Market Access

Many new prescriptions fail when a prior authorization (PA) is required because the PA process adds delays, paperwork, and administrative complexity that patients and providers often struggle to complete quickly. Long approval times, missing documentation, payer denials, and high out-of-pocket costs frequently cause patients to abandon therapy before the first fill. Without proactive outreach, automated PA support, and real-time visibility into case status, these friction points lead to significant drop-off at the start of treatment.

You can track step therapy (“fail-first”) across hundreds of health plans by bringing formulary and utilization management (UM) data into one standardized database. Pull policy details from payer documents and data vendors, then normalize them so every plan follows the same structure.

‍

Use coverage fields that clearly flag step therapy and prior authorization by plan, product, and line of business. This lets you map each drug to its required steps, compare rules across payers, and see access differences at national and regional levels.

‍

Then add two layers: policy-change alerts and outcome linkage. Alerts tell you when rules change, and linking to claims or patient journey data shows how step edits affect real outcomes like time-to-therapy, abandonment, and overall access.

You can detect where access restrictions increased in the last quarter by comparing the following payer-level utilization management metrics against the previous quarter.

• prior authorization rates
• step therapy enforcement
• rejection patterns
• formulary status changes

‍

Claims data, formulary intelligence sources, and hub activity can reveal spikes in PA volume, growing denial rates, or new fail-first requirements. By automatically highlighting plans where these indicators moved negatively, analytics makes it easy to pinpoint which payers tightened access and quantify the impact on prescriptions and patient progression.

You can identify which prior authorization criteria cause the most rejections by analyzing denial codes, documentation requirements, and clinical criteria within payer PA responses to see which rules are most frequently unmet.

‍

By linking rejection reasons to specific criteria, such as

• missing step therapy evidence
• unsupported diagnosis codes
• insufficient lab results
• incomplete paperwork

‍

you can pinpoint the drivers of failed PAs. Layering this with claims and HUB data highlights which criteria create the biggest barriers, helping teams target education, documentation improvements, and payer-specific strategies.

You can quantify the impact of prior authorizations and step edits by comparing prescription volume for patients exposed to these requirements against those who are not, using claims and hub data to track drop-off at each step of the access process.

‍

By measuring the following and linking them to TRx, NBRx, and first-fill outcomes, you can calculate how many prescriptions were delayed or lost due to UM (Utilization Management) barriers (access rules that can delay or block a fill, even when a prescription is written).

‍

• PA submission rates
• approval times
• denial patterns
• step therapy failures

‍

This analysis highlights the true volume impact and identifies where targeted interventions can recover starts.

You can measure how long it takes for patients to start therapy under different payers by tracking the timestamps between key steps in the access journey and comparing these intervals across plans:

‍

• Prescription
• Benefits verification
• Prior authorization
• Approval
• First fill

‍

Claims, hub, and specialty pharmacy data reveal where delays occur, such as

• slow PA processing
• high denial rates
• long patient outreach cycles

‍

By benchmarking each payer’s average time-to-therapy and highlighting outliers, analytics makes it easy to identify which plans create the biggest access bottlenecks.

Hub teams that handle prior authorizations (PAs) see big volume swings when payers change policies or when competitors launch and trigger formulary reviews. If you understaff, patients drop off. If you overstaff, you burn budget.

‍

Time-series forecasting of PA volume by payer and therapy lets you predict these spikes and dips. This supports dynamic staffing models: adding capacity when PA volume is expected to surge and scaling back when it falls, so you maintain service levels and control costs at the same time.

Prior authorization approval rates can range from 30% to 80% depending on the payer, but many manufacturers only see aggregate denial data monthly or quarterly. Real-time tracking requires pulling together data from hub services, specialty pharmacies, and claims reversals into one view.

‍

Tellius can connect these different data sources through its semantic layer, giving daily visibility into PA approval rates by payer and flagging when approval performance starts to deteriorate so teams can respond quickly.

Yes, if you treat payer policy PDFs as a real data source. In access, rules often change in the documents first. Claims and dashboards usually show it later. In 2026, you need change detection fast.

