Common-cause Variation Research Articles

Variation in financial protection and its association with health expenditure indicators: an analysis of low- and middle-income countries.

BackgroundAn insight into variation in financial protection among countries and the underpinning factors associated with the variations observed will help to inform public health policy and practice.MethodSecondary datasets from Global Health Expenditure Database and World Bank Development Indicators collected between 2000 and 2016 were used. Financial protection was measured in 75 low- and middle-income countries (LMICs) using the sustainable development goals framework. Funnel plot charts were used to explore the variation, and regression models were used to measure associations.ResultFifty-three (67%) countries were within the 99% control limits indicating common-cause variation; 11 countries were above the upper control limit and 15 countries were below the lower control limit. In the fully adjusted model, country, spending on health relative to their economy had the strongest association with the variation in catastrophic spending. Every 1% increase in health spending relative to gross domestic product (GDP) was found to be associated with a reduction of 0.13% in the number of people that incurred catastrophic health spending.ConclusionThere is substantial variation in financial protection, as measured by the number of people that incurred catastrophic health spending, in LMICs; a proportion of this could be explained by the difference in GDP and external health expenditure.

Country Characteristics and Variation in Diabetes Prevalence among Asian Countries – an Ecological Study

ObjectivesTo determine the variation in diabetes prevalence across Asian countries and its relationship with the quality of health system and socioeconomic characteristics of the country.MethodologyAn ecological analysis was conducted using publicly available data from the World Bank, the World Health Organization and the International Diabetes Federation. Geographical variation in diabetes prevalence across countries was examined using control charts while the relationships between country-level determinants and diabetes prevalence were investigated using linear regression analysis.ResultsThe control chart shows special-cause variation in diabetes prevalence in 21 (58%) of the Asian countries; nine countries were below the 99.8% control limits while twelve were above it.Fifteen (42%) countries suggest common-cause variation. Three country characteristics independently associated with diabetes prevalence were hypertension prevalence (OR 0.39, 95% CI 0.22 to 0.55; p-value<0.001), obesity prevalence (OR 0.15, 95% CI 0.13 to 0.18; p-value<0.001), and quality of health care governance (OR 0.18, 95% CI 0.04 to 0.34; p-value=0.02).ConclusionsThere is a considerable geographical variation in diabetes prevalence across Asian countries. A substantial part of this variation could be explained by differences in the quality of health care governance, hypertension prevalence and obesity prevalence.

Variation in neonatal mortality and its relation to country characteristics in sub-Saharan Africa: an ecological study

BackgroundA substantial reduction in neonatal mortality is the main priority to reduce under-five mortality. A clear understanding of the variation in neonatal mortality and the underlying causes is important for...

VARIATION IN NEONATAL MORTALITY AND ITS RELATION TO COUNTRY CHARACTERISTICS IN SUB-SAHARAN AFRICA

BackgroundA substantial reduction in neonatal mortality is the main priority to reduce under-five mortality. A clear understanding of the variation in neonatal mortality and the underlying causes is important for...

Development and field application of a multivariate statistical process control framework for health-monitoring of transportation infrastructure

We present a two-part multivariate statistical process control framework to support health-monitoring of transportation infrastructure. The first part consists of estimation of regression and ARIMA–GARCH models to explain, predict, and control for common-cause variation in the data, i.e., changes that can be attributed to usual operating conditions, including traffic loads, environmental effects, and damage when present throughout the data. The second part of the framework consists of using multivariate control charts to simultaneously analyze the standardized innovations of the aforementioned models in order to detect possible special-cause or extraordinary events, such as unique/infrequent traffic, weather, or the onset of damage. The proposed approach revolves around construction of T2 control charts as a framework to jointly monitor the evolution and contemporaneous correlation of a set of measurements. The approach provides significant practical/computational advantages over individual analysis of multiple structural properties, and addresses technical problems stemming from ignoring the relationships among them.To illustrate the framework, we analyze strain and displacement data from the monitoring system on the Hurley Bridge (Wisconsin Structure B-26-7). Data were collected from 1 April 2010 to 28 June 2012. Analysis of 7 measurement sequences collected over the 27month planning horizon revealed 6 possible special-cause events. In terms of outlier interpretation, we use Mason–Young–Tracy Decomposition to establish the contribution of (subsets of) the measurements. Also, we link the most significant special-cause events, in terms of magnitude and duration, to unusual changes in weather and traffic. To conclude, we compare the proposed approach and benchmark the empirical results with Principal Component Analysis, perhaps, the most common alternative appearing in the literature.

