Abstract:
lactation counsellors were trained to advise mothers of partially breast-fed infants who were admitted to hospital because of diarrhoea, so that they could start exclusive breast-feeding during their hospital stay. infants (n = 250) up to 12 weeks of age were randomized to intervention and control groups. mothers in the intervention group were individually advised by the counsellors while mothers in the control group received only routine group health education. during follow-up at home by the counsellors a week later, only the mothers in the intervention group were counselled. all the mothers were evaluated for infant feeding practices at home two weeks after discharge. among the 125 mother-infant pairs in each group, 60% of mothers in the intervention group were breast-feeding exclusively at discharge compared with only 6% in the control group (p < 0.001); two weeks later, these rates rose to 75% and 8% in the intervention and control groups, respectively (p < 0.001). however, 49% of mothers in the control group reverted back to bottle-feeding compared with 12% in the intervention group (p < 0.001). thus, individual counselling had a positive impact on mothers to start exclusive breast-feeding during hospitalization and to continue the practice at home. maternal and child health facilities should include lactation counselling as an integral part of their programmes to improve infant feeding practices.

Abstract:
A rapid, low cost, accurate point-of-care (POC) device to detect influenza virus is needed for effective treatment and control of both seasonal and pandemic strains. We developed a single-use microfluidic chip that integrates solid phase extraction (SPE) and molecular amplification via a reverse transcription polymerase chain reaction (RT-PCR) to amplify influenza virus type A RNA. We demonstrated the ability of the chip to amplify influenza A RNA in human nasopharyngeal aspirate (NPA) and nasopharyngeal swab (NPS) specimens collected at two clinical sites from 2008–2010. The microfluidic test was dramatically more sensitive than two currently used rapid immunoassays and had high specificity that was essentially equivalent to the rapid assays and direct fluorescent antigen (DFA) testing. We report 96% (CI 89%,99%) sensitivity and 100% (CI 95%,100%) specificity compared to conventional (bench top) RT-PCR based on the testing of n = 146 specimens (positive predictive value = 100%(CI 94%,100%) and negative predictive value = 96%(CI 88%,98%)). These results compare well with DFA performed on samples taken during the same time period (98% (CI 91%,100%) sensitivity and 96%(CI 86%,99%) specificity compared to our gold standard testing). Rapid immunoassay tests on samples taken during the enrollment period were less reliable (49%(CI 38%,61%) sensitivity and 98%(CI 98%,100%) specificity). The microfluidic test extracted and amplified influenza A RNA directly from clinical specimens with viral loads down to 103 copies/ml in 3 h or less. The new test represents a major improvement over viral culture in terms of turn around time, over rapid immunoassay tests in terms of sensitivity, and over bench top RT-PCR and DFA in terms of ease of use and portability.

Abstract:
We show that the disagreement coefficient of certain smooth hypothesis classes is $O(m)$, where $m$ is the dimension of the hypothesis space, thereby answering a question posed in \cite{friedman09}.

Abstract:
We report a disposable and highly effective polymeric microfluidic viral sample concentration device capable of increasing the concentration of virus in a human nasopharyngeal specimen more than one order of magnitude in less than 30 min without the use of a centrifuge. The device is fabricated using 3D maskless xurography method using commercially available polymeric materials, which require no cleanroom operations. The disposable components can be fabricated and assembled in five minutes. The device can concentrate a few milliliters (mL) of influenza virus in solution from tissue culture or clinical nasopharyngeal swab specimens, via reduction of the fluid volume, to tens of microliters (mL). The performance of the device was evaluated by nucleic acid extraction from the concentrated samples, followed by a real-time quantitative polymerase chain reaction (qRT-PCR). The viral RNA concentration in each sample was increased on average over 10-fold for both cultured and patient specimens compared to the starting samples, with recovery efficiencies above 60% for all input concentrations. Highly concentrated samples in small fluid volumes can increase the downstream process speed of on-chip nucleic acid extraction, and result in improvements in the sensitivity of many diagnostic platforms that interrogate small sample volumes.

