Table 1

 Crude and adjusted relative risks (RR) obtained from Cox regression analyses predicting failure of new users to persist with statin treatment after two years.

Crude RR95% CIAdjusted* RR95% CI
*Adjusted for all variables listed.
†Significantly different from reference category; RR >1 indicates that the variable increases failure to persist on statin treatment; RR <1 indicates that the variable decreases failure to persist on statin treatment.
‡Chronic disease score: 0–1, no or minimal co-morbidity; 2–4, low and average co-morbidity; >4, high co-morbidity.
Characteristics at index date
    Female0.980.93 to to 1.07
Age group (year)
    45–640.830.76 to 0.920.84†0.76 to 0.93
    ⩾650.810.73 to 0.890.82†0.74 to 0.91
Type of statin
    Atorvastatin0.900.84 to 0.970.90†0.83 to 0.96
    Fluvastatin1.231.13 to 1.341.27†1.12 to 1.39
    Pravastatin0.980.91 to 1.060.990.91 to 1.08
Characteristics 1 year before index date
Antihypertensive treatment
    Yes0.870.82 to 0.930.91†0.85 to 0.98
Psychotropic agents
    Yes1.111.04 to 1.181.13†1.06 to 1.20
Antidiabetic drugs
    Yes1.020.95 to to 1.14
‡Chronic disease score
    2–40.920.86 to 0.990.990.91 to 1.07
    >40.890.83 to 0.960.960.88 to 1.05
Hospitalisation for ischaemic heart disease in 2 years before index date
    Yes0.760.68 to 0.850.78†0.70 to 0.88