Biomedical subjects

Robert H Lyles

Publications and source records attributed to Robert H Lyles.

At least 19 recordsLinked to original sources

PubMed2007

A practical approach to computing power for generalized linear models with nominal, count, or ordinal responses.

Data analysts facing study design questions on a regular basis could derive substantial benefit from a straightforward and unified approach to power calculations for generalized linear models. Many current proposals for dealing with binary, ordinal, or count outcomes are conceptually or computationally demanding, limited in terms of accommodating covariates, and/or have not been extensively assessed for accuracy assuming moderate sample sizes. Here, we present a simple method for estimating conditional power that requires only standard software for fitting the desired generalized linear model for a non-continuous outcome. The model is fit to an appropriate expanded data set using easily calculated weights that represent response probabilities given the assumed values of the parameters. The variance-covariance matrix resulting from this fit is then used in conjunction with an established non-central chi square approximation to the distribution of the Wald statistic. Alternatively, the model can be re-fit under the null hypothesis to approximate power based on the likelihood ratio statistic. We provide guidelines for constructing a representative expanded data set to allow close approximation of unconditional power based on the assumed joint distribution of the covariates. Relative to prior proposals, the approach proves particularly flexible for handling one or more continuous covariates without any need for discretizing. We illustrate the method for a variety of outcome types and covariate patterns, using simulations to demonstrate its accuracy for realistic sample sizes.

Computer Simulation↗

PubMed2007

Improving point predictions of random effects for subjects at high risk.

The prediction of random effects corresponding to subject-specific characteristics (e.g. means or rates of change) can be very useful in medical and epidemiologic research. At times, one may be most interested in obtaining accurate and/or precise predictions for subjects whose characteristic places them in a tail of the distribution. While the typical posterior mean predictor dominates others in terms of overall mean squared error of prediction (MSEP), its tendency to 'overshrink' has motivated research into alternatives emphasizing other criteria. Here, we specifically target MSEP within a certain region (e.g. above a known cut-off for high risk or a specified percentile of the random effect distribution), and we consider minimizing this quantity with and without constraints on overall MSEP efficiency. We use the normal-theory random intercept model to derive prediction methods with potential to yield markedly better performance for subjects in the specified region, given a well-controlled and (if desired) modest concession of overall MSEP. Criteria geared toward classification as well as overall and regional prediction unbiasedness are also provided. We evaluate the proposed techniques and illustrate them using repeated measures data on fasting blood glucose from type 2 diabetes patients. A simulation study verifies that theoretical properties and relative performances of the proposed predictors are essentially maintained when calculating them in practice based on estimated mixed linear model parameters. Straightforward extensions to incorporate covariates and additional random effects are briefly outlined.

Bias↗

PubMed2006

Inference for case-control studies when exposure status is both informatively missing and misclassified.

In case-control studies, it is common for a categorical exposure variable to be misclassified. It is also common for exposure status to be informatively missing for some individuals, in that the probability of missingness may be related to exposure. Procedures for addressing the bias due to misclassification via validation data have been extensively studied, and related methods have been proposed for dealing with informative missingness based on supplemental sampling of some of those with missing data. In this paper, we introduce study designs and analytic procedures for dealing with both problems simultaneously in a 2x2 analysis. Results based on convergence in probability illustrate that the combined effects of missingness and misclassification, even when the latter is non-differential, can lead to naïve exposure odds ratio estimates that are inflated or on the wrong side of the null. The motivating example comes from a case-control study of the association between low birth weight and the diagnosis of breast cancer later in life, where self-reported birth weight for some women is supplemented by accurate information from birth certificates.

Adult↗

PubMed2006

North American white mitochondrial haplogroups in prostate and renal cancer.

