Multidimensional hierarchical sum aggregations can be formulated as matrix multiplications involving dummy matrices which can be referred to as model matrices. In contrast to standard model matrices, all categories of all variables must be included. For this purpose, the R package SSBtools includes functionality to create model matrices in two alternative ways, by model formulas or by so-called hierarchies. The latter means a coding of hierarchical relationships, and this can be done in several ways. Tree -shaped hierarchies are not required. The internal standard in the package is a parent -child coding. Functionality to find hierarchies automatically from the data is also included. The model matrix functionality is applied in several R packages for statistical disclosure control. This enables general implementation of methods and a flexible user interface. This paper describes the model matrix and hierarchy functions in SSBtools, as well as the methods and functions behind it.

To protect tabular data through cell suppression, efficient algorithms are essential. Gaussian elimination can be used for secondary cell suppression to prevent exact disclosure. A beneficial feature of this method is that all tables created from the same microdata can be handled simultaneously. This paper presents a solution to the issue where suppressed zeros in frequency tables cannot protect each other. In magnitude tables, it outlines how the algorithm can be tailored to provide protection against singleton contributors using their own data for disclosure.

Cell suppression is a widely used statistical disclosure control method for tabular data. Commonly, several linked tables are suppressed simultaneously. After publication, additional tables may be requested. In many contexts, new tables mean new ways of grouping and aggregating data that has already been published. The suppression of the new tables must be coordinated with the tables that have already been disseminated. A certain type of synthetic decimal numbers has proven to be very useful for this purpose. Based on the aggregation of these decimal numbers, one can decide whether a cell should be suppressed or not. An aggregation summing up to a whole number means the same as non-suppression. This article describes the theoretical basis for such decimal numbers. This is based on standard methodology from ordinary linear regression. The method is illustrated by a small example. In addition, two practical applications at Statistics Norway are presented, where one involves large hierarchical and linked tables where more than 50000 unique cells were primarily suppressed.

Recent law changes in Norway have forced Statistics Norway to re-evaluate the use of SDC methods for protecting its publications. Previous distinctions regarding the sensitivity of variables are no longer included in the law. Instead, publications are required by law to ensure that no identifiable individual can have attributes directly or indirectly disclosed. Common heuristics (such as small count primary suppression) used in frequency table suppression do not directly target disclosure, but rather rely on established rule-of-thumb. Such methods can lead to the suppression of a large number of cells and simultaneously not provide adequate protection from disclosure. Motivated by risk measures within microdata, we discuss a new approach aimed directly at detecting those cells that can lead to disclosure in frequency tables. In this context we define the notion of direct disclosure: the scenario where an attacker, with full knowledge of themselves (or a coalition), can disclose attributes of specific individuals. Finally, we present an implementation where the above method for primary suppression is combined with an alternative method to secondary suppression which aims to guarantee no linear dependency between published and suppressed cells. The secondary suppression method is implemented using Gaussian elimination and shows promising results in practice, striking a balance between computational complexity and number of suppressed cells. The presented methods are available in the R package GaussSuppression. Suppression of Directly Disclosive Cells in Frequency Tables Daniel P. Lupp∗, Øyvind Langsrud∗∗ ∗ Statistics Norway, daniel.lupp@ssb.no ∗∗ Statistics Norway, oyvind.langsrud@ssb.no Abstract. Recent law changes in Norway have forced Statistics Norway to re-evaluate the use of SDC methods for protecting its publications. Previous distinctions regarding the sensitivity of variables are no longer included in the law. Instead, publications are required by law to ensure that no identifiable individual can have individual attributes directly or indirectly disclosed. Common heuristics (such as small count primary suppression) used in frequency table suppression do not directly target disclosure, but rather rely on established rule-of-thumb. Such methods can lead to the suppression of a large number of cells and simultaneously not provide adequate protection from disclosure. Motivated by risk measures within microdata, we discuss a new approach aimed directly at detecting those cells that can lead to disclosure in frequency tables. In this context we define the notion of direct disclosure: the scenario where an attacker, with full knowledge of themselves (or a coalition), can disclose attributes of specific individuals. Finally, we present an implementation where the above method for primary suppression is combined with an alternative method to secondary suppression which aims to guarantee no linear dependency between published and suppressed cells. The secondary suppression method is implemented using Gaussian elimination and shows promising results in practice, striking a balance between computational complexity and number of suppressed cells. The presented methods are available in the R package GaussSuppression. Recent law changes in Norway have forced Statistics Norway to re-evaluate the use of SDC methods for protecting its publications. Previous distinctions regarding the sensitivity of variables are no longer included in the law. Instead, publications are required by law to ensure that no identifiable individual can have individual attributes directly or indirectly disclosed. Common heuristics (such as small count primary suppression) used in frequency table suppression do not directly target disclosure, but rather rely on established rule-of-thumb. Such methods can lead to the suppression of a large number of cells and simultaneously not provide adequate protection from disclosure. Motivated by risk measures within microdata, we discuss a new approach aimed directly at detecting those cells that can lead to disclosure in frequency tables. In this context we define the notion of direct disclosure: the scenario where an attacker, with full knowledge of themselves (or a coalition), can disclose attributes of specific individuals. Finally, we present an implementation where the above method for primary suppression is combined with an alternative method to secondary suppression which aims to guarantee no linear dependency between published and suppressed cells. The secondary suppression method is implemented using Gaussian elimination and shows promising results in practice, striking a balance between computational complexity and number of suppressed cells. The presented methods are available in the R package GaussSuppression.

