Tiered Review Protocol¶

A review-routing protocol — instantiates Oversight Span Calibration

Routes items into fixed review lanes by risk band, each lane staffed at a defined level of scrutiny and authority, so the scarcest reviewers see only the heaviest cases.

Tiered Review Protocol operationalizes graded oversight into a small set of named lanes. Its defining move is discreteness: rather than a continuous dial of review depth, it defines a handful of tiers — say low, moderate, and high — each with a pre-assigned reviewer, a fixed depth of scrutiny, and a bounded decision authority — and routes every item into exactly one. The lightest tier may clear on a single sign-off or a self-certification; the middle tier goes to a peer-review cell; the top tier goes to the full committee or senior authority. The point is to reserve the scarcest, highest-judgment reviewers for the cases that genuinely need them, while giving the mass of ordinary items a fast, defined, lower-cost path that never touches the bottleneck.

Example¶

A university's ethics board cannot convene the full committee for every study — most are surveys and record reviews that pose minimal risk. Its tiered review protocol sorts each protocol into one of three lanes, mirroring the well-established exempt / expedited / full-board structure of human-subjects review.^[1] Minimal-risk studies are handled in an exempt lane by a single administrator; moderate-risk studies go to an expedited lane reviewed by one or two experienced members acting as a peer-review cell; only greater-than-minimal-risk studies — a new drug, a vulnerable population — reach the full board. Each lane has a fixed authority: the exempt reviewer may approve but not waive consent; only the full board may.

The full committee, the scarcest oversight resource, now meets to consider a handful of genuinely hard protocols instead of a hundred routine ones. Throughput rises and, paradoxically, so does quality — the heavy cases get undivided expert attention because the light ones were routed away from the experts entirely.

How it works¶

Define the lanes. Fix a small number of tiers, each with a named reviewer level, a scrutiny depth, and a decision authority — what that lane may approve alone and what it must pass up.
Set the routing rule. Specify the criteria (a risk band, a value threshold, a change type) that deterministically place each item in exactly one lane.
Staff each lane to its level. Light lanes get fast single reviewers or peer cells; the top lane gets the committee or senior authority — the resource you are trying to protect.
Provide a lane-jump path. If an item in a light lane turns out harder than its routing suggested, it is bumped up rather than forced through the wrong level.

Tuning parameters¶

Number of tiers — few lanes are simple but coarse; many lanes fit better but add routing overhead and boundary disputes.
Lane thresholds — where the cut lines fall between tiers. Set high and the top lane starves; set low and the committee floods again.
Peer-cell vs. hierarchy for the middle — whether moderate items go to a peer-review cell or up a management line. Peers preserve throughput; hierarchy preserves consistency.
Lane authority width — how much each tier may decide alone. Wide authority speeds the light lanes but risks under-scrutinized approvals.
Lane-jump sensitivity — how easily an item is escalated to a higher lane mid-review, trading throughput for safety.

When it helps, and when it misleads¶

Its strength is that it protects the bottleneck resource by construction: the expensive reviewers are structurally shielded from cheap work, and each item gets a scrutiny level matched to its lane. Because the lanes and their authorities are explicit, the protocol is also auditable — you can check whether items were routed honestly.

Its weaknesses live at the boundaries. A crude routing rule mis-sorts items, and a genuinely dangerous case dressed as routine sails down the light lane with no one qualified looking — so the routing criterion is the whole ballgame. Rigid tiers also invite gaming the lane: an item is described so as to qualify for the faster, lighter tier, turning tiering into a way to dodge scrutiny rather than allocate it — the protocol's version of running the system backwards. And boundary cases pile up in disputes about which lane they belong in. The discipline is a reliable routing input (often a real risk grade), a low-friction lane-jump so mis-routing is recoverable, and periodic sampling of the light lanes to confirm nothing hazardous is slipping through them.

How it implements the components¶

Tiered Review Protocol fills the routing-and-authority slice of the archetype — the structure of who reviews what and with what power:

peer_review_cell — the middle lane is typically a peer-review cell: qualified peers, not a manager, review moderate-risk items, keeping the hierarchy from becoming the bottleneck.
authority_boundary — each tier carries an explicit decision authority — what that lane may approve on its own and what it must pass to a higher lane — which is what makes the lanes safe as well as fast.

It routes and bounds authority but does not itself grade the risk that decides an item's lane — that is Risk-Based Review — nor set the default-ignore threshold for routine work, which is Management by Exception; an item that outgrows even its top lane is carried onward by Escalation System.

Instantiates: Oversight Span Calibration — this protocol turns graded oversight into concrete lanes that protect the scarcest reviewers.
Consumes: Risk-Based Review supplies the risk grade the routing rule reads.
Sibling mechanisms: Risk-Based Review · Escalation System · Management by Exception · Sample Audit Review · Delegation Framework · Supervision Ratio Model · Caseload Cap · Span-of-Control Design · Lead or Deputy Role · Management Layer Design · Oversight Dashboard

References¶

[1] U.S. human-subjects research review is organized into exempt, expedited, and full-board tiers under the Common Rule, with review level and required scrutiny set by the study's risk. It is a canonical, correctly-tiered review protocol: the full board's limited capacity is reserved for greater-than-minimal-risk studies. ↩