Local influence in estimating equations

Local influence diagnostics based on estimating equations as the role of a gradient vector derived from any fit function are developed for repeated measures regression analysis. Our proposal generalizes tools used in other studies (Cook, 1986: Cadigan and Farrell, 2002), considering herein local influence diagnostics for a statistical model where estimation involves an estimating equation in which all observations are not necessarily independent of each other. Moreover, the measures of local influence are illustrated with some simulated data sets to assess influential observations. Applications using real data are presented. (C) 2010 Elsevier B.V. All rights reserved.

THE INFLUENCE OF HIV-1 SUBTYPES C, CRF31_BC AND B ON DISEASE PROGRESSION AND INITIAL VIROLOGIC RESPONSE TO HAART IN A SOUTHERN BRAZILIAN COHORT

Background: Although most HIV-1 infections in Brazil are due to subtype B, Southern Brazil has a high prevalence of subtype C and recombinant forms, such as CRF31_BC. This study assessed the impact of viral diversity on clinical progression in a cohort of newly diagnosed HIV-positive patients. Methods: From July/2004 to December/2005, 135 HIV-infected patients were recruited. The partial pol region was subtyped by phylogeny. A generalized estimating equation (GEE) model was used to examine the relationship between viral subtype, CD4+ T cell count and viral load levels before antiretroviral therapy. Hazard ratio (Cox regression) was used to evaluate factors associated with viral suppression (viral load < 50 copies/mL at six months). Results: Main HIV-1 subtypes included B (29.4%), C (28.2%), and CRF31_BC (23.5%). Subtypes B and C showed a similar trend in CD4+ T cell decline. Comparison of non-B (C and CRF31_BC) and B subtypes revealed no significant difference in the proportion of patients with viral suppression at six months (week 24). Higher CD4+ T cell count and lower viral load were independently associated with viral suppression. Conclusion: No significant differences were found between subtypes; however, lower viral load and higher CD4+ T cell count before therapy were associated with better response.

Comparison of quality of life after stereotactic body radiotherapy and surgery for early-stage prostate cancer

Background: As the long-term efficacy of stereotactic body radiation therapy (SBRT) becomes established and other prostate cancer treatment approaches are refined and improved, examination of quality of life (QOL) following prostate cancer treatment is critical in driving both patient and clinical treatment decisions. We present the first study to compare QOL after SBRT and radical prostatectomy, with QOL assessed at approximately the same times pre- and post-treatment and using the same validated QOL instrument. Methods: Patients with clinically localized prostate cancer were treated with either radical prostatectomy (n = 123 Spanish patients) or SBRT (n = 216 American patients). QOL was assessed using the Expanded Prostate Cancer Index Composite (EPIC) grouped into urinary, sexual, and bowel domains. For comparison purposes, SBRT EPIC data at baseline, 3 weeks, 5, 11, 24, and 36 months were compared to surgery data at baseline, 1, 6, 12, 24,and 36 months. Differences in patient characteristics between the two groups were assessed using Chi-squared tests for categorical variables and t-tests for continuous variables. Generalized estimating equation (GEE) models were constructed for each EPIC scale to account for correlation among repeated measures and used to assess the effect of treatment on QOL. Results: The largest differences in QOL occurred in the first 16 months after treatment, with larger declines following surgery in urinary and sexual QOL as compared to SBRT, and a larger decline in bowel QOL following SBRT as compared to surgery. Long-term urinary and sexual QOL declines remained clinically significantly lower for surgery patients but not for SBRT patients. Conclusions: Overall, these results may have implications for patient and physician clinical decision making which are often influenced by QOL. These differences in sexual, urinary and bowel QOL should be closely considered in selecting the right treatment, especially in evaluating the value of non-invasive treatments, such as SBRT.

