Chapter 8 Health Statistics
Introduction A special language that expresses outcomes of mathematical calculations is known as statistics To understand statistics, one must first build a statistical vocabulary Because of the variety and complexity of statistics, the need for a common understanding takes on a great importance
Statistics Overview STATISTICAL TYPES Descriptive statistics Can have a narrow or very wide focus Used to characterize or summarize a given population Inferential statistics Reaching conclusions based on data from a sample Patterns are modeled to address randomness and uncertainty
Statistics Overview STATISTICAL TYPES Applied statistics Use of statistics in real-life situations Both descriptive and inferential statistics are forms of applied statistics Mathematical statistics Theoretical basis of statistics Vital statistics Data on human events (births, deaths, marriages, etc.)
Statistics Overview STATISTICAL TYPES Public health surveillance statistics Similar in concept to vital statistics Focus on illnesses, conditions, and diseases Demographic statistics Study of human populations Look at the size of populations How populations change over time
Statistics Overview STATISTICAL LITERACY Statistical measurements used to draw conclusions Miles per gallon Miles per hour Mortality rates Morbidity rates Test scores Statistical literacy Determining if a study or conclusion is credible
Statistics Overview STATISTICAL BASICS Math concepts Percentage Rounding Ration Rate Fraction
Statistics Overview STATISTICAL BASICS Measures of central tendency Common measures: mean median, mode Other concepts Standard deviation Statistically significant t-test Null hypothesis
Statistics Overview DATA COLLECTION Aggregate data collection Specific data combined into larger groupings Can be used to describe a bigger concept Retrospective data collection Converting paper records to computerized abstract Concurrent data collection Data gathered at the time they are entered Uses modern interface and transmission standards
Statistics Overview DATA COLLECTION Seven broad categories of patient data Dates Counts Test results Diagnoses Procedures Treatment outcomes Assessments
Statistics Overview STATISTICAL FORMULAE Daily inpatient census Number of patients present at the official census-taking time daily Inpatient service day Services received by one patient during one 24-hour period
Statistics Overview STATISTICAL FORMULAE Inpatient service day (cont.)
Statistics Overview STATISTICAL FORMULAE Length of stay Number of calendar days between admission and discharge
Statistics Overview RATE FORMULAE Rate formulae in hospitals and health care settings Number of times something happens Divided by the number of time it could have happened General pattern applied to the context of everyday life Provides a basic understanding of how the formula works Helps its application in the health care setting make sense
Statistics Overview RATE FORMULAE Morbidity rates Prevalence rates Incidence rates Case fatality rates
Statistics Overview RATE FORMULAE Mortality rates
Statistics Overview RATE FORMULAE Rates commonly referred to using percentage form
Data Presentation CLASSIFICATIONS OF DATA Discrete data Specific values are distinct and separate Points between values are not considered valid Continuous data Values that lie within a certain range Categorical data Values belonging to the set can be sorted into categories Categories do not overlap
Data Presentation CLASSIFICATIONS OF DATA Nominal data Ordinal data Values are classified Logical ordering of values is not required Ordinal data Ordered data set Differences between values are not important Interval data Both ordered and constant Contains no natural zero as a point of reference
Data Presentation PRESENTATION METHODS Bar graph Line graph Pie chart Used for frequency of categorical data Line graph Similar to bar graph Frequently used to display time trends Pie chart Represents frequency of data Is a circle divided into sections
Data Presentation PRESENTATION METHODS Histogram Frequency polygon Vertical axis contains continuous intervals for categories Conveys how a single variable is used continuously Frequency polygon Resembles a histogram Well-suited to superimposing data within one graph Should only be used to display numerical values
Regression Analysis INVESTIGATING RELATIONSHIPS Regression analysis Investigates relationships between variables Models relationships and determines their magnitude Determines correlation between at least one variable and another Correlation Mutual relation or interdependence
Regression Analysis INVESTIGATING RELATIONSHIPS Strength Weakness Can establish correlations Weakness Cannot always prove causation Predictive power Can be used to solve a problem and predict the future outcome
Regression Analysis VARIABLES Simple regression or univariate regression Only one variable will provide the answer Bivariate regression Two variables will provide the answer Multivariate regression Numerous variables are used to provide the answer
Regression Analysis REGRESSION ANALYSIS MODELS Pearson product-moment correlation Determines whether a relationship exists between variables Both variables must be numbers (interval or ratio) ANOVA test (F-test) Compare more than two groups Early warning system to test a hypothesis
Regression Analysis REGRESSION ANALYSIS MODELS Chi-square Measure relationships between two variables Variables are categorical Used in qualitative studies Determines degree of association among qualitative variables
Regression Analysis GROUPER PROGRAMS Grouper program examples Groupers Diagnosis related groups (DRGs) All patient refined-diagnosis related groups (APR-DRGs) Ambulatory payment classification groups (APCs) Groupers Place patients into similar categories Provide predictive values for patient stays
Health Information Management Statistics Most commonly used statistical formulas Inpatient service days Occupancy ratios Mortality rates Length of stay Discharge days Frequency statistics Traditionally captured in acute care environments
Health Information Management Statistics PRODUCTIVITY Labor analytics Determining staff for a work area Enough staff Too many staff Working with a staffing shortage Labor ratio Shows how much earned revenue is spent on labor expenses
Health Information Management Statistics PRODUCTIVITY Productivity tracking of medical transcription How quickly recorded physician reports are typed Productivity tracking of coding How quickly health records are coded
Health Information Management Statistics STATISTICAL TOOLS Control charts A graph with statistically generated upper and lower control limits Used to measure key processes over time Software for control chart management Statistical Process Control (SPC) Used in the Six Sigma process Aid in creation and use of control charts
Health Information Management Statistics STATISTICAL TOOLS
Health Information Management Statistics STATISTICAL TOOLS Trend charts Graphical presentation of data Shows patterns or shifts according to time Differ from control charts because they focus on time patterns Can be created with software applications
Health Information Management Statistics STATISTICAL TOOLS
Health Information Management Statistics STATISTICAL TOOLS Process capability analysis A process control technique Ensures continued improvement to the process Should not be used if the control chart shows a wide range of variability
Health Information Management Statistics STATISTICAL TOOLS
Summary Statistics are categorized by type, with descriptive statistics playing the largest role in health care Relationships among data variables are the focus of regression analysis Pearson product-moment correlation, ANOVA test, and chi-square are all used in regression analysis Statistical tools solve process problems, and improve productivity