AZIS R. DABAS

Healthcare strategy
Care, growth + capital

Operating recordMandates
Index
Let’s talk

AZIS R. DABAS / RESEARCH & EXECUTIVE JUDGMENT

Evidence becomes strategy
at the point of decision.

A result matters because it changes what an institution should fund, build, coordinate or stop. These analyses connect scholarly evidence, care delivery and enterprise economics—with the method, limits and decision visible together.

Layered translucent sheets and archival stone slabs framing a precise beam of light
Visual essay / The discipline between evidence and action

Evidence informs the decision when study design, mechanism and operating context are examined together.

01 / THE RESEARCH DESK

What changes the decision?

Six substantial research theses connect recent scientific findings to concrete business and allocation decisions. Nine focused briefs remain in the collection: eight draw on 2025–2026 research, and one revisits a foundational 2019 allocation-bias study. These are original editorial publications by Azis R. Dabas, not journal publications.

15 of 15 publications

Conceptual editorial artwork: Layered translucent sheets and archival stone slabs framing a precise beam of light.

Oncology and precision diagnostics / RESEARCH THESIS

Precision Oncology’s Next Investment Is the Pathway Between the Report and Treatment

Recent evidence supports genomically matched treatment in selected advanced cancers, but a report is only an intermediate product. Service-line investment should target the measured losses between testing, interpretation, patient discussion and feasible treatment.

Original editorial analysis · 3 peer-reviewed sources
Newest source: September 3, 2026
Thesis → evidence → business case → allocation decision
Conceptual editorial artwork: Copper pathways between ivory columns converge on a narrow passage into an open courtyard.

Acute care and community delivery / RESEARCH THESIS

Hospital at Home Must Buy Community Capacity Before It Can Release Hospital Capacity

Recent trials make a credible case for selected patients receiving acute care at home, but they do not make savings automatic. The operating decision is whether a health system can finance reliable community care and turn released hospital capacity into measurable value.

Original editorial analysis · 4 peer-reviewed sources
Newest source: June 22, 2026
Thesis → evidence → business case → allocation decision
Conceptual editorial artwork: Six carved stone terraces joined by a continuous copper channel and a fitted bridge.

Women's health / RESEARCH THESIS

The Postpartum Handoff Is a Business Model: Pay for Continuity Beyond Delivery

Recent evidence supports a practical investment thesis for women's health: fund the transfer from maternity services into continuing care, and judge monitoring by the decisions it enables. The first-year business case should stand on completed transitions and measured blood pressure outcomes, with long-term savings treated as an unproven upside.

Original editorial analysis · 3 peer-reviewed sources
Newest source: August 7, 2026
Thesis → evidence → business case → allocation decision
Conceptual editorial artwork: Copper pathways between ivory columns converge on a narrow passage into an open courtyard.

Aging and dementia / RESEARCH THESIS

Dementia Care Needs a Capacity Contract: Finance the Household, Not Just the Referral

Trials and workforce studies challenge the assumption that more navigation automatically lowers burden or spending. A durable home-care business should purchase reliable response capacity, track family time alongside formal costs, and make the distinction between societal value and cash savings explicit.

Original editorial analysis · 4 peer-reviewed sources
Newest source: December 19, 2025
Thesis → evidence → business case → allocation decision
Conceptual editorial artwork: Sculptural apothecary shelves connected by copper to a small architectural neighborhood.

Cardiometabolic health / RESEARCH THESIS

Obesity Pharmacotherapy Is Becoming a Continuity Business

Recent maintenance and discontinuation evidence changes the commercial unit of obesity care: the valuable product is a durable, clinically supervised treatment pathway. Payers and pharmacy operators should measure continuity, patient burden and total episode cost before treating prescription volume as success.

Original editorial analysis · 3 peer-reviewed sources
Newest source: May 13, 2026
Thesis → evidence → business case → allocation decision
Conceptual editorial artwork: An ivory sculptural balance supported by a copper fulcrum between architecture and new growth.