PURPOSE: While the mitochondrion is known to be a key mediator of apoptosis, there has been little inquiry into the inheritance pattern of mitochondria in patients with cancer. We compared the mtDNA haplotype in patients with prostate and renal cancer to that in controls to determine if there is an association between mitochondrial genotype and cancer. MATERIALS AND METHODS: Haplotyping was performed using polymerase chain reaction/digest identification of key polymorphic sites in the mitochondrial genome. A total of 121 and 221 white men with renal and prostate cancer, respectively, were identified following pathological confirmation of cancer, while 246 white controls were selected randomly from a bank of cadaveric organ donor DNA. Statistical analysis was performed and ORs were calculated. RESULTS: Mitochondrial haplogroup U was a highly significant risk factor for prostate and renal cancer vs controls (16.74% and 20.66% vs 9.35%, Fisher's exact test p = 0.019 and 0.005, respectively). The association remained statistically significant in renal cancer even after Bonferroni adjustment for multiple comparisons. Haplogroup U carried an OR of 1.95 for prostate cancer and an OR of 2.52 for renal cancer. CONCLUSIONS: The inheritance of mitochondrial haplogroup U is associated with an approximately 2-fold increased risk of prostate cancer and 2.5-fold increased risk of renal cancer in white North American individuals. Therefore, individuals with this mitochondrial haplotype are in a high risk group. Because mitochondrial haplogroup U is found in 9.35% of the white United States population, there are more than 20 million individuals in this high risk group.

Adolescent↗

PubMed2006

New contralateral vesicoureteral reflux following dextranomer/hyaluronic Acid implantation: incidence and identification of a high risk group.

PURPOSE: To our knowledge the incidence of NCVUR following the endoscopic treatment of VUR with Dx/HA has not been reported previously. We evaluated the outcomes in a group of patients to determine the incidence, and to attempt to identify risk factors. MATERIALS AND METHODS: A total of 126 children with primary unilateral VUR underwent unilateral Dx/HA implantation at our institutions. The incidence of NCVUR was determined by postoperative VCUG. Indications for surgery, patient age and gender, preoperative grade of VUR and volume of Dx/HA injected were assessed as possible risk factors for NCVUR. RESULTS: Of the patients 96 (76.2%) were female, and mean age was 4.8 years. The principal indications for Dx/HA implantation were persistent reflux in 56 patients (44.4%) and primary therapy in 51 (40.5%). At followup VCUG 17 patients (13.5%) had NCVUR. No variable independently appeared to influence the incidence of NCVUR. Statistical analysis suggests that females younger than 5 years have an increased incidence of NCVUR (13 of 62, or 21% vs 4 of 64, or 6.3% of the remaining patients, p = 0.016). CONCLUSIONS: NCVUR occurred in approximately 13% of our patients. Patients with higher preoperative VUR grade or a lower number of preoperative VCUGs and those undergoing treatment as primary therapy did not have an increased incidence. Girls younger than 5 years had the highest incidence of NCVUR, and initial bilateral injection may be a consideration for this group. Further effort directed at identifying the etiology and risk factors for NCVUR is needed.

Adolescent↗

PubMed2005

Assay validation for left-censored data.

In laboratory validation studies, it is often important to assess agreement between two assays, based on different techniques. Oftentimes, both assays have lower limits of detection and thus measurements are left censored. For example, in studies of Human Immunodeficiency Virus (HIV), the branched DNA (bDNA) assay was developed to quantify HIV-1 RNA concentrations in plasma. Validation of newer assays, such as the RT-PCR (reverse transcriptase polymerase chain reaction) involves assessing agreement of measurements obtained using the two techniques. Both bDNA and RT-PCR assays have lower limits of detection and thus new statistical methods are needed for assessing agreement between two left-censored variables. In this paper, we present maximum likelihood and generalized estimating equations approaches to evaluate agreement between two assays that are subject to lower limits of detection. The concordance correlation coefficient is used as an agreement index. The methodology is illustrated using HIV RNA assay data collected as part of a long-term HIV cohort study.