In a Eurostat-granted project within Statistics Norway, the cell key method (CKM) and the small count rounding method (SCR) implemented in the r package SmallCountRounding are to be compared. CKM perturbs all values without additivity constraints. SCR perturbs a number of inner cells so that small frequencies are avoided in the aggregated data to be published. A fair comparison is challenging since the methods are very different. Inspired by synthetic data methods, one approach is to generate expected inner cell frequencies from the perturbed tables. These expected frequencies can be viewed as the best guess of individual level data, although the frequencies are not whole numbers. Since SCR is additive, expected frequencies can be generated by a variant of iterative proportional fitting (IPF). This estimation coincides with log-linear modelling. This approach is not directly applicable to CKM. The data needs to be additivity restored first and we can perform this by least squares estimation. The results are then maximum likelihood estimates under the assumption of gaussian noise. This is a simplification compared to using the actual discrete noise distribution. A modification is needed to ensure non-negativity. Afterwards, expected inner cell frequencies can be obtained by IPF. No extra information has been added during this post-processing, but individual disclosive information is easier available. Appropriate measures of risk and utility can be made based on the post-processed data and the original data. We will discuss possible measures that can be used to compare CKM, SCR, as well as other methods. In addition, they can be used as guidance for fine-tuning each method. Fair risk-utility comparison of tabular perturbation methods by post-processing to expected frequencies Øyvind Langsrud∗, Daniel P. Lupp∗∗ ∗ Statistics Norway, oyvind.langsrud@ssb.no ∗∗ Statistics Norway, daniel.lupp@ssb.no Abstract. In a Eurostat-granted project within Statistics Norway, the cell key method (CKM) and the small count rounding method (SCR) implemented in the R package In a Eurostat-granted project within Statistics Norway, the cell key method (CKM) and the small count rounding method (SCR) implemented in the R package SmallCountRounding are to be compared. CKM perturbs all values without additivity constraints. SCR perturbs a number of inner cells so that small frequencies are avoided in the aggregated data to be published. A fair comparison is challenging since the methods are very different. Inspired by synthetic data methods, one approach is to generate expected inner cell frequencies from the perturbed tables. These expected frequencies can be viewed as the best guess of individual level data, although the frequencies are not whole numbers. Since SCR is additive, expected frequencies can be generated by a variant of iterative proportional fitting (IPF). This estimation coincides with log-linear modelling. This approach is not directly applicable to CKM. The data needs to have additivity restored first, which we can achieve by least squares estimation. The results are then maximum likelihood estimates under the assumption of Gaussian noise. This is a simplification compared to using the actual discrete noise distribution. A modification is needed to ensure non-negativity. Afterwards, expected inner cell frequencies can be obtained by IPF. No extra information has been added during this post-processing, but individual disclosive information is more easily available. Appropriate measures of risk and utility can be made based on the post-processed data and the original data. In this paper we discuss possible measures that can be used to compare CKM, SCR, as well as other methods. In addition, they can be used as guidance for fine-tuning each method.

