Table 1. Association between demographic factors and VL relapse.
| Risk factor | Cured cohort | Relapse cohort | Unadjusted odds ratio (95% CI) | p |
| Sex (n = 8537) | 8418 | 119 | ||
| Male | 4745 (56.4) | 83 (69.7) | 1.8 (1.2–2.6) | 0.003 |
| Female | 3673 (43.6) | 36 (30.3) | - | - |
| Castea (n = 8486) | 8368 | 118 | ||
| Scheduled caste | 2453 (29.3) | 32 (27.1) | 1.4 (0.7–2.7) | 0.32 |
| Other backward class | 4624 (55.3) | 74 (62.7) | 1.7 (0.9–3.2) | 0.08 |
| General Category | 1291 (15.4) | 12 (10.2) | - | - |
| Living in MSF-supported blocks (n = 8537) | 8418 | 119 | ||
| No | 4297 (51.0) | 56 (47.1) | 0.9 (0.6–1.2) | 0.39 |
| Yes | 4121 (49.0) | 63 (52.9) | - | - |
| Previous relapse (n = 8537) | 8418 | 119 | ||
| Multiple | 18 (0.2) | 1 (0.8) | 3.9 (0.1–25.4) | 0.23c |
| Single | 299 (3.6) | 4 (3.4) | 0.95 (0.3–2.5) | 1.00c |
| No | 8101 (96.2) | 114 (95.8) | - | - |
| Age, years (n = 8537) | 8418 | 119 | ||
| <5 | 574 (6.8) | 17 (14.3) | 3.6 (1.8–7.1) | <0.001 |
| 5 to <15 | 3229 (38.4) | 43 (36.1) | 1.6 (0.9–2.9) | 0.10 |
| 15 to <30 | 1935 (23.0) | 16 (13.4) | - | - |
| 30 to <45 | 1490 (17.7) | 21 (17.6) | 1.7 (0.9–3.3) | 0.11 |
| ≥45 | 1190 (14.1) | 22 (18.5) | 2.2 (1.2–4.3) | 0.013 |
| Treatment location (n = 8537) | 8418 | 119 | ||
| Treatment camps (ambulatory) | 162 (1.9) | 0 | NA | - |
| Primary health center (ambulatory) | 1372 (16.3) | 13 (10.9) | 0.6 (0.3–1.1) | 0.10 |
| Hospital (inpatient) | 6884 (81.8) | 106 (89.1) | - | - |
| Season of treatment (n = 8537) | 8418 | 119 | ||
| March–June | 3566 (42.4) | 53 (44.5) | 1.0 (0.7–1.6) | 0.84 |
| July–October | 2527 (30.0) | 33 (27.7) | 0.9 (0.6–1.5) | 0.74 |
| November–February | 2325 (27.6) | 33 (27.7) | - | - |
| Year of treatmenta , b (n = 7471) | 7359 | 112 | ||
| 2011 | 1692 (23.0) | 31 (27.7) | 1.4 (0.8–2.2) | 0.22 |
| 2010 | 1614 (21.9) | 25 (22.3) | 1.1 (0.7–1.9) | 0.61 |
| 2009 | 1468 (19.9) | 21 (18.8) | 1.1 (0.6–1.8) | 0.84 |
| 2007–2008 | 2585 (35.1) | 35 (31.3) | - | - |
a The remaining data were missing or incorrectly coded in the database.
b The data from years 2007 and 2008 were combined, as operations began mid-2007. Patients treated in 2012 were excluded, as there would be limited time for relapse to present.
c Fisher's Exact test.