Computer Simulation↗

PubMed2005

Allelic imbalances of chromosomes 8p and 18q and their roles in distant relapse of early stage, node-negative breast cancer.

INTRODUCTION: Identification of breast cancer patients at risk for postoperative distant relapse is an important clinical issue. Existing pathological markers can predict disease recurrence only to a certain extent, and there is a need for more accurate predictors. METHODS: Using 'counting alleles', a novel experimental method, we determined allelic status of chromosomes 8p and 18q in a case-control study with 65 early stage, node negative, invasive ductal carcinomas (IDCs). The association between allelic imbalance (AI) of both chromosomal markers and distant relapses was examined. RESULTS: Eighty percent of tumors contained 8pAI and sixty-eight percent of tumors contained 18qAI. However, none of the tumor samples retained both chromosome 8p and 18q alleles. More importantly, tumors with 8pAI but not 18qAI were more likely to have distant relapse compared to tumors with 18qAI but not 8pAI. CONCLUSION: Our finding suggests that differential allelic loss of chromosomes 8p and 18q may represent subtypes of early stage IDC with different tumor progression behaviors.

Adult↗

PubMed2005

Estimation of the intervention effect in a non-randomized study with pre- and post-mismeasured binary responses.

In non-randomized clinical studies, the regression phenomenon can confound interpretation of the effectiveness of an intervention. The regression effect arises due to daily variation and/or misclassification of the biologic marker used in selection as well as in the assessment of the intervention effect. We consider a scenario in which the selection criterion for a subject's participation in the study is such that he/she must have a positive diagnostic test at screening. The disease status is then reassessed at the end of intervention. Thus, two repeated measurements of a binary disease outcome are available, with only selected subjects having a second measurement upon follow-up. We propose methods for estimating the change in event probability resulting from implementing the intervention while adjusting for the misclassification that produces the regression effect. We extend this approach to estimation of both the placebo and intervention effects in placebo-controlled studies designed with a misclassified binary outcome. Analyses of two biomedical studies are used for illustration.

Acoustic Impedance Tests↗

PubMed2005

Sentinel lymph node biopsy in the management of cutaneous head and neck melanoma.

Sentinel lymph node biopsy has revolutionized the surgical management of primary malignant melanoma. Most series on sentinel lymph node mapping have concentrated on extremity and truncal melanomas. The head and neck region has a rich and unpredictable lymphatic system. The use of sentinel lymph node mapping in the management of head and neck melanoma is evaluated. The authors conducted a retrospective review of patients treated for clinical stage I and stage II malignant melanoma of the head and neck with dynamic lymphoscintigraphy and gamma probe-guided sentinel lymph node biopsy. One hundred thirty-two patients (99 male patients and 33 female patients) were identified. The primary melanoma sites were the scalp (n = 54), ear (n = 14), face (n = 37), and neck (n = 27). Primary tumor staging was as follows: T1, 11; T2, 38; T3, 39; and T4, 44. Dynamic lymphoscintigraphy visualized sentinel lymph nodes in 128 patients (97 percent). In 71 cases (55 percent), a single draining nodal basin was identified, and in 57 cases there were multiple draining nodal basins (two basins, 55; three basins, two). Sentinel lymph nodes were successfully identified in 176 of 186 nodal basins (95 percent). Positive sentinel lymph nodes were identified in 22 patients (17.6 percent). Sentinel lymph node positivity by tumor staging was as follows: T2, 10.8 percent; T3, 19.4 percent; and T4, 26.8 percent. Completion lymphadenectomy revealed residual disease in seven patients (33.3 percent). Sentinel lymph node mapping for head and neck melanoma can be performed with results comparable to those of other anatomical sites.

Adolescent↗

PubMed2005

Extending McNemar's test: estimation and inference when paired binary outcome data are misclassified.

