Product specialization in international trade : a further investigation

In an influential paper, Schott [Schott. Peter K. (2004). “Across-product versus within-product specialization in international trade.” Quarterly Journal of Economics, 119 (2): 647–678] makes two empirical observations about U.S. imports. (1) The United States is increasingly sourcing the same product (however narrowly defined) from both developed and developing countries. That is, ‘across-product specialization’ has been decreasing. (2) The unit values of these multiple-sourced products are positively and significantly correlated with the capital and skill abundance of exporters and with the capital–labor ratios used by exporters. That is, endowments-driven ‘within-product specialization’ has been increasing. We show that both these observations extend to the imports of Brazil, India and Japan. However, our main finding is that observation (1) is largely driven by two factors. First, China is the dominant low-wage exporter of multiple-sourced products. Second, the most developed countries remain the primary exporters of multiple-sourced products. The U.S. case is the most extreme of our four importers: When China is deleted from the U.S. import data there is no trend in across-product specialization and rich exporters are increasing their trade share of multiple-sourced products. Since deleting China has no theoretical justification, these results must be viewed not as a contradiction of Schott's work but as a way of deepening our understanding of his empirical results.

Identification of tRNAs incorporated into wild-type and mutant human immunodeficiency virus type 1

We have identified the tRNAs which are incorporated into both wild-type human immunodeficiency virus type 1 strain IIIB (HIV-1IIIB) produced in COS-7 cells transfected with HIV-1 proviral DNA and mutant, noninfectious HIV-1Lai particles produced in a genetically engineered Vero cell line. The mutant proviral DNA contains nucleotides 678 to 8944; i.e., both long terminal repeats and the primer binding site are absent. As analyzed by two-dimensional polyacrylamide gel electrophoresis, both mutant and wild-type HIV-1 contain four major-abundance tRNA species, which include tRNA(1,2Lys), tRNA(3Lys) (the putative primer for HIV-1 reverse transcriptase) and tRNA(Ile). Identification was accomplished by comparing the electrophoretic mobilities and RNase T1 digests with those of tRNA(3Lys) and tRNA(1,2Lys) purified from human placenta and comparing the partial nucleotide sequence at the 3' end of each viral tRNA species with published tRNA sequences. Thus, the absence of the primer binding site in the mutant virus does not affect tRNA(Lys) incorporation into HIV-1. However, only the wild-type virus contains tRNA(3Lys) tightly associated with the viral RNA genome. The identification of the tightly associated tRNA as tRNA(3Lys) is based upon an electrophoretic mobility identical to that of tRNA(3Lys) and the ability of this RNA to hybridize with a tRNA(3Lys)-specific DNA probe. In addition to the four wild-type tRNA species, the mutant HIV-1-like particle contains two tRNA(His) species and three tRNA-sized species that we have been unable to identify. Their absence in wild-type virus makes it unlikely that they are required for viral infectivity.

Improved risk predictions via sparse imputation of patient conditions in electronic medical records

Electronic Medical Records (EMR) are increasingly used for risk prediction. EMR analysis is complicated by missing entries. There are two reasons - the “primary reason for admission” is included in EMR, but the co-morbidities (other chronic diseases) are left uncoded, and, many zero values in the data are accurate, reflecting that a patient has not accessed medical facilities. A key challenge is to deal with the peculiarities of this data - unlike many other datasets, EMR is sparse, reflecting the fact that patients have some, but not all diseases. We propose a novel model to fill-in these missing values, and use the new representation for prediction of key hospital events. To “fill-in” missing values, we represent the feature-patient matrix as a product of two low rank factors, preserving the sparsity property in the product. Intuitively, the product regularization allows sparse imputation of patient conditions reflecting common comorbidities across patients. We develop a scalable optimization algorithm based on Block coordinate descent method to find an optimal solution. We evaluate the proposed framework on two real world EMR cohorts: Cancer (7000 admissions) and Acute Myocardial Infarction (2652 admissions). Our result shows that the AUC for 3 months admission prediction is improved significantly from (0.741 to 0.786) for Cancer data and (0.678 to 0.724) for AMI data. We also extend the proposed method to a supervised model for predicting of multiple related risk outcomes (e.g. emergency presentations and admissions in hospital over 3, 6 and 12 months period) in an integrated framework. For this model, the AUC averaged over outcomes is improved significantly from (0.768 to 0.806) for Cancer data and (0.685 to 0.748) for AMI data.

