbidity	Intensified % (n=2,797)	Non-intensified % (n=4,062)	Prevalence ratio [95% CI]
Diabetes, type 2 (E11)	37.8%	38.6%	1.02 [0.96–1.08]
Chronic kidney disease (N18)	34.4%	36.1%	1.05 [0.98–1.12]
Heart failure (I50)	27.7%	27.7%	1.00 [0.93–1.08]
COPD (J44)	13.9%	17.5%	1.26 [1.12–1.41]
Dementia (F03)	6.3%	8.5%	1.36 [1.14–1.62]
Alzheimer disease (G30)	1.6%	2.7%	1.71 [1.21–2.41]
Coronary artery disease (I25)	48.7%	51.8%	1.06 [1.01–1.12]
Atrial fibrillation (I48)	29.0%	28.9%	1.00 [0.92–1.08]

Table 7. Intensified vs non-intensified (docetaxel-inclusive union cohorts). Prevalence ratio = non-intensified ÷ intensified. Bold PRs exclude 1.

Disparities — bone-density screening (≤12 mo)

Group	ADT cohort n	Screened n	Rate	RR vs White
White	26,289	3,479	13.2%	1.00 (ref)
Black	5,521	656	11.9%	0.90 [0.83–0.97]
Hispanic	1,660	235	14.2%	1.07 [0.95–1.21]
Asian	1,284	145	11.3%	0.85 [0.73–1.00]

Disparities — treatment intensification

Group	Mets1 n	Intensified n	Rate	RR vs White
White	5,708	2,339	41.0%	1.00 (ref)
Black	1,043	409	39.2%	0.96 [0.88–1.04]
Hispanic	385	152	39.5%	0.96 [0.85–1.09]
Asian	266	133	50.0%	1.22 [1.08–1.38]

Figures pca_figures.py

The manuscript figures, produced by pca_figures.py from the same workbook and referenced live — re-run it to update them here.

**Figure 2.** Central DXA at any point among men ≥70 initiating ADT (denominator 35,851). Seven in ten men were never centrally screened.

**Figure 3.** Cumulative central vs any DXA uptake after ADT initiation, to five years, with sustained-exposure and survival-controlled sensitivity analyses.

**Figure 8.** Treatment intensification by index year: union (ARPI or docetaxel), ARPI, docetaxel, and triplet, with 95% Wilson CIs.

**Figure 7.** Composition of the metastatic cohort (n = 7,711) once docetaxel is counted: ARPI-only, docetaxel-only, triplet, and not intensified.

**Figure 9.** ARPI agent share within intensified men (pooled). Agents are not mutually exclusive.

**Figure 10.** Comorbidity prevalence among non-intensified vs intensified men (prevalence ratios, log scale).

**Figure 4.** Pre-existing (pre-ADT) bone disease among promptly screened vs never-screened men (prevalence ratios, log scale). Every marker lies right of 1.0.

**Figure 6.** Three bone-health outcomes in men with vs without pre-existing bone disease at ADT initiation (risk ratios, log scale).

**Figure 5.** Bone-protective initiation and fragility-fracture incidence, screened vs never-screened non-metastatic men (incident-user cohort). Screened men both treat and fracture more — confounding by indication.

**Figure 12.** Screening and intensification by race, as the risk ratio vs White men. Screening rates diverge by race; intensification rates do not.

**Figure 11.** Treatment intensification by metastatic site — visceral vs bone-only. Rates are identical (RR 1.00).

Methods, pitfalls & limitations

Query-builder pitfalls — each produced a wrong number before it was caught

Never put a time constraint on a group that serves as a relationship anchor — it changes what “first instance” means and slides the index date forward, inflating counts.
Instance rules and term filters must sit on separate ANDed groups (an instance rule on a filtered term collapsed 4,201 → 217).
Groups containing CANNOT-HAVE terms cannot take an instance rule; keep washout and instance-count groups separate.
Relationship direction inverts easily — the washout you want is “From 1 Year Before → To Same Day” on a CANNOT-HAVE group.
AND inside a CANNOT-HAVE block means “had both” — every operator in an exclusion list must be OR.
Read the description the tool returns for every code — CPT 55250 is vasectomy and inflated the ADT cohort by ~19,000.
ICD-O is not ICD-10-CM — the Oncology tab returns sparse tumour-registry codes.
The global age filter is current age, not age at event — apply Age-at-Event on the C61 term.
A restriction can never raise a count; if a number goes up after tightening, something else loosened.
Name and star every query as you build it — unnamed queries are unrecoverable within a day.

Limitations

Aggregate counts only — no patient-level adjustment beyond the marginal comparisons, so residual confounding (especially unmeasured frailty in the intensification comparison) cannot be excluded. The network is weighted toward U.S. academic organizations and is predominantly White. Incident-user status is approximate, and the metastatic denominator grew over the study period as organizations were added — but because the intensification estimate is a proportion, shared accrual cancels.

References

44 of 46 entries carry a verified DOI (NCCN and a Science news item are cited by URL).