Comparison of quality of life after stereotactic body radiotherapy and surgery for early-stage prostate cancer

Background: As the long-term efficacy of stereotactic body radiation therapy (SBRT) becomes established and other prostate cancer treatment approaches are refined and improved, examination of quality of life (QOL) following prostate cancer treatment is critical in driving both patient and clinical treatment decisions. We present the first study to compare QOL after SBRT and radical prostatectomy, with QOL assessed at approximately the same times pre- and post-treatment and using the same validated QOL instrument. Methods: Patients with clinically localized prostate cancer were treated with either radical prostatectomy (n = 123 Spanish patients) or SBRT (n = 216 American patients). QOL was assessed using the Expanded Prostate Cancer Index Composite (EPIC) grouped into urinary, sexual, and bowel domains. For comparison purposes, SBRT EPIC data at baseline, 3 weeks, 5, 11, 24, and 36 months were compared to surgery data at baseline, 1, 6, 12, 24,and 36 months. Differences in patient characteristics between the two groups were assessed using Chi-squared tests for categorical variables and t-tests for continuous variables. Generalized estimating equation (GEE) models were constructed for each EPIC scale to account for correlation among repeated measures and used to assess the effect of treatment on QOL. Results: The largest differences in QOL occurred in the first 1-6 months after treatment, with larger declines following surgery in urinary and sexual QOL as compared to SBRT, and a larger decline in bowel QOL following SBRT as compared to surgery. Long-term urinary and sexual QOL declines remained clinically significantly lower for surgery patients but not for SBRT patients. Conclusions: Overall, these results may have implications for patient and physician clinical decision making which are often influenced by QOL. These differences in sexual, urinary and bowel QOL should be closely considered in selecting the right treatment, especially in evaluating the value of non-invasive treatments, such as SBRT.

Predictors of remission in the treatment of major depressive disorder: real-world evidence from a 6-month prospective observational study

BACKGROUND: This study examined potential predictors of remission among patients treated for major depressive disorder (MDD) in a naturalistic clinical setting, mostly in the Middle East, East Asia, and Mexico. METHODS: Data for this post hoc analysis were taken from a 6-month prospective, noninterventional, observational study that involved 1,549 MDD patients without sexual dysfunction at baseline in 12 countries worldwide. Depression severity was measured using the Clinical Global Impression of Severity and the 16-item Quick Inventory of Depressive Symptomatology Self-Report (QIDS-SR16). Depression-related pain was measured using the pain-related items of the Somatic Symptom Inventory. Remission was defined as a QIDS-SR16 score ≤5. Generalized estimating equation regression models were used to examine baseline factors associated with remission during follow-up. RESULTS: Being from East Asia (odds ratio [OR] 0.48 versus Mexico; P<0.001), a higher level of depression severity at baseline (OR 0.77, P=0.003, for Clinical Global Impression of Severity; OR 0.92, P<0.001, for QIDS-SR16), more previous MDD episodes (OR 0.92, P=0.007), previous treatments/therapies for depression (OR 0.78, P=0.030), and having any significant psychiatric and medical comorbidity at baseline (OR 0.60, P<0.001) were negatively associated with remission, whereas being male (OR 1.29, P=0.026) and treatment with duloxetine (OR 2.38 versus selective serotonin reuptake inhibitors, P<0.001) were positively associated with remission. However, the association between Somatic Symptom Inventory pain scores and remission no longer appeared to be significant in this multiple regression (P=0.580), (P=0.008 in descriptive statistics), although it remained significant in a subgroup of patients treated with selective serotonin reuptake inhibitors (OR 0.97, P=0.023), but not in those treated with duloxetine (P=0.182). CONCLUSION: These findings are largely consistent with previous reports from the USA and Europe. They also highlight the potential mediating role of treatment with duloxetine on the negative relationship between depression-related pain and outcomes of depression.

Outcomes of cervical human papillomavirus (HPV) infections among mothers in the Finnish Family HPV Study

To understand the natural history of cervical human papillomavirus (HPV)-infections, more information is needed on their genotype-specific prevalence, acquisition, clearance, persistence and progression. This thesis is part of the prospective Finnish Family HPV study. 329 pregnant women (mean age 25.5 years) were recruited during the third trimester of pregnancy and were followed up for 6 years. The outcomes of cervical HPV infections were evaluated among all the mothers participating in the study. Generalized estimating equation (GEE)-models and Poisson regression were used to estimate the risk factors of type-specific acquisition, clearance, persistence and progression of Species 7 and 9 HPV-genotypes. Independent protective factors against incident infections were higher number of life-time sexual partners, initiation of oral contraceptive use after age 20 years and becoming pregnant during FU. Older age and negative oral HR-HPV DNA status at baseline were associated with increased clearance, whereas higher number of current sexual partners decreased the probability of clearance. Early onset of smoking, practicing oral sex and older age increased the risk of type-specific persistence, while key predictors of CIN/SIL were persistent HR-HPV, abnormal Pap smear and new sexual partners. HPV16, together with multiple-type infections were the most frequent incident genotypes, most likely to remain persistent and least likely to clear. Collectively, LR-HPV types showed shorter incidence and clearance times than HR-HPV types. In multivariate models, different predictors were associated with these main viral outcomes, and there is some tentative evidence to suggest that oral mucosa might play a role in controlling some of these outcomes.