Abstract:
Given a Gaussian Markov random field, we consider the problem of selecting a subset of variables to observe which minimizes the total expected squared prediction error of the unobserved variables. We first show that finding an exact solution is NP-hard even for a restricted class of Gaussian Markov random fields, called Gaussian free fields, which arise in semi-supervised learning and computer vision. We then give a simple greedy approximation algorithm for Gaussian free fields on arbitrary graphs. Finally, we give a message passing algorithm for general Gaussian Markov random fields on bounded tree-width graphs.

Abstract:
We consider the problem of choosing a density estimate from a set of distributions F, minimizing the L1-distance to an unknown distribution (Devroye, Lugosi 2001). Devroye and Lugosi analyze two algorithms for the problem: Scheffe tournament winner and minimum distance estimate. The Scheffe tournament estimate requires fewer computations than the minimum distance estimate, but has strictly weaker guarantees than the latter. We focus on the computational aspect of density estimation. We present two algorithms, both with the same guarantee as the minimum distance estimate. The first one, a modification of the minimum distance estimate, uses the same number (quadratic in |F|) of computations as the Scheffe tournament. The second one, called ``efficient minimum loss-weight estimate,'' uses only a linear number of computations, assuming that F is preprocessed. We also give examples showing that the guarantees of the algorithms cannot be improved and explore randomized algorithms for density estimation.

Abstract:
We consider the problem of computing L1-distances between every pair ofcprobability densities from a given family. We point out that the technique of Cauchy random projections (Indyk'06) in this context turns into stochastic integrals with respect to Cauchy motion. For piecewise-linear densities these integrals can be sampled from if one can sample from the stochastic integral of the function x->(1,x). We give an explicit density function for this stochastic integral and present an efficient sampling algorithm. As a consequence we obtain an efficient algorithm to approximate the L1-distances with a small relative error. For piecewise-polynomial densities we show how to approximately sample from the distributions resulting from the stochastic integrals. This also results in an efficient algorithm to approximate the L1-distances, although our inability to get exact samples worsens the dependence on the parameters.

Abstract:
In order to counter the common perception that molecular diagnostics are too complicated to work in low resource settings, we have performed a difficult sample preparation and DNA amplification protocol using instrumentation designed to be operated without wall or battery power. In this work we have combined a nearly electricity-free nucleic acid extraction process with an electricity-free isothermal amplification assay to detect the presence of Clostridium difficile (C. difficile) DNA in the stool of infected patients. We used helicase-dependent isothermal amplification (HDA) to amplify the DNA in a low-cost, thermoplastic reaction chip heated with a pair of commercially available toe warmers, while using a simple Styrofoam insulator. DNA was extracted from known positive and negative stool samples. The DNA extraction protocol utilized an air pressure driven solid phase extraction device run using a standard bicycle pump. The simple heater setup required no electricity or battery and was capable of maintaining the temperature at 65°C±2°C for 55 min, suitable for repeatable HDA amplification. Experiments were performed to explore the adaptability of the system for use in a range of ambient conditions. When compared to a traditional centrifuge extraction protocol and a laboratory thermocycler, this disposable, no power platform achieved approximately the same lower limit of detection (1.25×10？2 pg of C. difficile DNA) while requiring much less raw material and a fraction of the lab infrastructure and cost. This proof of concept study could greatly impact the accessibility of molecular assays for applications in global health.

Abstract:
objective: to evaluate vaccine effectiveness and to assess risk factors for measles in dhaka, bangladesh. method: a case-control study, involving 198 cases with 783 age-matched neighbourhood controls and 120 measles cases with 365 age-matched hospital controls, was conducted in 1995-96 in three large hospitals in dhaka. findings: measles vaccine effectiveness was estimated at 80% (95% confidence interval (ci) = 60-90%) using neighbourhood controls; very similar results were obtained using hospital controls. visits to a health facility 7-21 days before onset of any symptoms were associated with increased risk of measles compared with neighbourhood (adjusted odds ratio (or) = 7.0, 95% ci = 4.2-11.6) or hospital (adjusted or = 1.7, 95% ci = 1.01-2.8) controls. cases were more likely than controls to come from a household where more than one child lived (adjusted or = 1.6, 95% ci = 1.1-2.5 versus neighbourhood controls; adjusted or = 1.8, 95% ci = 1.02-3.0 versus hospital controls). conclusions: to improve measles control in urban dhaka missed immunization opportunities must be reduced in all health care facilities by following who guidelines. for measles elimination, more than one dose of vaccine would be required.