NWAU Periscope
Your activity, from the extracts your data warehouse already sends to IHACPA, priced in NWAU by the National Efficient Price Determination and forecast to the end of the year, with the coding behind it checked and audited.
Getting activity in
Open Warehouse activity and choose Import. Pick the stream (admitted acute, subacute, mental health, non-admitted or emergency), the date the warehouse produced the extract, and the file. The columns are the national minimum data sets’; Template downloads them, and a column with one of their usual names is found without renaming.
Preview reads the file and stores nothing: how many rows will be accepted, which belong to another stream, which cannot be read and why, and which hospitals are not mapped yet. Import then stores it. Send the same episodes again in a later extract and they are updated: an uncoded episode becomes coded when an extract brings its DRG.
The patient identifier is turned into a one-way key as the file is read and is never stored. Names, Medicare numbers and addresses are not asked for.
Mapping hospitals and held rows
Each establishment identifier in an extract is mapped once to one of your hospitals, under Mappings. Rows for an establishment that is not mapped are held, not dropped: Held & files lists them with a Map button, and they are priced the moment the mapping is saved.
Every file is listed there too, with the rows it could not read and the reason for each.
The forecast
Forecast shows, for the year and the network or hospital you pick:
- Coded NWAU to date: what the coded episodes are worth.
- The coding backlog: separated episodes not yet coded, estimated at the coded mix of similar episodes, and how many the next four weeks should code.
- The projected year end, with its range, against your target. Beside it, what the year would project from coded activity alone, so the difference the backlog makes is plain.
Below them: each month, how quickly your coding catches up after separation (drawn from successive extracts), how the backlog was priced, every stream, and the assumptions the forecast rests on.
Activity & NWAU and funding scenarios
Imported admitted episodes are part of Activity & NWAU: NWAU by month against the phased target, by hospital, specialty and DRG, and where each figure came from. Funding scenarios reprices the same episodes under a different determination or a changed assumption.
The coding audit
Where the warehouse can send each episode’s documents, Audit recodes a random sample with Code Captain, blind to the coding that was submitted: a coding summary is withheld and every code in the documents masked before it reads them. The documents arrive with the identifiers to redact, which are used once and never stored.
Each difference between the submitted coding and the recode is a question for an auditor: the submitted code was right, the recode was right, or neither. Only those decisions count. Accuracy is reported both ways (codes missed and codes the record does not support), by coder, from the random sample only. Each recode is a Code Captain run and counts against your allowance.
Reading the warehouse directly
Where your warehouse allows it, Feeds reads a view on a schedule instead of waiting for a file: the connection is set up once by your administrator, and a feed names the view, its stream, which of its columns are which, and how often to read it. Preview shows what the next read would bring, as counts and reasons; Pull now reads it straight away, and Runs lists every read.
If something looks wrong
- Rows are held: their establishment is not mapped. Map it from Held & files.
- An episode is not priced: its DRG or class is not in that year’s determination, or its version is not the one the year prices. The reason is on the row.
- The projection looks too low: check the coding backlog. A month with many uncoded separations is estimated, and the estimate is only as good as the coded mix of similar episodes.
Still stuck? Email support@healthos.net.au
