Mandolin onboarding: how do we define success metrics like time-to-therapy, backlog reduction, PA turnaround time, and denial rate?

Most infusion and specialty-drug teams don’t need another dashboard; they need proof that a new system is actually moving the needles that matter: faster starts, fewer denials, and a back office that isn’t drowning in work. During Mandolin onboarding, we define those success metrics up front—using your real workflows, volumes, and payer mix—so you can measure impact in time-to-therapy, backlog reduction, prior auth turnaround, and denial rate from day one.

This guide walks through exactly how we define, baseline, and track those metrics during Mandolin onboarding, and how we tie them back to both patient access and financial performance.

How we approach success metrics in Mandolin onboarding

Our onboarding framework is simple:

Define the operational outcomes that matter most
For most teams, that’s:
- Time-to-therapy (or time-to-first-infusion / time-to-first-fill)
- Backlog days and capacity
- Prior authorization turnaround time
- Denial rate and avoidable write-offs
Baseline each metric using your current workflows
We pull from your existing referral logs, EHR timestamps, work queues, and denial reports to capture “before Mandolin” performance.
Map metrics to specific Mandolin workflows
Every improvement must be traceable to a concrete workflow we’re automating: intake, benefits investigation, OOP estimation, medical policy review and prior auth, claims statusing, and appeals.
Set targets and review cadence
We align with your leadership team on what “success” looks like by 90 days and by 6–12 months, then build those targets into regular performance reviews.

From there, we treat each metric like an operational contract: Mandolin’s AI agents do the work; we show you the logs, timestamps, and outcomes.

Time-to-therapy: defining and measuring patient access gains

What we mean by “time-to-therapy”

Time-to-therapy is the elapsed time from referral received to first dose administered (or first prescription filled), depending on your care model.

During onboarding, we define the exact start and end points for your organization, commonly:

Start timestamp:
- Date/time referral is first received (fax, portal, email, or EHR message), or
- Date/time referral is first visible to your intake/authorization team in your system of record
End timestamp:
- Date/time of first infusion encounter, or
- Date/time first prescription is dispensed/filled

We then segment by therapy, site of care, payer, and line of business when that level of detail is available.

How we baseline time-to-therapy

Before go-live, we typically:

Pull a 3–6 month lookback of referrals for your top therapies
Capture:
- Referral receipt date
- Date benefits were cleared
- Date prior auth was approved (if required)
- Date of first administered dose / first fill
Calculate:
- Average time-to-therapy
- Median and 90th percentile (to understand outliers and delay risk)
- Differences by payer and therapy

Example:
A national AIC customer saw multi-day delays between referral receipt and data reaching the EHR. Manually, documents took ~20 minutes each and up to 3 days to get fully processed. With Mandolin’s AI agents, the same workflow dropped to 3 minutes per document and under 2 hours end-to-end, contributing directly to faster patient starts.

How Mandolin impacts time-to-therapy

Mandolin compresses time-to-therapy by removing lag at each step:

Intake & onboarding:
AI agents read and interpret referrals, lab reports, and clinical notes from fax/portals in minutes—not days—and enter data into your EHR in near real time.
Benefits investigation & OOP estimation:
Agents log into payer portals, run full benefits checks, apply site-specific fee schedules, GPO and 340B pricing, and co-pay assistance, and produce precise out-of-pocket estimates without waiting for staff availability.
Medical policy review & prior auth:
Agents pull payer policies, check requirements against clinical documentation, assemble PA packages, and submit via portal/fax/phone through the channels your payers actually require.

By the end of onboarding, we’re not just reporting “Mandolin is faster.” We’re able to show:

Average time from referral receipt to:
- Intake completed
- Benefits cleared
- PA submitted and approved
Resulting reduction in time-to-therapy, with a breakdown by payer/therapy

Backlog reduction: eliminating days of work in the queue

What we mean by backlog

Backlog is the combination of:

Open referrals/prescriptions not yet fully intake-completed and in the EHR
Outstanding benefit investigations / PA cases not yet actioned
Aging claims and appeals not yet touched within your standard window

Operationally, we define backlog in two ways:

Volume-based: number of items waiting (referrals, PA cases, claims)
Time-based: “backlog days” = how many days of work are sitting in the queue at current throughput

How we baseline backlog

Before Mandolin, we typically see:

Manual teams processing 10–20 minutes per document
Multi-day delays from fax receipt to EHR entry
Backlogs measured in days of prescriptions or referrals waiting for basic intake

Example:
One infusion operation processed 200–300 new prescriptions per day manually, spending 10–12 minutes per Rx and carrying a 4-day prescription backlog. During onboarding, we codified that as:

Backlog days = 4
New prescriptions/day = 200–300
FTEs dedicated to that queue = 2–3 full-time staff

That became the baseline for measuring improvement.

