Clinical Epidemiology Study Who benefits? How do we know? B Lynn Beattie MD FRCPC Professor Emeritus Div Geri Med, Dept Med, UBC Medical Director UBCH.

Clinical Epidemiology Study Who benefits? How do we

know?

B Lynn Beattie MD FRCPCProfessor Emeritus

Div Geri Med, Dept Med, UBCMedical Director UBCH CARD

Low-income subset of the cohort from the Utilization and Cost Study

Policy begins

Low-income people*

ChEI use

No use: Historical Controls*On MSP premium subsidy

70% increase among poor

a) Impact on Low-income Cohort (n = 24,253)

…vs 30% increase in all BC

Rate of contacts with physicians: Low-income cohort vs all of BC

Low Income:Policy cohortHistorical control

All BC:Policy cohortHistorical control

No impact

BC linked data for 2006-2009

Rate of entry to long-term or palliative care:Low-income vs all of BC

Low Income:Policy cohortHistorical control

No impact

BC linked data for 2006-2009

Rate of hospitalizations: Low-income cohort vs all of BC

Low Income:Policy cohortHistorical control

No impact

BC linked data for 2006-2009

b) Special Authority data: Changes in Clinical Measures

New users of ChEI (naïve)

Continuing users (non-naïve)

Problem: No SMMSE data collected before policy.

Initial Special Authority Form

Renewal Special Authority Form –Overall Patient Assessment Rating

OPAR

Change in SMMSE scores over 6 mo

SMMSE Change

First SMMSE Low scorers, who score below 10 at 6 months, do not submit SA forms

High scorers cannot score much higher

Naïve: new users of ChEIs (n = 1094)

Middle of graph is relatively free of bias

Change in SMMSE scores over 6 mo

SMMSE Change

First SMMSE Low scorers, who score below 10 at 6 months, do not submit SA forms

High scorers cannot score much higher

Non-naïve: Continuing users (n = 1584)

Middle of graph is relatively free of bias

Changes in SMMSE scores

naïve

continuingSMMSE Change

+ 0.28

- 0.22

Difference: + 0.5(95% CI: 0.3-0.7)

SMMSE Change

Naïve improved by half a point more than Non-Naive.

Change in GDS by first SMMSE

naïve

continuing

Difference:0.05 (95% CI: 0.0-0.09)

GDS Change

+ 0.03

- 0.03

GDS Change

Naïve achieved 5% of a GDS point more than Non-Naïve

OPAR compared by first SMMSE

naïve

continuing

Difference: 0.23 (95% CI: 0.15-0.33)

OPAR

0.49

0.26

OPAR

Naive achieved a quarter of a point more on OPAR than Non-Naive

ConclusionsIn real-world usage of ChEIs, there is evidence of clinical improvement, based on measurements by frontline physicians.

This is consistent with the pivotal trials, most of which were 6 months RCTs.

Value of OPAR will be looked at further.

We look forward to the evidence on longer-term effects of ChEIs.