Data Quality Checks Research Articles

Population-based cancer survival in the United States: Data, quality control, and statistical methods.

Robust comparisons of population-based cancer survival estimates require tight adherence to the study protocol, standardized quality control, appropriate life tables of background mortality, and centralized analysis. The CONCORD program established worldwide surveillance of population-based cancer survival in 2015, analyzing individual data on 26 million patients (including 10 million US patients) diagnosed between 1995 and 2009 with 1 of 10 common malignancies. In this Cancer supplement, we analyzed data from 37 state cancer registries that participated in the second cycle of the CONCORD program (CONCORD-2), covering approximately 80% of the US population. Data quality checks were performed in 3 consecutive phases: protocol adherence, exclusions, and editorial checks. One-, 3-, and 5-year age-standardized net survival was estimated using the Pohar Perme estimator and state- and race-specific life tables of all-cause mortality for each year. The cohort approach was adopted for patients diagnosed between 2001 and 2003, and the complete approach for patients diagnosed between 2004 and 2009. Articles in this supplement report population coverage, data quality indicators, and age-standardized 5-year net survival by state, race, and stage at diagnosis. Examples of tables, bar charts, and funnel plots are provided in this article. Population-based cancer survival is a key measure of the overall effectiveness of services in providing equitable health care. The high quality of US cancer registry data, 80% population coverage, and use of an unbiased net survival estimator ensure that the survival trends reported in this supplement are robustly comparable by race and state. The results can be used by policymakers to identify and address inequities in cancer survival in each state and for the United States nationally. Cancer 2017;123:4982-93. Published 2017. This article is a U.S. Government work and is in the public domain in the USA.

Plug-in style results data quality checking system for mine remote sensing monitoring based on AE

Plug-in style results data quality checking system for mine remote sensing monitoring based on AE

Fog-based automatic true north detection for absolute magnetic declination measurement

Abstract. Absolute magnetic measurements are of great importance in magnetic observatories. They allow not only instrument calibration but also data quality checking. They require the vertical and the geographic or true north as reference directions, usually determined by means of a level and by pointing an azimuth mark, respectively. We present here a novel system able to measure the direction of the magnetic field and of the vertical and true north. A design of a north seeker is proposed taking into account sensor bias as well as misalignment errors. Different methods are derived from this model and measurement results are presented. A measurement test at high latitude is also shown.

National Assessment of Data Quality and Associated Systems-Level Factors in Malawi.

Routine health data can guide health systems improvements, but poor quality of these data hinders use. To address concerns about data quality in Malawi, the Ministry of Health and National Statistical Office conducted a data quality assessment (DQA) in July 2016 to identify systems-level factors that could be improved. We used 2-stage stratified random sampling methods to select health centers and hospitals under Ministry of Health auspices, included those managed by faith-based entities, for this DQA. Dispensaries, village clinics, police and military facilities, tertiary-level hospitals, and private facilities were excluded. We reviewed client registers and monthly reports to verify availability, completeness, and accuracy of data in 4 service areas: antenatal care (ANC), family planning, HIV testing and counseling, and acute respiratory infection (ARI). We also conducted interviews with facility and district personnel to assess health management information system (HMIS) functioning and systems-level factors that may be associated with data quality. We compared systems and quality factors by facility characteristics using 2-sample t tests with Welch's approximation, and calculated verification ratios comparing total entries in registers to totals from summarized reports. We selected 16 hospitals (of 113 total in Malawi), 90 health centers (of 466), and 16 district health offices (of 28) in 16 of Malawi's 28 districts. Nearly all registers were available and complete in health centers and district hospitals, but data quality varied across service areas; median verification ratios comparing register and report totals at health centers ranged from 0.78 (interquartile range [IQR]: 0.25, 1.07) for ARI and 0.99 (IQR: 0.82, 1.36) for family planning to 1.00 (IQR: 0.96, 1.00) for HIV testing and counseling and 1.00 (IQR: 0.80, 1.23) for ANC. More than half (60%) of facilities reported receiving a documented supervisory visit for HMIS in the prior 6 months. A recent supervision visit was associated with better availability of data (P=.05), but regular district- or central-level supervision was not. Use of data by the facility to track performance toward targets was associated with both improved availability (P=.04) and completeness of data (P=.02). Half of facilities had a full-time statistical clerk, but their presence did not improve the availability or completeness of data (P=.39 and P=.69, respectively). Findings indicate both strengths and weaknesses in Malawi's HMIS performance, with key weaknesses including infrequent data quality checks and unreliable supervision. Efforts to strengthen HMIS in low- and middle-income countries should be informed by similar assessments.