Beta2-agonists use during pregnancy and the risk of congenital malformations

Selon les lignes directrices de traitement de l'asthme pendant la grossesse, les beta2-agonistes inhalés à courte durée d’action (SABA) sont les médicaments de choix pour tous les types d’asthme [intermittent, persistant, léger, modéré et sévère] comme médicaments de secours rapide et dans la gestion des exacerbations aiguës. D’autre part, les beta2-agonistes inhalés à longue durée d’action (LABA) sont utilisés pour les patients atteints d'asthme persistant, modéré à sévère, qui ne sont pas entièrement contrôlés par des corticostéroïdes inhalés seuls. Malgré que plusieurs études aient examinées l’association entre les LABA, les SABA et les malformations congénitales chez les nouveau-nés, les risques réels restent controversés en raison de résultats contradictoires et des difficultés inhérentes à la réalisation d'études épidémiologiques chez les femmes enceintes. L'objectif de cette étude était d'évaluer l'association entre l'exposition maternelle aux SABA et LABA pendant le premier trimestre de grossesse et le risque de malformations congénitales chez les nouveau-nés de femmes asthmatiques. Une cohorte de grossesses de femmes asthmatiques ayant accouchées entre le 1er janvier 1990 et le 31 décembre 2002 a été formée en croisant trois banques de données administratives de la province de Québec (Canada). Les issues principales de cette étude étaient les malformations congénitales majeures de touts types. Comme issues secondaires, nous avons considéré des malformations congénitales spécifiques. L'exposition principale était la prise de SABA et/ou de LABA au cours du premier trimestre de grossesse. L'exposition secondaire étudiée était le nombre moyen de doses de SABA par semaine au cours du premier trimestre. L'association entre les malformations congénitales et la prise de SABA et de LABA a été évaluée en utilisant des modèles d’équations généralisées (GEE) en ajustant pour plusieurs variables confondantes reliées à la grossesse, l’asthme de la mère et la santé de la mère et du foetus. Dans la cohorte formée de 13 117 grossesses de femmes asthmatiques, nous avons identifié 1 242 enfants avec une malformation congénitale (9,5%), dont 762 avaient une malformation majeure (5,8%). Cinquante-cinq pour cent des femmes ont utilisé des SABA et 1,3% ont utilisé des LABA pendant le premier trimestre. Les rapports de cotes ajustées (IC à 95%) pour une malformation congénitale associée à l'utilisation des SABA et des LABA étaient de 1,0 (0,9-1,2) et 1,3 (0,9-2,1), respectivement. Les résultats correspondants étaient de 0,9 (0,8-1,1) et 1,3 (0,8-2,4) pour les malformations majeures. Concernant le nombre moyen de doses de SABA par semaine, les rapports de cotes ajustées (IC à 95%) pour une malformation congénitale était de 1.1 (1.0-1.3), 1.1 (0.9-1.3), et 0.9 (0.7-1.1) pour les doses >0-3, >3-10, and >10 respectivement. Les résultats correspondants étaient de 1.0 (0.8-1.2), 0.8 (0.7-1.1), et 0.7 (0.5-1.0) pour les malformations majeures. D'autre part, des rapports de cotes (IC à 95%) statistiquement significatifs ont été observés pour les malformations cardiaques (2.4 (1.1-5.1)), les malformations d'organes génitaux (6.8 (2.6-18.1)), et d'autres malformations congénitales (3.4 (1.4 à 8.5)), en association avec les LABA pris pendant le premier trimestre. Notre étude procure des données rassurantes pour l’utilisation des SABA pendant la grossesse, ce qui est en accord avec les lignes directrices de traitement de l’asthme. Toutefois, d'autres études sont nécessaires avant de pouvoir se prononcer sur l’innocuité des LABA pendant la grossesse.

