POLYSOMNOGRAPHY

Sleep lab study scoring software that keeps the scorer, the epochs and the physician review on one study

Sleep lab study scoring is the epoch by epoch reading of a polysomnogram, in which a technologist stages sleep and marks respiratory events so a physician can interpret the study. Medicare treats a sleep study as 6 or more hours of recording with physician review, and requires the laboratory physician to review the entire raw data recording for every patient studied. Scoring is also the least reproducible step: in a study of 70 polysomnograms read by 10 technologists, only 1.5 percent of stage N1 epochs were scored unanimously. Neurobird holds the study, the scorer and the review on one record.

Sleep technologist at a scoring workstation with a night of polysomnogram traces on screen, used to show the epoch by epoch judgment calls that turn a recording into the numbers a physician will sign
The recording is objective. The staging on top of it is a judgment, made one epoch at a time.
6 hoursof recording a Medicare sleep study needs
1.5 percentof stage N1 epochs scored unanimously
15 eventsper hour, the AHI that opens CPAP coverage

Neurobird Sleep Lab Study Scoring in short

  • Hold the scoring queue by study rather than by whoever happens to be logged into the acquisition system
  • Attach the scorer's credential and its expiry date to the study they actually read
  • Compare a rescore against the original and report the agreement instead of quietly overwriting it
  • 6 hours Recording length Medicare uses to describe a sleep study or polysomnogram, with physician review, interpretation and report, and the laboratory physician must review the entire raw data recording for every patient studied. CMS Article A53019
  • 15 events Apnea hypopnea index per hour that supports an initial 12 week period of CPAP, or an index of 5 to 14 events per hour with documented sleepiness, cognitive or mood symptoms, hypertension, ischemic heart disease or stroke. CMS NCD 240.4
  • Pricing runs 179 to 1380 US dollars per month across 3 tiers. Early access is free.
The daily reality

What sleep laboratories and diagnostic sleep centers actually deal with

Two scorers read the same night and the report goes out without anyone knowing how far apart they were.
The physician review is a signature on a report rather than a record that the raw data was actually opened.
Scorer credentials, the ones the payer asks about, are in a folder in the manager's office and nowhere in the study.

Why it stays broken

A sleep laboratory buys its scoring tools from whoever made its amplifiers, so the epoch data lives inside a vendor system built to display a waveform rather than to manage a queue of studies and the people reading them. The regulator does not care which software you use, only that a credentialed person scored it and a qualified physician reviewed the whole recording, and neither of those facts is something a waveform viewer was designed to prove. So the audit trail ends up as a spreadsheet beside the scoring station.

In 70 polysomnograms scored by 10 experienced technologists, intraclass correlation with the group average ranged from 0.18 to 0.96 depending on the technologist and the stage.

A sleep study, also known as polysomnography, is a test that measures and records different body functions while you sleep.

MedlinePlus, sleep study, source
What it does

Neurobird Sleep Lab Study Scoring Inspection Tracking Platform

One record per study: the recording and its duration, the technologist who scored it and what they were credentialed to do that night, the rescore if there was one, the agreement between the two reads, and the physician review that closed it.

  • 1Hold the scoring queue by study rather than by whoever happens to be logged into the acquisition system
  • 2Attach the scorer's credential and its expiry date to the study they actually read
  • 3Compare a rescore against the original and report the agreement instead of quietly overwriting it
  • 4Record the physician review of the raw recording as an event with a time, not as a signature line

What changes with Neurobird Sleep Lab Study Scoring?

The same work, read left to right: how it runs today, and how it runs once the record is in one place.

Sleep lab study scoring: current practice compared with Neurobird Sleep Lab Study Scoring
TodayWith Neurobird Sleep Lab Study Scoring
Two scorers read the same night and the report goes out without anyone knowing how far apart they wereHold the scoring queue by study rather than by whoever happens to be logged into the acquisition system
The physician review is a signature on a report rather than a record that the raw data was actually openedAttach the scorer's credential and its expiry date to the study they actually read
Scorer credentials, the ones the payer asks about, are in a folder in the manager's office and nowhere in the studyCompare a rescore against the original and report the agreement instead of quietly overwriting it

Who is this for?

The same epochs, three very different operations.

Hospital sleep laboratory

You score in house

A handful of beds and a small scoring team. You need the queue, the credentials and the physician review in one place when the surveyor asks how a study was read.

Independent sleep center

You send scoring out

Contract scorers reading overnight across time zones. You need turnaround, agreement between readers and the credential file for each one on the study itself.

Home sleep testing service

You score high volume

Type III and Type IV recordings at volume, with a physician who has to interpret every one. You need the exceptions surfaced rather than buried in a batch.

What the evidence says about scoring a sleep study

The Medicare documentation and coverage rules a sleep study has to satisfy, and the published measurement of how far apart two experienced scorers can be on the same night.