Health Science Research with Primary Care Routine Data From Electronic Patient Records: the BeoNet Registry

Reliable current information is needed for daily patient care and the health system, but comprehensive data, especially from primary care, are scarce and/or hard to access. The BeoNet Register-Database (BNR) aims to depict quality of medical care in Germany using routine data and with no additional documentation or costs for GPs. Registry design, data structure and database use will be demonstrated using chronic obstructive lung disease (COPD) as an example. The BNR is moving through pilot- and roll-out phases. Participating practices provide standardized information about (1) master data (2) practice profile (3) infrastructure and documentation systems. The routinely documented data will be pseudonymized before transmission via standardized secure interfaces to the registry database. Practices can participate in studies in cooperation with BNR if informed consent is obtained from the patients. The registry data will be linked with questionnaires on health care utilization, quality of life and disease-specific parameters. Researchers and physicians can access quasi anonymized data in specific datasets extracted from the registry. Regular automatic data quality checks and feedback from GPs, patients and researchers enhance the quality of the registry. Performance and data analysis are continuously optimized. The data security concept has been approved and ethical approval is on hand. GPs, pediatricians and pneumologists from Hanover, Munich and Heidelberg participated in the pilot phase. The registry database currently holds 98497 patient IDs and related information. BNR contains more than the standard billing data collected by health insurers. The data are applicable to a wide range of health practices, health care and economics and disease development questions. Questionnaires have been generated, tested and are being distributed. First descriptive analyses of prevalence, age and gender have been performed. Quality and validity checks have been implemented. 2016 marks the roll-out phase of the BNR. Further recruitment of practices throughout Germany, data linkage with additional questionnaires, recruitment of patients for health care research projects, quality checks and data analysis of disease-specific costs and health-related quality of life in COPD are progressing with the overall aim of improving pulmonary healthcare strategies. The floor is now open for retro- and prospective cross-sectional and longitudinal studies.

A Comparison of Data Quality Assessment Checks in Six Data Sharing Networks.

Objective:To compare rule-based data quality (DQ) assessment approaches across multiple national clinical data sharing organizations.Methods:Six organizations with established data quality assessment (DQA) programs provided documentation or source code describing current DQ checks. DQ checks were mapped to the categories within the data verification context of the harmonized DQA terminology. To ensure all DQ checks were consistently mapped, conventions were developed and four iterations of mapping performed. Difficult-to-map DQ checks were discussed with research team members until consensus was achieved.Results:Participating organizations provided 11,026 DQ checks, of which 99.97 percent were successfully mapped to a DQA category. Of the mapped DQ checks (N=11,023), 214 (1.94 percent) mapped to multiple DQA categories. The majority of DQ checks mapped to Atemporal Plausibility (49.60 percent), Value Conformance (17.84 percent), and Atemporal Completeness (12.98 percent) categories.Discussion:Using the common DQA terminology, near-complete (99.97 percent) coverage across a wide range of DQA programs and specifications was reached. Comparing the distributions of mapped DQ checks revealed important differences between participating organizations. This variation may be related to the organization’s stakeholder requirements, primary analytical focus, or maturity of their DQA program. Not within scope, mapping checks within the data validation context of the terminology may provide additional insights into DQA practice differences.Conclusion:A common DQA terminology provides a means to help organizations and researchers understand the coverage of their current DQA efforts as well as highlight potential areas for additional DQA development. Sharing DQ checks between organizations could help expand the scope of DQA across clinical data networks.

