Confidence Intervals For Relative Risks Research Articles

Risks of noncancer diseases in the cohort of personnel involved in the disposal of nuclear munitions

The aim was to estimate the risk of non-tumor disease incidence in the cohort of workers involved in the disposal of nuclear munitions as a function of non-radiational and radiological factors. Materials and Methods. The retrospective cohort study included 8078 nuclear industry employees involved in nuclear munitions disposition and covered the period from recruitment to December 31, 2018. Risk analysis for non-tumor diseases was performed based on Poisson regression using the AMFIT module of the EPICURE software. Relative risk estimates were obtained for the following factors: body mass index, blood pressure, smoking index, alcohol consumption status, external gamma and internal alpha radiation dose. 95% confidence intervals for relative risk were calculated by the maximum likelihood method. Results were considered statistically significant at p < 0.05. Results. In the study cohort of workers, 2168 cases of coronary heart disease, 2961 cases of cerebrovascular disease, 2995 cases of arterial hypertension, 742 cases of chronic obstructive pulmonary disease and 556 cases of diabetes mellitus were registered at the end of the observation period. Overweight and obesity, elevated blood pressure, smoking and alcohol abuse statistically significantly increased the risk of non-tumor disease among workers in the study cohort. The incidence of non-tumor diseases statistically significantly increased with increasing age of workers. Discussion. The results of the present study are consistent with data from numerous large-scale international studies, previous cohort studies of nuclear industry workers, and meta-analyses. Conclusion. The risk of non-tumor disease among workers involved in the disposal of nuclear munitions was statistically significantly influenced by both well-known non-radiation and radiation risk factors.

Addressing Social Needs in the Clinical Setting: Description of Needs Identified in a Quality Improvement Pilot Across 3 Community Hospital Service Areas.

Describe patient social needs in a large midwestern health care system. Health care systems are increasingly interested in addressing social needs in the clinical setting for improved health outcomes and lower costs. This retrospective cross-sectional analysis of health records data was a population health pilot project which tested a digital platform to screen and refer patients for social needs across three community hospitals serving Chicago and its South Suburbs. The analysis reports on social needs identified based on the referrals made. Electronic health record data were linked to the referral data to describe patient clinical and demographic characteristics. A total of 2909 patients had at least 1 social need identified. The mean age was 49 years ±21.7 and 65.7% were women. Only 33% of patients had 1 social need identified, indicating that if a social need is identified for patients, they most likely have social needs from other categories as well. Patients of color, specifically African American patients, were 14% more likely to have at least 3 social need categories identified (95% confidence interval of relative risk: 1.02-1.29). Financial stability was the most commonly reported social need (50.7%), followed by food and nutrition (43%). Transportation (22%) and housing (21%) were also commonly reported. This population health initiative reveals that many patients accessing health care have multiple social needs that can be identified at the point of care, and patterns of need-based referral vary by patient clinical and demographic characteristics.

New Confidence Intervals for Relative Risk of Two Correlated Proportions.

Biomedical studies, such as clinical trials, often require the comparison of measurements from two correlated tests in which each unit of observation is associated with a binary outcome of interest via relative risk. The associated confidence interval is crucial because it provides an appreciation of the spectrum of possible values, allowing for a more robust interpretation of relative risk. Of the available confidence interval methods for relative risk, the asymptotic score interval is the most widely recommended for practical use. We propose a modified score interval for relative risk and we also extend an existing nonparametric U-statistic-based confidence interval to relative risk. In addition, we theoretically prove that the original asymptotic score interval is equivalent to the constrained maximum likelihood-based interval proposed by Nam and Blackwelder. Two clinically relevant oncology trials are used to demonstrate the real-world performance of our methods. The finite sample properties of the new approaches, the current standard of practice, and other alternatives are studied via extensive simulation studies. We show that, as the strength of correlation increases, when the sample size is not too large the new score-based intervals outperform the existing intervals in terms of coverage probability. Moreover, our results indicate that the new nonparametric interval provides the coverage that most consistently meets or exceeds the nominal coverage probability.