Eating marshmallows reduces ileostomy output: a randomized crossover trial

AIM: Anecdotally, many ostomates believe that eating marshmallows can reduce ileostomy effluent. There is a plausible mechanism for this, as the gelatine contained in marshmallows may thicken small bowel fluid, but there is currently no evidence that this is effective. METHOD: This was a randomized crossover trial. Adult patients with well-established ileostomies were included. Ileostomy output was measured for 1 week during which three marshmallows were consumed three times daily, and for one control week where marshmallows were not eaten. There was a 2-day washout period. Patients were randomly allocated to whether the control or intervention week occurred first. In addition, a questionnaire was administered regarding patient's subjective experience of their ileostomy function. RESULTS: Thirty-one participants were recruited; 28 completed the study. There was a median reduction in ileostomy output volume of 75 ml per day during the study period (P = 0.0054, 95% confidence interval 23.4-678.3) compared with the control week. Twenty of 28 subjects (71%) experienced a reduction in their ileostomy output, two had no change and six reported an increase. During the study period, participants reported fewer ileostomy bag changes (median five per day vs six in the control period, P = 0.0255). Twenty of 28 (71%) reported that the ileostomy effluent was thicker during the study week (P = 0.023). Overall 19 (68%) participants stated they would use marshmallows in the future if they wanted to reduce or thicken their ileostomy output. CONCLUSION: Eating marshmallows leads to a small but statistically significant reduction in ileostomy output.

Economic evaluation of a group-based exercise program for falls prevention among the older community-dwelling population

BACKGROUND: Falls among older people are of growing concern globally. Implementing cost-effective strategies for their prevention is of utmost importance given the ageing population and associated potential for increased costs of fall-related injury over the next decades. The purpose of this study was to undertake a cost-utility analysis and secondary cost-effectiveness analysis from a healthcare system perspective, of a group-based exercise program compared to routine care for falls prevention in an older community-dwelling population.

METHODS: A decision analysis using a decision tree model was based on the results of a previously published randomised controlled trial with a community-dwelling population aged over 70. Measures of falls, fall-related injuries and resource use were directly obtained from trial data and supplemented by literature-based utility measures. A sub-group analysis was performed of women only. Cost estimates are reported in 2010 British Pound Sterling (GBP).

RESULTS: The ICER of GBP£51,483 per QALY for the base case analysis was well above the accepted cost-effectiveness threshold of GBP£20,000 to £30,000 per QALY, but in a sensitivity analysis with minimised program implementation the incremental cost reached GBP£25,678 per QALY. The ICER value at 95% confidence in the base case analysis was GBP£99,664 per QALY and GBP£50,549 per QALY in the lower cost analysis. Males had a 44% lower injury rate if they fell, compared to females resulting in a more favourable ICER for the women only analysis. For women only the ICER was GBP£22,986 per QALY in the base case and was below the cost-effectiveness threshold for all other variations of program implementation. The ICER value at 95% confidence was GBP£48,212 in the women only base case analysis and GBP£23,645 in the lower cost analysis. The base case incremental cost per fall averted was GBP£652 (GBP£616 for women only). A threshold analysis indicates that this exercise program cannot realistically break even.

CONCLUSIONS: The results suggest that this exercise program is cost-effective for women only. There is no evidence to support its cost-effectiveness in a group of mixed gender unless the costs of program implementation are minimal. Conservative assumptions may have underestimated the true cost-effectiveness of the program.