Important: The POPLINE website will retire on September 1, 2019. Click here to read about the transition.

Your search found 2 Results

  1. 1
    069874

    Use and interpretation of pediatric anthropometry in epidemiologic studies.

    Sullivan K; Gorstein J; Trowbridge F; Yip R

    [Unpublished] 1991. Presented at the Society for Epidemiologic Research 24th Annual Meeting, Buffalo, New York, June 11-14, 1991. 12, [18] p.

    Health workers use anthropometry to determine the nutritional status of children. The accepted international growth reference curves provide the bases for the indices which include weight for height (W/H), height for age (H/A) and weight for age (W/A). Health workers must interpret these indices with caution, however. For example, W/H and H/A represent different physiological and biological processes while W/A combines the 2 processes. Further Z-scores, percentiles, or percent of median may be used as the scale for the indices and each scale has different statistical features. Specifically, Z-scores and percentiles acknowledge smoothed normalized distributions around the median, but the percent-of-median ignores the distribution around the median. Some researchers suggest using Z-scores rather than percentiles or percent-of-median since statisticians can interpret them more clearly and can calculate the proportion of children in the reference population who fall above or below a cut off point more easily. This cutoff should be only used to screen children who are likely to be malnourished since not all children below a cutoff are indeed malnourished. Some researchers have identified a leading limitation of the CDC/WHO based indices. A disjunction exists where the 2 smoothed based curves based on a population of <36 month old children from Ohio (longitudinal data) and another population of 2-18 year old children (cross sectional health surveys) meet. Further there is a reduction in age specific prevalences at 24 months. Thus some researchers recommend that anthropometry data be presented on an age specific basis, if age information is accurate. They further suggest that, if comparing data from different geographic areas, researchers should standardize age to have a summary measure. If age is not known the W/H summary measure should include 2 groups: <85 cm and =or+ 85 cm.
    Add to my documents.
  2. 2
    065543

    AIDS in Africa [letter]

    Lauwers M; Ndoluvualu-Namata N; Goubau P; Desmyter J

    NEW ENGLAND JOURNAL OF MEDICINE. 1991 Mar 21; 324(12):848.

    Dr. Goodgame pleads for more openness in discussing the diagnosis of AIDS with the patient. On the other hand, he believes testing for HIV antibodies is largely unnecessary for diagnosis in Uganda, which has 1 of the highest prevalences in the world. Given, however, that the WHO clinical AIDS definition has a positive predictive value of 73% in Ugandan patients (or 83% if cough due to tuberculosis is excluded), 27% of patients in whom there is a clinical suspicion will be erroneously told they have AIDS--"dreadful and at times almost unbearable" news. In other parts of Africa with a lower prevalence this may be even less acceptable. In Gemena, northern Zaire, we evaluated the WHO clinical Aids definition, as modified by Colebunders et al., in 166 patients in 1988-1989. The positive predictive value was 61% (67% if patients with tuberculosis were excluded). This means a wrong diagnosis of AIDS in 1 of every 3 patients. The HIV seroprevalence in this population was 7.9%, as measured in a group of 340 healthy pregnant women. Another problem is the lack of sensitivity of the clinical case definition of AIDS, leading to the possible exclusion of 30-46% of African patients with HIV-related disease in the absence of testing for HIV antibodies. Many patients with AIDS would thus escape detection until they were ill enough to meet the diagnostic criteria. If a standard of care for patients with AIDS is to be achieved in Africa, as Dr. Goodgame proposes, correctly identifying the patients early in the course of the disease is necessary, and we do not believe this is possible without laboratory confirmation. We are aware of the problems that may arise when anti-HIV testing is introduced, and the questions raised (e.g. Who will be tested? What will be done when a positive result is found?) should be thoroughly discussed with the local health team before the test is introduced. In addition, screening of blood donors should have absolute priority over diagnostic testing if a choice has to be made because of the dearth of reagents. (full text)
    Add to my documents.