McNemar's test is popular for assessing the difference between proportions when two observations are taken on each experimental unit. It is useful under a variety of epidemiological study designs that produce correlated binary outcomes. In studies involving outcome ascertainment, cost or feasibility concerns often lead researchers to employ error-prone surrogate diagnostic tests. Assuming an available gold standard diagnostic method, we address point and confidence interval estimation of the true difference in proportions and the paired-data odds ratio by incorporating external or internal validation data. We distinguish two special cases, depending on whether it is reasonable to assume that the diagnostic test properties remain the same for both assessments (e.g., at baseline and at follow-up). Likelihood-based analysis yields closed-form estimates when validation data are external and requires numeric optimization when they are internal. The latter approach offers important advantages in terms of robustness and efficient odds ratio estimation. We consider internal validation study designs geared toward optimizing efficiency given a fixed cost allocated for measurements. Two motivating examples are presented, using gold standard and surrogate bivariate binary diagnoses of bacterial vaginosis (BV) on women participating in the HIV Epidemiology Research Study (HERS).

Biometry↗

PubMed2005

Limited health care access impairs glycemic control in low income urban African Americans with type 2 diabetes.

Limited access to health care is associated with adverse outcomes, but few studies have examined its effect on glycemic control in minority populations. Our observational cross-sectional study examined whether differences in health care access affected hemoglobin A1c (HbA1c) levels in 605 patients with diabetes (56% women; 89% African American; average age, 50 years; 95% with type 2 diabetes) initially treated at a municipal diabetes clinic. Patients who had difficulty obtaining care had higher A1c levels (9.4% vs. 8.7%; p=0.001), as did patients who used acute care facilities (9.5%; p<0.001) or who had no usual source of care (10.3%; p<0.001) compared with those who sought care at doctors' offices or clinics (8.6%). In adjusted analyses, HbA1c was higher in persons who gave a history of trouble obtaining medical care (0.57%; p=0.04), among persons who primarily used an acute care facility to receive their health care (0.49%; p=0.047), and in patients who reported not having a usual source of care (1.08%; p=0.009). Policy decisions for improving diabetes outcomes should target barriers to health care access and focus on developing programs to help high-risk populations maintain a regular place of health care.

Black or African American↗

PubMed2005

Diabetes management by residents in training in a municipal hospital primary care site (IPCAAD 2).

PURPOSE: Since diabetes is largely a primary care problem but we know little about management by residents in training--the primary care practitioners of the future--we examined surrogate outcomes reflective of their performance. METHODS: A seven-week observational study was conducted in a typical training site- a municipal hospital internal medicine resident "continuity" (primary care) clinic in a large, academic, university-affiliated training program. We evaluated control of glucose, blood pressure, and lipids; screening for proteinuria; and use of aspirin relative to national standards. RESULTS: Five hundred fifty-six (556) patients were 72% female and 97% African-American, with mean age 63 years, duration of diabetes 12 years, and BMI 34 kg/m2. Patients were managed largely with diet alone (22%) or oral agents alone (40%); 7% used oral agents and insulin in combination, and 30% insulin alone. Hemoglobin A1c (mean 8.2%) was above goal (<7.0%) in 61% of patients. Low density lipoprotein cholesterol (mean 128 mg/dL) was above goal (<100) in 76% of patients, but high density lipoprotein (mean 53 mg/dL) was at goal in 46%, and triglycerides (mean 138 mg/dL) were at goal in 85%. Diastolic pressure (mean 75 mm Hg) was at goal (<85) in 77% of patients, but systolic pressure (mean 143) was at goal (<130) in only 25% of patients. An average of only 53% of the patients had urine protein screening per 12 months, and use of aspirin was documented for only 39% of patients. CONCLUSIONS: Patients with type 2 diabetes in a typical internal medicine resident primary care clinic frequently do not achieve national standard of care goals. Since skills and attitudes developed in residency are likely to carry over into later practice, local diabetes educators may need to work with medical faculty to develop new interventions to improve postgraduate medical education in diabetes management.

