| Clark RF et al, 1993, USA | 54 patients who attended with crotalid (rattlesnake) envenomations observed for signs of infection | Prospective observational study | Presence of infection | 3% patients without antibiotics developed an infection, compared with 22% (2/9) who were receiving antibiotics | Follow up involved telephone consultation as well as direct observation |
| Data available for 41 | | | | Very small numbers |
| Kerrigan KR et al, 1997, Ecuador | 114 patients–59 randomly assigned to receive antibiotics | PRCT | Presence of infection as shown by abscess formation | 10.2% of treated patients developed abscesses compared with 5.5% of the untreated p=0.558 | No blinding |
| | | | | Not controlled for other variables, for example, antivenin administration or surgical debridement |
| | | | | Initiation of antibiotic treatment delayed |
| Blaylock RS, 1999, South Africa | 363 patients presenting with snake bites–both venomous and non-venomous (12%) (310 files available) requiring admission to hospital with swelling | Prospective trial | % receiving antiobiotics | 84.8% of patients received no antiobiotics | Not PRCT |
| | | | | ?Comparable groups |
| | | Compared the length of stay in the two groups, (antibiotics v no antibiotics) further subdivided and analysed depending on main symptomatology–swelling, weakness | No difference in length of stay between the groups–this is interpreted as reflecting no difference in infection rate | Extrapolation and interpretation of results is controversial |
| | | | | Very little raw result data provided |