Search

A stepwise multiple logistic regression was computed to assess which of the nutritional variables differentiate the healthy and diseased group of participants in the Forsyth Specialized Caries Center (n = 275). Variables considered as candidates for the model included the consumption per week of sugars, starch, cheese, fruits and fruit juices, noncariogenic foods, and dairy products. Two variables--sugars and cheese--were statistically significant by the stepwise procedure. Increased intake of sugar was associated with being in the root caries group, whereas high intake of cheese was negatively associated with root caries. Thus, cheese seems to have a protective effect after sugar intake is controlled for. Odds ratios were computed to quantify the influence of the variables. An increase of two exposures of sugar per day corresponded with an odds ratio of 1.26. The odds ratio continues to increase with sugar intake so that an increase to five exposures per day gives an odds ratio of 1.79. This model is consistent with past in vitro studies suggesting that cheese protects against caries formation when sugar intake is controlled for and that increasing frequency of sugar intake increases the odds of root caries.

The possible associations betweenperiodontitis subject age and the distribution of Actinobacillus actinomycetemcomitansand Porphyromonas gingivaliswere examined using an extensive data bank of subgingival plaque specimens analyzed using DNA probes. The results suggest that A. actinomycetemcomitansis strongly related to subjects in the youngest age group (10 to 19 years) with decreasing prevalence and concentration levels in older age groups. In contrast, P. gingivalisshowed a reverse relationship to subject age with highest prevalence and concentration levels in older subjects (30 years and older) compared to subjects in younger age groups. Statistical testing of these relationships were highly significant. (P< 0.0001). J Periodontol 1991; 62:490–494.

Prevalence of bacterial resistance to antibiotics was found to be substantially different in two separate regions of the world. The average percent resistant to individual antibiotics was nearly three times greater, and the percent of isolates resistant to six or more antibiotics 14 times greater among isolates at the Hôpital St Joseph in Paris than among those at the Peter Bent Brigham Hospital in Boston or at any of six US hospitals. Differences were not due to culture sampling or susceptibility testing methods used in the hospitals. Differences were nearly as great between isolates from patients recently admitted to the two hospitals, suggesting differences in the bacterial flora of their communities. Similar systematic comparisons of resistance prevalence in different parts of the world might help to define optimal antibiotic usage practices.(JAMA 239:1518-1523, 1978)

AbstractThe antibiotic resistance of Klebsiella pneumoniaeisolates from 12 medical centers worldwide, over a 1- to 6-year period, were tested. Clinical isolates of K. pneumoniaewere resistant to ampicillin and carbenicillin. Resistance to other antibiotics was less frequent with isolates of K. pneumoniaefrom 5 of 6 US centers than with those from 6 centers outside the US. In nearly all of the centers, resistance to sulfamethoxazole-trimethoprim, gentamicin, tobramycin, or chloramphenicol was more frequent in isolates of K. pneumoniaethan in those of Escherichia coli,while the reverse was true for resistance to tetracycline. Resistance to multiple antibiotics declined gradually in isolates of K. pneumoniaeat one center, but rose abruptly again with dissemination of a new plasmid.

A survey of clinical isolates from a hospital laboratory showed that Escherichia colicould be grouped into three classes of beta-lactam-antibiotic resistance by results of routine susceptibility testing to ampicillin, cephalothin, and carbenicillin. E. colihighly resistant to ampicillin and carbenicillin but not to cephalothin (class I) were found to have one of two levels of R factor-mediated, periplasmic-β-lactamase which resembled RTEMand was located behind a permeability barrier to penicillins but not to cephalosporins. This permeability barrier appeared to act synergistically with the β-lactamase in producing high levels of resistance to penicillins. E. colihighly resistant to ampicillin and cephalothin but not carbenicillin (class II) were found to have a β-lactamase with predominantly cephalosporinase activity which was neither transferable nor releasable by osmotic shock. E. colimoderately resistant to one or to all three of these antibiotics (class III) were found to have low levels of different β-lactamases including a transferable β-lactamase which resembled R1818. Thus, different mechanisms producing resistance to β-lactam antibiotics could be deduced from the patterns of resistance to ampicillin, cephalothin, and carbenicillin found on routine susceptibility testing. E. coliof class I were much more prevalent than the other classes and the proportion of E. colithat were class I increased with duration of patient hospitalization. The incidence of class I E. colirose only slightly over the past 7 years and that of class II E. coliremained constant despite increased usage of both cephalothin and ampicillin. These observations emphasize that the properties of the apparently limited number of individual resistance mechanisms that exist in a bacterial flora, such as their genetic mobility and linkages and the spectrum of their antibiotic inactivating enzymes and permeability barriers, may govern the effect that usage of an antibiotic has upon the prevalence of resistance to it and to other antibiotics.

Automatic graphic display of the data output increases the effectiveness of computer analysis of results of hospital antimicrobial-susceptibility tests. The system we used plots the frequency distribution of inhibition-zone diameters around any antibiotic disk for selected groupings of bacterial isolates. It tallies the group's relative frequency of isolation and its percentage susceptible to the antibiotic. These plots allow comparison of differences due to variables, such as time and source of isolation. They permit detection of distinctive subgroups among the hospital's flora and of associated resistance to multiple antibiotics. They can contribute to quality control of the susceptibilitytesting procedure and of the process of bacterial classification itself. They provide for the practicing physician a concise and graphic statistical basis for antibiotic selection.