How Mandolin impacts backlog

Mandolin’s AI agents actively work the queue—just like your best back-office specialist:

Pulling faxes as they arrive
Navigating payer portals to complete downstream steps
Documenting every action with logged, traceable records

For the customer above, once Mandolin went live:

Intake time dropped to ~3 minutes per document
Backlog dropped from 4 days to 0 days
Prescriptions were processed in real time—no queue buildup

During onboarding, we set clear backlog targets, such as:

Goal: Reduce prescription backlog from 4 days to same-day processing within 60–90 days
Metric: Backlog days = (total open items ÷ average daily completed items)

We then report progress weekly: number of items in queue, average age of items, and the associated reduction in staff-hours required.

Prior authorization turnaround time: from “sent” to “approved”

What we mean by PA turnaround time

For Mandolin onboarding, we define prior authorization turnaround time as:

Time from PA-ready documentation (or referral receipt, depending on your preference) to payer decision (approval or initial denial).

We align with you on the precise start point:

Option A (most common):
Time from PA submitted (via portal/fax/phone) to PA decision
Option B (end-to-end):
Time from referral received to PA decision, capturing delays caused by missing documentation or policy checks

We usually track both, because it’s the internal delays prior to submission that often hide the real opportunity.

How we baseline PA turnaround

Before Mandolin, we:

Pull a sample of PA cases for key drugs and payers
Capture:
- Date referral received
- Date PA determined to be required
- Date PA submitted
- Date PA approved (or denied)
Calculate:
- Internal PA prep time (referral → submission)
- Payer response time (submission → decision)
- Total PA turnaround time

We also flag avoidable rework: PAs sent incomplete, missing labs, or misaligned with current medical policy, resulting in avoidable delays.

How Mandolin impacts PA turnaround

Mandolin compresses the internal clock:

Medical policy review:
AI agents pull the relevant payer policy, compare requirements to the patient’s clinical data, and identify missing labs or documentation before submission.
PA assembly & submission:
Agents compile all required documentation, populate payer-specific forms, and submit via the mandated channel—portal, fax, or phone.
Follow-up & statusing:
Agents log into portals or call payers to check status, document updates, and escalate when responses lag.

Because every step is logged and timestamped, we can show:

Average time from referral to PA submission
Average time from submission to decision
Reduction in internal PA turnaround time post-implementation

In practice, when intake and benefits work drop from 20 minutes per doc and up to 3 days of lag to 3 minutes and under 2 hours of turnaround, you reclaim days that used to sit between “referral in” and “PA out.”

Denial rate: measuring avoidable denials, not just raw counts

What we mean by denial rate

Denial rate can be defined in several ways. During onboarding, we standardize it so that you can clearly attribute change to Mandolin-driven workflows.

Common definitions we align on:

Claim-level denial rate:
Denied claims ÷ total claims submitted for target therapies / sites
Dollar-weighted denial rate:
Denied dollars ÷ total billed dollars for those therapies
Avoidable denial rate:
Denials related to:
- Missing/incorrect authorization
- Eligibility/benefits discrepancies
- Documentation mismatches to medical policy
- Timely filing failures

We focus heavily on avoidable denials, because that’s where Mandolin’s AI agents have direct, measurable impact.

How we baseline denial rate

Before go-live, we:

Pull 3–6 months of claims and remittance data for your specialty therapies
Categorize denials into:
- Authorization-related
- Eligibility/coverage
- Coding/charge issues
- Clinical/medical necessity
- Administrative (timely filing, missing info)
Calculate:
- Overall denial rate
- Avoidable denial rate
- Denial rate by payer and therapy

We also track the operational overhead: how many staff are touching denials, how many appeals are filed, and the average time spent per denial.

