Skip to main content
. 2017 Apr 7;46(6):2067–2077. doi: 10.1093/ije/dyx014

Table 2.

Association of proposed instrument and outcomes of other patients who saw the GP on the same day as they issued a smoking cessation therapy to an index patient (n = 101861)

Robust linear regression
Risk difference*100 95% confidence interval
Lower Upper
Male 0.77 0.02 1.52
Age (years)a −0.35 −0.65 −0.05
Number of GP visits in previous yeara −0.35 −0.61 −0.08
Diagnoses in the previous year
    Autism 0.00 −0.02 0.02
    Bipolar −0.01 −0.05 0.03
    Current smoker 0.28 −0.51 1.07
    Dementia 0.00 −0.12 0.13
    Depression −0.04 −0.42 0.35
    Eating disorder −0.02 −0.05 0.02
    Hyperkinetic disorder 0.00 −0.01 0.01
    Learning disability −0.01 −0.06 0.05
    Neurotic disorder 0.56 0.29 0.84
    Other behavioural disorder −0.01 −0.04 0.01
    Personality disorder 0.02 −0.03 0.07
    Schizophrenia 0.00 −0.05 0.05
    Alcohol misuse 0.08 −0.06 0.22
    Probable self-harm −0.01 −0.01 0.00
    Drug misuse 0.03 −0.04 0.11
    Fractures 0.03 −0.15 0.21
    Any psychiatric illness 0.36 −0.10 0.81
    Chronic disease 0.12 −0.34 0.59
Prescriptions in the previous year
    Antidepressant 0.44 −0.28 1.17
    Antipsychotic 0.17 −0.18 0.52
    CNS stimulant 0.02 −0.02 0.05
    Dementia medication 0.00 −0.12 0.13
    Hypnotic anxiolytic −0.02 −0.40 0.35
    Lithium 0.07 −0.01 0.16

Robust standard errors clustered by physician reported.

*Mean differences reported. Each outcome was defined as an event in the year after the index prescription.