The effects of antifungal therapy on the recurrence of aspergillus infection after pulmonary aspergilloma resection: a study protocol for a single-center, prospective, non-blind, randomized, 24-month, parallel group study

BackgroundIn recent years, the incidence of pulmonary aspergilloma has increased. The harm of aspergilloma is life-threatening massive hemoptysis, and the conventional treatment is surgical treatment. However, whether the antifungal treatment after surgery is required and the course of treatment before and after surgery are still unclear.MethodsIn this study, patients with pulmonary aspergilloma confirmed pathologically after surgery will be selected as subjects to conduct a single-center, randomized, parallel grouping, prospective, 2-year clinical study. Through regular visits, the recurrence of aspergillus infection, quality of life, lung function indicators, safety of antifungal therapy and other indicators were recorded to evaluate the recurrence risk of aspergillus infection and safety of antifungal agents. Cox proportional risk regression model was used to analyze the influencing factors of antifungal therapy on aspergillus infection recurrence after aspergillus bulbectomy. Cox multiple regression model was used for optimal model fitting, and regression coefficient (β), relative risk (RR) and 95% confidence interval of RR were calculated.DiscussionThe study will explore whether antifungal therapy could improve the quality of life, reduce the recurrence of aspergillus infection, and ultimately improve the prognosis of patients with aspergilloma. The study results will provide high-quality evidence-based medical evidence for the formulation, revision and optimization of international and domestic clinical guidelines and expert consensus on chronic aspergillus lung disease, effectively improve the clinical treatment effect of aspergilloma, and form the latest concept of diagnosis and treatment of aspergilloma.Trial registration: The trial was registered on the Chinese Clinical Trial Registry website (https://www.chictr.org.cn/showprojen.aspx?proj=33231). Registration number: ChiCTR1800019990.

Estimating Relative Risk When Observing Zero Events-Frequentist Inference and Bayesian Credibility Intervals.

Relative risk (RR) is a preferred measure for investigating associations in clinical and epidemiological studies with dichotomous outcomes. However, if the outcome of interest is rare, it frequently occurs that no events are observed in one of the comparison groups. In this case, many of the standard methods used to obtain confidence intervals (CIs) for the RRs are not feasible, even in studies with strong statistical evidence of an association. Different strategies for solving this challenge have been suggested in the literature. This paper, which uses both mathematical arguments and statistical simulations, aims to present, compare, and discuss the different statistical approaches to obtain CIs for RRs in the case of no events in one of the comparison groups. Moreover, we compare these frequentist methods with Bayesian approaches to determine credibility intervals (CrIs) for the RRs. Our results indicate that most of the suggested approaches can be used to obtain CIs (or CrIs) for RRs in the case of no events, although one-sided intervals obtained by methods based on deliberate, probabilistic considerations should be preferred over ad hoc methods. In addition, we demonstrate that Bayesian approaches can be used to obtain CrIs in these situations. Thus, it is possible to obtain statistical inference for the RR, even in studies with no events in one of the comparison groups, and CIs for the RRs should always be provided. However, it is important to note that the obtained intervals are sensitive to the method chosen in the case of small sample sizes.

Association of Financial Worry and Material Financial Risk with Short-Term Ambulatory Healthcare Utilization in a Sample of Subsidized Exchange Patients.

Financial burden can affect healthcare utilization. Few studies have assessed the short-term associations between material (debt, trouble paying rent) and psychological (worry or distress about affording future healthcare) financial risks, and subsequent outpatient and emergency healthcare use. Worry was defined as concerns about affording future healthcare. Examine whether worry about affording healthcare is associated with healthcare utilization when controlling for material risk and general anxiety DESIGN: Longitudinal observational study PARTICIPANTS: Kaiser Permanente members with exchange-based federally subsidized health insurance (n = 450, 45% response rate) MAIN MEASURES: Survey measures of financial risks (material difficulty paying for medical care and worry about affording healthcare) and general anxiety. Healthcare use (primary care, urgent care, emergency department, and outpatient specialty visits) in the 6 months following survey completion. Emergency department and primary care visits were not associated with material risk, worry about affording care, or general anxiety in individual and pooled analyses (all 95% confidence intervals (CI) for relative risk (RR) included 1). Although no individual predictor was associated with urgent care use (all 95% CIs for RR included 1), worry about affording prescriptions (relative risk (RR) = 2.01; 95% CI 1.14, 3.55) and general anxiety (RR = 0.38; 95% CI 0.15, 0.95) were significant when included in the same model, suggesting the two confounded each other. Worry about affording healthcare services was associated with fewer specialty care visits (RR = 0.40; 95% CI 0.25, 0.64) even when controlling for material risk and general anxiety, although general anxiety was also associated with more specialty care visits (RR = 1.98; 95% CI, 1.23, 3.18). Screening for both general anxiety and financial worry may assist with specialty care utilization. Identifying these concerns may provide more opportunities to assist patients. Future research should examine interventions to reduce worry about cost of care.