Using Tablet Computers to Implement Surveys in Challenging Environments

Computer-assisted personal interviewing (CAPI) has increasingly been used in developing countries, but literature and training on best practices have not kept pace. Drawing on our experiences using CAPI to implement the Local Governance Performance Index (LGPI) in Tunisia and Malawi and an election study in Jordan, this paper makes practical recommendations for mitigating challenges and leveraging CAPI’s benefits to obtain high quality data. CAPI offers several advantages. Tablets facilitate complex skip patterns and randomization of long question batteries and survey experiments, which helps to reduce measurement error. Tablets’ global positioning system (GPS) technology reduces sampling error by locating sampling units and facilitating analysis of neighborhood effects. Immediate data uploading, time-stamps for individual questions, and interview duration capture allowed real time data quality checks and interviewer monitoring. Yet, CAPI entails challenges, including costs of learning new software; questionnaire programming; and piloting to resolve coding bugs; and ethical and logistical considerations, such as electricity and Internet connectivity.

Towards Data Value-Level Metadata for Clinical Studies.

While several standards for metadata describing clinical studies exist, comprehensive metadata to support traceability of data from clinical studies has not been articulated. We examine uses of metadata in clinical studies. We examine and enumerate seven sources of data value-level metadata in clinical studies inclusive of research designs across the spectrum of the National Institutes of Health definition of clinical research. The sources of metadata inform categorization in terms of metadata describing the origin of a data value, the definition of a data value, and operations to which the data value was subjected. The latter is further categorized into information about changes to a data value, movement of a data value, retrieval of a data value, and data quality checks, constraints or assessments to which the data value was subjected. The implications of tracking and managing data value-level metadata are explored.

Demonstrating Effect of Rainfall Characteristics on Wheat Yield: Case of Sinana District, South Eastern Ethiopia

Demonstrating Effects of Current and Projected Rainfall Characteristics on Wheat yield has been investigated in Sinana district for the period 1995-2016. Data on rainfall and crop yield for the period 1995-2016 were obtained from National Meteorological Agency and Sinana District Agricultural Offices, respectively. Following data quality checking, rainfall data (current and future), correlation and regression studies were analyzed using Statistical software like Instat V3.36 and SPSS V20. Downscaling the output of CSIRO-Mk3.6.0 GCM model (daily rainfall data) for RCP8.5 emission scenario using a web based software tool (Marksim GCM) for the period 2020-2049 were used to determine seasonal total rainfall and ascertain its impact on yield. This study used mean, coefficient of variation, correlation and regression analysis to ascertain the relation, cause and effect relationship between rainfall characteristics and wheat yields. The results indicated that the mean onset date of the main rainy season (JJAS) for Robe and Sinana station was June 30 and July 3. Furthermore, the results of Pearson Correlation Coefficients indicated that kiremt rainfall total (JJAS) had moderate positive relationship (r = 0.499) with wheat yield in the study area. It was also observed that nearly fifty percent of total variance of crop yield is explained jointly by kiremt rainfall total and rainy day (R2 value was 47.9%). The result of projected wheat yield indicated that there will be a slight decrease in wheat yield (qt/ha) after 2030 years due to the impact of expected weakening of kiremt rainfall total. Taking in to account the above findings, it could be suggested that the farmers’ community will be encouraged encouraged to utilize timely climate information issued from National Meteorological Agency of Ethiopia (NMA) and other centers for farm level decision to enhance their crop production.

Альтернативные оценки доходов населения в регионах России

Average income per capita is one of the main indicators of the standard of living. This indicator is widely used in research devoted to poverty, interregional income inequality, and is also the main reference point of success of pursued regional policies. A number of researchers note poor quality of official data on household income and call into question their reliability in some regions of Russia. The main reasons include the following: the scale of shadow economy, especially at the level of some regions; insufficient state financing of respective research; and outdated statistical methods of data processing. The author offers the method of quality checking of official statistical data on income per capita by means of a number of indirect indicators. Application of this method also allows receiving alternative estimates of the average per capita income of the population in regions of the country. On the basis of data for 2013-2014 it is shown that in a number of regions official statistical data are underestimated by more than 20%. The received alternative estimates of the average per capita income can be used in further research in place of official indicators.

