Overview of the Guide
- Arjun Radhakrishnan (Unlicensed)
Overview
The PCOC Data Quality Report (DQR) contains information relating to the extracts submitted PCOC extracts.
It includes three main areas:
Scope Details
Report summary
Data Item Errors.
The Data Item Errors are, then, split further into five sections for correction or review.
•Section 3 - Missing clinical items
•Section 5 - Internal items (PCOC only)
This document provides an explanation of each of the areas and sections included in the PCOC DQR.
If corrections are made, data should be extracted from the IT system again and resubmitted to PCOC to ensure the most up-to-date data is included in reports with the highest data quality possible.
Scope Details
The scope details for a Data Quality Report can be found in the top left-hand corner of page 1.
It includes:
•Report Date: The date that the PCOC extract was submitted and therefore date the DQR was produced.
•Service ID: Unique service identifier assigned by PCOC
•Service Name: The Name of the palliative care service
•Date Range: The time period/data scope (i.e. start and end dates) that will be applied to data from the extracts to determine which will be included in the DQR and the outcome report
•Episodes: The number of inpatient, community, and ambulatory episodes that are within the date range (Not only those that are valid)
•IT system: The IT System that the service uses to collect and extract PCOC information
•Date Range: The time period/date scope (i.e. start and end dates) that will be applied to data from the extracts to determine which will be included in the DQR and the outcome report
Please check Episodes and IT systems in particular to make sure these are correct
Report Summary
This section gives a summary of the volume of data submitted, It includes:
Level: Relates to the type of extract submitted and corresponding ‘data-level information’. For version 3 of the PCOC dataset, patient, episode and phase extracts are required. Assessment and profile extracts are optional.
Filename: The filenames of each extract. These files often include the date and time the extract was created.
Total, In Range and Valid data:
Total: the total number of records included in the extract file
In range: the number of records that are determined to be In Range of as specified by the ‘Scope details’ section
Valid: the number of records that are Valid after going through PCOC data quality checking
If there are a large number of errors in the DQR, please check that the extracts:
Are submitted to the correct submission field (i.e. Patient to Patient, Episode to Episode, Phase to Phase)
All the extracts were created on the same date (see red circled section below), if not, create a new set of extracts and resubmit
Data Item Errors
This area of the DQR includes errors found in the checking process. Data Item Errors are split into 5 sections:
•Section 1 - Data Item Errors
•Section 2 - Missing Items
•Section 3 - Missing Clinical Items
•Section 4 - For Review
•Section 5 - Internal Items (PCOC only)
These sections each have the following details, listed left to right, for each recorded error:
Item on the table | Definition |
|---|
Item on the table | Definition |
|---|---|
Patient | The patient ID that the error relates to |
Ep Start | The episode start date/Episode ID of the episode. |
Ph Start | The phase start date of the phase |
Level | The data level/extract that this error relates to e.g. patient, episode or phase |
Data Item | The specific data item where there is an error |
Description | A description of the specific error that is occurring (e.g. missing value) |
Value | The response provided for that item in the data extract |
Severity | Indicates how this error may affect the reporting of this data item |
Line | Refers to the line that the record sits in the data extract file |
Type of error severity
Record errors invalidate an entire record. This means that the episode and all associated phases are removed from the dataset. You can see how many additional phases are removed in section 5 of the data quality report.
Variable errors invalidate the response to that item and only the response is excluded from the report
Variable missing - the response for that variable is missing and therefore cannot be included in the dataset
Variable warning - the response for that variable is for review only. The variable is not incorrect but we frequently see missing or inconsistent responses and so it is flagged for review. If this response is correct, no further action is required.