Taxanes for Ovarian Cancer: Progress Report Rosemary Tate Information Projects Team December 2000.

Taxanes for Ovarian Cancer:Progress Report

Rosemary Tate

Information Projects Team

December 2000

Background Issues

• Recent research evidence has shown improved survival times with paclitaxel (Taxol) for ovarian cancer patients

• High cost • Variation in HA funding• Yorks Network (Aug 98), Northern (May 99)• National Cancer Guidance (July 99)• NICE (May 00)

Aims and Objectives (1)

• To quantify proportion of patients with a newly diagnosed, histologically confirmed ovarian cancer who received chemotherapy in 1998 compared with 2000

• To determine the proportion of patients who received taxane-based chemotherapy as first/second/third line chemotherapy

Aims and Objectives (2)

• To describe the patterns of yearly taxane prescribing across the different Health Authorities of N & Y

• To identify the circumstances under which taxanes were given, e.g. in a cancer centre/unit, as part of a trial/protocol

• To investigate reasons for any variation in the use of taxanes

Data Required

• Complete regional coverage• Diagnosis • Patient-based• Chemotherapeutic agents prescribed• Chemotherapeutic line (1st, 2nd, 3rd)• Clinical trial, name, arm

Data Sources Investigated

• Registry data• Regional Pharmacy

– Medicare Audit UK database

• Northern Gynaecological Oncology Group (NGOG) database

• Clinician’s own data• Trust/hospital Pharmacy data

Registry Ovarian Cancer Data(September 2000, excluding certain paths., private & ER hospitals))

HistologicallyConfirmed

1998 1999 2000 TOTAL

Complete 438 154 8 600

All 459 535 534 1528

% ofComplete/All

95% 29% 2% 39%

All withChemo “Yes”

221 108 11 340

% of All withChemo “Yes”

48% 20% 2% 22%

N.B. Chemotherapeutic agents, trials and line data not recorded

Medicare Audit Data

• 95% UK coverage of overall prescribing patterns

• Covers several high-cost drugs, listed by specialty, Trust

• May be made up at one hospital, but prescribed at another

• Trials and line data not recorded

Paclitaxel Average Monthly Cost by Trust

£0

£20,000

£40,000

£60,000

£80,000

£100,000

£120,000

LE

EDS TEACHIN

G

GATESHEAD

RO

YAL HULL

NEW

CASTLE C

ITY

SOUTH T

EES

BRADFO

RD

HUDDERSFIE

LD

CARLIS

LE

HARRO

GATE

NO

RTH TEES

NO

RTHALLERTON

PIN

DERFIELD

S & P

ONTE

SCARBO

RO & N

E YORKS

SOUTH D

URHAM

SOUTH T

YNESIDE

W

EST CUM

BRIA

2000

1999

1998

MEDICARE AUDIT DATA

Medicare Data - Missing TrustsAiredale )

Calderdale )

Dewsbury )

Newcastle upon Tyne Hospitals ) do not

North Durham ) prescribe

North Lakeland ) taxanes?

Northumbria )

Priority Healthcare Wearside )

Sunderland )

Tees and North East Yorks )

York )

North East Lincs ) not in our

Scunthorpe and Goole ) region?

NGOG Database

• 16 larger Northern hospitals submit data• 11 smaller hospitals and NCCT have not• Patient-based data from Jan 1997• Diagnostic, FIGO staging, Trials data

included• Chemotherapy type unreliably recorded• Clinician compliance disappointing• Changing to data entry at clinical interface

Clinicians’ patients receiving Taxol 1998-2000

• List of 11 Northern clinicians’ named patients with dates prescribed and costs

• 30/53 patients identified on Registry database as having ovarian cancer diagnosis

• Cost per patient varied from £1,026 to £11,437 (Mean £5690, SD £2960)

• Administered for mean of 67 days and up to 121 days

Treatment pathway of patients receiving Taxol 1998-2000 as recorded

by Registry • 9/30 (30%) patients had surgery immediately

after diagnosis• 6/30 (20%) had chemotherapy 1 to 7 months

after surgery• 5/30 (17%) had Taxol 5 months to 3.5 years

after first chemotherapy• 1/30 (3%) had post-op radiotherapy• 21/30 (70%) had no treatment

Clinicians’ patients receiving Taxol 1998-2000

• Patients prescribed Taxol 7 months to 12.5 years after diagnosis, mean 3.5 years

• 12/30 (40%) who received Taxol 1998 - 2000 were diagnosed 1998 - 2000

• 1/30 (3%) chemotherapy date = Taxol date (7 months after diagnosis)

• 15/30 (50%) patients now dead, given Taxol in year preceding death

Trust/Hospital Pharmacy Data Held

• 30/31 (97%) Pharmacies returned completed questionnaires

• 18/30 (60%) reported that they prescribed taxanes 1998 - 2000– 8 (44%) had paper records only– 3 (17%) had electronic records only– 6 (33%) had paper and electronic records– 1 (6%) had no records of taxane prescribing

Trust/Hospital Pharmacy Data Items (N = 18)

• 87% Patient name• 76% Chemotherapy type• 74% Clinician name• 36% Trials details• 31% NHS No• 28% Patient diagnosis• 20% Chemotherapy line no.• 6% Clinical stage

Hospital Type of SoftwareAiredale General Ascribe program for

asepticsCumberland Infirmary Filemaker Pro (can be

converted to Access)North Tees Ascribe after Oct 99Pinderfields Unix Pharmacy system,

hoping for Access archivesoon

QEH, Gateshead AscribeSouth Tyneside PICKScunthorpe MMTSouth Cleveland Excel (1998 only)Sunderland Royal Excel (after 1998)St James’s MAGIC

Pharmacy Departmental Software

Pharmacy Department Comments (13/18 or 72%)

• missing information can be obtained from clinician involved (28%)

• could cross-reference with other data sources to cope with lack of diagnosis/staging info. but would involve “mammoth task of paper sorting” (22%)

• only one consultant prescribes taxanes for ovarian cancer in our hospital (17%)

Pharmacy Department Comments (continued)

• patients receiving taxanes referred to Cancer Centre (6%)

• “patient numbers very small” (6%)

• “regional database on treatment would be helpful” (6%)

• taxanes only used with small number of private patients (6%)

Summary of Data SourcesData Source Registry Medicare NGOG Clinician’s

OwnPharmacy

DepartmentsCompleteRegionalCoverage

100% 95% 25%* 10%* 97%

ReliableDiagnosisInformation

100% 0% 100% 100%# 28%

Patient-basedData

100% 0% 100% 100% 87%

ReliableTaxolInformation

0% 100% 0% 100% 76%

Reliable LineInformation

0% 0% 0% 72% 20%

Reliable TrialInformation

0% 0% 100% 0% 36%

OVERALLAVERAGE

50% 33% 54% 64% 57%

* estimated# when used in conjunction with Registry database

Further Possibilities• Leeds/Yorkshire clinicians’ data• Medical records

– sampling 2 x 6 months’ of data?• Change inclusion criteria to taxanes prescribed

1998 - 2000 rather than patients diagnosed 1998 - 2000

• Prospective regional data collection of pharmacy data

• MUST be able to identify ALL patients who received taxanes in study sample