Kidney care / RESEARCH THESIS

Kidney Prevention Belongs in the Capital Plan Before the Next Dialysis Expansion

Recent kidney research strengthens the case for identifying risk and delivering longitudinal prevention. The capital-allocation challenge is to finance that capability without treating biomarker improvement as booked dialysis savings or weakening capacity for people who already need kidney replacement therapy.

Original editorial analysis · 3 peer-reviewed sources
Newest source: September 23, 2026
Thesis → evidence → business case → allocation decision

Healthcare AI

Ambient AI needs a workflow-specific purchase decision.

A randomized scribe trial makes the buying question more precise: which tool improves which part of the work, for which clinicians, at what total cost?

Three-arm pragmatic randomized clinical trial · November 26, 2025Evidence → mechanism → operating decision

Population health

The target variable is a resource-allocation decision.

A foundational algorithm audit shows why a model can predict spending accurately while directing care-management resources away from people with greater need.

Observational audit of a deployed commercial risk algorithm · October 25, 2019Evidence → mechanism → operating decision

Wireless sensing

A larger training set does not validate every home.

Recent sleep research illustrates the distance between respiratory-model performance, radar measurement and a dependable home-care service.

Model development and validation with transfer to a clinical radar cohort · October 22, 2025Evidence → mechanism → operating decision

Social care

A benefit offer is only the beginning of an intervention.

A produce-prescription trial gives population-health leaders a concrete reason to examine uptake, delivery and the objective purchased before promising medical savings.

Pragmatic randomized clinical trial · February 16, 2026Evidence → mechanism → operating decision

Behavioral health

Behavioral-health integration needs a plan for month thirteen.

Longer follow-up from a collaborative-care trial shifts the operating question from initial engagement to sustained improvement and a practical route back into treatment.

Secondary longitudinal analysis of a randomized controlled trial · January 2026 issue; published online September 2025Evidence → mechanism → operating decision

Pharmacy

Pharmacy value depends on the barrier being removed.

A recent discharge trial and a foundational coverage trial show why medication access, transition safety and utilization savings need distinct investment cases.

Two randomized trials testing different medication-support mechanisms · March 17, 2026; foundational comparison published December 1, 2011Evidence → mechanism → operating decision

Value-based care

A longer savings horizon still needs an access test.

Longitudinal ACO evidence supports patient capital, while leaving leadership responsible for quality, equitable access and the participant's own economics.

Difference-in-differences observational analysis of Medicare spending · April 28, 2025Evidence → mechanism → operating decision

Payer-provider strategy

Integration should earn its place at the handoff.

Hospital-operated Medicare Advantage evidence shows why network strategy needs an episode-level account of utilization, outcomes and the value of ownership.

Cross-sectional observational analysis with adjustment and inverse-probability weighting · August 31, 2026Evidence → mechanism → operating decision

AI economics

Productivity becomes value only when the operating model changes.

Workplace evidence suggests that AI can spread useful expertise unevenly. Healthcare leaders still need to demonstrate how that benefit survives review, handoffs and the commercial contract.

Observational study of staggered workplace deployment · February 4, 2025; May 2025 journal issueEvidence → mechanism → operating decision

Explore the complete 81-article archive ↗

02 / MAKE THE ASSUMPTIONS VISIBLE

The mechanism behind the claim.

All rates in the equations are proportions: 25% = 0.25. Chart axes and tables display percentages.

Four reproducible analytical exhibits connect signal quality, delivery capacity and economic return. All chart values are computed illustrative scenarios, not observed study results or client outcomes. Data and formulas are provided beside each chart.

