← Drug Development Suite Schedule, Scope & Resource · Vitalis Therapeutics Inc.

Resource Plan

Download Excel
109
Headcount
102
Working pool
191k
Hours
20.1
Peak concurrent FTE
Contents
  1. What This Plan Commits
  2. The Roster by Function
  3. Headcount Is Not Effort
  4. Phasing by Stage
  5. What the Internal Roster Actually Buys
  6. How a Person Is Actually Assigned
  7. The Plan's Blind Spots
  8. Resource Risks
  9. A Correction Recorded

1. What This Plan Commits

A resource plan is not a staffing wish-list. It is the commitment that the work in the work breakdown structure can be delivered by a stated number of people, working a stated number of hours, at a stated cost, in a stated sequence — and that all four of those statements reconcile to each other and to the Program Budget. This plan carries three numbers, and they answer three different questions.

NumberValueThe question it answers
Headcount109 peopleHow many individuals are named against this program at some point in its life. It is a roster, not a room. It is the wrong number for capacity planning and the right number for onboarding, training, system access and GxP qualification.
Hours191,420 hoursHow much effort the program consumes. This is the number that reconciles to cost, because internal labor is charged by the hour against a blended rate.
Concurrent FTE20.1 at peak (Stage 4)How much of the organization is occupied at once. This is the number a functional head actually negotiates over, and it is nowhere near the headcount.
Those three numbers describe the same program and they are 109, 191,420 and 20.1.

Most resourcing arguments in a development program are two people using two different ones without saying which. A functional head asked to “give” four people to a program is answering a concurrent-FTE question; a program director quoting a 109-person roster is answering a headcount question. They are not disagreeing about the facts.

2. The Roster by Function

Eleven lines: 10 working functions that carry deliverables and appear in the work breakdown structure, plus the 7-seat Development Committee, which carries none. The working pool is 102.

FunctionHeadcountHoursHrs / headBlended rateInternal costScope
Clinical Operations2451,9002,162$190$9,861,000Sites, monitoring, enrolment, clinical supply
Technical Operations / CMC1630,2001,887$180$5,436,000Process, analytical methods, scale-up, validation
Clinical Development1217,9001,491$200$3,580,000Protocols, medical monitoring, clinical study reports
Biostatistics & Data Management1222,6001,883$180$4,068,000SAP, EDC, database lock, analysis datasets
Regulatory Affairs1018,9001,890$210$3,969,000Strategy, submissions, agency interaction, labeling
Pharmacovigilance & Drug Safety814,2001,775$180$2,556,000Case processing, signal detection, DSUR
Quality Assurance (GxP)712,8001,828$180$2,304,000Audits, CAPA, GxP compliance, inspection readiness
Nonclinical Safety & Pharmacology59,4001,880$180$1,692,000Toxicology, pharmacology, study oversight
Program Management Office47,2001,800$180$1,296,000Governance, planning, risk, reporting
Market Access & Commercial45,0001,250$170$850,000Payer evidence, pricing, launch readiness
Development Committee (oversight only)71,320188$300$396,000Gate decisions, funding release, condition closure
Total109191,4201,756$188$36,008,000

The blended rate is inclusive of salary, benefits and overhead recovery. It is not a salary and should not be read as one. Regulatory Affairs carries the highest rate at $210 because the function is senior-weighted — there is no junior version of the person who signs a commitment to the agency. Clinical Development runs at $200 on only 12 people because medical monitors are physicians. The Committee's $300 is the highest rate in the program and buys the fewest hours: 1,320 hours across 8 years, or 188 hours per member, which is roughly one working month each spread across the entire program.

Read the hours-per-head column, not the headcount column.

Market Access & Commercial has 4 people and 1,250 hours per head — the thinnest allocation in the program, against a roster average of 1,756. Clinical Operations has 24 people at 2,162. Both facts are invisible if you read only the first column, and the second is the one that predicted the outcome.

3. Headcount Is Not Effort

Converting the hours envelope into concurrent effort is arithmetic, and the result is the single most useful number in this document.

StageWindowDurationHoursConcurrent FTE
Stage 1Mar 2022 – Dec 202210.1 months17,1809.8
Stage 2Jan 2023 – May 202417.0 months23,8408.1
Stage 3Jun 2024 – Jun 202625.0 months31,3207.2
Stage 4Jul 2026 – Sep 202827.0 months94,18020.1
Stage 5Oct 2028 – Jun 20299.0 months24,90016.0
109 people are named against this program. At its busiest it occupies 20.1 full-time equivalents.

That ratio — roughly 5 named people per concurrent seat — is not slack and it is not double-counting. It is what a matrix organization is. A regulatory affairs specialist who spends a fifth of her time on VitaFlow is one headcount, 0.2 FTE, and fully accountable for every submission she touches. The program cannot reduce her to 0.2 of a person for the purposes of qualification, training, system access, or an inspection interview.