Aligning service processes to the nature of care in hospitals: an exploratory study of the impact of variation

Operational management (OM) approaches typically aim to reduce the variation in processes by removing clearly identifiable causes of variation -the so-called special causes of variation - leading to improved efficiency and quality. In healthcare, OM must deal not only with special-cause variation, but also with the type of variation that cannot be eliminated: inherent or common-cause variation. Using an exploratory database analysis of four hospitals, this article investigates whether hospital care processes can be assigned to different groups defined by their kind and size of variation, resulting in better alignment of type of care and organization of care. A detailed analysis of the length-of-stay of All Patient Refined Diagnosis Related Groups suggests groups with high and low within-group variation, which might indicate that there are groups of patients with inherently different degrees of variation in length-of-stay due to illness and treatment patterns. As is well-known, hospital care can be divided into sequential and iterative processes. Some patients groups may be classified as high-variation in one hospital but as low-variation in another, which clearly shows that deliberate choices in the design of the operational system of hospitals - i.e. special-cause variation - must be taken into account when analyzing the length-of-stay performance of hospitals. Furthermore, separating common-cause from special-cause variation will increase the likelihood of identifying the right type of process (sequential versus iterative) and business model for the right type of patients.

Increasing recruitment rates in an inpatient clinical research study using quality improvement methods.

One important benefit of successful patient recruitment is increased generalizability of findings. We sought to optimize enrollment of children admitted with asthma as part of a population-based, prospective, observational cohort study with the goal of enrolling at least 60% of all eligible and staffed patients. Quality improvement methods were used to improve cohort recruitment. Weekly meetings with study staff and study leadership were held to plan and discuss how to maximize recruitment rates. Significant initial variability in recruitment success prompted the team to use small-scale tests of change to increase recruitment numbers. A number of tests were trialed, focusing primarily on reducing patient refusals and improving recruitment process efficiency. Recruitment rates were calculated by dividing eligible by enrolled patients and displayed using annotated Shewhart control charts. Control charts were used to illustrate week-to-week variability while also enabling differentiation of common-cause and special-cause variation. The study enrolled 774 patients, representing 54% of all eligible and 59% of those eligible for whom staff were available to enroll. Our mean weekly recruitment rate increased from 55% during the first 3 months of the study to a statistically significant sustained rate of 61%. This was sustained given numerous obstacles, such as departing and hiring of staff and adding a second recruitment location. Implementing quality improvement methods within a larger research study led to an increase in the rate of recruitment as well as the stability in recruitment rates from week-to-week.

On the meaning of the special-cause variation concept used in the quality discourse – And its link to unforeseen and surprising events in risk management

In quality management, ‘common-cause variation’ and ‘special-cause variation’ are key concepts used to control and improve different types of processes. In this paper we study the meaning of these concepts, having a special focus on the latter concept: how is the special-cause concept linked to ideas and concepts used in risk and uncertainty management, to reflect unforeseen and surprising events? In the quality discourse it is common to refer to two possible mistakes when confronting the variation: (i) to react to an outcome as if it were from a special cause, when actually it came from common causes of variation; and (ii) to treat an outcome as if it were from common causes of variation, when actually it came from a special cause. However, at the point of decision making it is difficult or impossible to know what the “true” state is. It is also appropriate to ask whether such a true state does in fact exist. In the paper we discuss these issues, the main aim of the paper being to improve our understanding of some of the fundamental concepts used in risk and quality management.

Analysis of common-cause and special-cause variation in the deterioration of transportation infrastructure: A field application of statistical process control for structural health monitoring

We present a statistical process control framework to support structural health monitoring of transportation infrastructure. We contribute an integrated, generally-applicable (to various types of structural response data) statistical approach that links the literatures on statistical performance modeling and on structural health monitoring. The framework consists of two parts: The first, estimation of statistical models to explain, predict, and control for common-cause variation in the data, i.e., changes, including serial dependence, that can be attributed to usual operating conditions. The ensuing standardized innovation series are analyzed in the second part of the framework, which consists of using Shewhart and Memory Control Charts to detect special-cause or unusual events.We apply the framework to analyze strain and displacement data from the monitoring system on the Hurley Bridge (Wisconsin Structure B-26-7). Data were collected from April 1, 2010 to June 29, 2011. Our analysis reveals that, after controlling for seasonal effects, linear trends are significant components of the response measurements. Persistent displacement may be an indication of deterioration of the bridge supports. Trends in the strain data may indicate changes in the material properties, i.e., fatigue, sensor calibration, or traffic loading. The results also show that autocorrelation and conditional heteroscedasticity are significant sources of common-cause variation. Use of the control charts detected 43 possible special-cause events, with approximately 50% displaying persisting effects, and 25% lasting longer than one week. Analysis of traffic data shows that unusually heavy loading is a possible cause of the longest special-cause event, which lasted 11days.