Academic Medical Centers↗

PubMed2004

Expression of e-cadherin in high-risk breast cancer.

PURPOSE: E-cadherin expression is diverse, and differences in patient characteristics may produce variability in expression. Whereas some studies have indicated that downregulation of e-cadherin, associated with loss of cellular adhesiveness, was correlative with poor prognosis and metastasis, other studies have failed to confirm this. The present study uses a highly homogenous population of patients at high-risk for breast cancer, on the basis of ethnic and socio-economic status, to examine the relationship between e-cadherin and other prognostic markers in breast cancer. METHODS: Immunohistochemical staining was undertaken for estrogen (ER) and progesterone (PR) receptors, epidermal growth factor receptor 2 (Her-2), p53, vascular endothelial factor (VEGF), and hypoxia inducible factor 1alpha (HIF-1alpha) and the levels of these markers was compared to e-cadherin expression in a high-risk African-American patient population. RESULTS: E-cadherin expression persisted into the later stagers of the disease, and was strongly associated with Her-2 and HIF-1alpha expression, but not p53, ER/PR or VEGF. CONCLUSIONS: In contrast to other studies on heterogeneous populations, e-cadherin is preserved in aggressive tumors in this high-risk population. The ethnic and socio-economic risk stratification needs to be accounted for in studies correlating markers and prognosis.

Adult↗

PubMed2004

Surgical margin and Gleason score as predictors of postoperative recurrence in prostate cancer with or without chromosome 8p allelic imbalance.

BACKGROUND: Identification of prostate cancer patients at risk for postoperative disease recurrence is an important clinical issue. Existing pathological markers can predict disease recurrence only to a certain extent, and there is a need for more accurate predictors. METHODS: Using "counting alleles," a novel experimental method, we determined allelic status of chromosome 8p in 107 prostatectomy specimens. Statistical analyses examined the association between pathologic predictors (Gleason score, stage, surgical margin, etc.) and cancer recurrence in patients with and without 8p allelic imbalance (8p AI). RESULTS: 8p AI cancers were more likely to recur in the presence of a positive surgical margin, whereas recurrence of 8p retaining tumors was associated with the Gleason score, but not with the surgical margin. CONCLUSIONS: Our findings suggest that chromosome 8p allelic status affects the predictive value of "traditional" markers of prostate cancer recurrence. If confirmed by larger studies, these results may have important clinical implications.

Age Factors↗

PubMed2004

Correlation and expression of p53, HER-2, vascular endothelial growth factor (VEGF), and e-cadherin in a high-risk breast-cancer population.

BACKGROUND: Many genetic traits common to aggressive breast carcinoma have been identified; yet little is known about the interrelationships of such traits during tumor development, especially in women prone to aggressive cancer. This study examined the expression of four biological markers associated with poor prognosis at each stage of breast cancer progression in primary tumors from women of lower economic status and assessed the relationship between these markers. METHODS: Archived primary breast tumors from 77 patients were assessed by immunohistochemical analysis for expression of human epidermal growth receptor 2 (HER-2), p53, vascular endothelial growth factor (VEGF), and e-cadherin, and the relationships between the expressions of these molecules were studied. RESULTS: Twenty-two (29%) patients had advanced (stage III or IV) disease. HER-2, VEGF, e-cadherin, and p53 signal were positive for 31 (40%), 58 (75%), 63 (82%), and 37 (48%) of patients, respectively. Among the markers tested, only p53 exhibited a significant association between expression and stage of the disease ( P = 0.012). Expression of e-cadherin was positively associated with HER-2 overexpression ( P = 0.004), and high levels of HER-2 occurred with strongly positive e-cadherin tumors. Marginally significant positive associations were observed between HER-2 and p53 signal ( P = 0.06), and between disease stage and e-cadherin expression ( P = 0.08). CONCLUSION: The significant tendency toward expression of e-cadherin in conjunction with HER-2 overexpression in breast cancer is a novel finding. The association of p53 with more advanced stages of cancer emphasizes it as a key participant in metastatic processes in breast cancer. Many genetic traits common to aggressive breast carcinoma have been identified; yet little is known about the interrelationships of such traits during tumor development, especially in women prone to aggressive cancer. This study examined the expression of four biological markers associated with poor prognosis at each stage of breast cancer progression in primary tumors from women of lower economic status and assessed the relationship between these markers.

