Karen Smith
Acknowledgement
Ŷt = β0 + β1 x timet + β2 x interventiont + β3 x
time_after_interventiont + et
Yt is the outcome
time indicates the number of quarters from the start of the series
intervention is a dummy variable taking the values 0 in the pre-
intervention segment and 1 in the post-intervention segment
time_after_intervention is 0 in the pre-intervention segment and
counts the quarters in the post-intervention segment at time t
β0 estimates the base level of the outcome at the beginning of the
series
β1 estimates the base trend, i.e. the change in outcome per quarter
in the pre-intervention segment
β2 estimates the change in level in the post-intervention segment
β3 estimates the change in trend in the post-intervention segment
et estimates the error
Threats to Validity
6000
Co-proxam ol w ithdraw al
announced
5000
Prescription items dispensed per quarter (thousands)
co-proxam
4000 NSAIDs
paracetam
co-codam
tramadol
3000
co-dydram
codeine
dihydroco
2000
1000
0
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Year (quarters)
90
80 Co-proxamol withdrawal
announced
70 co-proxa
60
Number of deaths
other
analges
50
co-proxa
best fit w
40 announc
co-proxa
30 best fit w
announc
20
10
0
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Year (quarters)
Prescriptions
Pre-intervention Post-intervention
Co-proxamol 3050.1 (139.9) <0.001 -45.9 (7.7) <0.001 -554.8 (74.9) <0.001 -46.8 (16.8) 0.01
Cocodamol 1349 (12.4) <0.001 34.1 (0.8) <0.001 300.5 (53.6) <0.001 30.7 (6.4) <0.001
Codeine 204.8 (4.2) 0.007 9.6 (0.2) <0.001 20.5 (11.6) 0.089 3.5 (1.2) 0.007
Codrydamol 1055.2 (6.1) <0.001 -1.1 (0.3) 0.004 148.7 (35.8) <0.001 -4.2 (4.1) 0.316
Dihydrocodeine 686.4 (31.1) <0.001 -1.5 (1.4) 0.291 -29.7 (2.4) <0.001 -0.8 (2.4) 0.731
NSAIDs 4652.4 (47.2) <0.001 28.4 (3) <0.001 -622.9 (70.2) <0.001 -66.2 (8.5) <0.001
Paracetamol 1493.4 (56.1) <0.001 42.1 (3.0) <0.001 232 (66.6) 0.001 23 (8.2) 0.01
Tramadol 47.2 (51.4) 0.365 31.4 (2.7) <0.001 -41.9 (6.8) <0.001 16.3 (5.2) 0.004
Suicides
Pre-intervention Post-intervention
Co-proxamol 81.0 (4.5) <0.001 -1.194 (0.3) <0.001 -28.3 (4.9) <0.001 0.6 (0.6) 0.355
Other analgesics 51.3 (2.8) <0.001 -0.3 (0.2) 0.095 6.4 (6.0) 0.297 -0.3 (0.6) 0.724
All drugs except co- 221.2 (7.3) <0.001 -0.5 (0.5) 0.299 21.6 (12.8) 0.100 -5.4 (1.4) <0.001
proxamol and other
analgesics
All drugs 353.7 (10.2) <0.001 -2.0 (0.7) 0.008 0.004 (18.1) 1.000 -4.9 (1.7) 0.007
All causes 1319.0 (22.5) <0.001 -4.8 (1.4) 0.002 12.8 (34.8) 0.716 -5.4 (4.1) 0.192
Estimating Absolute Effect
• The model may be used to estimate the absolute effect of the
intervention. This is the difference between the estimated
outcome at a certain time after the intervention and the
outcome at that time if the intervention not taken place.
• For example, to estimate the effect of the intervention at the
midpoint of the post-intervention period (when time = 34.5
and time_after_intervention = 6.5), we have
Mean quarterly estimated Mean quarterly number post Mean quarterly change (95% CI)
number pre announcement announcement
Prescriptions (x1000)
Suicides, Open