System for Verifiable CT Radiation Dose Optimization Based on Image Quality. Part II. Process Control System

To evaluate the effect of an automated computed tomography (CT) radiation dose optimization and process control system on the consistency of estimated image noise and size-specific dose estimates (SSDEs) of radiation in CT examinations of the chest, abdomen, and pelvis. This quality improvement project was determined not to constitute human subject research. An automated system was developed to analyze each examination immediately after completion, and to report individual axial-image-level and study-level summary data for patient size, image noise, and SSDE. The system acquired data for 4 months beginning October 1, 2011. Protocol changes were made by using parameters recommended by the prediction application, and 3 months of additional data were acquired. Preimplementation and postimplementation mean image noise and SSDE were compared by using unpaired t tests and F tests. Common-cause variation was differentiated from special-cause variation by using a statistical process control individual chart. A total of 817 CT examinations, 490 acquired before and 327 acquired after the initial protocol changes, were included in the study. Mean patient age and water-equivalent diameter were 12.0 years and 23.0 cm, respectively. The difference between actual and target noise increased from -1.4 to 0.3 HU (P < .01) and the standard deviation decreased from 3.9 to 1.6 HU (P < .01). Mean SSDE decreased from 11.9 to 7.5 mGy, a 37% reduction (P < .01). The process control chart identified several special causes of variation. Implementation of an automated CT radiation dose optimization system led to verifiable simultaneous decrease in image noise variation and SSDE. The automated nature of the system provides the opportunity for consistent CT radiation dose optimization on a broad scale.

Understanding and managing variation: three different perspectives

Presentation Managing variation is essential to quality improvement. Quality improvement is primarily concerned with two types of variation – common-cause variation and specialcause variation. Common-cause variation is random variation present in stable healthcare processes. Special-cause variation is an unpredictable deviation resulting from a cause that is not an intrinsic part of a process. By careful and systematic measurement, it is easier to detect changes that are not random variation. The approach to managing variation depends on the priorities and perspectives of the improvement leader and the intended generalizability of the results of the improvement effort. Clinical researchers, healthcare managers, and individual patients each have different goals, time horizons, and methodological approaches to managing variation; however, in all cases, the research question should drive study design, data collection, and evaluation. To advance the field of quality improvement, greater understanding of these perspectives and methodologies is needed [1].

Statistical process control charts for attribute data involving very large sample sizes: a review of problems and solutions

The use of statistical process control (SPC) charts in healthcare is increasing. The primary purpose of SPC is to distinguish between common-cause variation which is attributable to the underlying process,...

Can simulation improve the traditional method of root cause analysis: A preliminary investigation

The Joint Commission on Accreditation of Healthcare Organizations recommendations for conducting root cause analysis (RCA) include identifying "root causes" and "common-cause variation" rather than "proximate causes" and "special-cause variation" to create interventions. Simulation for health care RCA is a novel technique but has not been compared with traditional RCA methods. All of the RCAs of adverse events conducted at Tulane Hospital between September 2010 and September 2011 were reviewed. A case of missed postprocedural, preoperative hemorrhage resulting in death was chosen. Hospital records were analyzed to identify the presumed root causes. A simulation of the event was developed and conducted. Six test subjects (preoperative and postanesthesia care unit nurses) participated in the simulation. Root causes identified by simulation analysis were compared with those identified by traditional RCA. In 2 of 6 simulations, the adverse event was duplicated. The root cause identified by standard RCA technique was inattention to signs of bleeding in the patient/ lack of appropriate monitoring of the patient by nursing staff ("special-cause variation"). Simulation-based RCA revealed that the root cause was not only inadequate monitoring, but also the lack of physical presence of physicians in the care environment ("common-cause variation"). Simulation-based RCA identified root causes more amenable to intervention. This study demonstrates that simulation-based RCA can identify additional root causes amenable to making health care interventions when compared with traditional RCA.

