In This Issue: Adjustments and Claim Metrics

Issue summary and other notes...

Article 1 Claims Adjustments. Measuring, tracking, and improving claims turnaround.

Article 2 Decision Support in Claims. Best practices for setting up and managing a reporting environment in claims.

Notes

Our fixation on events (last month's sales, budget cuts, earnings...) is actually part of our evolutionary programming. The irony is that, today, the primary threats to our survival, both of our organizations and of our societies, come not from sudden events but from slow, gradual processes: decline in product quality, obsolute phyiscal capital, environmental decay... -- Peter Senge (The Fifth Discipline)

If you have any trouble sounding condescending, find a Unix user to show you how it's done. Scott Adams

We don't have a monopoly. We have market share. There's a difference. Steve Balmer

“What worries me is the possibility that we’ll create a world that is much more economically efficient – but that is much less satisfying to live in… Tom Malone MIT professor

The Misunderstood World of Claims Adjustments

Claims adjustments are usually not the focus of process improvement efforts. That’s probably the way it should be. Read on and see why.

An adjusted claim is a claim that reaches final status (paid or denied) and is then later adjusted (typically manually). The most common adjustment is for a provider to inquire (call) regarding the paid amount for a particular claim, and for that amount to be adjusted.

Claims operations tend to not to focus their process improvement resources on adjustments for four reasons:

1. If a claims operation is subdivided into activities, adjustments typically only represent about ten to twenty percent of total claims work (with the other eighty to ninety percent divided between data entry, manual claims adjudication, and member/provider calls). Process improvement efforts tend to focus on larger opportunities.
2. The member service and provider relations impacts of adjustments tend not to be significant. First, few of the calls are from members. Second, those from providers are, arguably, part of any bill paying process – everyone would be happier if they were gone but they’re not going to substantially damage relations.
3. The data to understand adjustments is spread across systems (calls in the call center system and adjustments in the claims system) and time periods and as a result is often difficult to link and analyze. In addition, adjustments are typically manually logged by the person taking the call or adjusting the claim. The reason code assigned by that person is characterized by the accuracy problems typically associated with manually selected reason codes. Finally, it is not uncommon for providers to both call regarding a particular claim and to resubmit the claim – adding the complexity of potential duplicate claims to the mix.
4. Many benefit plans measure quality based on a review for accuracy of a sample of paid claims – rather than measuring quality based on the number of claims adjusted.

Adjustments aren’t as hopeless as they might appear. A few quick tips might help jump start your efforts.

1. Get a better handle on the biggest reasons for adjustments (assuming you either don’t have reason codes or the ones you have are not useful) by having two or three examiners tick sheet one day worth of adjustments to track their reasons. Look for the big stuff — if they adjust 20 claims a day, 10 to 15 of them are driven by three big root causes.
2. Look at manually paid claims. People make more mistakes than claims systems do. Look for patterns in the mistakes made (i.e. benefit plan related).
3. Create a database of all the claims adjusted in a particular week and look at how many were associated with duplicates. For the larger providers, work to both understand adjustment patterns and reduce resubmission.

A Yardstick for Claims Operations

The right operating measures put you on track to be a world class claims operation. Here's the ClaimHeader shortlist...

It has been a busy 20 years for measurement. Technologies and concepts have raced ahead. The balanced scorecard. Data warehouses and marts. Desktop analysis tools. In spite of this evolution, designing an appropriate set of measures is not a trivial task. With today’s technologies, the biggest challenge typically ends up being what NOT to measure.

How did we pick our claims process operating measures? We started with the characteristics of a good measure:

1. Drives appropriate operating behavior. FedEx leads the way here by measuring how many packages they failed to deliver in one day — an operating measure which focuses attention on getting packages to the customer in one day (process efficiency and customer service).
2. Easy to manage. Again the FedEx example applies. Assuming the appropriate technology is in place, establishing a clearly defined one-day target and measuring it is a manageable task.
3. Linked to company financial performance. At the end of the day, things get measured for one reason, bottom line impact. Maybe the biggest trick of measurement is to understand how and when operating measures impact the bottom line.

We then divided the claims operation into three measurement categories and used the above criteria to pick our measures. We focused on the primary processes of a claims operation and left out other processes (data entry, enrollment and benefit plan set-up — a set of measures for those secondary processes will likely show up in a forthcoming issue). We also left out many derivative measures (i.e. cost per claim) because they’re generally not directly operational.

Production

1. Backlog (daily measure): Total claim volume in inventory (TAT is a derivative measure of backlog). Used for daily staffing and resource decisions.
2. Productivity (daily measure): Total volume of claims per day per examiner/analyst. Most claims operations have individual-level production goals which roll-up to operation-level goals.
3. AA Rate (weekly measure): Less actionable than backlog or productivity – increasing your AA rate is a significant long-term project. Still AA rate is the a fundamental measure of claims process efficiency. The higher a plan’s AA Rate the more efficiently (i.e. cheaply) it processes claims.

Quality

1. Adjustment Rate (weekly measure): To understand what to do about adjustments, see the article above.
2. Error Rate (monthly measure): Use a sample of paid claims to measure system and examiner/analyst accuracy.

Customer Service

1. Speed of answer and dropped calls (daily measure): How long do members and providers have to wait and how many get tired of waiting?
2. Calls per provider/member (monthly measure): Understand who is calling, and whether they’re calling more or less often.