The Role of Selective Digestive Tract Decontamination on Mortality and Respiratory Tract Infections: Results
When all studies are combined, 2,270 patients were available for analysis. There were 1,165 control patients (51.3 percent) and 1,105 patients treated with SDD (48.7 percent). Three hundred five of the 1,165 control patients died yielding a mortality rate of 0.262. This was not significantly different (p=0.291) from the mortality rate for the SDD-treated patients (0.243, representing 268 deaths out of 1,105 patients). These mortality rates yielded a risk difference of 0.019 (95 percent confidence interval of —0.016 to 0.054). Figure 1 shows mortality risk differences along with their corresponding 95 percent confidence intervals for the individual studies as well as for the cumulative study population. To assess the observed power for this patient sample size, a x2 approximation was used with an alpha error of 0.05, yielding a beta error of 0.16 (power=0.84).
Mortality related to acquired nosocomial infection was examined for the seven studies reporting these data.”’ Forty-two of the 415 control patients died as a result of acquired infection yielding a mortality rate of 0.101. This was significantly different (p=0.007) from the acquired infection mortality rate for SDD-treated patients (0.050, representing 18 deaths out of 363 patients). This mortality rate yielded a risk difference of 0.051 (95 percent confidence interval of 0.015 to 0.089) in detail http://antimicrobialmed.com. This difference was highly dependent on the results of one study which when removed from the analysis resulted in a nonsignificant difference between control and SDD groups for mortality related to acquired infection (p=0.074). Fourteen of the analyzed studies (Table 1) included information on the number of individual patients developing nosocomial pneumonia. Two of the studies reported only total respiratory infections for both patient groups so that individual group infection rates could not be determined. A total of 2,128 patients were available for this part of the analysis. There were 1,097 patients in the control group (51.6 percent), of whom 240 developed at least one episode of nosocomial pneumonia (rate=0.219). The SDD-treated group consisted of 1,031 patients (48.4 percent), of whom 76 developed at least one episode of nosocomial pneumonia (rate=0.074, p<0.0001 compared with the control group). The acquired pneumonia risk difference (Pc—Pt) was 0.145 (95 percent confidence interval of 0.116 to 0.174). The risk differences for nosocomial pneumonia in the individual studies and the cumulative population, along with the corresponding 95 percent intervals, are shown in Figure 2.
Figure 1. Mortality risk differences (Pc—Pt) and 95 percent confidence intervals for individual studies and the cumulative study population (lowermost point). Study risk differences follow the order presented in Table 1 starting with reference 29 at the top.
Figure 2. Acquired pneumonia risk differences and 95 percent confidence intervals for individual studies and the cumulative study population (lowermost point). Study risk differences follow the order presented in Table 1 starting with reference 29 at the top.
Category: Respiratory Symptoms
Tags: acquired infections, gram-positive bacteria, pneumonia, respiratory infection, tracheobronchitis