Discordant Observations Research Articles

Detection of Discordant Content Uniformity Observations and Compendial Compliance

AbstractA discordant observation is a data point whose value is drastically different from that of the rest of the members in the data set. In the context of content uniformity experiments, however, a discordant observation arises in two ways: (i) when the value of an observation is markedly distant from that of the other data points even though it is within the required compendial range, and (ii) when the value of an observation is outside the permissible compendial range. Several statistical tests for detecting one or more discordant observations are presented. Since discordancy distorts the symmetricity of the data, several tests of symmetricity are provided. Tests for detection of group discordancy induced by discordant samples are also included. The compendial requirements are explained in statistical terms. The impact of discordant observations on compendial compliance requirements is assessed. The statistical basis of the construction of compendial limits as well as the assumptions implicit in the ...

Finding suspected causes of measurement error in multivariate environmental data

AbstractEnvironmental data are usually multivariate, with the variables conforming to some correlation structure. Occasionally, measurements which do not conform in structure or magnitude may occur in one or more variables. It is important (1) to characterize these discordancies in terms of the disturbed variables and the direction and magnitude of the anomalous error and (2) to associate each discordant observation with a specific cause of measurement error in order to prevent further mismeasurement. We describe a procedure for identifying suspected causes of discordant observations in otherwise multinormal data sets. Variables are assigned to groups, each of which is associated with a specific cause of measurement error. Discordant observations are identified with the generalized distance test or the multivariate kurtosis test. Suspected causes of measurement error are identified by repeating the tests with one of the groups of variables omitted in each analysis. The procedures are evaluated with simulated data sets having a correlation structure similar to that of a large environmental data set.

Flow Cytometric and Image Analyses of Colorectal Adenocarcinomas: A Comparative Study with Clinical Correlations

Quantitative DNA measurements were performed in 183 colorectal carcinomas by image and flow cytometric analyses of paraffin-embedded tissue. Flow cytometric analysis yielded more diploid tumors compared with image analysis, which identified more tetraploid tumors. Histogram patterns were concordant in 115 tumors (66%); the discordant cases were primarily tumors interpreted as diploid by flow cytometric analysis but were aneuploid or tetraploid by image analysis. Linear regression analysis of DNA indices of concordant samples showed good correlation but only moderate correlation for the entire group. Both techniques revealed more aneuploid tumors in the distal colon and rectum than in the proximal colon. Diploid tumors were associated with a better prognosis; however, tetraploid tumors behaved like aneuploid tumors by flow cytometric analysis but like diploid tumors by image analysis. When stratified by stage, the prognostic value of diploid tumors was seen in stages A and B disease by image analysis only and in stage C disease by flow cytometric analysis only, possibly because of the small cohort size. The S-phase fraction (mean value, 16.8% +/- 9.9%) was higher in aneuploid than in diploid tumors, but no relationship to prognosis was seen. Flow cytometric and image analyses are useful to study ploidy of colorectal carcinoma from archival material. However, important discordant observations reflecting differences in characteristics of the two techniques should be considered, depending on which technique is used.

Prediction of future observations from Pareto population in the presence of an outlier

In this article a method of calculating prediction intervals for the order statistics from the Pareto population based on samples which contain a discordant observation is presented.

Use of single-gene reassortant viruses to study the role of avian influenza A virus genes in attenuation of wild-type human influenza A virus for squirrel monkeys and adult human volunteers.