and Paper The microdata underlying tabular frequency data can be represented by the inner frequency table obtained by crossing all the main dimensional variables. This paper will discuss approaches to postprocessing the results of tabular SDC techniques to obtain releasable inner frequencies, which may not be whole numbers. As suggested in a recent publication and implemented in the r-package RegSDC, suppressed cells may be replaced by decimal numbers, in a way so that they add correctly up to non-supressed marginals. A similar method could be used in cases without cell suppression, but where only some marginals are considered releasable. This method (Method A) can be viewed as ordinary regression, modified to handle collinearity, with the inner cells as the unknown parameters and with the releasable marginals as response observations. Method A can also be interpreted as linear calibration weighting. An improvement of this approach is to ensure non-negative values by applying nonnegative least squares with ridge penalty (Method B). A more natural choice in this case, would be to obtain inner frequencies as fitted values from log-linear modelling. Considering situations where the marginals are perturbed in a non-additive way, ordinary log-linear modelling is not possible, but Method B is useful. Log-linear fitting may be performed afterwards to obtain results in in accordance with such modelling assumptions. Another situation is when only some of the releasable marginals are perturbed. Then, it is advantageous to modify Method B by incorporating weights (Method C). The perturbed marginals are weighted down, so that the resulting inner frequencies add up almost exactly to the non-perturbed marginals. The methodology is applicable when the marginals are perturbed by Laplace noise to ensure differential privacy. Although, in this case, one may consider least absolute deviation instead of least squares. The generated frequencies serve several purposes: They can be used internally or published directly. Additivity of marginals can be restored. One can also use the results to generate public microdata. Then, all frequencies must be whole numbers. A possibility is to round up or down by comparing the decimals to draws from the uniform distribution. This is an alternative to synthetic data drawn as a multinomial sample according to cell probabilities, as in the r-package synthpop. The methodology is tested using example data of various size and complexity. For an efficient implementation, making use of the r-package glmnet seems promising. Releasable inner cell frequencies by post-processing protected tabular data Øyvind Langsrud∗ ∗ Statistics Norway, Department of methodology and data collection, Oyvind.Langsrud@ssb.no, github.com/olangsrud Abstract. The microdata underlying tabular frequency data can be represented by the inner frequency table obtained by crossing all the main dimensional variables. This paper will discuss approaches to post-processing the results of tabular SDC techniques to obtain releasable inner frequencies, which may not be whole numbers. As suggested in a recent publication and implemented in the r-package RegSDC, suppressed cells may be replaced by decimal numbers, in a way so that they add correctly up to non-supressed margins. A similar method could be used in cases without cell suppression, but where only some margins are considered releasable. This method (Method A) can be viewed as ordinary regression, modified to handle collinearity, with the inner cells as the unknown parameters and with the releasable margins as response observations. Method A can also be interpreted as linear calibration weighting. An improvement of this approach is to ensure non-negative values by applying non-negative least squares with ridge penalty (Method B). A more natural choice in this case, would be to obtain inner frequencies as fitted values from log-linear modelling. Considering situations where the margins are perturbed in a non-additive way, ordinary log-linear modelling is not possible, but Method B is useful. Log-linear fitting may be performed afterwards to obtain results in in accordance with such modelling assumptions. Another situation is when only some of the releasable margins are perturbed. Then, it is advantageous to modify Method B by incorporating weights (Method C). The perturbed margins are weighted down, so that the resulting inner frequencies add up almost exactly to the non-perturbed margins. The methodology is applicable when the margins are perturbed by Laplace noise to ensure differential privacy. Although, in this case, one may consider least absolute deviation instead of least squares. The generated frequencies serve several purposes: They can be used internally or published directly. Additivity of margins can be restored. One can also use the results to generate public microdata. Then, all frequencies must be whole numbers. A possibility is to round up or down by comparing the decimals to draws from the uniform distribution. This is an alternative to synthetic data drawn as a multinomial sample according to cell probabilities, as in the r-package synthpop. The methodology is tested using example data of various size and complexity. For an efficient implementation, making use of the r-package glmnet seems promising. The microdata underlying tabular frequency data can be represented by the inner frequency table obtained by crossing all the main dimensional variables. This paper will discuss approaches to post-processing the results of tabular SDC techniques to obtain releasable inner frequencies, which may not be whole numbers. As suggested in a recent publication and implemented in the r-package RegSDC, suppressed cells may be replaced by decimal numbers, in a way so that they add correctly up to non-supressed margins. A similar method could be used in cases without cell suppression, but where only some margins are considered releasable. This method (Method A) can be viewed as ordinary regression, modified to handle collinearity, with the inner cells as the unknown parameters and with the releasable margins as response observations. Method A can also be interpreted as linear calibration weighting. An improvement of this approach is to ensure non-negative values by applying non-negative least squares with ridge penalty (Method B). A more natural choice in this case, would be to obtain inner frequencies as fitted values from log-linear modelling. Considering situations where the margins are perturbed in a non-additive way, ordinary log-linear modelling is not possible, but Method B is useful. Log-linear fitting may be performed afterwards to obtain results in in accordance with such modelling assumptions. Another situation is when only some of the releasable margins are perturbed. Then, it is advantageous to modify Method B by incorporating weights (Method C). The perturbed margins are weighted down, so that the resulting inner frequencies add up almost exactly to the non-perturbed margins. The methodology is applicable when the margins are perturbed by Laplace noise to ensure differential privacy. Although, in this case, one may consider least absolute deviation instead of least squares. The generated frequencies serve several purposes: They can be used internally or published directly. Additivity of margins can be restored. One can also use the results to generate public microdata. Then, all frequencies must be whole numbers. A possibility is to round up or down by comparing the decimals to draws from the uniform distribution. This is an alternative to synthetic data drawn as a multinomial sample according to cell probabilities, as in the r-package synthpop. The methodology is tested using example data of various size and complexity. For an efficient implementation, making use of the r-package glmnet seems promising.