Comparison between Time-Limited,Venetoclax-Based and Continuous Bruton Thyrosine Kinase Inhibitors-Based Therapy in the Upfront Treatment of Chronic Lymphocytic Leukemia (CLL):a Systematic Review and Network Meta-Analysis

Background: Targeted agents (TAs) have shown impressive activity in the upfront treatment of chronic lymphocytic leukemia (CLL). However, TAs have rarely been compared in head-to-head clinical trials. With this background, a systematic literature review and network meta-analysis (NMA) was performed to estimate the relative efficacy of TAs approved by the FDA and/or EMA for upfront therapy of CLL (i.e., ibrutinib, acalabrutinib, and venetoclax). Methods: A systematic search of MEDLINE, EMBASE, BioSciences Information Service, and the Cochrane Library databases was conducted. Eligible studies consisted of randomized controlled trials (RCTs) assessing the efficacy or safety of TAs in previously untreated CLL patients. Outcomes considered were hazard ratios for progression-free survival (PFS), odds ratios for overall response rate (ORR) and adverse event rates. A given treatment was considered more effective than another one when a 95 % upper confidence interval (CI) for relative risk (RR) did not cross the value 1.0 (equivalent to a Bayesian probability for this pairwise comparison p≥97.5%). Results: Among relevant RCTs, 6 met criteria of low risk for bias according to the Cochrane Handbook for Systematic Reviews of Interventions and were selected for analysis. Three studies were excluded because they lacked a common comparator arm (i.e., RESONATE2, ALLIANCE,and ECOG-ACRIN). Three trials were suitable for the base-case network analysis (i.e., ILLUMINATE, ELEVATE-TN, and CLL14). In aggregate, these trials included1336 patients and evaluated the combination of ibrutinib-obinutuzumab (IO) (ILLUMINATE trial;n=113), venetoclax-obinutuzumab (VO) (CLL14 trial;n=216) and acalabrutinib (A) single agent (ELEVATE-TN trial; n=179). Chlorambucil-obinutuzumab (CO) was the control arm across these studies (n=504). Since results of A plus obinutuzumab (AO)(n=179) in the ELEVATE-TN trial were based on a post-hoc analysis they were not included in the NMA. In terms of PFS, fixed-effect analyses comparing VO to IO (RR 1.52[0.82-2.81]), A to IO (RR 0.87 [0.47-1.61]) and A to VO (RR 0.57[0.32-1.03]) revealed that the upper limit of 95% CI for RR did exceed the 1.0 value (Fig 1). This implies a lack of significant difference in PFS for IO, VO, and acalabrutinib. Similarly, no differences with respect to ORR were found in the indirect comparison of different TAs: VO vs. IO (RR 0.98 [0.61-1.59]), A vs. IO (RR 0.90[0.55-1.48]) and A vs. VO (RR 0.92[0.60-1.40]). The analysis of treatment side effects was performed comparing in aggregate all adverse events (AEs). No differences in the frequency of AEs was observed across different TAs: VO vs. IO (RR 1.00 [0.63-1.58]), A vs. IO (RR 1.01[0.62-1.63]) and A vs. VO (RR 1.01[0.68-1.52]). The same applied when the analysis was restricted to events with grade 3-4 toxicity: VO vs. IO (RR 1.05[0.64-1.73]), A vs. IO (RR 0.73[0.43-1.24]) and A vs. VO (RR 0.69[0.44-1.09]). Conclusions: This systematic review and network meta-analysis did not identify significant differences in PFS between BTKi-and time-limited venetoclax-based treatments in CLL upfront therapy. Further trials are needed to ascertain the pros and cons of different targeted treatments. Meanwhile, treatment selection in routine clinical practice should be based on drugs' safety, cost, availability, and treatment objectives. Figure Disclosures Molica: Gilead: Membership on an entity's Board of Directors or advisory committees; Janssen: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Roche: Membership on an entity's Board of Directors or advisory committees; Abbvie: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau.

