Ian Painter

Quality control and quality assurance in genotypic data for genome-wide association studies

Genome‐wide scans of nucleotide variation in human subjects are providing an increasing number of replicated associations with complex disease traits. Most of the variants detected have small effects and, collectively, they account for a small fraction of the total genetic variance. Very large sample sizes are required to identify and validate findings. In this situation, even small sources of systematic or random error can cause spurious results or obscure real effects. The need for careful attention to data quality has been appreciated for some time in this field, and a number of strategies for quality control and quality assurance (QC/QA) have been developed. Here we extend these methods and describe a system of QC/QA for genotypic data in genome‐wide association studies (GWAS). This system includes some new approaches that (1) combine analysis of allelic probe intensities and called genotypes to distinguish gender misidentification from sex chromosome aberrations, (2) detect autosomal chromosome aberrations that may affect genotype calling accuracy, (3) infer DNA sample quality from relatedness and allelic intensities, (4) use duplicate concordance to infer SNP quality, (5) detect genotyping artifacts from dependence of Hardy‐Weinberg equilibrium test P‐values on allelic frequency, and (6) demonstrate sensitivity of principal components analysis to SNP selection. The methods are illustrated with examples from the “Gene Environment Association Studies” (GENEVA) program. The results suggest several recommendations for QC/QA in the design and execution of GWAS. Genet. Epidemiol. 34: 591–602, 2010.

Bayesian statistics in genetics: a guide for the uninitiated

Statistical analyses are used in many fields of genetic research. Most geneticists are taught classical statistics, which includes hypothesis testing, estimation and the construction of confidence intervals; this framework has proved more than satisfactory in many ways. What does a Bayesian framework have to offer geneticists? Its utility lies in offering a more direct approach to some questions and the incorporation of prior information. It can also provide a more straightforward interpretation of results. The utility of a Bayesian perspective, especially for complex problems, is becoming increasingly clear to the statistics community; geneticists are also finding this framework useful and are increasingly utilizing the power of this approach.

A simulation study comparing aberration detection algorithms for syndromic surveillance

BackgroundThe usefulness of syndromic surveillance for early outbreak detection depends in part on effective statistical aberration detection. However, few published studies have compared different detection algorithms on identical data. In the largest simulation study conducted to date, we compared the performance of six aberration detection algorithms on simulated outbreaks superimposed on authentic syndromic surveillance data.MethodsWe compared three control-chart-based statistics, two exponential weighted moving averages, and a generalized linear model. We simulated 310 unique outbreak signals, and added these to actual daily counts of four syndromes monitored by Public Health – Seattle and King Countys syndromic surveillance system. We compared the sensitivity of the six algorithms at detecting these simulated outbreaks at a fixed alert rate of 0.01.ResultsStratified by baseline or by outbreak distribution, duration, or size, the generalized linear model was more sensitive than the other algorithms and detected 54% (95% CI = 52%–56%) of the simulated epidemics when run at an alert rate of 0.01. However, all of the algorithms had poor sensitivity, particularly for outbreaks that did not begin with a surge of cases.ConclusionWhen tested on county-level data aggregated across age groups, these algorithms often did not perform well in detecting signals other than large, rapid increases in case counts relative to baseline levels.

GWASTools: an R/Bioconductor package for quality control and analysis of Genome-Wide Association Studies

GWASTools is an R/Bioconductor package for quality control and analysis of genome-wide association studies (GWAS). GWASTools brings the interactive capability and extensive statistical libraries of R to GWAS. Data are stored in NetCDF format to accommodate extremely large datasets that cannot fit within Rs memory limits. The documentation includes instructions for converting data from multiple formats, including variants called from sequencing. GWASTools provides a convenient interface for linking genotypes and intensity data with sample and single nucleotide polymorphism annotation.