Visplause: Visual Data Quality Assessment of Many Time Series Using Plausibility Checks

Trends like decentralized energy production lead to an exploding number of time series from sensors and other sources that need to be assessed regarding their data quality (DQ). While the identification of DQ problems for such routinely collected data is typically based on existing automated plausibility checks, an efficient inspection and validation of check results for hundreds or thousands of time series is challenging. The main contribution of this paper is the validated design of Visplause, a system to support an efficient inspection of DQ problems for many time series. The key idea of Visplause is to utilize meta-information concerning the semantics of both the time series and the plausibility checks for structuring and summarizing results of DQ checks in a flexible way. Linked views enable users to inspect anomalies in detail and to generate hypotheses about possible causes. The design of Visplause was guided by goals derived from a comprehensive task analysis with domain experts in the energy sector. We reflect on the design process by discussing design decisions at four stages and we identify lessons learned. We also report feedback from domain experts after using Visplause for a period of one month. This feedback suggests significant efficiency gains for DQ assessment, increased confidence in the DQ, and the applicability of Visplause to summarize indicators also outside the context of DQ.

Towards improved quality of soil morphology and analytical data in Australia: starting the discussion

The development and implementation of a national data schema for soil data in Australia over the last two decades, coupled with advances in information technology, has led to the realisation of more comprehensive state and national soil databases. This has facilitated increased access to soil data for many purposes, including the creation of many digital soil-mapping products, such as the Soil and Landscape Grid of Australia. Consequently, users of soil data have a growing need for clarity concerning the quality of the data; many new users have little understanding of the varying quality of the data. To date, statements about the quality of primary soil data have typically been qualitative and/or judgemental rather than explicit. The consequences of poor-quality primary data and of the lack of a coding system for data quality are growing with increased usage and with demand for soil data at the regional to national scale. Pillar 4 of the Global Soil Partnership and the National Soil Research, Development and Extension Strategy both identify the need to improve the quality of soil data. Various international standards do exist with respect to the quality of soil data but these tend to focus on general principles and quality-assurance frameworks rather than the detail of describing data quality. The aim of this paper is to stimulate a discussion in the Australian soil science community on how to quantify and describe the quality of primary soil data. We provide examples of the data quality issues and propose a framework for structured data-quality checking procedures and quality coding of soil morphological and analytical data in Australia.

RAINBIO: a mega-database of tropical African vascular plants distributions.

The tropical vegetation of Africa is characterized by high levels of species diversity but is undergoing important shifts in response to ongoing climate change and increasing anthropogenic pressures. Although our knowledge of plant species distribution patterns in the African tropics has been improving over the years, it remains limited. Here we present RAINBIO, a unique comprehensive mega-database of georeferenced records for vascular plants in continental tropical Africa. The geographic focus of the database is the region south of the Sahel and north of Southern Africa, and the majority of data originate from tropical forest regions. RAINBIO is a compilation of 13 datasets either publicly available or personal ones. Numerous in depth data quality checks, automatic and manual via several African flora experts, were undertaken for georeferencing, standardization of taxonomic names and identification and merging of duplicated records. The resulting RAINBIO data allows exploration and extraction of distribution data for 25,356 native tropical African vascular plant species, which represents ca. 89% of all known plant species in the area of interest. Habit information is also provided for 91% of these species.

Neuronal Oscillations with Non-sinusoidal Morphology Produce Spurious Phase-to-Amplitude Coupling and Directionality.

Neuronal oscillations support cognitive processing. Modern views suggest that neuronal oscillations do not only reflect coordinated activity in spatially distributed networks, but also that there is interaction between the oscillations at different frequencies. For example, invasive recordings in animals and humans have found that the amplitude of fast oscillations (>40 Hz) occur non-uniformly within the phase of slower oscillations, forming the so-called cross-frequency coupling (CFC). However, the CFC patterns might be influenced by features in the signal that do not relate to underlying physiological interactions. For example, CFC estimates may be sensitive to spectral correlations due to non-sinusoidal properties of the alpha band wave morphology. To investigate this issue, we performed CFC analysis using experimental and synthetic data. The former consisted in a double-blind magnetoencephalography pharmacological study in which participants received either placebo, 0.5 or 1.5 mg of lorazepam (LZP; GABAergic enhancer) in different experimental sessions. By recording oscillatory brain activity with during rest and working memory (WM), we were able to demonstrate that posterior alpha (8–12 Hz) phase was coupled to beta-low gamma band (20–45 Hz) amplitude envelope during all sessions. Importantly, bicoherence values around the harmonics of the alpha frequency were similar both in magnitude and topographic distribution to the cross-frequency coherence (CFCoh) values observed in the alpha-phase to beta-low gamma coupling. In addition, despite the large CFCoh we found no significant cross-frequency directionality (CFD). Critically, simulations demonstrated that a sizable part of our empirical CFCoh between alpha and beta-low gamma coupling and the lack of CFD could be explained by two-three harmonics aligned in zero phase-lag produced by the physiologically characteristic alpha asymmetry in the amplitude of the peaks relative to the troughs. Furthermore, we showed that periodic signals whose waveform deviate from pure sine waves produce non-zero CFCoh with predictable CFD. Our results reveal the important role of the non-sinusoidal wave morphology on state of the art CFC metrics and we recommend caution with strong physiological interpretations of CFC and suggest basic data quality checks to enhance the mechanistic understanding of CFC.