‍

1. Collect the latest policy docs: Formularies, medical policies, UM updates, bulletins—from approved sources.
   
2. Turn PDFs into usable text: Use document parsing (and OCR only if scanned) so tables + sections don’t get mangled.
   
3. Extract structured “policy facts” with NLP/LLM: Pull fields like:
   

• PA required (yes/no)
• step therapy sequence
• clinical criteria (diagnosis, labs, prior drugs tried)
• effective date + line of business
  
  

4. Reconcile across sources (so one PDF doesn’t mislead you): Cross-check with vendor feeds + hub/SP signals (denials, time-to-therapy) + internal intel. If sources disagree, flag it instead of guessing.
   
   
5. Agentic monitoring: detect + summarize changes automatically: When a new doc appears, the system re-extracts rules, compares to the old version, and flags meaningful changes.

‍

What you get:

• A change log: “Plan X added step therapy for Drug Y”
• A rules table you can join to outcomes (PA approvals, abandonment, TRx/NBRx impact)
• An audit trail: which document text supported the change

‍

That’s the “AI/agentic” upgrade: you stop reading PDFs manually and start getting alerts when access rules actually change.

It often takes 2-3 months to assess the market access impact of a new competitor because the data that reflects real payer and patient behavior, such as

‍

• adjudicated claims
• rejection patterns
• formulary updates
• copay changes
• PA trends

‍

lags behind the launch date.

‍

Payers typically take weeks to update policies, pharmacies adopt changes at different speeds, and early claims don’t show mature utilization or switching behavior. Until enough data accumulates across these sources, analytics teams cannot reliably evaluate how a competitor’s entry affects access, pull-through, and patient conversion.

Similar formulary positions can yield different results across payers because real-world access depends on more than the listed coverage tier. It’s shaped by

‍

• variations in utilization management rules
• PA approval rates
• step therapy enforcement
• pharmacy networks
• patient cost-sharing
• operational efficiency of each plan

‍

Even when two payers label coverage the same way, differences in adjudication behavior, benefit design, and member demographics can produce very different speed-to-therapy, abandonment, and overall access outcomes.

You can find the payers driving a TRx or NBRx decline by breaking the drop into payer-level contributions and comparing this month to prior periods.

‍

Step 1: Attribute the decline by payer

• Split TRx/NBRx by payer/plan
• Rank payers by how much they contributed to the total drop

‍

Step 2: Diagnose what changed inside those payers. For the top “negative” payers, check shifts in:

• New starts (NBRx)
• Approvals vs. rejections
• Utilization management (UM) barriers, like prior authorization (PA) and step therapy (fail-first)

‍

Step 3: Layer in the likely reasons. Overlay payer contribution with signals like:

• formulary or tier changes
• claim rejection spikes
• higher patient cost-share (copay increases)
  
  

This gives you both who caused the decline and why, so you can target follow-up actions with the specific payers that matter most.

Yes. You can identify which payers drive real prescriptions by looking at what happens inside each plan, not just how many lives are covered.

‍

What to measure by payer/plan

• TRx and NBRx volume (actual scripts and new starts)
• Approval rates (how often patients get through)
• Patient progression (how many move from script → approval → first fill → refills)

‍

Where the insight comes from
Claims + hub data show:

• which payers convert eligible patients into starts
• how often PA or step therapy blocks patients
• where adherence and persistence stay strong (or drop)

‍

How to interpret it
Compare real prescription flow to the plan’s covered lives:

• Some payers have fewer lives but high pull-through (they truly drive volume)

Others look good on paper (lots of covered lives) but produce low real-world scripts because restrictions or cost-share block uptake

For each big payer, you need to know three basics:

‍

• which tier your drug is on
• what extra rules apply (like prior auth or step therapy)
• what patients are likely to pay out of pocket.

‍

These details change often, and payer documents are usually hard to read and don’t always show every restriction clearly. So in real life, teams pull information from three places:

‍

• public formulary files
• what account and field teams hear in payer/HCP meetings
• win–loss notes from contracting.

‍

The real goal is to see your relative position: being on tier 2 can be an advantage if competitors are on tier 3, but a weakness if they are on tier 1. Tellius can bring all of this data into one view and build simple scorecards by payer and region.

