5. ACTIVITY, PERFORMANCE AND
EFFICIENCY (continued)
5.2.6 Could the Department comment on
the progress of the Outpatient Commissioning Dataset? How much
has been spent on the project, by year? How much is planned to
be spent, by year? When are data likely to be published and in
what form? Could the Department comment on the quality of these
data? Could the Department comment on the possibility of making
diagnosis and procedure mandatory in data submission? (Q73)
ANSWER
1. The Outpatient Commissioning Dataset
(CDS) flowed for the first time midway through 2001-02. Therefore,
the dataset is relatively new. Like all new datasets, data quality
and coverage were initially patchy. The advent of Payment by Results
and the 18-week wait programme have raised the importance of the
dataset significantly. Work is being undertaken to improve quality
by:
— Publishing the data, so NHS organisations
can view each others' figures and compare quality; and
— The national data improvement programme
(NDIP). This is led by the Information Centre (IC) with the aim
of improving data quality and fitness for purpose by working with
NHS organisations and by influencing future dataset developments
to ensure that they facilitate the submission of good quality
data from the NHS.
2. Table 73 shows the estimated historic
and planning on selected projects around the outpatient CDS.
3. In July 2006, analysis of the record
level information from the Outpatient Commissioning dataset was
published for the first time by the Information Centre's Hospital
Episode Statistics (HES) team. The published data covered 2004-05
and 2003-04 and were labelled as "experimental". The
"experimental" status is used within Official Statistics
to cover data which contain a great deal of useful and usable
material but which are indicative and are not yet sufficiently
robust to be definitive.
4. The publication can be downloaded from
the IC website at the following address:
http://www.ic.nhs.uk/pubs/hesoutpatients2003to2005/hespub/file
5. The quality of the data is discussed
in section 4 of the publication referenced above. The main findings
on quality are:
— National attendance counts show close
agreement between these data and existing DH aggregate returns
and in many areas the data provide a good basis for national comparison;
— At local level the new data need
to be treated with greater caution due to a few large local variations
in completeness and recording practices;
— Data items that are well completed
(above 90%) include age, referral source and personal information
around the patient that is essential for tracking the patient's
journey through the NHS; and
— Data items that are poorly completed
are mainly those that are not mandated, such as information on
whether the patient did not attend as well as procedure and diagnosis
information.
6. In addition to publishing the analysis,
customers have the opportunity to request the raw data, so that
they can perform further analysis of the data and provide the
IC with comments around their quality. A specific outpatient data
quality publication is planned for late autumn 2006, that will
explore quality issues in more detail based on this user feedback.
7. The possibility of mandating procedure
and diagnosis is currently being actively discussed and explored.
Clearly there would be a large cost to the NHS in additional staff
time and enhanced systems to enable this information to be consistently
recorded. However, having these items mandatory would make the
dataset much richer and facilitate both the monitoring of the
18 week wait target and the funding of outpatient procedures and
diagnoses under Payment by Results (PbR).
The current plan is that NHS providers will
be paid for outpatient activity under Version 4 Healthcare Resource
Groups under the PbR tariff from 2008-09. Procedure and, probably,
diagnosis information will be needed to create these HRGs. Therefore,
if NHS organisations do not start submitting the information at
this point, they will not get paid, so this will produce a powerful
incentive for this information to be submitted, regardless of
whether its submission is mandated.
| 