Illustrative signal: the base-rate effect

Illustrative signal: the base-rate effect; numerical values and method below
Illustrative Bayesian arithmetic over event prevalence 0.1–10%, with sensitivity 90%, specificity 95%, and 10,000 people. PPV is true positives divided by all positive alerts. At 1% prevalence the expected counts are 90 true positives and 495 false positives, giving 15.4% PPV. Expected counts may be fractional; this is not measured device performance.
Inspect data and equations
Selected computed values; full data in CSV.
prevalence pcttrue positivesmissed eventsfalse positivestrue negativesalertspositive predictive value pct
0.191499.59,490.5508.51.77
2.6234264879,25372132.45
5.145951474.59,015.5933.549.17
7.6684764628,7781,14659.69
109001004508,5501,35066.67

true_positives = population * prevalence * sensitivity

false_positives = population * (1 - prevalence) * (1 - specificity)

ppv = true_positives / (true_positives + false_positives) * 100

Download complete chart data · CSV

Illustrative care: capacity and completion

Illustrative care: capacity and completion; numerical values and method below
Illustrative one-period care planning. Expected demand is 10,000 × 60% × 70% × 75% = 3,150 services. Completed services are the lesser of expected demand and staffed hours divided by one hour per completion. Unmet expected demand is the difference. Adding capacity alone does not establish demand, clinical benefit, or financial return.
Inspect data and equations
Selected computed values; full data in CSV.
staffed hoursexpected demandcompletion capacitycompleted servicesunmet expected demand
03,150003,150
1,2503,1501,2501,2501,900
2,5003,1502,5002,500650
3,7503,1503,7503,1500
5,0003,1505,0003,1500

demand = eligible * reach_rate * consent_rate * completion_rate

completed = min(demand, staffed_hours / hours_per_completion)

unmet = demand - completed

Download complete chart data · CSV

Illustrative AI: annual cash and adoption

Illustrative AI: annual cash and adoption; numerical values and method below
Illustrative annual scenario before one-time setup. Net released hours equal adopted tasks × (8 minutes saved − 2 minutes review) ÷ 60; multiply by $60/hour and 35% realization to estimate cash benefit. Buyer cash deducts the $180,000 annual supplier fee and $90,000 other buyer operating cost. Supplier delivery contribution deducts $0.40 per adopted task and $50,000 fixed delivery cost from that same fee. Buyer cash is $504,000 × adoption − $270,000; supplier contribution is $130,000 − $96,000 × adoption. Supplier fees cancel when the accounts are combined. Other buyer costs exclude review time already deducted. These are assumptions, not study results.
Inspect data and equations
Selected computed values; full data in CSV.
adoption pctadopted tasksnet released hourscapacity value after reviewmodeled cash benefitbuyer annual net cashsupplier annual delivery costsupplier annual delivery contributioncombined annual net cash
00000-270,00050,000130,000-140,000
2560,0006,000360,000126,000-144,00074,000106,000-38,000
50120,00012,000720,000252,000-18,00098,00082,00064,000
75180,00018,0001,080,000378,000108,000122,00058,000166,000
100240,00024,0001,440,000504,000234,000146,00034,000268,000

buyer_annual_net_cash = 504000 * adoption - 270000

supplier_annual_delivery_contribution = 130000 - 96000 * adoption

combined_annual_net_cash = 408000 * adoption - 140000

Download complete chart data · CSV

Illustrative AI: five-year cumulative cash

Illustrative AI: five-year cumulative cash; numerical values and method below
Illustrative undiscounted cash at 65% adoption, using the annual assumptions in the adoption chart. The buyer generates $57,600 annually and the supplier $67,600 in annual delivery contribution. Separate time-zero setup costs are $120,000 for the buyer and $30,000 for the supplier. Buyer cumulative cash after n years is $57,600n − $120,000; supplier cumulative delivery cash is $67,600n − $30,000. At five years the results are $168,000 and $308,000. Year zero includes setup and no operating cash. Lines connect year-end observations and do not imply within-year cash timing. Setup costs are internal costs, not another transfer between the parties. No discounting, tax, growth, working-capital change, residual value, or unallocated corporate expense is modeled.
Inspect data and equations
Selected computed values; full data in CSV.
yearbuyer cumulative net cashsupplier cumulative delivery cashcombined cumulative net cash
0-120,000-30,000-150,000
1-62,40037,600-24,800
2-4,800105,200100,400
352,800172,800225,600
4110,400240,400350,800
5168,000308,000476,000

