Sequential drug decision problems in long-term medical conditions: a case Study of Primary Hypertension Eunju Kim ba, ma, msc

5.6.2Age-dependent systolic blood pressure and blood pressure lowering effect

Age and gender-related changes in SBP were estimated using the regression coefficients developed by Wills et al^[309]. Wills et al estimated the life course trajectories of SBP from childhood to elderly by fitting the longitudinal data from eight UK population based cohorts to a non-linear function. Different regression coefficients were estimated for different age groups. The hypertension SDDP model used the regression coefficients in Equation 5.4, which covers the age range between 55 and 77.
137.4+(1.54*Age)+(-0.046*Age^2) for men, where the intercept age is 55.

136.7+(1.482*Age)+(-0.348*Age^2) for women, where the intercept age is 55. Equation 5.4.

The hypertension SDDP model is based on the assumption that the difference in SBP lowering effect mainly contributes the treatment benefit^{[310, 311]}. However, there are different perspectives to the blood pressure lowering effect between antihypertensive drugs. Some systematic reviews demonstrated that the blood pressure lowering effects of antihypertensive drugs are not significantly different to each other^{[247, 248]}. In contrast, some evidence shows that certain drugs have a significantly better blood pressure lowering effect than others^{[312, 313]}. In order to populate the clinical decision rule based on SBP in SDDP modelling, the hypertension SDDP model accepted the latter even though the difference in the blood pressure lowering effect is small.

Four systematic reviews that conducted a meta-analysis to compare the SBP lowering effect of five major antihypertensive drugs were initially identified (see Table ‎5.)^{[282, 313-315]}. The SBP lowering effect in the hypertension SDDP model was based on Wright et al’s systematic review, which provides the weighted mean difference and 95% confidence interval (CI) between active drug arm and placebo arm at one year or the earliest time after one year^[313], and Wald et al’s systematic review, which presents mean placebo-subtracted SBP reduction (and 95% CI) over a period of 4-12 weeks^[315]. Law et al’s review was excluded because of the short time period, which was 24 hours^[282]. Baguet et al’s review was also excluded because the final outcome of meta-analysis was not controlled by placebo^[314].

Table ‎5.. Comparison between systematic reviews on SBP lowering effect of antihypertensive drugs

		Law et al, 2003^[282]	Baguet et al, 2007^[314]	Wald et al, 2009^[315]	Wright et al, 2009^[313]
No of studies included		354 trials (56,000 participants) between 1966 and 2000.	80 trials (10,818 patients) between 1973 and 2007.	42 trials (10,968 participants) between 1966 and 2008.	24 trials (58,040 participants) between 1966 and 2008.
Types of included studies		Randomised, placebo-controlled trials followed up more than 2 weeks (Crossover studies were included).	Randomised, double-blinded trials (Crossover studies were excluded).	Randomised-controlled trials (Crossover studies were included).	Randomised-controlled trials followed up at least one year.