This paper presents a unified framework for regression-based statistical disclosure control for microdata. A basic method, known as information preserving statistical obfuscation (IPSO), produces synthetic data that preserve variances, covariances and fitted values. The data are then generated conditionally according to the multivariate normal distribution. Generalizations of the IPSO method are described in the literature, and these methods aim to generate data more similar to the original data. This paper describes these methods in a concise and interpretable way, which is close to efficient implementation. Decomposing the residual data into orthogonal scores and corresponding loadings is an essential part of the framework. Both QR decomposition (Gram–Schmidt orthogonalization) and singular value decomposition (principal components) may be used. Within this framework, new and generalized methods are presented. In particular, a method is described by means of which the correlations to the original principal component scores can be controlled exactly. It is shown that a suggested method of random orthogonal matrix masking can be implemented without generating an orthogonal matrix. Generalized methodology for hierarchical categories is presented within the context of microaggregation. Some information can then be preserved at the lowest level and more information at higher levels. The presented methodology is also applicable to tabular data. One possibility is to replace the content of primary and secondary suppressed cells with generated values. It is proposed replacing suppressed cell frequencies with decimal numbers, and it is argued that this can be a useful method.

Statistics Norway is modernising the production processes. An important element in this work is a library of functions for statistical computations. In principle, the functions in such a methods library can be programmed in several languages. A modernised production environment demand that these functions can be reused for different statistics products, and that they are embedded within a common IT system. The embedding should be done in such a way that the users of the methods do not need to know the underlying programming language. As a proof of concept, Statistics Norway soon has established a methods library offering a limited number of methods for macro-editing, imputation and confidentiality. This is done within an area of municipal statistics with R as the only programming language. This paper presents the details and experiences from this work. The problem of fitting real word applications to simple and strict standards is discussed and exemplified by the development of solutions to regression imputation and table suppression.

This paper presents a newly developed R-package for calculation of variances of estimates based on data from several waves of a repeated survey with partly overlapping samples. Development of the package is a part of on-going work on quality improvements of the Labour Force Survey in Norway, which is quarterly and based on a rotating panel. The package can, for example, be used to calculate variances of net changes of annual averages of unemployment rates for persons aged 20-64. The methodology is based on linearly calibrated weights (as calculated by the packages ReGenesees and survey) and residuals from the corresponding regression modelling. These computations may be done separately for each wave. The functionality is generic and the user can specify any calibration model and any linear combination of (quarterly) estimates. Linearization is used to calculate variances of rates. The main method assumes that all relevant population totals can be computed from register data, but situations where totals are unknown for some of the calibration variables are also handled.