Self-assessed preoperative level of habitual physical activity predicted postoperative complications after colorectal cancer surgery: A prospective observational cohort study

IntroductionThere is a growing interest in physical activity in relation to recovery after surgery. One important aspect of measuring recovery after surgical procedures is postoperative complications. The aim of this study was to determine if there is an association between the preoperative level of habitual physical activity and postoperative complications in patients undergoing elective surgery for colorectal cancer. Materials and methods115 patients scheduled for elective surgery due to colorectal cancer between February 2014 and September 2015 answered a questionnaire regarding physical activity and other baseline variables. Physical activity was assessed using the Saltin-Grimby physical activity level scale. Complications within 30 days after surgery were classified according to Clavien-Dindo, and the Comprehensive Complications Index (CCI) was calculated. Primary outcome was difference in CCI and key secondary outcome was risk for CCI ≥20. ResultsPhysically inactive individuals had a CCI that was 12 points higher than individuals with light activity (p = 0.002) and 17 points higher than regularly active individuals (p = 0.0004). Inactive individuals had a relative risk for a CCI ≥20 that was 65% higher than for individuals reporting light activity (95% confidence interval (CI) for relative risk (RR) = 1.1–2.5) and 338% higher than for regularly active individuals (95% CI for RR = 2.1–9.4). ConclusionSelf-assessed level of habitual physical activity before colorectal cancer surgery was associated with fewer postoperative complications measured with CCI, in a dose-response relationship.

Risk factors for return visits and rehospitalizations to the child emergency psychiatric unit: A retrospective study over 2 years at Saint-Étienne University Hospital

PurposeThis retrospective study aimed to achieve a better understanding of risk factors leading children and adolescents hospitalized in an emergency psychiatric ward to return visits, and to propose preventive devices. MethodFrom January 2, 2010 through February 29, 2012, 180 children and adolescents younger than 17 years were hospitalized in a total of 261 stays in the emergency psychiatric ward of University hospital of Saint-Étienne (France). We assessed clinical and sociodemographic characteristics of these patients and traced any of their return visits to the same unit through December 31, 2012. Risk factors for patients’ repeated visits were calculated using multivariate analysis, and the cumulative incidence of returns using the Kaplan-Meier method for censored data. We used confidence interval of relative risk, considering 0.05 to reflect significance. ResultsOver the 2 years of the study, 77 (42.8%) of the 180 patients revisited the emergency psychiatric ward; 62 (80.7%) of these required further hospitalizations. Multivariate analysis linked the patients’ psychiatric history (RR=2.5) and pursuit of vocational education (RR=4) with the risk of return. Return visits rose from 27.2% at 6 months to 41.2% at 2 years. ConclusionKnowledge of risk factors would allow implementation of secondary or tertiary preventive devices. Students could undergo early screening of psychiatric pathologies using mobile screening teams which would save money, avoid hospitalizations, and when necessary, facilitate both hospitalization and return visits.

Distribution of postpartum blood loss: modeling, estimation and application to clinical trials

BackgroundThe loss of large amounts of blood postpartum can lead to severe maternal morbidity and mortality. Understanding the nature of postpartum blood loss distribution is critical for the development of efficient analysis techniques when comparing treatments to prevent this event. When blood loss is measured, resulting in a continuous volume measure, often this variable is categorized in classes, and reduced to an indicator of volume greater than a cutoff point. This reduction of volume to classes entails a substantial loss of information. As a consequence, very large trials are needed to assess clinically important differences between treatments to prevent postpartum haemorrhage.MethodsThe authors explore the nature of postpartum blood loss distribution, assuming that the physical properties of blood loss lead to a lognormal distribution. Data from four clinical trials and one observational study are used to confirm this empirically. Estimates of probabilities of postpartum haemorrhage events ‘blood loss greater than a cutoff point’ and relative risks are obtained from the fitted lognormal distributions. Confidence intervals for relative risk are obtained by bootstrap techniques.ResultsA variant of the lognormal distribution, the three-parameter lognormal distribution, showed an excellent fit to postpartum blood loss data of the four trials and the observational study. A measurement quality assessment showed that problems of digit preference and lower limit of detection were well handled by the lognormal fit. The analysis of postpartum haemorrhage events based on a lognormal distribution improved the efficiency of the estimates. Sample size calculation for a hypothetical future trial showed that the application of this procedure permits a reduction of sample size for treatment comparison.ConclusionA variant of the lognormal distribution fitted very well postpartum blood loss data from different geographical areas, suggesting that the lognormal distribution might fit postpartum blood loss universally. An approach of analysis of postpartum haemorrhage events based on the lognormal distribution improves efficiency of estimates of probabilities and relative risk, and permits a reduction of sample size for treatment comparison.Trial registrationThis paper reports secondary analyses for trials registered at Australian New Zealand Clinical Trials Registry (ACTRN 12608000434392 and ACTRN12614000870651); and at clinicaltrials.gov (NCT00781066).