6 hours

Recording length Medicare uses to describe a sleep study or polysomnogram, with physician review, interpretation and report, and the laboratory physician must review the entire raw data recording for every patient studied.

CMS Article A53019
15 events

Apnea hypopnea index per hour that supports an initial 12 week period of CPAP, or an index of 5 to 14 events per hour with documented sleepiness, cognitive or mood symptoms, hypertension, ischemic heart disease or stroke.

CMS NCD 240.4
10 seconds

Minimum duration for an apnea. A hypopnea also lasts at least 10 seconds and needs at least a 30 percent reduction in airflow or thoracoabdominal movement together with at least a 4 percent oxygen desaturation.

CMS NCD 240.4
1.5 percent

Share of stage N1 epochs that received a unanimous score from all six technologists reading the same recordings. For stage N3 the unanimous share was 3.8 percent.

J Clin Sleep Med 2018, PMC5786839
45.2 percent

Share of epochs scored as stage N1 by one technologist that no other technologist scored as stage N1. The same figure for stage N3 was 40.8 percent.

J Clin Sleep Med 2018, PMC5786839
70 studies

Polysomnograms scored independently by 10 trained technologists from five academic sleep laboratories. Intraclass correlation with the group average ran from 0.18 to 0.96 depending on stage and scorer.

J Clin Sleep Med 2018, PMC5786839
Interactive preview

The scoring queue, as your laboratory would work it

Assign a study and watch the credential check, the agreement figures and the review step move with it.

neurobird / sleep study scoring queue
0 of 4 study inspected

Tap an asset to log a scan and timestamp the inspection.

How does Neurobird Sleep Lab Study Scoring work?

  1. Take the study in

    The recording, its total time, the order and the referring physician land as one record, so a study that fell short of 6 hours is visible before anyone starts scoring it.

  2. Assign and score

    The queue routes by credential and workload, the scorer is stamped on the study, and a second read is stored beside the first rather than on top of it.

  3. Review and release

    The interpreting physician's review of the raw data is recorded as its own step, and the report goes out with the scorer, the reviewer and the agreement figures attached to it.

Sleep lab study scoring software questions, answered

Key terms

What is sleep lab study scoring?
It is the epoch by epoch reading of a polysomnogram, in which a technologist assigns a sleep stage to each 30 second epoch and marks respiratory events, arousals and limb movements. The interpreting physician then reviews the raw recording and writes the report from the scored data.
What are the documentation requirements for a Medicare sleep study?
CMS describes a sleep study as 6 or more hours of recording with physician review, interpretation and report. The report has to carry the parameters monitored, start time and duration, total sleep time, sleep efficiency, time in each stage and the respiratory findings, and the laboratory physician must review the entire raw data recording.

Why does interscorer agreement matter?

Because the numbers on the report depend on who read it. In a published study of 70 polysomnograms read by 10 experienced technologists, only 1.5 percent of stage N1 epochs were scored the same way by everyone, and intraclass correlation with the group average ranged from 0.18 to 0.96.

Does this replace our acquisition system?

No. The amplifiers, the waveform display and the scoring tool stay where they are. This holds the queue around them: which study went to which scorer, what their credential was, how a rescore compared, and when the physician reviewed the recording.

Can it track scorer credentials?

Yes, and it ties them to the study rather than to a folder. The payer article names the credentials a scorer must hold, so the useful place to keep them is beside the studies that person actually read.

Why we are building this

We went looking for measurements that get reported as facts and are actually judgment calls, and sleep staging kept coming up. The published work is blunt about it: ten experienced technologists reading the same seventy nights agreed unanimously on 1.5 percent of stage N1 epochs. Nobody in a laboratory is surprised by that, and almost no laboratory has a way to see it happening in its own data. That is a record problem, and record problems are what we build. The sources are linked above. If we have the workflow wrong, tell us and we will fix it.

Neurobird Team neurobird.com

Where the requirement comes from

Primary sources, straight from the regulators.

How much does Neurobird Sleep Lab Study Scoring cost?

Priced per laboratory because the accreditation file and the scoring queue belong to the site rather than to a scorer. Studies, rescores and reviewers are unlimited on the upper tiers.

Single laboratory
$179
per month
  • Up to 4 beds
  • Scoring queue
  • Credential tracking
  • Physician review log
Request early access
Sleep center
$545
per month
  • Unlimited studies
  • Second read comparison
  • Interscorer agreement
  • Turnaround reporting
Request early access
Multi site group
$1,380
per month
  • Multiple laboratories
  • Contract scorer pool
  • Accreditation export
  • Priority support
Request early access
Opening 8 early access places for sleep laboratories and diagnostic sleep centers.

Get free early access

Early access means we load your bed list, your scorer roster and your current turnaround targets before you type anything, and the account stays free while we do it.

Straight answer on where this is: The software is in development. Nothing is purchasable today. Early access means you shape it and pay nothing while we build.

Prefer email? Write to office@neurobird.com and a person will reply. No autoresponder.