Public health communications and alert fatigue

BackgroundHealth care providers play a significant role in large scale health emergency planning, detection, response, recovery and communication with the public. The effectiveness of health care providers in emergency preparedness and response roles depends, in part, on public health agencies communicating information in a way that maximizes the likelihood that the message is delivered, received, deemed credible and, when appropriate, acted on. However, during an emergency, health care providers can become inundated with alerts and advisories through numerous national, state, local and professional communication channels. We conducted an alert fatigue study as a sub-study of a larger randomized controlled trial which aimed to identify the most effective methods of communicating public health messages between public health agencies and providers. We report an analysis of the effects of public health message volume/frequency on recall of specific message content and effect of rate of message communications on health care provider alert fatigue.MethodsHealth care providers enrolled in the larger study (n=528) were randomized to receive public health messages via email, fax, short message service (SMS or cell phone text messaging) or to a control group that did not receive messages. For 12 months, study messages based on real events of public health significance were sent quarterly with follow-up telephone interviews regarding message receipt and topic recall conducted 5–10 days after the message delivery date. During a pandemic when numerous messages are sent, alert fatigue may impact ability to recall whether a specific message has been received due to the “noise” created by the higher number of messages. To determine the impact of “noise” when study messages were sent, we compared health care provider recall of the study message topic to the number of local public health messages sent to health care providers.ResultsWe calculated the mean number of messages that each provider received from local public health during the time period around each study message and provider recall of study message content. We found that recall rates were inversely proportional to the mean number of messages received per week: Every increase of one local public health message per week resulted in a statistically significant 41.2% decrease (p < 0.01), 95% CI [0.39, .87] in the odds of recalling the content of the study message.ConclusionsTo our knowledge, this is the first study to document the effects of alert fatigue on health care providers’ recall of information. Our results suggest that information delivered too frequently and/or repetitively through numerous communication channels may have a negative effect on the ability of health care providers to effectively recall emergency information. Keeping health care providers and other first-line responders informed during an emergency is critical. Better coordination between organizations disseminating alerts, advisories and other messages may improve the ability of health care providers to recall public health emergency messages, potentially impacting effective response to public health emergency messages.

Changes to DA-CPR instructions: can we reduce time to first compression and improve quality of bystander CPR?

INTRODUCTION Dispatcher-assisted CPR (DA-CPR) can increase rates of bystander CPR, survival, and quality of life following cardiac arrest. Dispatcher protocols designed to improve rapid recognition of arrest and coach CPR may increase survival by (1) reducing preventable time delays to start of chest compressions and (2) improving the quality of bystander CPR. METHODS We conducted a randomized controlled trial comparing a simplified DA CPR script to a conventional DA CPR script in a manikin cardiac arrest simulation with lay participants. The primary outcomes measured were the time interval from call receipt to the first chest compression and the core metrics of chest compression (depth, rate, release, and compression fraction). CPR was measured using a recording manikin for the first 3 min of participant CPR. RESULTS Of the 75 participants, 39 were randomized to the simplified instructions and 36 were randomized to the conventional instructions. The interval from call receipt to first compression was 99 s using the simplified script and 124 s using the conventional script for a difference of 24s (p<0.01). Although hand position was judged to be correct more often in the conventional instruction group (88% versus 63%, p<0.01), compression depth was an average 7 mm deeper among those receiving the simplified CPR script (32 mm versus 25 mm, p<0.05). No statistically significant differences were detected between the two instruction groups for compression rate, complete release, number of hands-off periods, or compression fraction. DISCUSSION Simplified DA-CPR instructions to lay callers in simulated cardiac arrest settings resulted in significant reductions in time to first compression and improvements in compression depth. These results suggest an important opportunity to improve DA CPR instructions to reduce delays and improve CPR quality.

The Effect of Language Barriers on Dispatching EMS Response

Abstract Objective. The objective of this study was to investigate the effect of language barriers during medical 9-1-1 calls, on the time to dispatch and level of medical aid (Basic or Advanced Life Support). Methods. All 9-1-1 medical calls to two large call centers during one week for each of the months of August, October, December 2010 and February 2011, were reviewed for a notation of language barrier (LB). Non-language barrier calls were identified from the same time period such that there were an equal proportion of LB and non-LB calls by dispatch code and dispatcher. A total of 272 language barrier calls were identified. The computer-assisted dispatch (CAD) reports for the LB and non-LB calls were abstracted by research staff using a standard form, including: Start time of call, time to dispatch of BLS, time to dispatch of ALS, dispatch code, interpretation service use, on-scene upgrade to ALS, and on-scene downgrade to BLS. 9-1-1 recordings were abstracted for LB calls only to obtain information about use of interpreter services. Difference between LB and English speakers in time to assignment of BLS and ALS was examined using linear mixed effects models with log time as the outcome; language barrier, call center and dispatch code as fixed effects and dispatcher as a random effect. Results. The effect of language barrier on time to BLS assignment was, on average, 33% longer (p < 0.001) and time to ALS assignment 43% longer (P = 0.008). A majority of the effect was due to the effect of interpreter use, which increased time to BLS by 82% and 125% for ALS, when compared to non-language barrier calls. Data from the 9-1-1 recordings showed an average of 49 seconds between connecting to the service operator and connecting to the language interpreter. Language barrier calls were more likely to be up- and down-graded, only statistically significantly so for on-scene downgrades. Conclusion. Language barriers increase time to dispatch and the accuracy of the level of aid dispatched during medical emergency calls. Decreasing the time to connecting to an actual interpreter when using an interpretation service could minimize existing delays.