How Mandolin impacts denial rate

Mandolin reduces avoidable denials by:

Upstream accuracy:
Ensuring benefits investigations are complete and correct, and that site-of-care and coverage nuances are accurately captured.
Policy-aligned prior auths:
Reviewing medical policy and matching documentation before submission, so PAs are aligned with payer criteria at first pass.
Claims statusing:
Agents regularly log into portals or call payers to check claim status, identify issues early, and trigger appeals workflows with the right documentation.
Appeals preparation:
When denials do occur, agents gather medical policies, prior auth records, and clinical notes to assemble appeal packages more quickly and consistently.

We then track:

Change in overall denial rate
Change in avoidable denial rate for therapies and payers under Mandolin
Recovery of previously denied revenue through faster, better-supported appeals

Because every portal check, fax, and call is logged and traceable, you can tie denial improvements back to specific Mandolin workflows—not just a general “automation” effect.

Connecting success metrics to staffing and economics

During onboarding, we don’t just stop at operational metrics; we translate them into staffing and financial terms your CFO and COO care about.

Throughput and FTE impact

We map:

Documents/cases processed per day before vs. after Mandolin
Minutes per document/case and corresponding FTE equivalents

Example from a national AIC company:

Before Mandolin:
- 250 documents/day
- ~20 minutes per document
- Up to 3 days to enter into the EHR
- Over 100 FTEs required across sites
With Mandolin:
- 3 minutes per document
- Under 2 hours end-to-end turnaround
- Equivalent work with a fraction of the FTE load

Similarly, another operation:

Before Mandolin:
- 200–300 new prescriptions/day
- 10–12 minutes per Rx
- 4-day prescription backlog
- 2–3 full-time staff just to stay afloat
With Mandolin:
- 0-day backlog
- Real-time processing
- Freed staff for higher-value work

These throughput and FTE metrics are defined and baselined during onboarding so your team can see how time-to-therapy, backlog reduction, PA speed, and denial improvements translate into real staffing capacity.

Revenue and margin implications

For buy-and-bill and infusion operations, we also model:

Revenue preserved via fewer avoidable denials
Margin impacts driven by:
- Accurate application of site-specific fee schedules
- Correct use of GPO and 340B pricing
- Precise OOP estimates that reduce cancellations and non-starters

We don’t make abstract “cost savings” claims. We show:

X% reduction in avoidable denials
Y days less in the backlog (i.e., fewer days of billed-but-not-collected risk)
Z patients per month you can now onboard without adding staff

All of this is defined and agreed during onboarding so the success picture is clear.

How we operationalize these metrics in your Mandolin rollout

When you implement Mandolin, success metric definition isn’t an afterthought—it’s part of the rollout plan.

Step 1: Discovery and metric selection

We meet with your operations, revenue cycle, and finance leaders to:

Confirm your top priorities:
- Faster time-to-therapy?
- Clearing a referral backlog?
- Lowering denial rate?
- Supporting growth without adding FTEs?
Select the primary metrics:
- Time-to-therapy
- Backlog days and open volume
- PA turnaround time
- Denial rate (overall and avoidable)

Step 2: Data collection and baselining

We work within your existing systems—EHR, work queues, reports, and remits—to:

Pull historical data for each metric
Align on definitions and cutoffs for “before Mandolin”
Document baseline values and create a shared scorecard

No heavy integrations required; we operate in the same portals, faxes, and phone workflows your staff already live in.

Step 3: Go-live and early monitoring

As Mandolin’s AI agents begin working intake, benefits, PA, and claims:

Every action is logged and traceable
Timestamps allow us to calculate real-time cycle times
We compare early performance against baseline:
- Minutes per document
- Turnaround from fax to EHR
- Backlog trend lines

Step 4: 30/60/90-day reviews

We hold structured reviews to:

Report progress on time-to-therapy, backlog, PA turnaround, and denial rate
Surface payer- or therapy-specific patterns
Adjust workflows or policies where we see bottlenecks

By 90 days, you should have clear, data-backed answers to:

How much faster patients are starting therapy
How many backlog days have been eliminated
How PA turnaround times have shifted
How denial patterns have changed—especially avoidable denials

Mandolin’s value is measured in hard operational metrics, not promises. During onboarding, we define success in your language—time-to-therapy, backlog reduction, prior auth turnaround, and denial rate—then prove it with logged, auditable AI agent work across fax, portal, and phone.

Get Started