This is why resourcing a development program by headcount always looks affordable and always runs short. The headcount is genuinely available. The attention is what is scarce, and attention is what the FTE column measures.

The two lightest stages are worth a second look, because both are long. Stage 3 is the longest stage in the program at 25.0 months and the lightest at 7.2 FTE; Stage 2 runs 17.0 months at 8.1. Neither is an error. Stage 2 contains the IND submission and Phase 1, and most of its calendar is spent waiting: for the agency's thirty-day review to elapse, for Phase 1 cohorts to complete, for bioanalytical results to return from Kestrel. Stage 3 is Phase 2 — 8.3 of its 25 months are participants being treated, which consumes very little sponsor effort per unit of elapsed time and cannot be shortened by adding any.

Elapsed time and effort are not the same variable, and a plan that staffs to calendar rather than to effort will overstaff every waiting stage in the program. The inverse error is worse: a plan that reads Stage 3's 7.2 FTE as slack, and shortens the stage to match, is proposing to run a 36 weeks study in less than 36 weeks.

4. Phasing by Stage

Every function's hours, distributed across the five stages. Rows sum to the function's envelope; columns sum to the stage; the whole matrix sums to 191,420. Both directions are asserted at import — if a single cell is edited without its row and column being made good, nothing in this suite builds.

FunctionStage 1Stage 2Stage 3Stage 4Stage 5Total
Clinical Operations3,2009,80034,6004,30051,900
Technical Operations / CMC6,4004,2003,60013,4002,60030,200
Clinical Development6002,4004,1008,6002,20017,900
Biostatistics & Data Management3002,1004,40012,3003,50022,600
Regulatory Affairs1,9003,8002,4006,1004,70018,900
Pharmacovigilance & Drug Safety4001,6002,4007,6002,20014,200
Quality Assurance (GxP)1,3001,9001,8005,6002,20012,800
Nonclinical Safety & Pharmacology5,2002,9008004001009,400
Program Management Office9001,3001,2002,8001,0007,200
Market Access & Commercial2006002,3001,9005,000
Development Committee (oversight only)1802402204802001,320
Stage total17,18023,84031,32094,18024,900191,420
Share of program hours9.0%12.5%16.4%49.2%13.0%100.0%
Share of program money8.0%7.5%19.3%58.0%7.3%100.0%
Stage 4 takes 49% of the program's hours and 58% of its money.

The gap between those two percentages is the entire economics of drug development in one line. Stage 4 money is dominated by what is paid to Meridian, to the 260 sites, and to Kestrel — per-participant costs that scale with participants dosed, not with sponsor effort. Internal hours rise in Stage 4, but nothing like proportionally, because the sponsor is supervising a trial rather than running one.

The inverse holds in Stage 5: 13% of hours against 7% of money. Review and launch preparation is internally expensive and externally cheap. A stage's share of the budget tells you almost nothing about how much of the organization it will consume.

Three shapes in the matrix are worth naming, because each is a decision rather than an estimate. Nonclinical is front-loaded to the point of vanishing — 5,200 hours in Stage 1 falling to 400 in Stage 4. Judged on headcount it is the smallest technical function in the program; judged on when it matters, it is the function without which there is no IND. Technical Operations never goes quiet: it is the only function with material hours in all five stages, because CMC gates the filing on a clock that runs independently of the clinical program. And Market Access & Commercial does not appear at all in Stage 1 and barely in Stage 2 — which is the sequencing decision ยง7 returns to.

5. What the Internal Roster Actually Buys

The roster costs $36,008,000. The program's base budget is $217,000,000. Internal labor is therefore 16.6% of the program, and 83.4% is delivered by organizations that do not report to it.

AmountShare of base
Internal roster — salary, benefits, overhead recovery$36,008,00016.6%
External — CRO, CMO, central lab, sites, fees, advisors, counsel$180,992,00083.4%
Base program$217,000,000100.0%
16.6% of the money buys the capability to direct the other 83.4%.

That is the honest description of a sponsor-side roster, and it reframes what this plan is for. These 109 people do not manufacture drug substance, do not run the assays, do not enrol participants and do not write most of the documents in the submission. They specify, contract, oversee, verify and accept. The plan is a supervision capacity plan wearing a staffing plan's clothes.

Which is why under-resourcing it is so expensive and so hard to see. Cutting two people from 24 in Clinical Operations does not slow any trial activity down — Meridian's staff are unaffected. It reduces the number of monitoring reports anyone reads. The cost of that shows up in a finding, months later, in a document nobody connects back to a resourcing decision.

The corollary for estimating: internal labor does not scale with program value. It scales with the number of things that need supervising — sites, vendors, batches, submissions, datasets. A program twice this size run through the same CRO with the same number of sites would need a roster of very nearly the same shape. This is the reason the WBS records 64 work packages rather than costing by percentage of budget.