Exploring variations in under-5 mortality in Nigeria using league table, control chart and spatial analysis

Under-5 mortality rate (U5MR) is the probability of a child born in a specific year or period dying before reaching the age of 5. Nigeria has a high rate of U5MR and accounts for a significant proportion of the high U5MR in sub-Sahara Africa. There are differences in health and social practice in the different states in Nigeria, coupled with the differences in developmental priorities of each State government. There is therefore the need to identify the states with high and low U5MR, to further explore the risk factors and make recommendations for planning. This study investigates variation in U5MR in Nigeria using Shewhart's theory of variation and control charts. We used data from the birth histories included in the 2008 Nigerian Demographic and Health Survey to estimate U5MR using a synthetic cohort life table. We plotted control charts of the proportion of under-5 mortality for the 37 states (included federal capital Abuja) in Nigeria. The Local Indicators of Spatial Association was used as a measure of the overall clustering and is assessed by a test of a null hypothesis. On average, more than 1 in every 10 children born in Nigeria (159 per 1000 live births) does not survive to their fifth birthday. Kwara and Osun states had the lowest U5MR (less than 60 per 1000 live births), while Jigawa, Kano, Sokoto, Niger and Adamawa states had the highest U5MR (more than 200 per 1000 live births). There is a wide variation in the U5MR between the 37 states. The funnel plot identifies 27 (73%) states within the 99% control limits indicating common-cause variation. Four states were above the upper control limit (higher than the average) and six states were below the lower control limit (lower than the average), indicating special-cause variation. U5MR is high in Nigeria; the rates are in three digits. There is a wide variation in the U5MR in Nigeria, with 27% showing evidence of special-cause variation which merits further investigation to identify possible causes. However, the vast majority of states (73%) are consistent with common-cause variation.

CAN'T MISS: Conquer Any Number Task by Making Important Statistics Simple. Part 7. Statistical Process Control: x–s Control Charts

Statistical process control (SPC) can be thought of as the frequent monitoring of processes using inferential statistics. The feature that distinguishes SPC from the typical use of inferential statistics for analyzing populations is that in the former frequent samples are taken over time, whereas in inferential statistics a single sample is generaLLy taken before and after some intervention or treatment. An x-s control chart is used to monitor a continuous variable that reflects the output of a process. The x-s control chart is a graph that includes serial sample means (x) as the variables of interest, a centerline that represents the grand mean of the samples (x), and upper control limit (UCL) and lower control limit (LCL) that represent three standard errors (SEx) above and below the centerline. An x-s control chart is used to estimate with 99.7% confidence that the population mean of a continuous output variable was within the interval defined by the UCL and LCL during a period of baseline monitoring. It is further assumed that if the process remains stable, future population means wiLL remain between the control Limits for additional process outputs. Control charts allow the evaluation of both common- and special-cause variation. AnaLysis of the common-cause variation aLLows an assessment of the current process performance. Special-cause variation is identified when there is a sample mean that is beyond the UCL or LCL.

Incorporation of External Information into Multivariate PCA/PLS Models

Most processes are subject to changes of the operating conditions such as feed rate and composition, product specification, controller status, and so on. These operating condition changes and the resulting process variatioos largely contribute to the common-cause variation. Sometimes the common-cause variation disguises or distorts the relevant information to faults. If the fault effects are correlated, or small compared to the operating condition changes, then the correlation model would be likely insensitive to the faults. This difficulty can be avoided by using a new correlation model that includes only the process variations relevant for the fault detection and isolation. The new model called a hybrid correlation model is estimated by incorporating prior knowledge within an empirical correlation model. Various approaches to incorporating different types of prior knowledge into the correlation models, and using them for process monitoring and fault diagnosis are presented in this study. The proposed method is used for analyzing a real industrial dataset. It is shown that the hybrid correlation model outperforms the regular correlation model in analyzing the process abnormality and increases the sensitivity for process monitoring.

Phase I Analysis for Autocorrelated Processes

There has been recent interest in statistical process control for autocorrelated processes. Previous researchers have not distinguished models of autocorrelated common-cause variation from the actual behavior of baseline data. Standard estimators of common-cause parameters can be severely biased when assignable causes are present. A new estimation method given here overcomes this difficulty for the first-order autoregressive common-cause model. The method is illustrated with real data sets and assessed via simulation.