Various direct (citrate, malate and succinate) and indirect (pyruvate and glutamate) Krebs cycle substrates were added to ground beef in order to investigate their effect on the inter-conversion of Mb redox forms in aerobic and anaerobic packaging. Glutamate, malate, succinate, pyruvate, and citrate added (up to totally 0.1 mol/kg) to ground bovine M. semimembranosus mixed with either ground porcine or bovine fat, altered the myoglobin redox forms in aerobic and anaerobic packaging systems. In anaerobic packaging, a mixture of succinate and glutamate formed deoxymyoglobin rapidly and it remained in this state for 13 days. In aerobic packaging (75% O2), the highest oxymyoglobin level occurred with a molar ratio of (glutamate–malate) to citrate of 3:1. In this case, oxymyoglobin was more prevalent after 6–8 days of storage in aerobic condition than without addition of these compounds. Pyruvate induced metmyoglobin formation, acting as a pro-oxidant. Succinate presumed leading to FADH2; was most effective at converting metmyoglobin to deoxymyoglobin in anaerobic packaging. In aerobic packaging, NADH presumed formed by the oxidation of glutamate may maintain oxymyoglobin levels, but adding citrate as well is recommended. Overall, a combination of substrates relevant to mitochondrial oxygen consumption, improved meat color stability.

Abstract The WHO/UNICEF Baby‐friendly Hospital Initiative has been shown to increase breastfeeding rates, but uncertainty remains about effective methods to improve breastfeeding in community health services. The aim of this pragmatic cluster quasi‐randomised controlled trial was to assess the effectiveness of implementing the Baby‐friendly Initiative (BFI) in community health services. The primary outcome was exclusive breastfeeding until 6 months in healthy babies. Secondary outcomes were other breastfeeding indicators, mothers' satisfaction with the breastfeeding experience, and perceived pressure to breastfeed. A total of 54 Norwegian municipalities were allocated by alternation to the BFI in community health service intervention or routine care. All mothers with infants of five completed months were invited to participate (n = 3948), and 1051 mothers in the intervention arm and 981 in the comparison arm returned the questionnaire. Analyses were by intention to treat. Women in the intervention group were more likely to breastfeed exclusively compared with those who received routine care: 17.9% vs. 14.1% until 6 months [cluster adjusted odds ratio (OR) = 1.33; 95% confidence interval (CI): 1.03, 1.72; P = 0.03], 41.4% vs. 35.8% until 5 months [cluster adjusted OR = 1.39; 95% CI: 1.09, 1.77; P = 0.01], and 72.1% vs. 68.2% for any breastfeeding until 6 months [cluster adjusted OR = 1.24; 95% CI: 0.99, 1.54; P = 0.06]. The intervention had no effect on breastfeeding until 12 months. Maternal breastfeeding experience in the two groups did not differ, neither did perceived breastfeeding pressure from staff in the community health services. In conclusion, the BFI in community health services increased rates of exclusive breastfeeding until 6 months. © 2015 Blackwell Publishing Ltd

Krebs cycle substrates (KCS) can stabilise the colour of packaged meat by oxygen reduction. This study tested whether this reduction releases reactive oxygen species that may lead to lipid oxidation in minced meat under two different storage conditions.KCS combinations of succinate and glutamate increased peroxide forming potential (PFP, 1.18–1.32mmolperoxides/kg mince) and thiobarbituric acid reactive substances (TBARS, 0.30–0.38mgmalondialdehyde (MDA) equivalents/kg mince) under low oxygen storage conditions. Both succinate and glutamate were metabolised. Moreover, under high oxygen (75%) storage conditions, KCS combinations of glutamate, citrate and malate increased PFP (from 1.22 to 1.29mmolperoxides/kg) and TBARS (from 0.37 to 0.40mgMDA equivalents/kg mince). Only glutamate was metabolised.The KCS combinations that were added to stabilise colour were metabolised during storage, and acted as pro-oxidants that promoted lipid oxidation in both high and low oxygen conditions.