Using a Data Quality Framework to Clean Data Extracted from the Electronic Health Record: A Case Study.

Objectives:We examine the following: (1) the appropriateness of using a data quality (DQ) framework developed for relational databases as a data-cleaning tool for a data set extracted from two EPIC databases, and (2) the differences in statistical parameter estimates on a data set cleaned with the DQ framework and data set not cleaned with the DQ framework.Background:The use of data contained within electronic health records (EHRs) has the potential to open doors for a new wave of innovative research. Without adequate preparation of such large data sets for analysis, the results might be erroneous, which might affect clinical decision-making or the results of Comparative Effectives Research studies.Methods:Two emergency department (ED) data sets extracted from EPIC databases (adult ED and children ED) were used as examples for examining the five concepts of DQ based on a DQ assessment framework designed for EHR databases. The first data set contained 70,061 visits; and the second data set contained 2,815,550 visits. SPSS Syntax examples as well as step-by-step instructions of how to apply the five key DQ concepts these EHR database extracts are provided.Conclusions:SPSS Syntax to address each of the DQ concepts proposed by Kahn et al. (2012)1 was developed. The data set cleaned using Kahn’s framework yielded more accurate results than the data set cleaned without this framework. Future plans involve creating functions in R language for cleaning data extracted from the EHR as well as an R package that combines DQ checks with missing data analysis functions.

DATA ARCHIVING AND DISTRIBUTION OF LiDAR AND DERIVED DATASETS IN THE PHILIPPINES

LiDAR programs in the Philippines have been generating valuable resource and hazard information for most of the country at a substantial rate since 2012. Significant progress have been made due to the programs design of engaging 16 Universities and research institutions spatially distributed across the country. Because of this, data has been accumulating at a brisk rate which poses significant technical and logistic issues. While a central node, the University of the Philippines, Diliman, handles data acquisition, pre-processing, and quality checking, processing and ground validation are devolved to the various nodes. For this setup to be successful, an efficient data access and distribution system should be in place. <br><br> In this paper, we discuss the spatial data infrastructure and data access protocols implemented by the program. At the center of the data access and distribution operations is LiPAD or our LiDAR portal for archiving and distribution. LiPAD is built on open source technologies, established web standards, and protocols. At its back-end a central data archive has been established using state of the art Object Storage technology to store both raw, processed Lidar and derived data sets. Catalog of available data sets ranging from data acquisition foot prints, to DEM coverages, to derived products such as flood hazard, and crop suitability are viewable and accessible on the main site based on the popular GeoNode application. Data exchange is performed using varying protocols to address various logistical problems. Given the various challenges the program is successful in distributing data sets not just to partner processing nodes but to other stakeholders where main requesters include national agencies and general research and academic institutions.

DATA ARCHIVING AND DISTRIBUTION OF LiDAR AND DERIVED DATASETS IN THE PHILIPPINES

Abstract. LiDAR programs in the Philippines have been generating valuable resource and hazard information for most of the country at a substantial rate since 2012. Significant progress have been made due to the programs design of engaging 16 Universities and research institutions spatially distributed across the country. Because of this, data has been accumulating at a brisk rate which poses significant technical and logistic issues. While a central node, the University of the Philippines, Diliman, handles data acquisition, pre-processing, and quality checking, processing and ground validation are devolved to the various nodes. For this setup to be successful, an efficient data access and distribution system should be in place. In this paper, we discuss the spatial data infrastructure and data access protocols implemented by the program. At the center of the data access and distribution operations is LiPAD or our LiDAR portal for archiving and distribution. LiPAD is built on open source technologies, established web standards, and protocols. At its back-end a central data archive has been established using state of the art Object Storage technology to store both raw, processed Lidar and derived data sets. Catalog of available data sets ranging from data acquisition foot prints, to DEM coverages, to derived products such as flood hazard, and crop suitability are viewable and accessible on the main site based on the popular GeoNode application. Data exchange is performed using varying protocols to address various logistical problems. Given the various challenges the program is successful in distributing data sets not just to partner processing nodes but to other stakeholders where main requesters include national agencies and general research and academic institutions.

