The H2S water screening test and the membrane filtration faecal coliform count were compared with Escherichia coli counts for water samples collected from household water sources and domestic drinking water in rural Malaysia. Water samples were taken from 151 wells, 44 taps supplying water from the treated municipal supply and 192 domestic stored water supplies. E. coli were detected in 20% of the samples (42% of wells, 7% of tap water and 6% of drinking water). Excellent correlation (Spearman's rank correlation rs = 0.93) was found between the faecal coliform and E. coli counts for all sample types. The H2S method was poorly correlated whether read at 18 or 30 h. False positive rates were highest for well water, and false negative rates were highest for both well and drinking water samples, with low E. coli counts. The faecal coliform test was an excellent predictor of the presence of E. coli in these water samples, while the H2S test was very inadequate.
Since thrombi continue to incorporate fibrin during lysis we tested the effect of pretreatment with ancrod, a defibrinating agent from Malaysian pit viper venom, on thrombolysis with urokinase and streptokinase. Thrombi were induced by copper-coils in the carotid arteries of the dogs, weighed after 1 hour and inserted into the femoral arteries of the same animals. They were then exposed for 15 min to iv boluses of streptokinase 10,000 U/kg, urokinase 10,000 U/kg and urokinase 25,000 U/kg with or without pretreatment with ancrod. Ancrod depleted fibrinogen within 5 min and enhanced the lytic effect of streptokinase from 25 +/- 8% to 59 +/- 13% (p less than .05), urokinase 10,000 U/kg from 16 +/- 11% to 66 +/- 18% (p less than .01) and urokinase 25,000 U/kg from 27 +/- 17% to 85 +/- 8% (p less than .001) of the initial thrombus weight. Ancrod itself did not activate plasminogen to plasmin. We conclude that ancrod enhances thrombolysis probably by depleting fibrinogen and preventing new fibrin incorporation into the thrombus during lysis.