Traitement de la douleur chronique non cancéreuse à l’aide d’opioïdes : efficacité à long terme

La douleur chronique non cancéreuse (DCNC) est un phénomène complexe et des interventions multimodales qui abordent à la fois ses dimensions biologiques et psychosociales sont considérées comme l’approche optimale pour traiter ce type de désordre. La prescription d'opioïdes pour la DCNC a augmenté d’une façon fulgurante au cours des deux dernières décennies, mais les preuves supportant l'efficacité à long terme de ce type de médicament en termes de réduction de la sévérité de la douleur et d’amélioration de la qualité de vie des patients souffrant de DCNC sont manquantes. L'objectif de cette étude était d'investiguer dans un contexte de vraie vie l'efficacité à long terme des opioïdes pour réduire l’intensité et l’impact de la douleur et améliorer la qualité de vie reliée à la santé des patients souffrant de DCNC sur une période d’une année. Méthodes: Les participants à cette étude étaient 1490 patients (âge moyen = 52,37 (écart-type = 13,9); femmes = 60,9%) enrôlés dans le Registre Québec Douleur entre octobre 2008 et Avril 2011 et qui ont complété une série de questionnaires avant d'initier un traitement dans un centre multidisciplinaire tertiaire de gestion de la douleur ainsi qu’à 6 et 12 mois plus tard. Selon leur profil d'utilisation d'opioïdes (PUO), les patients ont été classés en 1) non-utilisateurs, 2) utilisateurs non persistants, et 3) utilisateurs persistants. Les données ont été analysées à l'aide du modèle d'équation d'estimation généralisée. Résultats: Chez les utilisateurs d’opioïdes, 52% en ont cessé la prise à un moment ou à un autre pendant la période de suivi. Après ajustement pour l'âge et le sexe, le PUO a prédit d’une manière significative l’intensité de la douleur ressentie en moyenne sur des périodes de 7 jours (p <0,001) ainsi que la qualité de vie physique (pQDV) dans le temps (p <0,001). Comparés aux non-utilisateurs, les utilisateurs persistants avaient des niveaux significativement plus élevés d'intensité de douleur et une moins bonne pQDV. Une interaction significative a été trouvée entre le PUO et le temps dans la prédiction de l’intensité de douleur ressentie à son maximum (p = 0,001), les utilisateurs persistants sont ceux rapportant les scores les plus élevés à travers le temps. Une interaction significative a aussi été observée entre le PUO et le type de douleur dans la prédiction de l'impact de la douleur dans diverses sphères de la vie quotidienne (p = 0,048) et de la mQDV (p = 0,042). Indépendamment du type de douleur, les utilisateurs persistants ont rapporté des scores plus élevés d'interférence de douleur ainsi qu’une moins bonne mQDV par rapport aux non-utilisateurs. Cependant, la magnitude de ces effets était de petite taille (d de Cohen <0,5), une observation qui remet en question la puissance et la signification clinique des différences observées entre ces groupes. Conclusion: Nos résultats contribuent à maintenir les doutes sur l'efficacité d’une thérapie à long terme à base d’opioïdes et remettent ainsi en question le rôle que peut jouer ce type de médicament dans l'arsenal thérapeutique pour la gestion de la DCNC.

Elasticity of substitution between capital and labor: a panel data approach

This paper estimates the elasticity of substitution of an aggregate production function. The estimating equation is derived from the steady state of a neoclassical growth model. The data comes from the PWT in which different countries face different relative prices of the investment good and exhibit different investment-output ratios. Then, using this variation we estimate the elasticity of substitution. The novelty of our approach is that we use dynamic panel data techniques, which allow us to distinguish between the short and the long run elasticity and handle a host of econometric and substantive issues. In particular we accommodate the possibility that different countries have different total factor productivities and other country specific effects and that such effects are correlated with the regressors. We also accommodate the possibility that the regressors are correlated with the error terms and that shocks to regressors are manifested in future periods. Taking all this into account our estimation resuIts suggest that the Iong run eIasticity of substitution is 0.7, which is Iower than the eIasticity that had been used in previous macro-deveIopment exercises. We show that this lower eIasticity reinforces the power of the neoclassical mo deI to expIain income differences across countries as coming from differential distortions.