In order to identify how different additives influenced lipid peroxidation formation, a sausage only using beef juice as pigment source and a standard beef-pork meat sausage were studied. The effects of different additives, including fish oil, myoglobin, nitrite, clove extract, and calcium sources on oxidation and sensory properties were examined. Both sausage systems were stored in 3 different manners prior to testing: (1) frozen immediately at -80 °C; (2) chilled stored for 2.5 weeks followed by fluorescent light illumination at 4 °C for another 2 wk; (3) frozen at -20 °C for 5 mo. The frozen group 3 showed the highest peroxide formation and thiobarbituric acid reactive substances (TBARS) for both sausage systems. Unpolar peroxides dominated in both systems. The clove extract could offset the peroxide formation from myoglobin/beef juice and/or fish oil, but the addition of clove flavor was recognized by the sensory panelists. Calcium addition reduced lipid peroxide formation. Added nitrite and fish oil seemed to interact to stimulate nitroso-myoglobin formation. Nitrite was identified to interact with clove addition and thereby, relatively speaking, increased TBARS. The 2 sausage systems generally ranked the additives similarly as pro- and antioxidants.

BACKGROUND:Randomized controlled psychosocial treatment studies on selective mutism (SM) are lacking.METHOD:Overall, 24 children with SM, aged 3-9 years, were randomized to 3 months treatment (n = 12) or wait list (n = 12). Primary outcome measure was the School Speech Questionnaire.RESULTS:A significant time by group interaction was found (p = .029) with significantly increased speech in the treatment group (p = .004) and no change in wait list controls (p = .936). A time by age interaction favoured younger children (p = .029). Clinical trail registration: Norwegian Research CouncilNCT01002196.CONCLUSIONS:The treatment significantly improved speech. Greater improvement in the younger age group highlights the importance of an early intervention.

We prospectively studied parental mental health after suddenly losing a son in a military training accident. Parents (N = 32) were interviewed at 1, 2 and 23 years after the death of their son. The General Health Questionnaire and Expanded Texas Inventory of Grief were self-reported at 1, 2, 5, and 23 years; the Inventory of Complicated Grief was self-reported at 23 years. We observed a high prevalence of psychiatric disorders at 1- and 2-year follow-ups (57% and 45%, respectively), particularly major depression (43% and 31%, respectively). Only one mental disorder was diagnosed at the 23-year follow-up. Grief and psychological distress were highest at 1- and 2-year follow-ups. Spouses exhibited a high concordance of psychological distress. Mothers reported more intense grief reactions than did fathers. The loss of a son during military service may have a substantial impact on parental mental health particularly during the first 2 years after death. Spouses' grief can be interrelated and may contribute to their psychological distress.

Summary Background Intestinal microbiota undergoes substantial development during the first 2 years of life, important for intestinal immunologic development and maturation influencing systemic immune responses. Objective We aimed to investigate, using a prospective study design, whether allergen‐specific IgE (sIgE) and atopic eczema are associated with variations in gut microbial colonization patterns in an unselected population during the first 2 years of life. Methods Faeces from 94 infants were repeatedly sampled from 10 days, 4 months, 1 and 2 years postnatal and analysed for 12 different bacterial species by quantitative real‐time PCR. Venous blood samples from the infants were collected at 2 years of age and were analysed for sIgE for 12 specific allergens. The temporal gut colonization patterns for 42 sIgE‐positive (sIgE⩾0.35 kU/L) and 52 sIgE‐negative children (sIgE<0.1 kU/L) were then compared. The association between colonization pattern and phenotype as atopic eczema according to UK Working Party (UKWP) criteria were also described. Results Subjects with atopic sensitization had lower levels of Escherichia coli at 4 months and 1 year, higher levels of Bifidobacterium longum at 1 year and lower levels of Bacteroides fragilis at 2 years. For E. coli and B. longum, the differences were only transient and had disappeared by 2 years of age. For other species, there were no differences in colonization patterns, and we found no association between colonization pattern and atopic eczema. Conclusions and Clinical Relevance We found temporal and transient variations in gut microbial colonization patterns associated with differences in sIgE sensitization at 2 years of age. A full understanding of the principles and mechanisms that underlie intestinal microbial colonization and diversity and host–microbiota relationships will be pivotal for the development of therapeutic approaches that manipulate the intestinal microbiota to maintain human health. [Registration number: ISRCTN28090297] Cite this as: O. Storrø, T. Øien, Ø. Langsrud, K. Rudi, C. Dotterud and R. Johnsen, Clinical & Experimental Allergy, 2011 (41) 1545–1554.

