Transcript Oral Presentation 1

Measuring improvements
in adherence
Jennie Connor
University of Auckland, New Zealand
Do fixed dose combination pills or unit-ofuse packaging improve adherence?
 Systematic review of randomised trials
 Effect on adherence and clinical outcomes
 Only a small number of trials found, most
poor quality
 Some benefit but evidence is weak
 What is a significant improvement?
 Heterogenous adherence measures
 How valid?
 How reliable?
 How useful in usual practice?
Methods for measuring adherence
Direct methods
Indirect methods
(look for evidence of the drug
in the body)
Pill counting
Medication event
monitoring (MEMS)
Prescribing/dispensing
records
Self-report
Diary
Adherence questionnaire
Appointment keeping
Therapeutic response
Test for drug, metabolite
or tracer in:
urine
blood
saliva
How valid?
= is it measuring what you intend it to?
 What are you trying to measure?
 Are they getting effective treatment? or
 Are they following the regimen (dose and
timing)?
 No ‘gold standard’ with which to compare
 Direct methods
 only measure recent ingestion of drug
 Pill counts and MEMS
 Show medications/caps have been removed
from the container
 Pill counts over-estimate adherence
 MEMS provides timing information, no doses
 Prescriptions/dispensing
 Little validity except discontinuation
 Greatly over-estimate adherence
 Self-report
 Response varies greatly with context,
relationship, and nature of questioning
 Social desirability and recency effects
 Generally over-estimates adherence. Good
specificity for nonadherence
 Formalised in questionnaires – can include
adherence-related behaviours,
barriers….e.g.PMAQ. Not for repeated use.
 Medication diaries – more reliable information
than recall. Details of timing.
 Appointment keeping: raises index of suspicion
 Therapeutic response: weak indicator
How reliable?
Hard to know without a good reference
standard
 Accuracy of self report varies between
settings
 PMAQ and other formal instruments are
designed and tested for reliability,
informal interview is not.
 Drug testing may vary between
individuals, or over time
How useful in practice?
 All methods have limitations: may need to
use more than one
 Consider the patient and clinician burden and
cost
 Self-report compares well with other measures
in many studies and is most readily available
 Need to think about a definition of clinically
significant non-adherence
 To distinguish those at high risk of treatment
failure, rather than e.g. “80% rule”
Improving adherence measurement
 In research
 Defining acceptable adherence: is there a
known threshold for effective treatment that
needs to be reached?
 Validation studies of measures – for this
condition, this population
 Multiple complementary measures and
composite measures - standardisation
 Distinguishing patterns of non-adherence –
erratic, unwitting, intelligent.
 What is a significant improvement in adherence
for this drug or condition?
Improving adherence measurement
 In practice
 Will usually need to use simple and cheap
methods
 Partnership with patient to improve adherence
and its measurement by self-report
 Assessment of adherence as part of routine of
care – a continuous process
 Context-specific combination of measures : little
place for routine pill counting: identifying
patterns of non-adherence as well as extent
 Consider poor attenders and poor responders
as higher risk