Adult↗

PubMed2004

Design and analytic considerations for single-armed studies with misclassification of a repeated binary outcome.

Due to logistics or prohibitive costs, clinical studies often rely upon a potentially misclassified binary outcome variable for assessing an intervention effect. We consider noncomparative single-armed studies that are sometimes necessary for ethical reasons, and we focus on the situation in which subjects are selected to receive the intervention contingent upon a positive screening test. Both initial misclassification at screening and a regression phenomenon impacting the error-prone follow-up outcome measure contribute to bias in the typical treatment effect estimate. We propose a study design involving the collection of internal validation data assuming the availability of a more demanding gold standard outcome measure. We pursue likelihood-based analysis and describe efficiency considerations relevant to two different treatment effect definitions. We identify four possible types of validation study observations, and discuss finding the optimal allocation into these four types in order to minimize the variance of the estimated treatment effect. The optimal allocation can be highly dependent upon whether a ratio or a difference measure is adopted to evaluate the intervention. The methods are illustrated numerically, and a real-life example motivating the proposed optimal design is provided.

Clinical Trials as Topic↗

PubMed2003

Missing data in the 2 x 2 table: patterns and likelihood-based analysis for cross-sectional studies with supplemental sampling.

Standard measures of crude association in the context of a cross-sectional study are the risk difference, relative risk and odds ratio as derived from a 2x 2 table. Most such studies are subject to missing data on disease, exposure, or both, introducing bias into the usual complete-case analysis. We describe several scenarios distinguished by the manner in which missing data arise, and for each we adjust the natural multinomial likelihood to properly account for missing data. The situations presented allow for increasing levels of generality with regard to the missing data mechanism. The final case, quite conceivable in epidemiologic studies, assumes that the probability of missing exposure depends on true exposure and disease status, as well as upon whether disease status is missing (and conversely for the probability of missing disease information). When parameters relating to the missing data process are inestimable without strong assumptions, we propose maximum likelihood analysis subsequent to collecting supplemental data in the spirit of a validation study. Analytical results give insight into the bias inherent in complete-case analysis for each scenario, and numerical results illustrate the performance of likelihood-based point and interval estimates in the most general case. Adjustment for potential confounders via stratified analysis is also discussed.

Bias↗

PubMed2003

Qualitative change in antibody responses of human immunodeficiency virus-infected individuals to pneumococcal capsular polysaccharide vaccination associated with highly active antiretroviral therapy.

Variable region gene family 3 (V(H)3) is the predominant immunoglobulin (Ig) gene family used in human antibodies to pneumococcal capsular polysaccharide (PPS). This study examined whether highly active antiretroviral therapy (HAART) restores the ability of human immunodeficiency virus (HIV)-infected individuals to generate a V(H)3-positive response to PPS. The IgM, IgG, and V(H)3 (represented by antibodies expressing the determinant recognized by the monoclonal antibody D12) responses to PPS were determined for first-time recipients of a 23-valent PPS vaccine, both receiving and not receiving HAART, and second-time vaccine recipients receiving HAART. The results showed that only the individuals receiving HAART manifested a V(H)3-(D12)-positive response to PPS, despite a similar IgG response in each group. There was also a negative correlation between HIV load and PPS response for the groups receiving HAART. These findings suggest that HAART may influence qualitative aspects of the PPS response by restoring expression of certain V(H)3 genes used in the normal PPS response.

Adult↗