6. How a Person Is Actually Assigned

Nothing in the tables above assigns anybody to anything. The roster is an envelope; assignment is a negotiation, and the rule that governs it is the same one set out in the organization chart.

QuestionWho decidesInstrument
Who does the work — which named individualFunction headLine management. The Program Director has no vote here and should not want one.
How the work is done — method, standard, reviewFunction headFunctional SOPs, which outlive the program.
What is produced and whenProgram DirectorCharter, work breakdown structure, milestone commitments, gate conditions.
How much effort is availableNegotiated, and escalated when it failsThis plan. Its whole purpose is to make the negotiation reference a number.

A resource plan without this table is a forecast pretending to be a commitment. With it, the plan does one specific job: when a function cannot supply the hours this document says the program needs, that shortfall becomes a visible disagreement between two named accountable people rather than a quiet slip that surfaces as a missed milestone two quarters later. The escalation path runs to the Development Committee, and the evidence tabled is this matrix against actuals.

The most common failure of a resource plan is that it is never used again after it is approved. A plan that is not re-baselined at each gate stops being a control and becomes a historical document that everyone cites and nobody checks. This plan is re-tabled at every gate alongside the budget, and the stage columns above are what get compared.

7. The Plan's Blind Spots

Three limits, each of which materialized in this program.

It does not resource the trial. 24 people in Clinical Operations oversee 260 independent institutions running the pivotals. The people who actually screen, consent, dose and follow participants are employed by those institutions and by Meridian, and appear nowhere in this document. When enrolment ran 76 participants behind curve, no line in this plan could be adjusted to fix it — which is precisely why the recovery levers were contractual and site-level rather than resourcing.

It does not see the vendor's staffing. The program contracts a deliverable and a service level from Meridian, not a headcount. That is the correct contracting posture, and it means the sponsor cannot tell from this plan whether the CRO has staffed adequately. The instrument that addresses it is oversight — the governance and audit rights in the CRO Oversight Plan — not a resource table.

It cannot correct a sequencing decision after the fact. Market Access & Commercial is allocated 0 hours in Stage 1 and 200 in Stage 2. The payer evidence that determines formulary position has to be designed into Phase 3 to exist at launch. By the time the function was resourced at anything like the level the launch needed, the Phase 3 protocol was final and the collection opportunity was closed.

Gate Condition GC-03 was the governance system detecting exactly this, and being unable to repair it.

A condition can require evidence. It cannot retroactively create the years of collection that would have produced the evidence. The resourcing decision that determined the access outcome was made in Stage 2, was visible in this table as two hundred hours, and was not contentious at the time because nobody reads a resource plan looking for the row with the smallest number in it.

The row with the smallest number is where a resource plan should be read first.

8. Resource Risks

RiskExposureResponse
Key-person concentration102 working staff across 10 functions leaves several single points of failure — most acutely in Regulatory Affairs (10 people carrying every agency commitment) and the Program Management Office (4).Named deputy for every control account owner; the gate pack cannot be tabled by one person.
Stage 4 concurrency20.1 FTE sustained across 27.0 months is the longest continuous demand in the program, and it overlaps the CMC validation peak.Phased against the CMC critical path rather than the clinical one; CMC hours are protected at gate review because CMC gates the filing independently.
Attrition through a {F.PROGRAM_YEARS:.0f}-year programThe program outlasts typical tenure. Turnover costs qualification and re-training, not merely recruitment — and in a GxP environment an unqualified person cannot perform the task at all.Documented handover as a deliverable, not a courtesy; training records maintained as essential documents in the TMF.
The plan not being re-baselinedThe most likely failure and the least dramatic. A stale resource plan is indistinguishable from a current one until it is needed.Re-tabled at every gate with the budget; stage columns compared to actuals.

9. A Correction Recorded

This plan was the artifact that exposed a defect in the suite, and the defect is worth recording because of the class of error it belongs to.

Until 29 July 2026 this program's roster existed in two places: a cost view used by the charter and the budget, and an organizational view used by the organization chart and the work breakdown structure. Both summed to exactly 109 people and 191,420 hours. Every automated check passed. They disagreed on every individual function. Regulatory Affairs was eleven people in one and nine in the other; Market Access & Commercial was nine and four.

The totals were right and the decomposition underneath them was fiction.

The checks in place asserted that each view summed to the roster. Both did. What no check asserted was that the two views described the same people — and because the organizational view was required to total 109, the seven oversight seats had to be absorbed into the working functions to make the arithmetic close. The assert was not the safety net. The assert was the cause.

Both views now derive from one roster, and the check that was missing is in place: they must agree function by function, not merely at the total. An identity that holds says nothing about whether the things being counted are the same things.

The correction changed distribution, not magnitude. Headcount, hours, internal labor cost and the 83.4% external share are unchanged; what changed is which function holds which share of them, and two claims elsewhere in the suite that had reasoned from the superseded numbers.