Winning better access is only half the job; you also need the field team to turn that access into prescriptions. The hard part is separating:

‍

• Growth driven by access changes (better formulary position), from
• Growth driven by field promotion and execution

‍

One useful approach is difference-in-differences analysis:

• Compare prescription trends before vs after an access change
• Across territories with different levels of field activity (call frequency, reach, programs)

‍

If territories with high field intensity grow much faster than similar territories with low field intensity after the same access win, that extra lift is likely pull-through driven by the field.

‍

Tellius Kaiya’s causal analysis can formalize this, helping you separate access-driven growth from promotion-driven growth and understand the mix. This, in turn, guides how you allocate budget and headcount between market access investments and field force resources.

Market access insights can be shared with field and brand teams through integrated dashboards, CRM alerts, and simplified payer profiles that translate complex access data into clear, actionable pull-through guidance. By embedding the following insights directly into field workflows, teams can tailor their messaging and resources for each account.

‍

• formulary changes
• PA trends
• rejection hotspots
• payer opportunities

‍

When insights are pushed in real time and aligned across brand, access, and field stakeholders, organizations can execute coordinated pull-through strategies that improve prescribing and patient access outcomes.

Yes. You don’t have to spend $500K+ per year on competitive formulary data if you use the information you already have access to. A lot of competitor access intel is available from:

‍

• Public payer documents (online formulary listings, medical policies, PDFs)
• Field intelligence (reps hearing about competitor coverage in calls)

‍

A cost-effective approach is to:

• Use web scraping to pull formulary data from payer sites
• Capture field feedback in CRM or simple forms
• Apply natural language processing (NLP) to parse and structure unstructured text

‍

Tellius can aggregate competitive access intelligence from multiple unstructured sources (such as field emails in Google Drive, Gong call transcripts where customers mention competitor coverage, and payer formulary PDFs) alongside your structured access data. Using NLP, it can extract and organize restrictions, tiers, and policy language into a unified view, so you can track competitor positioning without relying entirely on expensive syndicated subscriptions.

When PA approval drops fast, dashboards usually show the decline but not the cause. Leaders immediately ask: Which payers drove it? What changed? Which criteria are failing? Which HCP segments are impacted? Traditional investigation can take days because analysts have to pull hub data, denial codes, formulary updates, and then slice everything by payer, geography, specialty, and time.

‍

Agentic investigation workflows automate this. The system monitors access KPIs (PA approval, denials, time-to-therapy, abandonment). When a meaningful change happens, it automatically launches an investigation and returns an explanation quickly.

‍

What the workflow does:

• Triggers automatically: It detects significant movement and starts analysis without a ticket.
  
• Checks drivers in parallel:
  
  • Payer policy changes: New criteria, documentation requirements, step edits, or tighter rules.
    
  • Denial reason shifts: More “missing documentation,” “unsupported diagnosis,” “step therapy not met,” and similar patterns.
    
  • Segment concentration: Whether the problem is focused in certain specialties, regions, patient types, or HCP cohorts.
    
  • External context when available: Competitive events or guideline changes that often affect payer behavior.
    
• Quantifies contribution: It estimates which drivers explain most of the drop using denial-pattern statistics and segmentation.
  
• Recommends actions: It can suggest targeted fixes, like updating PA templates for a payer’s new lab requirement or focusing education on the practices with high incomplete submissions.
  
  

What you get: a short, defensible story of what changed, where it changed, why it changed, and what to do next, without waiting for a multi-day manual build.

You can forecast the portfolio impact of IRA Medicare price negotiations using scenario modeling on top of claims-based forecasting and patient-flow analytics. The goal is to estimate how negotiated Medicare prices could change revenue, volume, and access across products.