The transfer of six internal RNA segments from the avian influenza A/Mallard/New York/6750/78 (H2N2) virus reproducibly attenuates human influenza A viruses for squirrel monkeys and adult humans. To identify the avian influenza A virus genes that specify the attenuation and host range restriction of avian-human (ah) influenza A reassortant viruses (referred to as ah reassortants), we isolated six single-gene reassortant viruses (SGRs), each having a single internal RNA segment of the influenza A/Mallard/New York/6750/78 virus and seven RNA segments from the human influenza A/Los Angeles/2/87 (H3N2) wild-type virus. To assess the level of attenuation, we compared each SGR with the A/Los Angeles/2/87 wild-type virus and a 6-2 gene ah reassortant (having six internal RNA segments from the avian influenza A virus parent and two genes encoding the hemagglutinin and neuraminidase glycoproteins from the wild-type human influenza A virus) for the ability to replicate in seronegative squirrel monkeys and adult human volunteers. In monkeys and humans, replication of the 6-2 gene ah reassortant was highly restricted. In humans, the NS, M, PB2, and PB1 SGRs each replicated significantly less efficiently (P less than 0.05) than the wild-type human influenza A virus parent, suggesting that each of these genes contributes to the attenuation phenotype. In monkeys, only the NP, PB2, and possibly the M genes contributed to the attenuation phenotype. These discordant observations, particularly with regard to the NP SGR, indicate that not all genetic determinants of attenuation of influenza A viruses for humans can be identified during studies of SGRs conducted with monkeys. The PB2 and M SGRs that were attenuated in humans each exhibited a new phenotype that was not observed for either parental virus. Thus, it was not possible to determine whether avian influenza virus PB2 or M gene itself or a specific constellation of avian and human influenza A virus specified restriction of virus replication in humans.

Soluble factors produced by isolated first-trimester chorionic villi directly inhibit proliferation of T cells

Supernatants from human trophoblast cultures have been reported to be both suppressive and stimulatory of mitogen-induced T-lymphocyte transformation. These discordant observations have been attributed to methodologic differences among laboratories, but may instead relate to inadvertant cellular contamination during long-term culture. Retrieval of chorionic villi for genetic analysis has provided an opportunity to evaluate the immunoregulatory activity of the isolated villus preparation, unencumbered by extended culture techniques. Primary cytotrophoblast cultures were established from karyotypically and clinically normal pregnancies and the supernatants were recovered. Allogenic mononuclear cells, stimulated by phytohemagglutinin, were uniformly inhibited by trophoblast supernatant (mean inhibition, 34.3%). This inhibition was verified in one-way, mixed lymphocyte cultures and suppression of T cells occurred in a dose-dependent fashion over a range of trophoblast supernatant concentrations. No evidence of suppressor T-cell induction by trophoblast supernatant was noted with the use of both phenotypic studies of lymphocytes incubated with trophoblast supernatant and a functional suppressor T-cell inducer assay. We believe that the immunoregulatory properties of trophoblast supernatants identified with the use of these methods more closely approximate the paracrine function of the in situ villus.

An extra maternally derived X chromosome is deleterious to early mouse development

An extra copy of the X chromosome, unlike autosomes, exerts only minor effects on development in mammals including man and mice, because all X chromosomes except one are genetically inactivated. Contrary to this contention, we found that an additional maternally derived X (XM) chromosome, but probably not a paternally derived one (XP), consistently contributes to early death of 41,XXY and 41,XXX embryos in mice. Because of imprinted resistance to inactivation, two doses of XM remain active in the trophectoderm, and seem to be responsible for the failure in the development of the ectoplacental cone and extraembryonic ectoderm, and hence, from early embryonic death. Discordant observations in man indicating viability of XMXMXP and XMXMY individuals suggest that imprinting on the human X chromosome is either weak, unstable or erased before the initiation of X-inactivation in progenitors of extraembryonic membranes.

Geostatistical methods for detection of outliers in groundwater quality spatial fields

Detection of discordant groundwater quality observations is of primary importance. Because of the large amount of data to be analysed, methods of plausibility analysis need to be simple and not too time-consuming. Two geostatistical methods were studied, which allow the identification of outlying data points based on the pattern present in the remaining data. The two methods used, point kriging and the IRF- k method, differ in severity of assumptions. Both approaches were tested on concentration of chloride and on total hardness in groundwater and performed satisfactorily. They lend themselves well also to operational applications.

Likelihood ratio test for discordancy with slippage alternatives

A likelihood ratio test for discordancy in the sample is considered with slippage alternatives. It is shown for a wide class of univariate distributions that only the extreme observations in the sample need to be tested for discordancy. This result provides a firmer support to many commonly used discordancy tests that take only extreme observations as candidates. The problem of testing multiple discordant observations is also discussed.

