Table 2.
Cox models were used to estimate the effect of different anti-hypertensive therapies on infectious, inflammatory and structural outcomes
| Infectious (N = 21,523) | Inflammatory (N = 50,449) | Structural (N = 42,615) | |
|---|---|---|---|
| ACE-I vs Control | |||
| Model 1 (Initial Drug) | |||
| HR | 0.886 | 0.924 | 0.865 | 
| 95%CI | 0.859–0.886 | 0.906–0.942 | 0.847–0.885 | 
| p | <0.0001 | <0.0001 | <0.0001 | 
| NNT | 240 | 156 | 104 | 
| Model 2 (Drug duration) | |||
| Est | −0.0041 | −0.0034 | −0.031 | 
| p | <0.0001 | <0.0001 | <0.0001 | 
| ARB vs Control | |||
| Model 1 (Initial Drug) | |||
| HR | 0.957 | 0.970 | 0.900 | 
| 95%CI | 0.914–1.009 | 0.940–1.001 | 0.868–0.933 | 
| p | 0.1067 | 0.0568 | <0.0001 | 
| NNT | 637 | 395 | 140 | 
| Model 2 (Drug duration) | |||
| Est | −0.0050 | −0.0033 | −0.036 | 
| p | <0.0001 | <0.0001 | <0.0001 | 
Model 1 uses the first drug prescribed and follows the incidence of diagnosis up until the patient stops taking the medication. Model 2 takes into account the duration of the treatment in order to compare to emergence of disease. Diabetic-status, Flu Vaccine and Statin use are corrected for in each analysis comparing anti-hypertensive therapy. The number needed to treat (NNT) was also calculated for ACE-I and ARB use for infectious, inflammatory or structural outcomes