Table 2. Liver cancer risk due to use of potentially NDMA-contaminated valsartan drug products compared with uncontaminated valsartan.
| Hazard ratio [95% CI]*1 | Sample size/ Cancer cases | |
| Exposure to NDMA-contaminated valsartan | ||
| No exposure | 1.00 (ref) | 354 628/444 |
| Exposure | 1.16 [1.03; 1.31] | 385 167/736 |
| Exposure in dose categories | ||
| 0 to ≤ 90 DDD | 1.15 [0.98; 1.34] | 122 479/244 |
| > 90 to ≤ 170 DDD | 1.19 [1.02; 1.40] | 136 734/248 |
| > 170 DDD | 1.13 [0.97; 1.33] | 125 954/244 |
| NDMA exposure | ||
| Possible (contaminated valsartan batches < 75%) | 1.18 [1.01; 1.39] | 104 433/232 |
| Probable (contaminated valsartan batches ≥ 75%) | 1.15 [1.01; 1.31] | 280 734/504 |
| Long-term valsartan use*2 | ||
| No exposure | 1.00 (ref) | 61 236/102 |
| Exposure | 1.22 [0.80; 1.89] | 13 876/28 |
| Incidence rate per 100 000 person-years*3 | ||
| No exposure | 34.61 | |
| Exposure | 39.08 | |
*1 Lag time 1 year, fully adjusted for sex; age; polypharmacy (defined as prescription of five or more different drugs); prescription of low-dose acetylsalicylic acid (ASA), non-ASA non-steroidal anti-inflammatory drugs, 5α-reductase inhibitors, statins, spironolactone, glucocorticoids for systemic use, selective serotonin reuptake inhibitors, and hormone replacement therapy; the comorbidities diabetes, chronic obstructive pulmonary disease, congestive heart failure, and alcohol-related diseases; the Charlson comorbidity index (score); and prevalent valsartan use
*2 Long-term valsartan use is defined as valsartan prescription in at least nine quarters within the first 3 years of the study period.
*3 Standardized to the German population over 40 years in 2011 (e2)
DDD, Defined daily dose; NDMA, N-nitrosodimethylamine; 95% CI, 95% confidence interval