Agarwal R, Srinivas R, Aggarwal AN, Gupta D
Correspondence: Dr Ritesh Agarwal, riteshpgi@gmail.com
ABSTRACT
Introduction Immunosuppressive therapy has been shown to improve outcomes in patients with paraquat poisoning. The objective of this study was to evaluate the efficacy of immunosuppressive therapy in the management of lung injury due to paraquat poisoning.
Methods We searched the MEDLINE, OVID, and CINAHL databases for relevant studies published from 1980 to 2006. We included studies if (a) the study design was a randomised controlled trial, observational study with historical controls or observational study; (b) the study population included patients with paraquat poisoning, and received immunosuppressive therapy; and (c) the study provided data on mortality. We calculated the survival rate with 95 percent confidence intervals (CI) for observational studies, and relative risk and 95 percent CI for dichotomous outcomes.
Results 12 studies--four non-randomised, six non-randomised comparing historical controls, and two randomised controlled trials--had employed immunosuppressive therapy in the management of paraquat poisoning. The survival rate in the four non-randomised studies (39 patients) was 74.4 percent (95 percent CI 58.9-85.4). The relative risk of immunosuppressive therapy in decreasing mortality with paraquat poisoning was 0.55 (95 percent CI 0.39-0.77) and 0.6 (95 percent CI 0.27-1.34) for the non-randomised studies (comparing historical controls) and randomised controlled studies, respectively. There was significant heterogeneity and evidence of publication bias.
Conclusion One out of four patients (95 percent CI 3-5) were successfully treated with immunosuppressive therapy for paraquat poisoning. However, due to significant heterogeneity and publication bias, a large randomised controlled trial will be required to affirm the role of immunosuppression in paraquat poisoning.
Keywords: cyclophosphamide, gluco-corticoids, immunosuppressive therapy, lung injury, paraquat poisoning
Singapore Med J 2007; 48(11): 1000–1005