‍

These models usually account for:

• Eligibility timing (when a product becomes negotiation-eligible)
• Historical Medicare mix and utilization (how dependent the brand is on Medicare volume)
• Competitive dynamics (what competitors may do in response)
• Therapy switching risk (patients moving to alternatives)
• Patient cost and behavior effects (out-of-pocket changes that affect adherence and persistence)

‍

By running multiple price-and-volume scenarios, teams can:

• quantify financial exposure across the portfolio
• identify which assets are most vulnerable
• prioritize strategic responses (contracting, access strategy, portfolio timing) at the portfolio level

You can anticipate formulary changes before they hit market share by using predictive analytics to spot early signals in payer behavior.

‍

What the model looks at:

• Historical payer behavior (how each plan has changed policies in the past)
• Competitive access trends (where competitors are gaining preferred status)
• Contract events (renewals, rebate changes, new bids)
• Utilization management shifts (rising PA rates, new step edits, tighter criteria)

‍

Predictive models can then flag payers that are likely to:

• tighten restrictions
• add step therapy
• change preferred agents after competitor pricing moves or guideline updates

‍

By combining formulary intelligence with claims data and other market signals, teams can often see risk weeks to months earlier. This gives enough time to prepare proactive contracting and pull-through plans instead of reacting after share drops.

AI can forecast the impact of upcoming access changes by learning from past payer policy updates and how they affected performance.

‍

It looks at signals like:

• historical policy changes and timing
• denial and rejection patterns
• competitive moves
• utilization management trends (PA, step therapy, criteria tightening)

‍

Then it models how similar changes affected TRx/NBRx, access, and share in the past and simulates likely payer responses. This gives teams an early view of risk so they can plan contracting, messaging, and pull-through before performance drops.

AI agents can monitor payer behavior by continuously scanning:

‍

• formulary updates and policy language changes
• PA trends, approval/denial rates
• rejection patterns
• claims-based utilization shifts

‍

By comparing current activity to historical norms, the agent can flag early warning signs such as denial spikes, new step edits, shifting preferred products, or sudden policy changes. This gives market access and brand teams visibility weeks before the impact shows up in TRx/NBRx.

Agentic workflows prioritize pull-through by combining payer signals with regional prescribing patterns to focus teams where impact is highest.

‍

They continuously evaluate:

• regional payer performance and mix
• formulary and UM changes
• denial trends and friction points
• under-utilized “good access” opportunities

‍

Then they generate next-best actions per territory, such as:

• where to target PA education
• which messages to adjust for specific plans
• when to coordinate with patient services or hub teams

‍

This keeps field and brand teams focused on the payer opportunities that will move prescriptions, not just the loudest problems.

When specialty pharmacy data is incomplete, you can still measure access performance by triangulating across multiple sources to infer what’s happening between prescription and first fill.

‍

Use:

• claims data
• hub (HUB) case status and timestamps
• specialty pharmacy (SP) rejection feeds (where available)
• patient services activity

‍

Track patterns in:

• PA submission rates and approval times
• denial reasons and rejection spikes
• time-to-therapy by payer/region

‍

Linking these signals with formulary and payer trends helps you spot delays, quantify drop-off, and guide targeted pull-through, even inside the “data black hole.”

Patient Assistance Programs (PAPs) and copay support can easily consume 5-10% of gross revenue, but many companies still measure impact using very rough metrics. To optimize spend, you need analytics that track:

• PAP-to-paid conversion rates (how many assisted patients become paying patients)
• Patient lifetime value by channel and segment
• Competitive assistance levels (how generous competitors are)

‍

Tellius’s predictive models can estimate which patients are likely to convert from assistance to paid prescriptions, and which are likely to remain on free drug long-term. This allows you to focus PAP resources on high-value cases, improving both patient access and financial return instead of treating PAP as a blind cost center.

Not every payer is a good candidate for value-based contracts. Some lack the data, some lack interest, and pushing them wastes time. Predictive models can score “value-based readiness” by analyzing:

‍

• Past innovation history (have they done outcomes-based deals before?)
• Data and analytics capabilities
• Level of integration with provider networks and systems

‍

This helps you focus on progressive payers who can actually execute these contracts and avoid spending cycles on accounts that are not set up for outcomes-based agreements.

