Case Study 1 — Items and Segments: A Composite

Constructed. The organization, the vendor, and the figures are not real. The failure — two groups of competent people describing the same problem in two vocabularies and each concluding the other was being unhelpful — is ordinary, and it is the only case study in this book where nothing was wrong with anyone's understanding of the work.**

Background

Section 25.1 said something that reads like a footnote:

A rejection message naming an ITEM is describing the form. A rejection message naming a LOOP or SEGMENT is describing the 837P. They are the same problem in two vocabularies, and a biller needs both.

This is a practice that had one vocabulary and a vendor that had the other.

The composite

Constructed.

A practice implements a new practice management system. The implementation is competent. Claims go out.

A category of claims begins rejecting.

The billing office does what billing offices do: reads the rejection, identifies the problem, and calls the vendor's support line.

Biller: "Our claims are rejecting because item 17b isn't populated."

Support: "I don't see an item 17b. Can you tell me which loop and segment?"

Biller: "I don't know what a loop is. It's item 17b. The referring provider's NPI."

Support: "If you can send me the 837 and the 999, I can tell you what's failing."

Biller: "I don't have those. I have a rejection report from the clearinghouse."

Both people in that exchange are competent, are trying to help, and are describing the same field.

What happened over four months

(Constructed.)

The ticket was closed as "unable to reproduce."

The biller escalated. A second ticket was opened, described the same way, and routed to the same team. The vendor's position, reasonably, was that they could not diagnose a transaction they had not seen. The practice's position, reasonably, was that the rejection report said what was wrong and someone should fix it.

The claims kept rejecting. They were reworked by hand — someone opened each one, populated the field, resubmitted. (Constructed.) Several hours a week, for four months.

And a subset aged out of the payer's timely filing window, because a rejected claim was never adjudicated and the clock had been running the whole time — Chapter 24 §24.1's fifth row.

What was actually wrong

A mapping.

The system had a field for the referring provider. It was populating it into the wrong place in the 837P — into a loop the payer read as something else — with the result that the referring provider NPI segment arrived empty.

From the practice's side, item 17b was blank. From the vendor's side, the field contained data and the software was working.

Both observations were true. The data existed and was in the wrong container, and the two vocabularies described opposite halves of that sentence.

How it was resolved

A person who spoke both.

(Constructed.) The practice hired a biller from a larger organization, who had worked with 837 files directly. On her second week she asked to see one.

She opened the file, found the loop, saw it was empty, found the data sitting in an adjacent loop, and described the problem in the vendor's vocabulary in a single sentence.

The vendor fixed it in three days.

What it shows

First, this is a translation failure and nothing else. Nobody misunderstood claims. Nobody misunderstood the software. The biller knew exactly what was wrong and could not say it in a form the vendor could act on, and the vendor could have fixed it in three days at any point in those four months.

Second, the vendor's request was reasonable and unmeetable. "Send me the 837 and the 999" is the right question, and the practice had no way to produce either — the files existed in the clearinghouse and nobody at the practice knew they could be retrieved. A reasonable request to someone who cannot fulfill it functions as a refusal, however it is meant.

Third, the cost was rework plus timely filing. The rework was visible — someone was doing it, and it showed up as hours. The timely-filing losses were not, and Chapter 24's Case Study 2 established why a cost that lands in a different category than the effort tends not to be counted against it.

Fourth, one person's vocabulary was the entire fix. This is the ninth finding in this book that came from a person rather than a control, and it is unusual among them: she did not notice something nobody had noticed. Everyone had noticed. She translated it.

And fifth — the two vocabularies are not going away. Payers and clearinghouses describe problems by item; vendors and integrations describe them by loop and segment. Chapter 25 teaches one and Chapter 27 teaches the other, and the reason both are in this book is that a biller who has only one will eventually have this conversation.

The lesson

Learn both vocabularies, and learn how to get the file.

Four carry-forwards:

Know your item numbers. They are how payers and clearinghouses describe problems, and they are a memorization task rather than a reasoning one.

Know your loops and segments well enough to ask. Chapter 27 §27.4 covers them without the jargon. You do not need to write an 837. You need to be able to say which segment is empty.

Find out how to retrieve a submitted 837 and its acknowledgments. Your clearinghouse can produce them; most billing offices have never asked. It takes one call, once, and it is the difference between "our claims are rejecting" and "this segment is empty."

And when a vendor asks for something you cannot produce, say so immediately. "I don't have that — can you tell me how to get it?" Four months of this composite is two groups each waiting for the other to do something neither had said they could not do.

Discussion questions

1. Both parties were competent and the problem persisted four months. Whose responsibility was it to bridge the vocabularies? Argue it from the vendor's side first.

2. The practice could not produce the 837. Should a billing office be able to? What would it take, and what is the argument against?

3. The rework was visible and the timely-filing loss was not. Design the report that would have made the second one visible. What is its denominator?

4. The fix came from one person's vocabulary. Is that a hiring problem, a training problem, or neither?

5. This is the ninth finding in this book that came from a person rather than a control, and it is different from the other eight. Say how — and then say whether that difference makes it more or less troubling.