Meridian Claims: A Bottleneck That Hiring Alone Will Not Fix
Meridian Claims Services adjudicates motor claims for three insurers. Arrivals average 480 claims per working day, ranging from 380 to 610. The process runs through five steps with these daily capacities: registration 600, document verification 520, medical review 420, coding 480, payment release 560. Medical review is staffed by seven nurses handling 60 claims a day each; a fully loaded reviewer costs $85,000 a year. There are 3,360 claims in the system and the oldest is 19 days old. Last month the medical review team lead began batching work by insurer and releasing it in blocks on Tuesdays and Thursdays, arguing it 'keeps the queues tidy'. Roughly 22% of arriving claims are low value, injury-free and from claimants with clean histories; each still goes through medical review because no rules-based routing exists. Insurers pay a fixed fee per claim and require 90% of claims closed within 10 working days. The managing director must choose between six weeks of overtime at $48,000 and hiring two reviewers.
1. Identify the bottleneck and quantify the current position using flow measures. What is the effect of the twice-weekly batching, and what does it cost per additional claim cleared to add a reviewer versus using overtime?
Read the worked answer
Bottleneck. Medical review at 420 claims/day against arrivals of 480. Every other step has capacity above arrivals, so medical review sets the throughput of the whole process; capacity anywhere else is wasted. The gap is 60 claims/day, so the backlog grows by roughly 60 per working day.
Flow measures. Little's Law: lead time = work in process / throughput = 3,360 / 420 ≈ 8 working days, and rising daily. The oldest claim at 19 days already exceeds the 10-day SLA, so the 90% target is currently being missed and will worsen. Note that adding capacity downstream would only increase the queue in front of the bottleneck.
Batching. Releasing work twice a week creates artificial queueing: the bottleneck idles early in the week and surges on release days. It converts smooth capacity into bursts, which increases both average WIP and lead time without adding a single claim of capacity. 'Tidy queues' are a visual illusion - the real measure is age of work in process. Batching should stop and be replaced by daily pull.
Unit economics.
- One reviewer: $85,000 / (60 claims × ~250 working days) ≈ $5.70 per additional claim cleared - a permanent, sustainable capability.
- Overtime: $48,000 for 30 working days. Even at a generous 60 extra claims/day that is 1,800 claims, or about $26.70 each - roughly five times the unit cost, and it disappears when the budget runs out.
Overtime is defensible only as a short, targeted intervention to clear aged claims, never as a capacity strategy for permanent demand.
2. Work through the arithmetic of the two proposals. Does adding one reviewer solve the problem? What combination would meet the SLA, and in what order should the actions be taken?
Read the worked answer
One reviewer does not solve it. Capacity becomes 420 + 60 = 480, exactly equal to average arrivals. The backlog of 3,360 never clears; average lead time stays at about 7 working days and every demand spike goes straight into aged WIP. This is the trap in the question - one hire is the intuitive answer and it fails.
Two reviewers: capacity 540, surplus 60/day, backlog clears in 3,360 / 60 = 56 working days. That meets the mean but not the SLA: claims already 19 days old will still be many weeks old when they are finally reached, so the 90%-in-10-days standard is breached throughout the recovery period.
Capacity plus demand shaping. Divert the 22% simple claims - about 106 per day - to a rules-based fast lane. Arrivals at medical review fall to ~374, and their share of the backlog (about 740 claims) also leaves the queue.
- One reviewer: capacity 480 against 374 = surplus 106/day; the 2,620 remaining claims clear in about 25 working days.
- Even without a hire, 420 against 374 gives a surplus of 46/day, clearing in about 57 days - enough to stop the deterioration immediately while hiring completes.
Sequence:
- Stop batching and move to daily pull - free, immediate, reduces lead time.
- Stand up the fast lane with a quality-audit sample; agree the criteria with the insurers rather than imposing them silently.
- Hire the first reviewer now; hold the second contingent on whether the fast lane delivers the modelled diversion rate.
- Use a small, time-boxed overtime budget targeted only at claims aged over 8 days.
- Report WIP, p90 lead time and aged-claim counts weekly to the managing director.
3. What performance measures should govern the change, and what gaming behaviour should you anticipate and design against?
Read the worked answer
Choose measures that describe flow, not just effort.
- p90 lead time and the percentage of claims closed within 10 working days - the contractual measures.
- Aged WIP: count of claims over 8 and over 15 days, reviewed daily.
- First-pass accuracy and rework rate, because speed bought with errors simply relocates the queue.
- Diversion rate into the fast lane, with a monthly quality audit of a random sample of diverted claims.
- Bottleneck utilisation tracked weekly to detect the next constraint moving.
Anticipated gaming and the design response:
- Cherry-picking easy claims to hit a per-reviewer productivity target. Design response: never pay or report on claims-per-reviewer alone; measure p90 lead time at the process level.
- Re-categorisation. Nurses pushing complex claims into 'coding' or returning them to verification to move them off their queue. Design response: audit the rework and re-route rates by individual and by team, with a threshold that triggers review.
- Closing claims that are not really closed to satisfy the SLA clock, e.g. issuing a request-for-information and stopping the timer. Design response: define closure by contract, not by internal status, and report the count of claims in information-pending separately.
- Fast-lane drift. Gradually widening the diversion criteria until genuinely complex claims bypass clinical review. Design response: criteria change requires clinical sign-off and is version-controlled.
The general principle: metrics that are visible weekly and owned at the process level resist gaming better than individual productivity scores, and the managing director should review a sample of aged claims monthly rather than only the dashboard.