Yes. Modern AI and agentic analytics platforms can continuously scan your access data and push insights to you instead of making you hunt for them. They can automatically flag

• when a payer’s PA approval rate drops
• when abandonment spikes
• when NBRx jumps after a formulary win

‍

Platforms like Tellius can watch payer, territory, and HCP segments, then surface findings in plain language, such as

• “Most of this quarter’s NBRx growth came from Payer X after the tier upgrade,” or
• “Payer Y now has the highest abandonment in your top 10 accounts”

‍

They can also highlight under-utilized access: plans where you’re preferred but share is below expected, so teams know where to act first.

Market access teams have real constraints: hub staffing is limited, access specialists have limited bandwidth, and field time is finite. Meanwhile, payer barriers change quickly, and the “highest lives” payer is not always the payer with the biggest recoverable opportunity.

‍

AI optimization helps by making resource decisions based on expected patient recovery.

‍

What the system does:

• Scores opportunities continuously: It identifies where recovery is likely, such as payers with rising denials, high abandonment, or new criteria that practices are failing to meet.
  
• Balances real constraints: It allocates effort while respecting staffing limits, budget limits, service-level needs, and practical execution capacity.
  
• Forecasts volume so you can staff ahead of the spike: If PA volume is trending up for specific payers, forecasting can help you plan staffing instead of reacting after backlogs appear.
  
• Estimates marginal impact: It looks at diminishing returns so you can see when adding more effort stops paying off.
  
• Supports scenario planning: Leaders can test tradeoffs, like shifting staff from stable payers to emerging risk payers and seeing the expected portfolio impact.
  
  

You get a clear recommendation for where to deploy hub effort, payer strategy effort, and field enablement, based on which mix is most likely to improve starts and reduce abandonment under real-world constraints.

In an IRA environment, the best gross-to-net (GTN) platforms are the ones that put everything in one auditable system, including:

• contract terms and eligibility rules
• government pricing logic and IRA impacts
• rebates, chargebacks, and fees
• payer mix shifts and utilization trends
• scenario modeling (“what if price/rebate/mix changes?”)

‍

The key requirement is governed, traceable logic. Finance and market access teams need to simulate price, rebate, and volume scenarios and clearly see how IRA rules could change net revenue and portfolio performance, without relying on fragile spreadsheets.

A strong market access analytics platform should bring together the data needed to explain access performance end-to-end, including:

• claims, specialty pharmacy, hub, and formulary data
• a unified view of patient access, time-to-therapy, and payer performance

‍

It should support payer-level insight into:

• PA trends and approval/denial rates
• rejection drivers and step edits
• formulary changes and policy shifts
• real-time TRx/NBRx movement by plan

‍

More advanced platforms also add:

• forecasting and “what-if” modeling
• AI-driven risk detection (early access deterioration signals)
• pull-through recommendations by region
• workflows that push insights to brand and field teams in a usable format

‍

The real test: it makes complex access data easy to understand, actionable, and tied to measurable outcomes.

Payer contracting is a balancing act between:

• Rebate levels
• Formulary position (preferred, non-preferred, excluded)
• Utilization management (PA, step therapy, quantity limits)

‍

All of these affect net revenue, not just list price. Many teams still try to model this in Excel, but as soon as you include multiple payers, multiple indications, and different contract structures, the spreadsheet becomes unmanageable.

‍

Advanced analytics platforms can simulate thousands of contracting scenarios at once and estimate how each one affects volume, access, and net revenue over time. This lets you optimize the rebate / access / volume trade-off for maximum net present value (NPV) instead of guessing with a few manual “what-if” Excel tabs.

You can measure impact by comparing before vs. after results on the access steps the initiative was meant to change.

‍

Track metrics like:

• PA approval rate and denial reasons
• rejection trends and abandonment signals
• time-to-therapy
• first-fill rate and conversion through the journey
• payer-level TRx/NBRx changes

‍

To make it credible, isolate the populations actually affected such as:

• the contracted payer segment
• the patients in the hub program
• the regions where the workflow changed

‍

Then link access improvements to prescription lift to estimate ROI and identify which initiatives truly improve access and sales outcomes

Market access data is messy: lots of plan types, codes, and policy rules. Conversational analytics helps by letting non-technical users ask questions in plain English, such as:

• “Which plans have the lowest approval rates?”
• “Where did denial reasons change this quarter?”

