Postgraduate Medical School

1
Postgraduate Medical School

• Exploring GPRD prescribing data in the context of
  NICE guidelines
• A feasibility study on the use of proton pump
  inhibitors
• Corinne de Vries, Helen Seaman, Richard Farmer

Department of Pharmacoepidemiology
2
(No Transcript)
3
Q How long does it take for NICE to develop
its guidance? A On average, a technology
appraisal takes 12-14 months. Clinical guidelines
take around 2 years to develop. It is important
to make sure that all the evidence on a topic has
been considered, and this takes time. Also we
believe that consultation with those who may be
using the guideline is vital for the development
of robust advice, and we therefore need to give
people a reasonable length of time to consider
draft documents.
4
Q Does the medical profession have to follow
NICE's decision? A Although NICE guidance
doesn't override the responsibility of healthcare
professionals to make appropriate decisions based
on the circumstances of individual patients,
healthcare professionals are expected to take our
guidance fully into account when exercising their
clinical judgement.
5
Since 1999,

• 55 Technology appraisals
• 11 Clinical guidelines
• have been completed

6
Can we evaluate prescribing in the context of
NICE?

• Prescribing before introduction
• What are the implications
• Prescribing after introduction
• Changes in prescribing patterns

7
NICE guidelines on PPIs

• Documented gastric / duodenal ulcer
• H pylori negative or still symptomatic
• NSAID-induced ulcer taking NSAIDs
• Severe gastro-oesophageal reflux
• E.g. oesophageal ulceration, Barretts oesophagus
• NOT in non-ulcer dyspepsia
• Titrate down
• Least expensive

8
GPRD data
registration
medical
patient
prevention
therapy
9
Patient data

• DOB, sex, family number
• morbidity - hospital, GP diagnosis
• symptoms
• cause date of death
• childbirth
• prescribing
• preventive care e.g. blood pressure, height
  weight, smoking, alcohol

10
Medical records

• Diagnoses / symptoms coded through OXMIS READ
• Event dates
• Hospital referrals outcomes
• Comments

11
Prescription details

• Prescribing date
• Drug prescribed
• Number of days (??)
• Amount dispensed
• number of packs / number of units
• Prescribed dose
• ? duration can be calculated

12
Size of the GPRD

• 700 practices with history
• 360 current practices
• England Wales
• 8 million patients
• 3 million current patients
• over 35 million patient years
• average 8 years longitudinal data

13
Study design

• 1992 1999
• gt6 months up to standard data
• Amongst individuals prescribed PPI
• NSAID users
• GI morbidity
• Age
• Sex
• Which products?

14
Study findings

• 233,759 people received a PPI script
• 109,667 males 46.9
• 124,092 females 53.1
• 8.1 scripts per user
• 86,392 (40.0) lt2 scripts
• 180,134 (77.1) lt10 scripts
• 28,137 (12.0) gt20 scripts

15
Age distribution of people prescribed PPIs
16
Number of scripts for PPIs per calendar year
17
GI ulcer

• Men 26,992 (24.6)
• Women 17,577 (14.2)
• Total 44,569 (19.1)

18
Reflux

• Men 1,429 (1.3)
• Women 989 (0.8)
• Total 2,418 (1.0)

19
NSAID use gt3 scripts

• Men 25,028 (22.8)
• Women 43,351 (34.9)

20
Dyspepsia, no GI ulcer lt4 NSAID scripts

• Men 25,559 (23.3)
• Women 26,563 (21.4)
• Total 52,122 (22.3)

21
No record of GI morbidity

• Men 6,804 (6.2)
• Women 5,234 (4.2)
• Total 12,038 (5.1)

22
Summary

• Omaprazole prescribing decreased
• Lansoprazole prescribing increased
• Not all had supporting evidence for PPI use
• Under recording?
• Not according to guidelines?
• GPRD can be used for NICE purposes

23
Acknowledgement

• Funding from National Centre for Health Outcomes
  Development (NCHOD)