Statistical Process Control and Quality Improvement

1. Introduction to Quality Concepts and Statistical Process Control. What is Quality? Definitions of Quality. The Need for SPC. Prevention Versus Detection. SPC Goals. The Basic Tools for SPC. Statistical Process Control Techniques. Applying SPC to an Existing Manufacturing Process. Designed Experiments. The Quality Toolbox. 2. Striving for Quality: Management's Problem and Management's Solution. Management's Problem. Management's Dilemma. Leadership by Management. Deming's Contribution to Quality. Deming's 14 Points for Management. Deming's Seven Deadly Diseases. Crosby's Approach. A Comparison of Deming's 14 Points and Crosby's 14 Steps. Which Way to Top Quality? Pitfalls in the Quest for Quality. Total Quality Management (TQM). The Malcolm Baldrige National Quality Award. Total Customer Satisfaction. ISO-9000. The Service Sector. 3. Introduction to Variation and Statistics. Measurement Concepts. Special-Cause and Common-Cause Variation. The Variation Concepts. Distributions and SPC Goals. Basic Statistical Concepts. Distributions and Three Standard Deviations. 4. Organization of Data: Introduction to Tables, Charts, and Graphs. Stemplots. Frequency Distributions and Tally Charts. Histograms. Pareto Charts. Flowcharts. Storyboards. The Cause-and-Effect Diagrams. Checksheets. Scatterplots. 5. Introduction to Probability and the Normal Probability Distribution. Probability. Compound Probability. Counting with Permutations and Combinations. The Binomial Distribution. The Hypergeometric Distribution. Probability Distributions. The Normal Probability Distribution. The Application of the Central Limit Theorem. 6. Introduction to Control Charts. The Control Chart Concept. Preparation for Control Charting. Control Charts and Run Charts. The Basic x-bar and R Charts. The x-bar and R Chart Procedure. The Continuation Control Chart. The Capability Analysis. Six-Sigma. 7. Additional Control Charts for Variables. The Median and Range Chart (x-bar and R). x-bar and s Charts. Coding Data. A Modified x-bar and R Chart for Small Sets of Data. The Nominal x-bar and R Chart. The Transformation x-bar and R Chart. Control Chart Selection. 8. Variables Charts for Limited Data. Precontrol or Rainbow Charts. A Compound Probability Application. Modified Precontrol for Tight Control. Charts for Individual Measurements. 9. Attributes Control Charts. The Four Types of Attributes Charts. The p Chart. The np Chart. The c Chart. The u Chart. SPC Applied to the Learning Process. Technology in SPC. 10. Interpreting Control Charts. The Random Distribution of Points. Freaks. Binomial Distribution Applications. Freak Patterns. Shifts. Runs and Trends. Time and Control Chart Patterns. Cycles. Grouping. Instability. Stable Mixtures. Stratification. Using Control Chart Patterns in Problem Solving. 11. Problem Solving. The Problem-Solving Sequence. Teamwork for Problem Solving. Brainstorming. Using Problem-Solving Tools. Mistake Proofing. Problem Solving in Management. JIT (Just-in-time). Problem Solving in the Classroom. 12. Gauge Capability. Preparations for a Gauge Capability Study. The Gauge Capability Procedure. Analysis of R and R with Accuracy and Stability: Maximum Possible Deflection. The Elimination of Gauge Variation From Process Variation. Indecisive Gauge Readings. 13. Acceptance Sampling. The Sampling Dilemma. Random Sampling. Operating Characteristic Curves. The Average Outgoing Quality Curve. MLT-STD-105D for Inspection by Attributes. The Average Proportion Defective. Vendor Certification and Control Chart Monitoring. Appendix A: Basic Math Concepts. Signed Numbers. Variables. Order of Operations. Inequalities. Using the Statistical Calculator. Appendix B: Charts and Tables. Formulas and Constants for Control Charts. The G Chart. The Normal Distribution Table (Tail Area). The Normal Distribution Table (Center Area). The Normal Distribution Table (Left Area). Process Capability. Appendix C: Glossary of Symbols. Appendix D: Lab Exercises. Answers to Odd Exercises. Index.

Estimating Common-Cause Sigma in the Presence of Special Causes

In statistical process monitoring we test process data for departures from some model of common-cause variation, usually independent, identically distributed (iid) normal. One task in the set-up phase is to estimate the common-cause standard deviation. Most comparisons of estimators have assumed normal or contaminated normal distributions. The usual “trends and oscillations” argument for the standard moving-range estimator has not been adequately investigated, nor have other special-cause scenarios where the data cannot be reduced to an iid subset. I compare five old estimators and one new one in this context. A new proposal for retrospective analysis is based on the results of these comparisons.

CHARTING AUTOCORRELATED DATA: GLTIDELINES FOR PRACTITIONERS

Data from many industrial processes are autocorrelated. Standard control-charting techniques are based on a model of independence and tend to produce out-of-control signals in the presence of positive autocorrelation. Thus, common-cause variation is ope..