Effects of bystander CPR following out-of-hospital cardiac arrest on hospital costs and long-term survival

BACKGROUND Bystander cardiopulmonary resuscitation (CPR) is associated with a greater likelihood of survival to hospital discharge after out-of-hospital cardiac arrest (OHCA). However the long-term survival benefits in relationship to cost have not been well-studied. We evaluated bystander CPR, hospital-based costs, and long-term survival following OHCA in order to assess the potential cost-effectiveness of bystander CPR. PATIENTS AND METHODS We conducted a retrospective cohort study of consecutive EMS-treated OHCA patients >=12years who arrested prior to EMS arrival and outside a nursing facility between 2001 and 2010 in greater King County, WA. Utstein-style information was obtained from the EMS registry, including 5-year survival. Costs from the OHCA hospitalization were obtained from the Washington State Comprehensive Hospital Abstract Reporting System. Cost effectiveness was based on hospital costs divided by quality-adjusted life years (QALYs) for a 5-year follow-up window. RESULTS Of the 4448 eligible patients, 18.5% (n=824) were discharged alive from hospital and 12.1% (n=539) were alive at 5 years. Five-year survival was higher in patients who received bystander CPR (14.3% vs. 8.7%, p<0.001) translating to an average 0.09 QALYs associated with bystander CPR. The average (SD) total cost of the initial acute care hospitalization was USD 19,961 (40,498) for all admitted patients and USD 75,175 (52,276) for patients alive at year 5. The incremental cost-effectiveness ratio associated with bystander CPR was USD 48,044 per QALY. CONCLUSION Based on this population-based investigation, bystander CPR was positively associated with long-term survival and appears cost-effective.

The Role of Self-Efficacy in Communication and Emergency Response in Chinese Limited English Proficiency (LEP) Populations:

Background. Failure to engage in emergency preparedness, response, and recovery contributes to the differential outcome experienced by limited English proficiency (LEP) populations. Little is known about how psychosocial factors influence LEP individuals’ perception of emergency and their process of understanding, collecting, and synthesizing information. The purpose of this exploratory study is to understand how LEP conceptualize an emergency situation to determine when help is needed. Methods. The authors conducted 4 focus groups with 36 adult Chinese LEP speakers living in Seattle. All discussions were audio-taped, translated, and transcribed. Coded text passages were entered into Atlas.ti for data management and model generation. Results. Perception of an emergency situation affects LEP individual’s ability to manage the crisis. Self-efficacy may be an important psychological variable that positively shapes an individual’s response to an emergency situation by improving their confidence to handle the crisis and ability to connect to resources. Response to emergency resulting from this series of information gathering, synthesis, and utilization may not always result in a positive outcome. Discussion. Self-efficacy in risk communication messages should be included to engage LEPs in emergency preparedness. Effective communication can increase LEPs’ awareness of emergency situations and connecting LEP individuals with existing community resources may enhance LEPs’ level of self-efficacy in emergencies.

Biosurveillance, Case Reporting, and Decision Support: Public Health Interactions with a Health Information Exchange

This paper describes support for three public health practice domains in demonstrations of a model health information exchange (HIE): biosurveillance, case reporting, and communication from public health to providers through integrated decision support. The model HIE implements interoperability through the use of existing semantic and syntactic standards specified as part of Integration Profiles to support specific data transfer use cases. We implemented these profiles in several public health applications using a service-orientated architecture approach. Methods were validated for each public health domain in national showcase demonstrations. We believe that this work has implications for the integration of public health functions into any HIE, regardless of its architecture, because our informatics methods support a distributed environment. This approach may be extended to strengthen development of the Public Health Grid, a project currently being led by the Centers for Disease Control and Prevention.