‍

The system translates the question into the right joins and filters behind the scenes, so users don’t need SQL or schema knowledge. They can also ask follow-up questions to drill deeper in the same flow instead of filing new report requests. In Tellius, this is delivered through a governed semantic layer so answers stay consistent.

It can, but only if it is built on market access semantics and governed logic. Access questions are hard because they require multiple joins and careful definitions across formulary status, claims, hub activity, and denial reasons.

‍

For a question like “Show me payers where we’re preferred but pull-through is low, then break down by denial reasons”, conversational access analytics works when it has three foundations:

1. A market access semantic layer
   The system must understand access terms and how they map to data, such as “preferred,” “pull-through,” “PA approval,” “step therapy,” and “category benchmark.”
   
2. Context retention for multi-step investigations
   Access teams ask follow-ups in sequence (“Which payers?” → “Which denial reasons?” → “Which regions and specialties?”). The system must keep filters and context consistent across steps.
   
3. Multi-source querying
   Good answers require pulling from formulary data, claims, hub systems, specialty pharmacy signals (when available), and sometimes qualitative field intel. The system needs to combine them in one investigation flow.
   
   

The real test is reliability. If payer IDs do not match across systems or definitions vary by team, conversational AI can produce inconsistent answers. When the semantic layer is governed and the data is integrated, conversational AI becomes a practical way to answer complex access questions during planning meetings instead of waiting for analyst tickets.

This is a common trust problem in access analytics. Payer policies can change mid-cycle, but vendor files may lag. This creates confusing situations where forecasts, decks, and “current status” reports disagree. Formulary data governance solves this by making coverage changes traceable and auditable.

‍

Four capabilities matter:

1. Immutable snapshots with provenance
   Each ingest captures the coverage status plus source, version, timestamps, effective dates, and metadata. This makes it clear which “version of truth” a report used.
   
2. Change detection and alerts
   When a new file arrives, the system compares it to prior versions and flags meaningful changes (tier moves, new PA, new step edits, criteria changes). It also estimates who is affected (lives, plans, segments).
   
3. Version-controlled reporting
   Teams can lock reviews and forecasts to a chosen snapshot so decks remain internally consistent, while still allowing “latest view” reporting with clear labels.
   
4. Multi-source reconciliation
   Stronger systems do not rely on a single vendor feed. They cross-check vendor data against public payer docs, field intelligence, and operational signals like denial reason spikes. If sources conflict, the platform flags it and shows which source is most recent.
   
   

As a result, you end up with fewer “which number is right?” debates, faster response to real policy changes, and a leadership narrative that is consistent because it is based on versioned, explainable access data.

Many contract models over-focus on covered lives and rebate math, but real performance depends on whether patients can actually start and stay on therapy. A stronger scenario model treats a contract as something that changes both coverage and the patient journey.

‍

A better simulation includes:

• Contract terms: rebate levels, eligibility rules, channels (commercial vs Medicare), exclusions, and any admin/guarantee clauses.
  
• Formulary position changes: tier movement plus the actual restrictions (PA, step therapy, quantity limits, specialty pharmacy rules).
  
• Access friction assumptions (by payer/segment):
  
  • PA submission and approval rate
  • common denial reasons and how they shift
  • abandonment risk (script written → never filled)
  • time-to-therapy changes (delays from BV/PA to first fill)
    
• Outcome and economics: TRx/NBRx lift estimates, pull-through, persistence where available, and the net impact on gross-to-net (rebates, fees, copay/PAP effects if included).
  
  

With agentic analytics, you get side-by-side contracting scenarios that show the real tradeoffs, like:

• Higher rebate → better tier, but stricter UM increases abandonment and delays starts
• Lower rebate → worse tier, but fewer friction points improves conversion and speed-to-therapy

‍

You also get a decision view that ranks options by net revenue and patient conversion. The key is that the model outputs both:

1. the expected changes in coverage + UM (tier, PA, step edits)
   
2. the downstream impact on approvals, time-to-therapy, abandonment, and ultimately realized scripts and net revenue.