#	Authors	Title / source	DOI
	mCSPC / mHSPC treatment intensification — pivotal trials
1	Fizazi K, et al.	Abiraterone plus prednisone in metastatic, castration-sensitive prostate cancer (LATITUDE) N Engl J Med 2017; 377(4):352-360	10.1056/NEJMoa1704174
2	James ND, et al.	Abiraterone for prostate cancer not previously treated with hormone therapy (STAMPEDE) N Engl J Med 2017; 377(4):338-351	10.1056/NEJMoa1702900
3	Chi KN, et al.	Apalutamide for metastatic, castration-sensitive prostate cancer (TITAN) N Engl J Med 2019; 381(1):13-24	10.1056/NEJMoa1903307
4	Davis ID, et al.	Enzalutamide with standard first-line therapy in metastatic prostate cancer (ENZAMET) N Engl J Med 2019; 381(2):121-131	10.1056/NEJMoa1903835
5	Armstrong AJ, et al.	ARCHES: ADT with enzalutamide or placebo in metastatic hormone-sensitive prostate cancer J Clin Oncol 2019; 37(32):2974-2986	10.1200/JCO.19.00799
6	Smith MR, Hussain M, et al.	Darolutamide and survival in metastatic, hormone-sensitive prostate cancer (ARASENS) N Engl J Med 2022; 386(12):1132-1142	10.1056/NEJMoa2119115
7	Sweeney CJ, et al.	Chemohormonal therapy in metastatic hormone-sensitive prostate cancer (CHAARTED) N Engl J Med 2015; 373(8):737-746	10.1056/NEJMoa1503747
8	James ND, et al.	Docetaxel, zoledronic acid, or both added to first-line hormone therapy (STAMPEDE) Lancet 2016; 387(10024):1163-1177	10.1016/S0140-6736(15)01037-5
9	Gravis G, et al.	ADT alone or with docetaxel in non-castrate metastatic prostate cancer (GETUG-AFU 15) Lancet Oncol 2013; 14(2):149-158	10.1016/S1470-2045(12)70560-0
10	Fizazi K, et al.	Abiraterone added to ADT and docetaxel in de novo mCSPC (PEACE-1) Lancet 2022; 399(10336):1695-1707	10.1016/S0140-6736(22)00367-1
	Real-world underuse of intensification
11	Raval AD, et al.	Real-world combination therapy use in mHSPC, US 2017-2023 JCO Oncol Pract 2025; 21(8):1174-1184	10.1200/OP-24-00690
12	Swami U, et al.	Physician specialty and underutilization of treatment intensification in mCSPC J Urol 2023; 209(6):1120-1131	10.1097/JU.0000000000003370
	ADT, bone loss, fracture & bone-targeted agents
13	Shahinian VB, et al.	Risk of fracture after androgen deprivation for prostate cancer N Engl J Med 2005; 352(2):154-164	10.1056/NEJMoa041943
14	Smith MR, et al.	Denosumab in men receiving ADT for prostate cancer (HALT) N Engl J Med 2009; 361(8):745-755	10.1056/NEJMoa0809003
15	Smith MR, et al.	Pamidronate to prevent bone loss during ADT for prostate cancer N Engl J Med 2001; 345(13):948-955	10.1056/NEJMoa010845
16	Smith MR, et al.	Zoledronic acid to prevent bone loss during ADT (nonmetastatic prostate cancer) J Urol 2003; 169(6):2008-2012	10.1097/01.ju.0000063820.94994.95
17	Michaelson MD, et al.	Annual zoledronic acid to prevent GnRH-agonist-induced bone loss J Clin Oncol 2007; 25(9):1038-1042	10.1200/JCO.2006.07.3361
18	Saylor PJ, Smith MR	Metabolic complications of androgen deprivation therapy for prostate cancer J Urol 2009; 181(5):1998-2006	10.1016/j.juro.2009.01.047
	DXA / bone-density screening underuse
19	Suarez-Almazor ME, et al.	Low rates of BMD measurement in Medicare patients initiating ADT Support Care Cancer 2014; 22(2):537-544	10.1007/s00520-013-2008-z
20	Shahinian VB, Kuo YF	Patterns of BMD testing in men receiving androgen deprivation for prostate cancer J Gen Intern Med 2013; 28(12):1586-1591	10.1007/s11606-013-2477-2
	Guidelines — bone health & advanced prostate cancer
21	Saylor PJ, et al.	Bone health and bone-targeted therapies for prostate cancer: ASCO endorsement of a CCO guideline J Clin Oncol 2020; 38(15):1736-1743	10.1200/JCO.19.03148
22	Saylor PJ, et al.	Bone health and bone-targeted therapies for prostate cancer: ASCO endorsement summary JCO Oncol Pract 2020; 16(3):143-146	10.1200/JOP.19.00778
23	Lowrance WT, et al.	Advanced prostate cancer: AUA/ASTRO/SUO guideline — part I J Urol 2021; 205(1)	10.1097/JU.0000000000001375
24	Lowrance WT, et al.	Advanced prostate cancer: AUA/ASTRO/SUO guideline — part II J Urol 2021; 205(1)	10.1097/JU.0000000000001376
25	Lowrance W, et al.	Updates to advanced prostate cancer: AUA/SUO guideline (2023 amendment) J Urol 2023; —	10.1097/JU.0000000000003452