BACKGROUND:Clinician-rated measures are used extensively in child and adolescent mental health services (CAMHS). The Health of the Nation Outcome Scales for Children and Adolescents (HoNOSCA) is a short clinician-rated measure developed for ordinary clinical practice, with increasing use internationally. Several studies have investigated its psychometric properties, but there are few data on its correspondence with other methods, rated by other informants. We compared the HoNOSCA with the well-established Achenbach System of Empirically Based Assessment (ASEBA) questionnaires: the Child Behavior Checklist (CBCL), the Teacher's Report Form (TRF), and the Youth Self-Report (YSR).METHODS:Data on 153 patients aged 6-17 years at seven outpatient CAMHS clinics in Norway were analysed. Clinicians completed the HoNOSCA, whereas parents, teachers, and adolescents filled in the ASEBA forms. HoNOSCA total score and nine of its scales were compared with similar ASEBA scales. With a multiple regression model, we investigated how the ASEBA ratings predicted the clinician-rated HoNOSCA and whether the different informants' scores made any unique contribution to the prediction of the HoNOSCA scales.RESULTS:We found moderate correlations between the total problems rated by the clinicians (HoNOSCA) and by the other informants (ASEBA) and good correspondence between eight of the nine HoNOSCA scales and the similar ASEBA scales. The exception was HoNOSCA scale 8 psychosomatic symptoms compared with the ASEBA somatic problems scale. In the regression analyses, the CBCL and TRF total problems scores together explained 27% of the variance in the HoNOSCA total scores (23% for the age group 11-17 years, also including the YSR). The CBCL provided unique information for the prediction of the HoNOSCA total score, HoNOSCA scale 1 aggressive behaviour, HoNOSCA scale 2 overactivity or attention problems, HoNOSCA scale 9 emotional symptoms, and HoNOSCA scale 10 peer problems; the TRF for all these except HoNOSCA scale 9 emotional symptoms; and the YSR for HoNOSCA scale 9 emotional symptoms only.CONCLUSION:This study supports the concurrent validity of the HoNOSCA. It also demonstrates that parents, teachers and adolescents all contribute unique information in relation to the clinician-rated HoNOSCA, indicating that the HoNOSCA ratings reflect unique perspectives from multiple informants.

Staphylococcus aureus is an important food poisoning bacterium. In food preservation, acidification is a well-known method. Permeant weak organic acids, like lactic and acetic acids, are known to be more effective against bacteria than inorganic strong acids (e.g., HCl). Growth experiments and metabolic and transcriptional analyses were used to determine the responses of a food pathogenic S. aureus strain exposed to lactic acid, acetic acid, and HCl at pH 4.5. Lactic and acetic acid stress induced a slower transcriptional response and large variations in growth patterns compared with the responses induced by HCl. In cultures acidified with lactic acid, the pH of the medium gradually increased to 7.5 during growth, while no such increase was observed for bacteria exposed to acetic acid or HCl. Staphylococcus aureus increased the pH in the medium mainly through accumulation of ammonium and the removal of acid groups, resulting in increased production of diacetyl (2,3-butanedione) and pyrazines. The results showed flexible and versatile responses of S. aureus to different types of acid stress. As measured by growth inhibition, permeant organic acid stress introduced severe stress compared with the stress caused by HCl. Cells exposed to lactic acid showed specific mechanisms of action in addition to sharing many of the mechanisms induced by HCl stress.

BACKGROUND: Pre-rigor fillets of Atlantic salmon were either superchilled or chilled prior to packaging in air or in modified atmosphere (MAP, 60% CO2/40% N-2) with a CO2 emitter, in 5.3 L high-density polyethylene trays with three to four layers of fillets (3.0-3.7 kg). All samples were stored at 0.1 degrees C for 28 days.RESULTS: Fillets stored in MAP had significantly lower bacterial growth compared to fillets stored in air, and MAP superchilled bottomfillets had lower bacterial counts compared to the corresponding chilled fillets. Samples superchilled prior to refrigerated storage in air had similar bacterial growth to ordinary chilled samples. Faster fillet softening during storage and higher liquid loss were observed in superchilled MAP samples.CONCLUSION: Combining short-term superchilling and MAP with a CO2 emitter prolonged the shelf-life of pre-rigor salmon fillets, which can improve sustainability throughout the value chain. The superchilling method needs to be optimized to avoid negative effects on texture and liquid loss. (C) 2009 Society of Chemical Industry