Confidence Intervals for Relative Risk by Likelihood Ratio Test

Confidence Intervals for Relative Risk by Likelihood Ratio Test

Extending the classification approach for comparing two active treatment arms to binary and time-to-event outcomes

ABSTRACTFor regulatory purposes, in a trial comparing two active treatments, a hypothesis such as noninferiority or superiority must be prespecified even when there is little known about how they compare against each other or when the objective is simply to identify the best. In this article, we extend an alternative classification methodology, the classification approach of Qu et al. (Statistics in Medicine, 30:3488-3495), to compare two active treatments when outcomes are binary and time-to-event variables. This method based on estimation approach instead of hypothesis testing can be useful when little prior information is available on which treatment has better efficacy. The entire decision space is divided into eight distinct possible outcomes based on predefined lower and upper non-inferiority margins, and the conclusion will be drawn according to the location of the confidence interval for relative risk or hazard ratio (or its logarithm transformation). We demonstrate theoretically that this method controls the misclassification rate at the specified level. We also illustrate the method by simulations and using data from a Phase 3 first-line nonsmall cell lung cancer study.

Prevention of generalized tonic-clonic seizures in refractory focal epilepsy: a meta-analysis.

Secondary generalized tonic-clonic seizures (SGTCS) are among the most severe forms of seizures, and the main risk factor for sudden unexpected death in epilepsy (SUDEP). Whether some antiepileptic drugs (AEDs) might be more efficacious than others on SGTCS in patients with drug-resistant focal epilepsy thus represents an important clinical issue for which no data are currently available. We performed a meta-analysis of randomized controlled trials of adjunctive AED in which information on efficacy outcomes (i.e., responder rate and/or frequency per 28 days relative to baseline) were available both for all seizure types and for SGTCS. The primary analysis evaluated the efficacy of AEDs on all types of seizure and on SGTCS by comparing the responder rates for AED and for placebo. Responder rate was available both for all seizure types and for SGTCS in 13 of the 72 eligible trials, evaluating 7 AEDs. Only three AEDs--lacosamide, perampanel and topiramate--showed greater efficacy than placebo. However, confidence intervals of relative risks overlapped for all AEDs but pregabalin, which demonstrated significantly lower efficacy than lacosamide, perampanel, and topiramate. Moreover, there was a nonsignificant trend toward a lower relative risk of responder rate for SGTCS than for all seizure types, which appeared related to a greater response to placebo for this outcome. Indirect comparison of AEDs using randomized placebo-controlled add-on trials does not support robust differences between AEDs to prevent SGTCS. Alternative designs for evaluation of therapeutic interventions in patients at risk for SGTCS-related complications are required.

Does Primary Closure Increase Surgical Site Infection after Intestinal Stoma Reversal? A Retrospective Cohort Study

This study was undertaken to test the veracity of the hypothesis that primary incision closure after intestinal stoma reversal in adult patients is associated with a greater risk of surgical site infection (SSI) than are open incisions. A retrospective cohort study was conducted at the Surgical Department of the Aga Khan University Hospital, Karachi, Pakistan. The study included adult patients who underwent elective loop and double-barreled intestinal stoma (ileostomy or colostomy) reversal through peristomal incisions between January 2005 and May 2011. Files were reviewed independently by two surgeons to establish main exposure (closed or open surgical sites) and outcome; i.e., SSI based on U.S. Centers for Disease Control and Prevention criteria. Sample size calculation prior to the study required 71 patients to be included in each exposure arm. Patients with closed surgical sites were relatively younger (mean 36±15 [standard deviation] years) than those with open surgical sites (41±15 years), with a male preponderance in both groups. Fifteen patients were found to have SSI: 3/71 (4.2%) in open and 12/71 (16.9%) in closed incisions. The risk of SSI in closed surgical sites was 5.8 times greater than in open sites (95% confidence interval for relative risk 1.5-22.5) after adjusting for gender, body mass index (BMI), site of stoma, malignant disease, and preoperative chemo-radiotherapy. The risk of SSI in closed incisions is greater than that in open incisions. It is suggested that incisions not be closed primarily in patients undergoing stoma reversal.

