Inflection points in longitudinal models: Tracking recovery and return to play following concussion.

The return to play (RTP) process may occur during longitudinal studies tracking recovery after concussion. This factor, which is often omitted within statistical designs, could affect the fit and overall interpretation of the statistical model. This article demonstrates the difference in results and interpretation between 2 linear mixed-model designs: (1) a between-group longitudinal (GROUP) analysis and (2) a between-group longitudinal model that used an inflection point to account for changes around the time of RTP (RTP analysis). These analyses were conducted on instrumented balance data collected on 23 concussed athletes and 25 controls over 8 weeks following concussion. Total sway area and the range of mediolateral acceleration were used as outcome measures. No significant findings were found in the GROUP design for either outcome measure. In contrast, the RTP analysis revealed significant effects of time (P = .007) and RTP change (P = .007), and group*time (P = .028) and group*RTP change (P = .022) interactions for total sway area, and effects of group (P = .011), time (P = .010), and RTP change (P = .014), and group*time (P = .013) and group*RTP change interactions (P = .013) for range of mediolateral acceleration. For both outcomes, the RTP model fit the data significantly better on comparison of likelihood ratios (P ≤ .027). These results suggest that allowing for an inflection point in the statistical design may assist understanding of what happens around clinically meaningful time points. The choice of statistical model had a considerable effect on the interpretation of findings, and provokes discussion around the best method for analyzing longitudinal datasets when important clinical time points like RTP exist.

Membrane filter method based on FC-5-bromo-4-chloro-3-indolyl-beta-D-glucuronide medium facilitates enumeration of Escherichia coli in foods and poultry carcass rinses.

Three enumeration methods for Escherichia coli in foods, the Health Protection Branch most-probable-number (MPN) method MFHPB-19, a hydrophobic grid membrane filter method MFHPB-26 (HGMF-indole), and a hydrophobic grid membrane filter method utilizing 5-bromo-4-chloro-3-indolyl-beta-D-glucuronide in a (modified) mFC agar (HGMF-FC-BCIG) were compared in 80 food samples that included naturally and artificially contaminated raw vegetables, mung bean and alfalfa sprouts, raw meats, and chicken carcass rinses. The number of samples confirmed as positive for E. coli were 44, 36, and 42 for the MPN, HGMF-indole, and HGMF-BCIG methods, respectively. By the MPN method, E. coli was detected in 3 samples at levels below the limits of detection of the HGMFs; but the MPN method was very time-consuming. With the HGMF-indole procedure E. coli was missed in 4 artificially contaminated samples. With the HGMF-FC-BCIG method E. coli was enumerated in 1 sample of bean sprouts missed by both the MPN and HGMF-indole procedures. High levels of indole-positive Klebsiella spp. in bean sprouts interfered with the HGMF-indole method, but the blue colonies of E. coli were easily observed in the HGMF-FC-BCIG method. Specificity of the HGMF-FC-BCIG method is high enough that routine confirmation should be unnecessary; however, confirmation of presumptive E. coli is easier since no lethal indole-staining step is involved. It appears to be a very simple method for quantifying E. coli in foods or carcass rinses.

Improved aerobic colony count technique for hydrophobic grid membrane filters.

The AOAC International official action procedure for performing aerobic colony counts on hydrophobic grid membrane filters (HGMFs) uses Trypticase soy-fast green FCF agar (FGA) incubated for 48 h. Microbial growths are various shades of green on a pale green background, which can cause problems for automated as well as manual counting. HGMFs which had been incubated 24 or 48 h at 35 degrees C on Trypticase soy agar were flooded underneath with 1 to 2 ml of 0.1% triphenyltetrazolium chloride (TTC) solution by simply lifting one corner of the filter while it was still on the agar and adding the reagent. Microbial growths on HGMFs were counted after color had been allowed to develop for 15 min at room temperature. With representative foods, virtually all colonies stained pink to red. Automated electronic counts made by using the MI-100 HGMF Interpreter were easier and more reliable than control HGMF counts made by the AOAC International official action procedure. Manual counting was easier as well because of increased visibility of the microbial growths. Except in the case of dairy products, 24-h TTC counts did not differ significantly from 48-h FGA counts, whereas the FGA counts at 24 h were always significantly lower, indicating that for many food products the HGMF TTC flooding method permits aerobic colony counts to be made after 24 h.

Increased sensitivity of the rapid hydrophobic grid membrane filter enzyme-labeled antibody procedure for Escherichia coli O157 detection in foods and bovine feces.

Several strains of Escherichia coli O157:H7 artificially inoculated into vegetables and dairy products were recovered on hydrophobic grid membrane filters and enumerated by an enzyme-labeled antibody assay. The mean of the recoveries from 12 fresh vegetables was 108.8%, whereas that from 10 dairy products was 93.2%. Modified tryptic soy broth at 43 degrees C with shaking at 100 rpm provided optimum recovery of the organism from meat, with a sensitivity of less than or equal to 1 CFU/g, which is 10 times more sensitive than direct plating. The method performed equally well with vegetable and dairy products. Tryptic soy broth, however, under the same conditions gave the best results for fecal samples. Of 22 asymptomatic dairy cattle, reported as having positive Brucella titers when assayed with polyclonal antibodies, eight were found to contain E. coli O157 in their feces as demonstrated by the enzyme-labeled antibody assay by using monoclonal antibodies. This finding may explain some of the false-positive Brucella tests.

Efficient Nondestructive Sampler for Carcasses and Other Surfaces.

Two versions of an electrically powered device (Rotorinser) to sample carcasses or other surfaces in situ for microbiological analysis and several different sampling protocols were evaluated against excision plus stomaching for ability to remove bacteria from pig skin and beef carcass tissue. Both devices sampled circular areas of approximately 14 cm2. Ten tissue samples were used for each set of conditions. Rotorinser bacterial removal efficiency was calculated as R/(R + S), where R is the Rotorinser count (CFU cm-2) and S is the count on stomached excised tissue after rotorinsing. Stomacher efficiencies were calculated as S1/(S1 + S2), where S1 is the first stomacher count of excised tissue and S2 is the count from a second stomaching. Both Rotorinsers were much better than traditional swabs. Rotorinser 1 gave removal efficiencies of 0.79 to 0.88 for beef, and 0.79 to 0.95 for pork. Prewetting surfaces for 5 min improved removal, but mixtures of enzymes did not. Rotorinser 2 applied with NaCl or NaCl-Tween 80 diluent for either 30 or 60 s was significantly better (0.93 and 0.98) than the stomacher (0.86) at removing aerobic mesophilic bacteria from pork skin. The Rotorinser causes negligible tissue damage and can be used on surfaces at any angle.

Rapid hydrophobic grid membrane filter-enzyme-labeled antibody procedure for identification and enumeration of Escherichia coli O157 in foods.

An O-antigen-specific monoclonal antibody, labeled by horseradish peroxidase-protein A, was used in a hydrophobic grid membrane filter-enzyme-labeled antibody method for rapid detection of Escherichia coli O157 in foods. The method yielded presumptive identification within 24 h and recovered, on average, 95% of E. coli O157:H7 artificially inoculated into comminuted beef, veal, pork, chicken giblets, and chicken carcass washings. In food samples from two outbreaks involving E. coli O157:H7, the organism was isolated at levels of up to 10(3)/g. The lower limit of sensitivity was 10 E. coli O157 per g of meat. Specific typing for E. coli O157:H7 can be achieved through staining with labeled H7 antiserum or tube agglutination.