buyer_cumulative_net_cash = year * 57600 - 120000

supplier_cumulative_delivery_cash = year * 67600 - 30000

combined_cumulative_net_cash = year * 125200 - 150000

Download complete chart data · CSV

Change the care-delivery assumptions ↗ · Run the buyer–supplier model ↗

03 / CAPITAL, BUSINESS MODELS & OPERATING REALITY

Follow the operating economics across the healthcare system

Healthcare executives should read the 2026 investment conversation as a debate over where capital can change the cost or reliability of care. The useful connecting question is who controls access, operational throughput, and payment. That framing reaches beyond AI into employer benefits, medicines, diagnostics, staffing, home care, and the systems joining them. The following interpretation combines distinct investor and analyst perspectives; it is neither a statement of consensus nor clinical evidence.

Capital is not expressing one uniform healthcare thesis. PitchBook's August public listings pair a cautious healthcare services outlook with a healthcare IT headline describing deal activity on a record-setting pace. These are different markets, and the inaccessible underlying reports prevent a detailed comparison. Still, the contrast is a useful executive prompt: a company's ability to attract buyers does not answer whether the care organization using its product has investment capacity. Procurement decisions should therefore include vendor ownership, integration plans, continuity, and future pricing, alongside immediate functionality. This is a diligence implication, not a finding from the reports.

The workflow opportunity is similarly wider than automating documentation. YC's Fall 2026 request on aging focuses attention on coordination, monitoring, accessible interfaces, and assistance at home. For an operator, the unit of analysis should be the entire journey: who notices a problem, who responds, which appointment is arranged, and whether information reaches the next caregiver. A product can make one step faster while shifting work elsewhere. Useful evaluation would therefore examine response times, missed handoffs, caregiver burden, and avoidable escalation, while recording implementation cost and the staff capacity needed to support the service.

Payment and distribution determine who can capture that value. a16z's September employer-plan thesis points toward challenger insurers, PBMs, and infrastructure businesses competing for benefit spending. Its January discussion also raises the possibility of more consumer purchasing outside insurance. Read together, these arguments suggest examining both the purchaser and the user. Employer savings, member affordability, provider reimbursement, and the vendor's revenue need not move together. A purchasing process should ask which party carries clinical and financial risk, how patients reach appropriate care, and whether a lower headline fee merely relocates costs.

Pharmacy and life science infrastructure belong in the same discussion. Morningstar's Cigna note distinguishes insurance margin progress from a PBM transition; its life science outlook anticipates a profit recovery after a prolonged reset. These perspectives direct attention to established businesses, purchasing channels, and operating cycles that can shape value even without a new consumer application. The executive question is how contracting, purchasing power, supply capacity, and reimbursement distribute the economic benefit. Equity forecasts remain forecasts, and a company's margin recovery alone cannot establish better access or lower costs for patients.

A practical board review can separate three propositions: the operational change, the financial mechanism, and the patient consequence. Each deserves its own evidence. Saving staff minutes matters only if capacity is actually redeployed; faster payment matters only after considering disputes and rework; broader access matters only if the service meets the needs of the population it reaches.

Finally, cheaper tasks do not automatically mean a cheaper system. a16z's February care-pricing argument explicitly connects expanded capacity with payment design. An executive can use that hypothesis to test utilization, downstream resource use, and outcomes together rather than assume that unit-cost reductions become net savings. Funding momentum, attractive margins, startup enthusiasm, and valuation estimates can identify questions worth investigating. They cannot establish clinical safety, effectiveness, equity, or causal savings. Those claims require appropriate primary studies and local evaluation with relevant comparators, patient populations, follow-up, and the full cost of implementation.

