Interval censored survival data and multistate compartmental models in the analysis of first appearance of Plasmodium falciparum parasites in infants

Research output: Contribution to journalJournal articleResearchpeer-review

A statistical model for interval censored data is described. Assuming a piecewise constant incidence enables us to analyse very unbalanced data in a generalized linear model. The distribution of age at first appearance of P. falciparum parasites in infants in Liberia has been estimated. A new graphical method for presentation of test results on all children was developed. In an illness--death model it is described how the proportion of undetected and detected malaria parasitemias depends on parasite rates and testing frequency. The incidence of detectable malaria parasitemia was 0.14 per month in infants under 4 months of age, and 0.60 per month in children over 4 months (p < 0.001). The congenital resistance to malaria in African infants living in a highly endemic area had largely disappeared by the age of 4 months; before this age children were partly protected.

Original languageEnglish
JournalStatistics in Medicine
Volume14
Issue number24
Pages (from-to)2727-36
Number of pages10
ISSN0277-6715
Publication statusPublished - 30 Dec 1995

    Research areas

  • Age Distribution, Bias, Humans, Immunity, Maternally-Acquired, Incidence, Infant, Infant, Newborn, Liberia/epidemiology, Linear Models, Malaria, Falciparum/immunology, Markov Chains, Mass Screening/methods, Seasons, Survival Analysis

ID: 202982319