PubMed Health⌕ Search

Biomedical subjects

E K Harris

Publications and source records attributed to E K Harris.

18 recordsLinked to original sources
PubMed1991

Statistical criteria for separate reference intervals: race and gender groups in creatine kinase.

Previously published data confirming differences in creatine kinase (EC 2.7.3.2) among various race and gender subgroups in the Los Angeles area have been re-examined with use of recently proposed statistical criteria for defining separate reference intervals. Results indicate that one criterion may be too lenient, whereas another is clearly too restrictive in suggesting the need for separate intervals. Further experience with other analytes in both large and small population samples would be helpful.

Creatine Kinase↗
PubMed1990

On dividing reference data into subgroups to produce separate reference ranges.

We consider statistical criteria for partitioning a reference database to obtain separate reference ranges for different subpopulations. Using general formulas relating population variances, sample sizes, and the normal deviate test for the significance of the difference between two subgroup means, we show that partitioning into separate ranges produces little reduction in between-person variability, even when the differences between means are highly significant statistically. However, when there is a clear physiological basis for distinguishing between certain subgroups, simulation studies show that partitioning may be necessary to obtain reference limits that cut off the desired proportions of low and high values in each subgroup. Guidelines based on these results are provided to help decide whether separate ranges should be obtained for a given analyte.

Analysis of Variance↗
PubMed1989

Generation and application of data on biological variation in clinical chemistry.

Most clinical chemical analytes vary in a random manner around a homeostatic set point. Replicate analyses of a series of specimens collected from a group of subjects allows estimation of analytical, within and between subject components of variation. The preferred experimental procedures and statistical methods for evaluation of data and analysis of variance are described; a detailed example is provided in the Appendix. The many uses of data on biological variation in clinical chemistry are reviewed, including setting analytical goals, deciding the significance of changes in serial results from an individual, evaluating the utility of conventional population-based reference values in patient management, and other applications.

Analysis of Variance↗
PubMed1988

Proposed goals for analytical precision and accuracy in single-point diagnostic testing. Theoretical basis and comparison with data from College of American Pathologists proficiency surveys.

Expressing total analytic variance as the sum of the squares of imprecision and inaccuracy, or bias, and applying the Cotlove rule recommended by the 1976 College of American Pathologists Conference on Analytical Goals in Clinical Chemistry, namely, that analytic variance should be less than one fourth of the appropriate biological variance, I derive a rule for maximum allowable imprecision in the context of single-point diagnostic testing that takes into account the bias of the test procedure. This rule may be expressed in terms of a population-based reference range (in particular, the range of test results shown in a group of healthy individuals) and the bias of the test method. The latter is required not to exceed one eighth (0.125) of the reference range. These concepts are applied to eight common analytes for which estimates of the biases of specific methods and of within-laboratory imprecision have been published for large numbers of laboratories participating in recent College of American Pathologists proficiency surveys. Results indicate that some methods widely used in 1978 fail to meet the minimum accuracy criterion, while others show negligible bias. Even neglecting bias, more recent data show that average within-laboratory imprecision is still too high for sodium, chloride, and calcium but acceptable for potassium, glucose, cholesterol, urea, and uric acid.

Academies and Institutes↗
PubMed1979

Temporal changes in the concentrations of serum constituents in healthy men. Distributions of within-person variances and their relevance to the interpretation of differences between successive measurements.

The distributions of within-person variances in the concentrations of 10 commonly assayed serum constituents have been derived from data on 37 healthy male subjects studied at weekly intervals over a period of five months. All 10 distributions appear to be of log-normal form. The relevance of the findings to the interpretation of differences between serial measurements in a given individual is discussed. Examples are given to show how the information on within-person variances for a particular analyte, organised into a simple graph, may be used to test medical opinions on threshold values for serial changes in the concentration of this analyte in a given individual. In this way, biological variability as well as analytical error may be taken into account quantitatively when assessing the significance of a difference between two serial measurements.

Adult↗
PubMed1979

Statistical principles underlying analytic goal-setting in clinical chemistry.

The Survey programs of the College of American Pathologists (CAP) have assessed current levels of analytic variance in many biochemical measurements, and a number of clinical chemists have proposed analytic goals. The practical importance of further reductions in analytic variance depends on the specific use of the laboratory test. Three general areas of application are described: 1) surveying a population to detect disease, 2) determining whether a particular individual's level of a given analyte is above or below a predefined alarm point, 3) monitoring an individual over a period of time to detect trends. Within each of these different contests, statistical methods are proposed for judging the practical effect of improvements in current levels of analytic precision, taking into account recent estimates of biological variation within the average individual and between individuals. As might be expected, reductions in analytic variance have greatest impact in those applications where biological variance is minimal. Such reductions will generally have little effect on the efficiency of a population survey but may be extremely valuable in decision-making concerning a particular hospital patient.

Analysis of Variance↗
PubMed1978

Gallium-67 citrate scanning in the staging of idiopathic pulmonary fibrosis: Correlation and physiologic and morphologic features and bronchoalveolar lavage.