Market sources and access scope

  1. a16z · Employers Are Shopping for a New Health Plan — 2026-09-10. Investor thesis. First-party full article retrieved. An investor's proposed market opportunity, with commercial interests. The article does not establish that challengers outperform incumbent plans or that automation improves clinical outcomes.
  2. a16z · Infinite Healthcare: What’s It Worth? — 2026-02-26. Investor thesis and economic scenarios. First-party full article retrieved. Prospective economic argument with illustrative scenarios, not a measured demonstration that AI reduces total healthcare spending or improves population health.
  3. a16z · Healthcare 2026: AI Doctors, GLP-1s, and Insurance Defection — 2026-01-27. Investor and industry commentator discussion. First-party episode description retrieved; audio not independently transcribed. Summary of a discussion rather than a population-level utilization study. No claim that cash payment is appropriate or accessible for every patient.
  4. Y Combinator · Requests for Startups — Fall 2026: AI for the Aging Population — Fall 2026. Startup opportunity request. First-party live page retrieved. An invitation to founders, not a validated market-size model or clinical intervention study. The live RFS page changes; preserve the edition label. Demographic and caregiver numbers were not independently verified and should not be reused as established facts.
  5. PitchBook · Healthcare Services Report: Gusting Macroeconomic Headwinds Impede Progress — 2026-08-14. Private market research; public report listing. First-party public title, date, and scope verified. Full report returned HTTP 403 and was not accessed.. Do not infer detailed deal counts, causal explanations, geography, return forecasts, or patient effects from the public listing. No full-report statistics are asserted here.
  6. PitchBook · Healthcare IT PE Update: Deal Counts on Pace to Surpass 2021 Record — 2026-08-17. Private market research; public report listing. First-party public title, date, and scope verified. Full report returned HTTP 403 and was not accessed.. A pace estimate in a headline is not a completed full-year outcome. The report's underlying sample, methodology, and figures were unavailable; no detailed numerical finding is reproduced.
  7. Morningstar · Cigna Earnings: Medical Insurance Margin Progress Constrained by PBM Business Model Transition — 2026-07-30. Equity analyst note; public indexed excerpt. Title, dated listing, and public excerpt retrieved through search. Direct article and index opens returned HTTP 403.. Company-focused investment analysis. It does not demonstrate lower patient drug costs, better clinical outcomes, or a universal trend across all PBMs.
  8. Morningstar · After a Long Reset Period, Life Science Stocks Look Set to Rebound — 2026-03-23. Equity analyst outlook; public indexed excerpt. Public first-party indexed excerpt and publication date verified; direct page open returned HTTP 403.. Profit expectations and valuation judgments are forecasts. They do not establish that a therapeutic program works, that laboratory demand has recovered everywhere, or that the forecast has materialized.
Artistic portrait of Azis R. Dabas

THE AUTHOR & THE STANDARD

An operating perspective.
A visible evidence trail.

I am a healthcare executive and founder whose operating work spans specialty pharmacy, a multi-site kidney-care network, claims intelligence and behavioral-health growth. My analysis examines how population outcomes, delivery capacity, incentives and capital allocation fit together.

Each research thesis and brief distinguishes what a study reports, what its design cannot establish and the operating decision I would take from it. Models are illustrative. Case outcomes are self-reported and separated from proposed architectures. Source affiliations identify the researchers, not the author.

Historical articles retain their original context. This selected review is not exhaustive. Investor commentary is a source of hypotheses, not evidence of clinical effectiveness. Corrections and material updates belong with the affected claim.

Executive profile and operating record ↗ · Discuss a mandate ↗

FROM EVIDENCE TO EXECUTIVE ACTION

What would this change in your organization?