Unadjusted postoperative outcomes following total hip arthroplasty comparing current and former smokers to nonsmokers (reference category for both groups).
| Current Smoker | Prior Smoker | |||||
|---|---|---|---|---|---|---|
| Outcome | No. Studies, No. Patients | Relative Risk (95% CI) | NNH (95% CI) | No. Studies, No. Patients | Relative Risk (95% CI) | NNH (95% CI) |
| Reoperation or revision | 4 studiesa, 4000 pts | 1.4 [0.62, 2.10] | ∞ (NE) | 3 studiesf, 4180 pts | 1.17 [0.84, 1.63] | 100 (NE) |
| Implant loosening | 2 studiesb, 238 pts | 1.25 [0.97, 1.62] | 34 (NE) | 1 studyg, 200 pts | 1.16 [0.83, 1.63] | 50 (NE) |
| Any postoperative complication | 2 studiesc, 2512 pts | 1.24 [1.01, 1.54] | 50 (20, ∞) | 1 studyh, 2221 pts | 1.32 [1.05, 1.66] | 34 (17, 100) |
| Deep infections | 1 studyd, 1185 pts | 3.42[0.69, 16.85] | 100 (NE) | 1 studyl, 1499 pts | 3.68 [0.95, 14.16] | 100 (NE) |
| Death | 1 studye, 1539 pts | 1.63 [1.06, 2.51] | 34 (20, 100) | 1 Studyl, 1400 pts | 1.69 [1.08, 2.64] | 34 (20, 100) |
NNH: number needed to harm, calculated as = 1/absolute risk reduction; NE: confidence interval not estimable since 95% CI includes both positive and negative numbers.
↵a Data from 4 studies: Espehaug49, Sadr Azodi50, Khan62, Meldrum51; Espehaug59 and Sadr Azodi50 were multivariable-adjusted; Khan62, and Meldrum50 were univariate analyses.
↵b Data from 2 studies: Malik63 and Nixon61 both univariate studies.
↵c Data from 2 studies: Sadr Azodi50 (multivariable-adjusted), Meldrum50 (univariate).
↵d Data from 1 study: Khan62 (univariate).
↵e Data from 1 study: Sadr Azodi50 (multivariable-adjusted).
↵f Data from 3 studies for reoperation: Espehaug49, Sadr Azodi50, Khan62, Espehaug49 and Sadr Azodi50 were multivariable-adjusted; Khan62 included univariate analysis.
↵g Data from 1 study: Malik63 (univariate).
↵h Data from 1 study: Sadr Azodi50 (multivariable-adjusted).
i Data from 1 study: Khan62 (univariate).
j Data from 1 study: Sadr Azodi50 (multivariable-adjusted).