Idiopathic pulmonary fibrosis is a fatal disorder characterized by interstitial fibrosis and parenchymal inflammation. Current concepts of this disease suggest that the inflammation precedes and probably induces the fibrotic state. To evaluate the extent and relative activity of the inflammatory process, we scanned patients with idiopathic pulmonary fibrosis using gallium-67, a radionuclide known to concentrate in regions of inflammation. To quantify the amount of isotope in the lung parenchyma, the 67Ga-index was developed, a parameter derived from estimates of the size of regional pulmonary uptake, the uptake intensity, and its texture. Evaluation of 67Ga scans in 30 patients with idiopathic pulmonary fibrosis and 19 control subjects demonstrated that the 67Ga-index in the group with idiopathic pulmonary fibrosis was significantly higher (P less than 0.001) than that in the control group. When compared with lung biopsy morphologic studies in 22 patients with idiopathic pulmonary fibrosis, the 67Ga-index correlated with the degree of interstitial cellularity (P less than 0.05) and the degree of alveolar cellularity (P less than 0.005). When compared with cellular analysis of bronchoalveolar lavage fluid in 17 patients with idiopathic pulmonary fibrosis, the 67Ga-index correlated with the differential percentage of neutrophils (P less than 0.05), but not lymphocytes, eosinophils, or macrophages. These studies indicate that 67Ga accumulates in the lungs of patients with idiopathic pulmonary fibrosis and is probably associated with the active inflammatory state. The associations of the 67Ga-index with morphologic features and bronchoalveolar lavage analysis suggest that quantitative evaluation of these scans may be useful in staging the activity of idiopathic pulmonary fibrosis and following responses to therapy.

Adult↗
PubMed1977

Comparison of estimates of long-term analytical variation derived from subject samples and control serum.

Variation in the assays of uniform control serum commonly are assumed to represent day-to-day analytical variation. To test this assumption, we compared the differences between results of serum aliquots assayed immediately for 12 constituents and frozen aliquots accumulated and assayed on a single day with the results of control serum variation from the same period. One aliquot of each weekly sample was stored frozen. Eleven subjects were sampled for 12 weeks. Storage at --20 degrees C for 15 weeks had a mild destructive effect on two enzymes in serum. The control serum data revealed significant linear trends in magnesium (upwards) and alkaline phosphatase (downwards) that substantially increased the respective variances. In the other 10 constituents tested, comparison of variances indicated that long-term (weeks) variation in control serum assays is similar to the difference of variation between aliquots assayed immediately and those frozen and assayed at the same time. For these constituents, this finding justifies the use of control serum to estimate long term analytical variation.

Adult↗
PubMed1977

Intra-individual variation of some serum constituents and its relevance to population-based reference ranges.

The advent of high-capacity multi-channel analyzers allows estimation of long-term variability in serum constituents of large numbers of subjects. By frozen storage of specimens with subsequent analysis in a single machine run, long-term analytical variation may be eliminated, thus sharpening the estimates of intra-individual variation. In the present study we used the Vickers M-300 analyzer to obtain the data for such estimates from 37 male volunteers, each bled once a week for 22 weeks. Secimens were analyzed in random order to eliminate any biasing effect of analytical drift during the 4-h machine run. Ten serum constituents were measured. Storage-induced linear trends were small or negligible during the period of specimen collection. Using the ratio of average within-subject variance to the variance among subjects as a guide, serum alkaline phosphatase was found to show the greatest individuality, sodium and potassium the least. Other constitutents showed varying degrees of individuality, but for all these analytes, the usual population-based reference ranges were found to be either insensitive or irrelevant to the study of concentration changes over time within most healthy subjects. Our results generally confirmed those of smaller but comparable earlier studies.

Adult↗
PubMed1976

Some theory of reference values. II. Comparison of some statistical models of intraindividual variation in blood constituents.

Three models of intraindividual variation are reviewed, and statistical methods for distinguishing among them are discussed. Application of these methods to short series of observations from healthy individuals indicates that, in the large majority of cases, a strictly homeostatic model is appropriate for such constituents as serum calcium and magnesium. In less closely controlled variables, e.g., serum cholesterol and uric acid, a nonstationary, "rndom walk" model appears moresuitable in most cases. A more general autoregressive model, which includes the other models as extreme cases, could be used to describe all degrees of homeostatic control. This model is more complex, however, and requires at least 10 observations to yield estimates of acceptable precision. Moreover, it is sensitive to fluctuations in within-batch analytical variance. When biological variance is small relative to analytical variance, all three models yield essentially the same predicated values. To illustrate their use, these models have been applied to four short individual series of cholesterol observations showing increasing amounts of intrapersonal variation over long periods of time. I suggest that when less than 10 observations over time are available, the strictly homeostatic model and the nonstationary model be used to derive a "critical range" for assessing future changes. When longer series are available, the more general model might replace the other two for this purpose, if analytical variation has remained reasonably stable (within +/- 20% of its average value) during the period of observation. Much more experience with the use of all three models in health monitoring programs would be highly desirable.

Analysis of Variance↗
PubMed1975

Some theory of reference values. I. Stratified (categorized) normal ranges and a method for following an individual's clinical laboratory values.

The conventional population-based normal range has recently been shown to be a generally defective reference criterion for assessing individual laboratory test results. Applying a previously derived formula to published data, we find that the use of age-, sex-specific normal ranges may fail to produce a substantial improvement in sensitivity over nonspecific ranges, even when age-sex differences in mean values are statistically significant. This occurs when the difference in means is not accompanied by a sufficient reduction in the variation among individuals within a given class. Turning therefore to comparison of an individual's current measurement with his own previous value(s), I suggest a simple statistical model that leads to sequential testing of each new observation against an exponentially weighted moving average of previous results. Estimates of biological and analytical components of variance are required. The ability of this method to detect trends in very short series is explored with the aid of computer-simulated laboratory data. A sample of these data is also used to illustrate the application of these estimation and testing procedures by means of a graph.

Adult↗