Vitamin A and Infant Mortality: Beyond Intention-to-Treat in a Randomized Trial

This paper investigates the effect of one dose of vitamin A on subsequent 4 month mortality in children under 6 months of age in a randomized, double-blind placebo-controlled community trial in Nepal. An earlier published intention-to-treat analysis showed no benefit, but ignored the information on actual receipt of treatment. Structural failure time models (Robins and Tsiatis, '91) use randomization based inference and incorporate compliance information which is possibly selective. The data presented here offer some new challenges for this approach: ward-based randomization induces correlation between survival outcomes; and the actual receipt of vitamin A dose is not always recorded. To tackle the problem of the clustered survival data we consider a robust version of the structural parameter vector estimator. A sensitivity analysis captures boundaries for the estimated structural parameters reflecting a range of potential values of children whose true receipt of treatment is unknown. The analysis suggests that the effect of vitamin A was beneficial in the beginning of the trial but towards the end of the trial there was a reversal of this effect.     

The NUDIST qualitative data analysis system

Most computer approaches to qualitative data analysis have concentrated on coding and retrieval of text. This paper describes a research project which set out to support a range of methods for the analysis of unstructured data, with emphasis on the building and testing of grounded theory. It resulted in software whose innovations include: a) No limit to the number of coding categories and sub-categories, and no limit to the number of times a given text passage can be coded; b) The use of separate document and indexing databases, interrelated and of unlimited flexibility; c) Comprehensive hypermedia-like browsing tools for both document and indexing databases; d) Ability to search for words and lexical patterns occurring in text and to combine this with indexing of the text; e) Ability to handle off-line textual and non-textual data as well as on-line data; f) Ability to record textual comments in indexing categories — a memoing facility for emerging ideas and categories; g) Support and exploitation of hierarchical indexing systems; h) Mechanisms for creating new indexing categories out of existing ones, relating them to the data documents, and using them for further analyses. New goals of the project are to provide a number of artificial intelligence based information structuring and reasoning facilities which can be used to aid the organization and retrieval of qualitative data, and to extend the present capabilities of the software to express and test new ideas, concepts, generalizations and hypotheses about the data.