AUTOMATION IN PRACTISE: The SWCIS Experience Presentated By: Tina Ball …………Head Of Registration Gill Christmas ……Head Of IT 4 th December 2002
Automation In Practise - SWCIS Where Did We Start From? Merging of 2 Registries – Wessex (part auto) South West (fully manual) Auto load of Clearnet for at least 8 years No auto load of either pathology or deaths data Deaths – 6 weeks to process
Automation In Practise - SWCIS Where Are We Now? Electronic Data Records Coverage Clearnet 700, 000 100% Pathology - 100, 000+ (exp) 83% Cancer Deaths 22, 000 100% X-regionals - 3, 500 - 95%
Automation In Practise - SWCIS Why Did We Go the ‘Auto’ Route? 1) Clearnet processing – 700, 000+ records a year to handle - capture of clinical data 2) Speed up Registration – more timely + improve availability of information 3) Flexibility to handle constantly expanding data sources
Automation In Practise - SWCIS The Auto Load Process No manual intervention Overnight Batch Processing – Manual intervention Recycled Records 65 % Clearnet 80 % Cancer Deaths Auto fill of demographics, some tumour details and all treatment (Clearnet) Rapid processing of pathology text
Auto loaded
Auto loaded
Automation In Practise - SWCIS Strip off keep Processing Decisions – the ‘Markup’ process Report : SWCRF Sort sequence : SWCRF Run Number : 52 Originally reported : 07/08/2000 08: 05 Report date : 09/08/2000 11: 33 [1 m[1 MOF-NINE SEVEN [2 m F 23/07/1956 Q 123456 THE STARSHIP THE ENTERPRISE Dr B Crusher QAH ITUQ AA 007007 K/99 POST MORTEM HISTOLOGY Sections from the greater omentum show diffuse infiltration by a moderately differentiated adenocarcinoma as seen previously ……………. . Dr Test SHO in Pathology Test Hospital ALPHA QUADRANT DCT/GCS/JM 1. 8. 00 SNOMED CODES: Seq 2 T 57000 - M 81403, [1 m [2 m Dr I Test 847 25/05/2002 01/06/2002 [1 m[1 m. BAGGINS BILBO [2 m M 23/07/1956 P 000000 Pathology word document
Automation In Practise - SWCIS Text added to assist Registry officer End Result Q 123456||19560723||F||THE STARSHIP, THE ENTERPRISE||OFNINE|SEVEN|20020525|# Event Date: 20020525: # Hospital ENT 02 : # Unit No: Q 123456 # Clinician Name: B CRUSHER # Specimen Source: QAH # Pathologist Name: DR I TEST # Snomed Codes: T 57000 - M 81403, # Pathology Text: POST MORTEM HISTOLOGY ^ Sections from the greater omentum show diffuse infiltration by a ^ moderately differentiated adenocarcinoma as seen previously ………………^ Dr I Test ^ Test Hospital^ P 000000|19560723|M|MIDDLE EARTH, THE LAND|BILBO|BAGGINS Decided to use PERL: Programming Extraction and Reporting Language
Morphology Unit Number ICD Site Birth Date Behaviour Surname Automation In Practise - SWCIS NHS Number Laterality Forename Address Post Code Data extraction Data clean & validation Patient/Tumour Match Clearnet Auto Perl Markup Database Definate Match & New In. Definate match Path, pas etc Manual Cancerbase Database Business Rules / Data Validation Pool Database
Automation In Practise - SWCIS Automated Validation Rules 35 reference tables within Cancerbase On line cross validation checking, examples are - AD later than Do. D - Site and Morph compatible - Site and Behaviour compatible - Stage Type and Stage Based on ONS and UKACR guidelines
Automation In Practise - SWCIS Burden On IT Dept National GP, Consultant Hospital Codes OPCS 4 Codes ICD 10 Codes NHS no CLEARNET: - no names or addresses – nsts run - different formats received within each field - cleaning, validation and de-duplication - scripts not static – data formats & content change OTHER DATA SOURCES: - separate script for each data source - again scripts not static – data formats & content change - multiple applications
Automation In Practise - SWCIS Burden On IT Dept - contd STAFF: - increase in IT staffing levels – faults/enhancements - availability & retention of skilled staff - on-going training costs
Automation In Practise - SWCIS 3 Recommendations BEFORE developing an automated Cancer Registration System …………… 1) Do not under-estimate the complexity of automating the cancer registration process 2) Do not under-estimate the development costs and on-going costs of support & maintenance 3) Test , test and test again WE WISH YOU LOTS OF LUCK !
Скачать презентацию AUTOMATION IN PRACTISE The SWCIS Experience Presentated By
461e9f9b9a2f04a1b64de0e69e268d56.ppt