26	(consensus)	Multidisciplinary consensus on osteoporosis/fragility fractures in prostate cancer on ADT World J Mens Health 2022; —	10.5534/wjmh.210061
27	NCCN	NCCN Clinical Practice Guidelines in Oncology: Prostate Cancer (cite version + date accessed) NCCN 2026; v3.2026	no DOI — nccn.org/guidelines
	Racial & ethnic disparities
28	Dess RT, et al.	Association of Black race with prostate cancer-specific and other-cause mortality JAMA Oncol 2019; 5(7):975-983	10.1001/jamaoncol.2019.0826
29	Nyame YA, et al.	Racial inequities in surgical care quality among Medicare men with localized prostate cancer Cancer 2023; —	10.1002/cncr.34681
30	Nyame YA, et al.	Isolating the drivers of racial inequities in prostate cancer treatment Cancer Epidemiol Biomarkers Prev 2024; 33(3):435	10.1158/1055-9965.EPI-23-0892
31	(RWE study)	Emerging racial disparities among Medicare & Veterans with mCSPC Prostate Cancer Prostatic Dis 2024; —	10.1038/s41391-024-00815-1
32	Noel SE, et al.	Racial and ethnic disparities in bone health and outcomes in the US J Bone Miner Res 2021; 36(10)	10.1002/jbmr.4417
33	Calikyan A, et al.	Osteoporosis screening disparities among ethnic and racial minorities: systematic review J Osteoporos 2023; 2023:1277319	10.1155/2023/1277319
	Real-world-evidence methodology & study design
34	Hernan MA, Robins JW	Using big data to emulate a target trial when a randomized trial is not available Am J Epidemiol 2016; 183(8):758-764	10.1093/aje/kwv254
35	Suissa S	Immortal time bias in pharmacoepidemiology Am J Epidemiol 2008; 167(4):492-499	10.1093/aje/kwm324
36	Ray WA	Evaluating medication effects outside of clinical trials: new-user designs Am J Epidemiol 2003; 158(9):915-920	10.1093/aje/kwg231
37	Lund JL, et al.	The active comparator, new-user study design in pharmacoepidemiology Curr Epidemiol Rep 2015; 2(4):221-228	10.1007/s40471-015-0053-5
38	Langan SM, et al.	RECORD-PE: reporting of studies using routinely-collected data for pharmacoepidemiology BMJ 2018; 363:k3532	10.1136/bmj.k3532
39	Benchimol EI, et al.	The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) PLoS Med 2015; 12(10):e1001885	10.1371/journal.pmed.1001885
	TriNetX network — methodology & critiques
40	Liu, et al.	On the reported methodology in TriNetX-based studies: impossible index event designs Eur J Epidemiol 2025; —	10.1007/s10654-025-01342-6
41	Williford SE, et al.	Re: impossible index event designs in TriNetX-based studies (correspondence) Eur J Epidemiol 2026; —	10.1007/s10654-026-01374-6
42	Nassar M, et al.	TriNetX and real-world evidence: critical review of strengths, limitations & bias ASIDE Intern Med 2025; 1(2):24-32	10.71079/ASIDE.IM.03222516
43	(review)	Comprehensive review of methodologies to use the TriNetX real-world-data platform Front Pharmacol 2025; 16:1516126	10.3389/fphar.2025.1516126
44	Wang J, et al.	Analyzing clinical laboratory data in retrospective cohort studies using TriNetX Biochem Med (Zagreb) 2025; 35(3):030502	10.11613/BM.2025.030502
45	Wang J, et al.	Bibliometric analysis of TriNetX utilization in Taiwan Tzu Chi Med J 2025; 37:452-456	10.4103/tcmj.tcmj_279_24
46	Joelving F	Medical students are using a popular research tool to pump out misleading studies Science (news) 2026; —	no DOI — science.org

About & data provenance

Data source. TriNetX US Collaborative Network — a federated network returning de-identified aggregate counts per contributing organization; counts of ≤10 are obfuscated. No patient-level records leave their source, so the study is IRB-exempt.

Snapshot. Aug 19, 2026 — network 61,114,901; 72 of 74 HCOs online. Every reported value traces by formula to one raw-counts sheet; nothing is rescaled.

Self-sourcing. This page is generated by build_hub.py from Henry - PCa Data Master.xlsx (numbers) and pca_figures/ (figures, from pca_figures.py). Change the spreadsheet or re-run the figures, re-run build_hub.py, and the page updates.

Compliance. Henry M. Blair, SUNY Downstate College of Medicine (research year at Maimonides), with Dr. Jeffrey P. Weiss as PI. TriNetX policy prohibits sharing data or analyses with outside institutions — this is a Downstate-only submission; Maimonides collaborators do not appear on the protocol or receive analyses.