Effect of phosphodiesterase inhibitors on survival of patients with chronic congestive heart failure

Controlled and uncontrolled hemodynamic and clinical studies have noted that the long-term treatment of patients with chronic heart failure with phosphodiesterase (PDE) inhibitors, such as amrinone, milrinone, enoximone and imazodan, may accelerate progression of the underlying disease and provoke serious ventricular arrhythmias. However, in an experimental model of chronic progressive left ventricular dysfunction, milrinone has been reported to reduce mortality to a degree comparable to that seen with the converting-enzyme inhibitors. These discordant observations suggest that either the deleterious hemodynamic and electrophysiologic effects of the PDE inhibitors are not translated into an adverse effect on mortality, or the animal model used to evaluate the effects of milrinone cannot be used to investigate the action of these drugs in human heart failure. Unfortunately, no trial has prospectively evaluated the effect of PDE inhibition on the survival of patients with heart failure. To address this need, the Prospective Randomized Milrinone Survival Evaluation (PROMISE Trial) has been launched in 75 to 90 clinical research centers in the United States and Canada. This study will enroll 750 patients with severe (class IV) heart failure, who have symptoms refractory to conventional therapy with digitalis, diuretics, converting-enzyme inhibitors and direct-acting vasodilators. Patients will be randomly assigned to additional treatment with either oral milrinone or placebo, and followed until death or to the conclusion of the study. The primary end point will be all-cause mortality, but the effect of milrinone on functional capacity will also be evaluated. The results of the study should define the place of PDE inhibitors in the treatment of chronic heart failure.

The effect of blood flow and diffusion impairment on pulmonary gas exchange: A computer model

An assumption of the classical three-compartment lung model is that the physiologic shunt fraction ( Q ̇ s Q ̇ t ) remains constant with changes in cardiac output ( Q ̇ t ). As a result, when Q ̇ t decreases, both the arterial O 2 tension ( P aO 2 ) and the mixed venous O 2 tension ( P vO 2 ) must also decrease. Yet observations in intact dogs with pulmonary edema and humans with the adult respiratory distress syndrome (ARDS) indicate that as Q ̇ t decreases, Q ̇ s Q ̇ t occasionally declines while the P aO 2 increases. The present study was conducted to examine these discordant observations using a computer model of pulmonary gas exchange with a diffusion impairment. A computer model of gas exchange was developed to specifically analyze the effect of changing Q ̇ t upon P aO 2 and Q ̇ s Q ̇ t in the presence of diffusion impairment. Using the classical three-compartment model of gas exchange, a diffusion impairment was inserted by a MIMIC subroutine. The results of this computer simulation indicated that in the presence of a diffusion impairment, lowering Q ̇ t caused P aO 2 to increase while Q ̇ s Q ̇ t and P vO 2 decreased. These changes indicated that a diffusion impairment might account for the experimental observations of the effects of Q ̇ t on P aO 2 in ARDS.

Techniques for dealing with discordant observations

To test any theory such as theories of motion–Newtonian or relativistic–of solar system objects, one must compare the predictions of theory with observation. But discordant observations habitually plague the reducer of astronomical data. To alleviate the baleful effects, particularly harmful when the observations are reduced by the method of least squares, of discordant data investigators almost invariably reject observations whose corresponding (0-C)'s or post-solution residuals exceed a cutoff. But techniques that are insensitive to the assumption that the observational errors are normally distributed, called robust estimation in the literature, have also been developed.

Masking Effect on Tests for Outliers in Exponential Models

Abstract The masking effect in cases of tests for outlier(s) is defined and quantified by the loss in power due to the presence of more than the anticipated number of discordant observations in the sample. This effect is illustrated in cases of some commonly used outlier tests for exponential samples—namely, Dixon-type tests and the Cochran test. A comparison between the performances of the modified Dixon-type test and the Cochran test is made. Tables for powers of these tests are presented for different sample sizes and different values of discordancy parameter. An illustrative example is presented to support the conclusion that Cochran and modified Dixon-type tests do not suffer from the masking effect in the presence of two outliers.