Inference on Survival Data with Covariate Measurement Error - An Imputation-based Approach

We propose a new method for fitting proportional hazards models with error-prone covariates. Regression coefficients are estimated by solving an estimating equation that is the average of the partial likelihood scores based on imputed true covariates. For the purpose of imputation, a linear spline model is assumed on the baseline hazard. We discuss consistency and asymptotic normality of the resulting estimators, and propose a stochastic approximation scheme to obtain the estimates. The algorithm is easy to implement, and reduces to the ordinary Cox partial likelihood approach when the measurement error has a degenerative distribution. Simulations indicate high efficiency and robustness. We consider the special case where error-prone replicates are available on the unobserved true covariates. As expected, increasing the number of replicate for the unobserved covariates increases efficiency and reduces bias. We illustrate the practical utility of the proposed method with an Eastern Cooperative Oncology Group clinical trial where a genetic marker, c-myc expression level, is subject to measurement error.

Effect of Health Risk Assessment and Counselling on Health Behaviour and Survival in Older People: A Pragmatic Randomised Trial.

BACKGROUND Potentially avoidable risk factors continue to cause unnecessary disability and premature death in older people. Health risk assessment (HRA), a method successfully used in working-age populations, is a promising method for cost-effective health promotion and preventive care in older individuals, but the long-term effects of this approach are unknown. The objective of this study was to evaluate the effects of an innovative approach to HRA and counselling in older individuals for health behaviours, preventive care, and long-term survival. METHODS AND FINDINGS This study was a pragmatic, single-centre randomised controlled clinical trial in community-dwelling individuals aged 65 y or older registered with one of 19 primary care physician (PCP) practices in a mixed rural and urban area in Switzerland. From November 2000 to January 2002, 874 participants were randomly allocated to the intervention and 1,410 to usual care. The intervention consisted of HRA based on self-administered questionnaires and individualised computer-generated feedback reports, combined with nurse and PCP counselling over a 2-y period. Primary outcomes were health behaviours and preventive care use at 2 y and all-cause mortality at 8 y. At baseline, participants in the intervention group had a mean ± standard deviation of 6.9 ± 3.7 risk factors (including unfavourable health behaviours, health and functional impairments, and social risk factors) and 4.3 ± 1.8 deficits in recommended preventive care. At 2 y, favourable health behaviours and use of preventive care were more frequent in the intervention than in the control group (based on z-statistics from generalised estimating equation models). For example, 70% compared to 62% were physically active (odds ratio 1.43, 95% CI 1.16-1.77, p = 0.001), and 66% compared to 59% had influenza vaccinations in the past year (odds ratio 1.35, 95% CI 1.09-1.66, p = 0.005). At 8 y, based on an intention-to-treat analysis, the estimated proportion alive was 77.9% in the intervention and 72.8% in the control group, for an absolute mortality difference of 4.9% (95% CI 1.3%-8.5%, p = 0.009; based on z-test for risk difference). The hazard ratio of death comparing intervention with control was 0.79 (95% CI 0.66-0.94, p = 0.009; based on Wald test from Cox regression model), and the number needed to receive the intervention to prevent one death was 21 (95% CI 12-79). The main limitations of the study include the single-site study design, the use of a brief self-administered questionnaire for 2-y outcome data collection, the unavailability of other long-term outcome data (e.g., functional status, nursing home admissions), and the availability of long-term follow-up data on mortality for analysis only in 2014. CONCLUSIONS This is the first trial to our knowledge demonstrating that a collaborative care model of HRA in community-dwelling older people not only results in better health behaviours and increased use of recommended preventive care interventions, but also improves survival. The intervention tested in our study may serve as a model of how to implement a relatively low-cost but effective programme of disease prevention and health promotion in older individuals. TRIAL REGISTRATION International Standard Randomized Controlled Trial Number: ISRCTN 28458424.

Evaluation of a Quality Improvement Intervention for Anesthetic Management of Acute Ischemic Stroke Patients Undergoing Endovascular Therapy

Thesis (Master's)--University of Washington, 2016-06