Report: Release of the International Life Sciences Institute Crop Composition Database Version 5

The International Life Sciences Institute Crop Composition Database (ILSI-CCDB) Version 5 was released to the public in October 2014, and is an open-access source of comprehensive nutritional composition data for six conventionally bred crops (canola, cotton, field corn, rice, soybean, and sweet corn). Some notable features include a substantially greater amount of data (842,500 data points, a seven-fold increase compared to Version 4) as well as data for three additional crops (canola, rice, and sweet corn). For each additional crop, the appropriateness of each nutritional component and analytical methodology were carefully evaluated by a working group consisting of members representing academia, government agencies, and the agricultural and food industries. Rigorous data validation and quality control processes were established. Literature references for the analytical methods represented in the database were standardized and consolidated, allowing faster data upload for data providers and increased clarity for database end-users. Data quality checks were conducted on all data to identify and correct any errors that may have occurred during upload and subsequent handling in the database. The result of these efforts is a database with increased utility and ease of use that provides a high quality representation of variability in crop nutritional composition.

A Visualization Approach to Air Pollution Data Exploration—A Case Study of Air Quality Index (PM2.5) in Beijing, China

In recent years, frequent occurrences of significant air pollution events in China have routinely caused panic and are a major topic of discussion by the public and air pollution experts in government and academia. Therefore, this study proposed an efficient visualization method to represent directly, quickly, and clearly the spatio-temporal information contained in air pollution data. Data quality check and cleansing during a preliminary visual analysis is presented in tabular form, heat matrix, or line chart, upon which hypotheses can be deduced. Further visualizations were designed to verify the hypotheses and obtain useful findings. This method was tested and validated in a year-long case study of the air quality index (AQI of PM2.5) in Beijing, China. We found that PM2.5, PM10, and NO2 may be emitted by the same sources, and strong winds may accelerate the spread of pollutants. The average concentration of PM2.5 in Beijing was greater than the AQI value of 50 over the six-year study period. Furthermore, arable lands exhibited considerably higher concentrations of air pollutants than vegetation-covered areas. The findings of this study showed that our visualization method is intuitive and reliable through data quality checking and information sharing with multi-perspective air pollution graphs. This method allows the data to be easily understood by the public and inspire or aid further studies in other fields.

The quality of organic market data: providing data that is both fit for use and convenient

Members of the organic supply chain need high-quality data to make correct investment decisions, but data with sufficient depth and quality are not widely available in Europe. The quality of available data is a key concern for both data collectors and data users. The aim of this study is to identify whether the commonly used quality attributes (accuracy, coherence, comparability, timeliness, punctuality, accessibility, relevance), which have been developed from the perspective of data collectors, are also appropriate from the perspective of end users of organic market data. A further aim is to assess whether the data quality needs of end users are being met by the existing data. The results of two surveys carried out in Europe, one of data collectors and one of end users, are presented. Sales data at retail level (values and volumes) are used as an illustrative example and the perceptions of end users are compared with the reported data collection approaches, quality checks and availability of data. Correlation analysis and principal component analysis were used to investigate the relationship between users’ perceptions of the data quality attributes and their overall perceptions of data quality. The findings suggest that data quality checks do help to improve the quality of data as perceived by end users but that people will use whatever data they can get, even if it has poor quality. This could have potentially negative consequences, such as a lack of confidence in the organic market, if important decisions are based on poor quality data. The analysis also suggests that the commonly used attributes represent two dimensions of data quality: ‘fitness for use’ which encompasses accuracy, relevance, comparability and punctuality; and ‘convenience’, which encompasses affordability, comparability, timeliness and accessibility. The attribute of comparability belongs to both dimensions as it contributes to both fitness for use and convenience. Data collectors wishing to improve the quality of their data should focus on enhancing fitness for use first and then on the convenience of their data for users.