A viscoelastic coupling model for the cyclic deformation due to periodically repeated Earthquakes at subduction zones

Vertical displacements due to periodic reverse faulting events in an elastic plate overlying a viscoelastic (Maxwell) half space are obtained and compared with the observed deformation cycle (coseismic strain release, postseismic transients, interseismic strain accumulation) from Japan. The viscoelastic effects, including the influence of buoyant restoring forces, are obtained using the method developed by Rundle, and plate convergence and strain accumulation are incorporated following the procedure suggested by Savage. The resulting deformation cycle is compared with that of an analogous elastic half‐space dislocation model in which postearthquake effects are due to transient aseismic slip below the coseismic fault. Cyclic deformation is similar but not identical for the two models, and observations from southwest Japan suggest the superiority of the viscoelastic coupling model. In particular, inclusion of the effects of steady state flow in the asthenosphere overcomes a defect of the elastic half‐space model and results in agreement with the observed interseismic movement pattern. Several aspects of the postseismic deformation, its landward migration, and its transition to the interseismic phase of the cycle are explained as well, but the short duration of near‐trench transients relative to those observed farther inland is not matched. The success of a buried slip model in explaining early postseismic near‐trench movements and asthenospheric flow in accounting for cumulative postearthquake transient motions suggests the existence of a transition zone between lithosphere and asthenosphere whose behavior is brittle/elastic in the short term and ductile for longer‐term deformation, and such a modification may reconcile remaining discordant observations. However, reasonable variations in coupling model parameters cannot account for observed differences in the deformation cycle in other parts of Japan, and these regional differences remain unexplained.

Immunology of sarcoidosis

Sarcoidosis is a mukisystem disorder of ~ndetermined etiology and is characterized by ~:he presence of exuberant noncaseating epkhelioid ce!i granulomas in invalved organs (19). immunologic aberrations i:~ sarcoidosis have been recognized since the early part of ~:he twentieth century, w?~en Schaumann propo,Jnded the concept that sarcoidosis was a sys:emic disease rather than a cutaneous disorder, it has been recognized that sarcoidosis is associated with abnormalities of ao,n ceil-mediated and humoral immunity (5, i6, 17), and during the past decade new insights into the pathogenesis of this disease and i s imn~tmologic aberraiions have been ek:cldmea. CeH-Medimed ImaaunRy The presence of we11-developed granutomas in invoived organs of patients wkh sarcoidosis would suggest the presence of an ongoing heightened celluIar immune response. However, a plethora of studies, conducted since the early part of this century, have consistentiy demonstrated a variable degree of depression of cutaneous delayed hypersensitivity reaction to tuberculin in patients witk active sarcoidosis (5, 16, 17). This depression is noi iimited to tuberculin but is generalized and ex~ends to aiI recaii antigens, including potent natural sensitizers such as Californian keyhole limpet hemocyanin, as we11 as potent synthetic chemical sensit:lets such as 2,4-dinitrochlorobenzene. The depression of delayed cutaneous hypersensitivity is neither permanent nor absolute. This is suggested by the following observations: a) delayed cutaneous nypcrsensmv~ty ~s generally regained with resolution of sarcoidosis; b) the deve!opment of clinlcal!y manifest tuberculosis leads zo initial acquisition or reacquisition of previousiy suppressed reactivity of tuberculin; and c) ~ransient t~berculin responsiveness can be transferred passively by intravenous administration of transfer factor or by intracutaneous injection of peripherai white blood cells (probably iymphocytes) from tuberct~lin responsive individuals (25). These observations abc,~t cutaneous delayed hypersensitivity, which is an in viva method for evaluating cell-mediated immunity, led to the in vitro assessme:~t of ceilmediated immunity. Extensive quan~l~a~v~ and qualitative evaluation of lymphocytes and other effector cei!s froth peripheral blood and involved organs, particularly the lung, has demonstrated numerous abnormalities, listed in Table I (i i). These studies have provided some new insights into t]'~e pathogenesis of sarcoidosis as weil as an explanation, although incomplete, for the seemingly discordant observations of active granuloma formation in the tissues and the apparent depression of celluiar immunity. Current concepts of the pathogenesis of sarcoidosis suggest that rcdis t r ibut ion or compar tmentalization of immune effector cells, particularly the iymphocytes, occurs

Sequential carcinoembryonic antigen levels in the therapy of metastatic breast cancer: a predictor and monitor of response and relapse.

Serial measurements of plasma carcinoembryonic antigen (CEA) levels were analyzed in 42 patients with advanced breast cancer undergoing systemic chemotherapy. Pretreatment CEA levels exceeded 5 ng/ml in 22 patients, and 19 of 22 serial assays uniformly heralded tumor regression as well as subsequent tumor relapse. A significant quantitative alteration in CEA levels was established as a minimum change of 20% within 8 weeks of therapy. In 13 of 15 patients responding to chemotherapy and in all patients with CEA levels higher than 35 ng/ml, this criterion was not abrogated, and there were no discordant observations. Rising CEA levels were correlated with subsequent progression of disease in all patients with elevated baseline levels at a minimum of 8 weeks before the progression was clinically evident. In advanced breast cancer the effectiveness of therapy and the development of tumor resistance may be monitored by serial plasma CEA levels, and specific quantitative criteria should be applied.