Synchronous primary cancers of the head and neck region and upper aero digestive tract: Defining high-risk patients

Patients with cancers in the head and region are at increased risk of developing synchronous primary cancers. To see the epidemiology of synchronous cancers of the head and region and identification of high-risk factors for the development of synchronous primary in the head and neck cancers. Data of head and neck cancer patients from January 2010 to December 2011 were obtained from the hospital cancer registry for retrospective analysis of patients with synchronous cancers. All synchronous malignancies were analyzed for distribution of sites, association with smoking history, stage of index head, and neck tumor and the average age of patients at presentation with synchronous cancers. The Chi-square test for association of upper aero digestive tract (UADT) and smoking and statistical formula of median for calculating the average age have been employed for analysis. Incidence of synchronous primaries has been found to be 1.33%, majority were seen at the oropharynx (39.2%) and 60.7% synchronous occurred at the esophagus, 0.81% of all head and neck cancers developed synchronous primary at the esophagus. Approximately, 65% of all synchronous primaries were in Stage III and Stage IV disease and 88.2% esophageal synchronous had Stage II disease. Association of UADT synchronous cancers with smoking is highly significant, relative risk = 1.95 95% confidence interval for relative risk 1.05-3.64 P = 0.00010981 (P < 0.05) and the average age is 62.4 years in males and 57.8 years in females. Patients who are at the high-risk for the development of synchronous primary tumors in the cancers of the head and neck region are patients with oropharyngeal carcinoma, smoking population, patients over the age of 62 years in males, and 57 years in females and in patients with higher staged index tumor.

Mechanical failure begins preferentially near resorption cavities in human vertebral cancellous bone under compression

The amount of bone turnover in the body has been implicated as a factor that can influence fracture risk and bone strength. Here we test the idea that remodeling cavities promote local tissue failure by determining if microscopic tissue damage (microdamage) caused by controlled loading in vitro is more likely to form near resorption cavities. Specimens of human vertebral cancellous bone (L4, 7 male and 2 female, age 70±10, mean±SD) were loaded in compression to the yield point, stained for microscopic tissue damage and submitted to three-dimensional fluorescent imaging using serial milling (image voxel size 0.7×0.7×5.0μm). We found the resulting damage volume per bone volume (DV/BV) was correlated with percent eroded surface (p<0.01, r2=0.65), demonstrating that whole specimen measures of resorption cavities and microdamage are related. Locations of microdamage were more than two times as likely to have a neighboring resorption cavity than randomly selected sites without microdamage (relative risk 2.39, 95% confidence interval of relative risk: 2.09–2.73), indicating a spatial association between resorption cavities and microdamage at the local level. Individual microdamage sites were 48,700 (40,100; 62,700) μm3 in size (median, 25th and 75th percentiles). That microdamage was associated with resorption cavities when measured at the whole specimen level as well as at the local level provides strong evidence that resorption cavities play a role in mechanical failure processes of cancellous bone and therefore have the potential to influence resistance to clinical fracture.

FC02-04 - Co-morbidity and mortality in Alzheimer's disease: a seven year follow-up

ObjectivesNeurodegeneration and cognitive deficits may influence the pattern of co-morbid disorders and their possible effects on mortality in late-onset Alzheimer's disease (AD). However, subjects with AD have to live long enough to experience and be diagnosed with dementia. We investigated whether the pattern of co-morbidity and its relevance for later death differed between hospitalised AD and elderly controls subjects.MethodsDiseases with a prevalence of more than 1% were compared between 634 hospitalised AD and 72244 control subjects aged above 70 years referred to the University of Birmimgham NHS Trust. Predictors of mortality within the seven year follow-up were identified using regression regression analyses. Confidence intervals of relative risks were used to compare the relevance of risk factors for later mortality between groups.ResultsSubjects with AD suffer more infections, brain diseases and neck of femur fractures than other hospitalised elderly patients. In contrast, cardiovascular diseases and diabetes mellitus were less prevalent in AD subjects in comparison with hospitalised controls. Diseases that might contribute to later mortality in AD were pneumonia, ischemic heart disease and gastroenteritis, but there were no significant differences in their impact on mortality compared to other hospitalised elderly control subjects.ConclusionsPatients with Alzheimer's disease have a different pattern of co-morbidity, but die from the same diseases as other hospitalised patients. Infections including pneumonia, and diseases that may occur secondary to neurodegeneration and confusion may need special attention in patients with Alzheimer disease who might not be able to identify or report the early symptoms.

