Table 2

Input parameters for Markov model

Expected (mean) valueProbability distribution around the mean
Cohort size50 000-
Patient characteristics
Gender: male82%β (516, 116)
Age
 Male58Normal (58, 0.40)
 Female57Normal (57, 0.88)
Quality of life
History of CHD0.74β (18 479, 6469)
MI0.67β (144, 71)
History of MI0.74β (18 479, 6469)
Stroke0.62β (159, 101)
History of stroke0.68β (115, 71)
Costs
Primary care costs*2,095γ (147, 14)
Cost per MI*11 877γ (400, 30)
Cost per stroke*18 801γ (400, 47)
Intervention effects
Change in SBP (mm Hg)†
 Male−7.7Normal (−7.7, 1.3)
 Female−8.2Normal (−8.1, 2.5)
Change in LDL-C (mmol/L)†
 Male−0.1Normal (−0.1, 0.1)
 Female−0.1Normal (−0.1, 0.1)
Change in proportion of cohort smoking‡
 Male−15%Normal (−15%, 0.04)
 Female−26%Normal (−26%, 0.07)
Intervention cost
 Database/IT infrastructure75 000
 Project manager130 000
 Clinical advisor26 000
 Patient enrolment846 354
 Messages480 000
 Additional (miscellaneous expenses)100 000
 Total cost1 657 354U (1 243 016, 2 071 693)
  • *For 55–64 male, for full primary care and hospital costs, see Appendix.

  • †Analysis of covariance was used and included randomised groups (usual care and TEXT ME) and baseline values.

  • ‡Test of proportions. For probability distributions around the mean, the first and second values in parentheses correspond to mean and SE in normal distribution and alpha and beta in beta (β), gamma (γ) and uniform (U) distributions.

  • CHD, coronary heart disease; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; SBP, systolic blood pressure; TEXT ME, Tobacco, Exercise and Diet Messages.