Does Io have an ammonia atmosphere?

A number of anomalous or discordant observations of Io can be explained if this Jovian satellite has an atmosphere containing ammonia. Such an atmosphere will be frozen at the unilluminated pole during the solstices, but will evaporate at the equinoctial seasons. The ammonia atmosphere will explain (1) the posteclipse brightenings and their observed times of occurrence and nonoccurrence; (2) the observed departure from a two-layer model heating curve upon emergency from the eclipse; (3) the discordant temperatures obtained at 10 and 20 micrometers; and (4) discordant temperatures obtained at 10 and 20 micrometers during the total phase of an eclipse by Jupiter.

The microwave spectrum of Mars: An analysis

A weighted least squares fit to the best available data on the Martian microwave spectrum indicates that the brightness temperature decreases from long to short wavelengths, rather than increasing as expected from the solution of the one-dimensional equation of heat conduction. Reasonable assumptions on the ratio of electrical to thermal skin depths, on internal heat sources, on ferromagnetic materials, on radiative conduction, on compaction with depth, and on surface rpughness all fail in reproducing the deduced spectrum. A thin near-surface layer of a material with high dielectric constant and high millimeter wave absorption is needed. Since Mars exhibits marked surface overturn, a condensible material, namely liquid water, seems indicated. A layer of liquid water some tens of microns thick, on the average, localized in the top few millimeters of a Martian epilith with refractive index ⋍ 1.6 fits the microwave spectrum, and the infrared and radar data as well. The origin of such a layer of liquid water and its possible exobiological significance are discussed. The distribution of water should be nonuniform over the disk and may help explain discordant microwave observations and the anomalous variation of infrared brigthness temperature with latitude. Further millimeter wave radio and radar studies of Mars are needed.

SPECTRUM VARIATIONS IN THE WOLF-RAYET ECLIPSING BINARY HD 211853

HD 211853 is a known spectroscopic and eclipsing binary, consisting of a WN6 star and a companion which has been classified by Morgan, Code and Whitford (1955) as BO: I. A spectrographic orbit was published by Hiltner (1945); a light curve was published in 1963 (Hjellming and Hiltner 1963), together with a revision in the period. The present paper is based on low-dispersion (128 Â/mm) spectrograms obtained in 1964 and 1966 at the Kitt Peak National Observatory, and measured on a Grant comparator. A further slight change in the period was found necessary to bring the present observations into phase with those of 1944: the revised period is 6.6881 days. Hiltner's circular orbit is confirmed; however, the orbital analysis of the present observations gives a somewhat larger semiamplitude for the emission. The present work gives a value of K of 285 km/sec ( see Figure 1 ) , compared to 245 km/sec published earlier. The differences may be due to the fact that the present work uses only the He n A 4686 line, rather than averaging several emission lines of He n, N v, and N iv. The value of y from A 4686 is 75 km/sec. Hiltner's absorption-line velocity curve did not seem to follow a sine curve; in the present work it is possible to fit a sine curve with a semi-amplitude of about 100 km/sec and y of 95 km/sec (see Figure 2), except for discordant observations on September 8, 1966 (phase 0^15).

BK ERIDANI: A NORMAL RR LYRAE STAR

where (U-B) and (B-V) are the observed values near minimum light, is well correlated with Preston's ( 1959 ) AS. Just as large values of AS ( indicating weak-line stars ) are absent in variables with periods shorter than 0H4, small values of $( U-B) were found to be absent in this range with the exception of BK Eridani. In the General Catalogue of Variable Stars, BK Eri is listed with a period of 0^3540. The value of S(U-B), based on two discordant observations obtained with the Crossley reflector at Lick Observatory in 1963, was found to be (M)06. Additional observations at Lick on one night in 1964 were insufficient to reject either the period or the 8( U B). In December 1966, new observations of BK Eri were obtained at the 36-inch reflector