Statistics

1
Statistics
Ah Mun Kuan Head of Statistics 28th September 2005
2
(No Transcript)
3
Focus on 4 areas

• 1. Data in National Reporting and Learning
  System (NRLS)

2. Number of notifications
3. Analytical and reporting tools
4. Types of statistical reports
4

• Data in
• National Reporting and Learning System
• NRLS

5
National Reporting and Learning System (NRLS)

• Electronic system for NHS organisations and staff
  to confidentially report patient safety incidents
  to a national database held at NPSA
• Link through Trusts Local Risk Management
  Systems (LRMS)
• Roll out started in November 2003
• Roll out completed in July 2005
• 91 of Trusts are connected to the NRLS via their
  chosen route LRMS and permanent eForm
• 9 of Trusts are working on converting from
  connection via the eForm to LRMS

6
NHS Trusts - 8 vendors Pharmacies
NHS Trusts Pharmacy -PCT
Public Pharmacy High street
Email Letters Fax Phone calls
7
Two types of NRLS data
Structured data Care setting Location Incident
type Degree of harm Specialty Contributing
factors etc
Textual data What happen (IN07) Preventative
action (PD13) Underlying causes or events
(IN11) Drug details (MD07) etc

Anonymisation Recoding
8
Anonymisation and Recoding

• Anonymisation -
• Remove identifiable information on database
  names, hospital number and date of birth
• Use Data Cleansing Software (Solcara) to tag
  identifiable data
• Data Engineers replace them with ltnamegt and
  ltnumbergt
• Testing for auto-anonymisation
• Study of 6643 over a 5 day period
• True ve report 6.5
• False -ve report 2.1
• False ve type 1 5.1 (text lose its meaning)

9

• Recoding
• - Assign other into relevant category
  (medication, specialty, device and location)
• - To improve quality of data

10

• Number of Notifications

11
Number of Notifications
2,000 a day
10,000 a week
40,000 a month
12
(No Transcript)
13
The reporting gap?
Occurrence
Incident occurrence
10.0
(Per C.Vincent et al)
Reporting Gap ?
5.0
Reporting rate
(NRLS)
0
time
14
(No Transcript)
15
Notification to NRLS
Others 1 ambulance service 2 community general
dental service 3 community optometry / optician 4
community pharmacy 5 general practice Learning
disabilities
16
Increase in reporting by care setting
17

• Analytical and Reporting Tools

18
Analytical and Reporting Tools
Autonomy
SAS (Enterprise guide)
Textual Data
Analyse structured data Run statistical
analysis Search for key words (exact match)
Free text search key words (similar
words) Clustering - restriction on 11,000
reports
19
CLUSTERS- DEVICES
20
TOTALITY
21
Current issues

• Duplication identification
• Susceptibility to spelling
• Matching on length

22

• Statistical Reports

23
Statistical Reports using NRLS data

• Monthly management Team report
• Patient Safety Observatory reports, bulletins
• Web based feedback to Trusts Phrase 1, Phrase 2
• Monthly monitoring of notification via eForm
• Medication reports
• Analysis of top 10 incidents to inform
  prioritisation
• Analysis to support the issue of Alert
• Data request services (90 requests from April -
  August)
• Freedom of Information Act request (8 January
  August)

24
Current development

• Streamline recoding of textual data
• Database development
• Comparative analysis using family group
• Clustering of key words (DIY and list of similar
  words)
• Automation of reports
• Policies in releasing of NRLS data