PW01-83 - Co-morbidity and mortality in vascular dementia: a naturalistic follow-up 2000 to 2007

IntroductionVascular Dementia (VD) patients have increased co-morbidity and higher mortality in comparison to other patients. Neuro-degeneration, cognitive deficits and underlying medical conditions may influence the pattern of co-morbid disorders and their possible effects on mortality in VD.AimTo investigate whether the pattern of co-morbidity and its relevance for later death differed between hospitalised VD and elderly controls.MethodDiseases with a prevalence of more than 1% at hospital admission were compared between 341 hospitalised VD and 72244 control subjects aged above 70 years referred to the University of Birmingham Hospital, UK. Risk factors i.e. co-morbidities that were predictors of mortality within the seven year follow-up were identified using logistic regression and cox regression analyses. Confidence intervals of relative risks were used to compare the relevance of risk factors for later mortality between groups.ResultsSubjects with VD suffer more with peripheral vascular disease, atrial fibrillation, type 2 diabetes mellitus, pneumonia, ischemic stroke and urinary tract infections than other hospitalised elderly patients. In contrast, myocardial infarction was less prevalent in VD subjects in comparison with hospitalised controls. The prevalence of hypertension, ischemic heart disease, angina and heart failure was not statistically different in the two groups.ConclusionPatients with Vascular Dementia have a different pattern of co-morbidity, but die from the same diseases as other hospitalised patients. Infections including pneumonia may need special attention in patients with vascular dementia who might not be able to identify or report the early symptoms.

A comparison of methods for the construction of confidence interval for relative risk in stratified matched‐pair designs

A stratified matched-pair study is often designed for adjusting a confounding effect or effect of different trails/centers/ groups in modern medical studies. The relative risk is one of the most frequently used indices in comparing efficiency of two treatments in clinical trials. In this paper, we propose seven confidence interval estimators for the common relative risk and three simultaneous confidence interval estimators for the relative risks in stratified matched-pair designs. The performance of the proposed methods is evaluated with respect to their type I error rates, powers, coverage probabilities, and expected widths. Our empirical results show that the percentile bootstrap confidence interval and bootstrap-resampling-based Bonferroni simultaneous confidence interval behave satisfactorily for small to large sample sizes in the sense that (i) their empirical coverage probabilities can be well controlled around the pre-specified nominal confidence level with reasonably shorter confidence widths; and (ii) the empirical type I error rates of their associated test statistics are generally closer to the pre-specified nominal level with larger powers. They are hence recommended. Two real examples from clinical laboratory studies are used to illustrate the proposed methodologies.

Estimating risk difference from relative association measures in meta-analysis can infrequently pose interpretational challenges

Objective Risk difference (RD) is often estimated from relative association measures generated by meta-analysis and a particular group's baseline risk. We describe a problematic situation in using this approach. Study Design and Setting We encountered a meta-analysis in which a confidence interval (CI) of relative risk (RR) overlapped 1.0; the point estimate favored treatment A, but when we used RR and median baseline risk to calculate a CI for RD, a greater portion of the CI favored treatment B (a result that some may find counterintuitive). We then calculated 10 different RDs from recently published meta-analyses in outcomes in which CIs of RR crossed 1.0 using three methods: estimation from RR, estimation from the odds ratio, and pooling RDs across trials. Results When RD is estimated from relative measures, the counterintuitive result occurred in 2 of 10 instances. This discordance of interpretation is because of the logarithmic transformation that makes CIs of relative measures asymmetric around their point estimates. Conclusion When RD is estimated from relative association measures that are nonsignificant and this counterintuitive situation occurs, it may be more appropriate to pool RD across studies. Pooling is particularly valid when baseline risks across studies are homogeneous.