Statistical Terms/Databases and Data Sets- 1

Descriptive Vs Inferential statistics (1)

Descriptive statistics describe the data that are either in the past or present of a population, while Inferential statistics help make an intelligent guess about what the data would be in the future from a sample.

Descriptive Vs Inferential statistics (2)

Descriptive statistics describes while inferential statistics predicts.

Descriptive Vs Inferential statistics (3)

Descriptive statistics measures a complete population, while inferential measures a sample of a population.

National Cancer Database (NCDB)

Maintained by American College of Surgeons, the Commission on Cancer, and American Cancer Society A nationwide oncology outcomes database for more than 1500 Commission-accredited cancer programs in the United States and Puerto-Rico

Healthcare Effectiveness Data and Information Set (HEDIS)

Maintained by NCQA HEDIS is a tool used to measure performance on important dimension of care and service in the majority of health plans.

Uniform Hospital Discharge Data Set (UHDDS)

Maintained by National Uniform Billing Committee (NUBC) A set of data elements to provide uniformity in hospital inpatient data, collected on a form called UB-04

Uniform Ambulatory Care Data Set (UACDS)

Maintained by National Uniform Claim Committee (NUCC) A set of data to provide uniformity in outpatient settings, collected on a form called the CMS-1500.

Data Elements for Emergency Department Systems (DEEDS)

Maintained by US Government (CDC) A set of data elements to provide uniformity in collection of emergency department data.

Minimum Data Set (MDS)

Maintained by US Government (CMS) MDS is part of a nationally mandated process for clinical assessment of all residents in Medicare or Medicaid-certified nursing homes.

Independent variables

are the ones that are chosen and varied by the researcher. They are often described as the cause of whatever response is measured.

Secondary data

come from sources other than the original recorder or reporter of the data. Summarized of abstracted items that may or may not be patient identifiable.

Primary purpose of recording data

communication which is necessary to deliver patient care

Knowledge

conclusion arrived at by using information to determine a truth.

Categories of data

departmental data, financial data, clinical data, administrative (demographic & socioeconomic) data.

Types of statistics

descriptive and inferential

Qualitative data or...

discrete data dichotomous data binary data categorical data.

Data

items, observations, or raw facts

Examples of data items in health care

patient age, sex, marital status, and diagnosis.

Benchmarking

process of comparing one's results to standards or references. (ex: Medicare's Hospital Compare website)

Qualitative Vs Quantitative data

qualitative data are represented by categories quantitative data are numerical.

discrete data

represented by totals for the categories and are represented by whole numbers (47 married patients, 8 widowed, 33 single, and 12 divorced).

inferential statistics

study of using mathematical models to predict future events and to draw conclusion about a sample of a population.

Business intelligence

the application of knowledge to make a decision. Final use of the collection, analysis, interpretation, and presentation of data.

Dependent variables

the response or effect

Dependent vs independent variables examples

weeks of gestation (independent)/weight in grams (dependent) While the weeks in gestation could influence the weight, it is impossible for the weight to influence the weeks in gestation.

Information

Data that have been organized to give it context and meaning

Quantitative data

1) Numeric data (ex: age, weight, height, time,..) 2) may be used for calculations to further describe their groups (patients, hours per week, or salaries)

Data set

1) Standard group of elements collected so that they can be compared with similar data, 2)collected either in a different time or from a different facility. Without a standardized, defined data set, comparison or benchmarking is difficult. (Ex: Without the means of comparing cases of an incidence in the general population, along with a defined data set that collects the information, it is impossible to determine whether the second set of incidence of the cases is connected to the source.)

Ordinal level

1) descriptive and have an inherent natural order 2) A type of nominal data in which the observations have an ordered value. 3) no arithmetic relationship among observations ex: patient survey satisfactions satisfaction ranging from least to most satisfied.

interval data

1) have a logical order or ranking, and the the intervals are meaningful, but the concept of a natural zero is missing. 2) A type of quantitative data without a natural zero, in which the values may be added or subtracted but multiplying and dividing them have no meaning. (ex: interval scale--we are able to add and subtract degrees on the scale, but no meaning with multiplying and dividing)

ratio level data

1) have a logical order or ranking, meaningful, consistent intervals, and include a zero in its order. 2) A type of quantitative data with a meaningful, natural zero in its order. 3) Observation may be multiplied or divided to produce meaningful results. (Ex: patient ages, weights, length of stay, billing charges, etc...)

Continuous data

1) represented by numbers that have values in between the whole numbers. 2) Data on a scale in which numbers have values in between the whole numbers. Ex: temperatures: 37.5; weights: time; salaries: $32,450.81, $49,341.28; )

Nominal level data

1)descriptive and have no logical orders. 2) A type of qualitative data without a mathematical relationship or order among observations. 3) No numerical value or not measured, but can still be counted. ex: listing of insurance carriers, patient's personal identifier (medical records number), gender, race, and ethnicity.

Qualitative data

1)groups of categories that can be counted or sorted out but not added together or has no numerical value. 2) Observations without a numerical value that can be sorted or counted into categories. EX: Likert scales in patient satisfaction surveys (least to most satisfied)

Variables

Any items or characteristics that is measured or counted.

Dependent vs independent variables

Independent variables are the suspected cause, while dependent variables are the presumed effect.

Differences of data in quantitative data

Interval data and ratio data have numerical value and can have some arithmetic relationships only ratio data have a true zero value, enabling all mathematical functions.

Outcome and ASSessment Information Set (OASIS)

Maintained by US government (CMS) OASIS is for use in Home-Health agencies (HHAs), state agencies, software vendors, professional associations, and other federal agencies in implementing and maintaining OASIS.

ORYX for Hospitals (National Hospital Quality Measures)

Maintained by the Joint Commission (TJC) ORYX is the TJC's performance measurement and improvement initiative, which integrates outcomes and other performance measure data into accreditation. Reported to quality check.

Two levels of qualitative data

Nominal level and Ordinal level data

descriptive stratistics

The analytical activities performed to explain what is and what was. discipline of collecting, interpreting, describing, and presenting data.

Values

The count or measurement of the observation. The individual observation of the variables being measured.

pie chart

Used to compare categories with one another in relation to the whole group.

table

Used to compare characteristics of items.

Histogram

Used to illustrate a frequency distribution

Data analytics

The inspection and evaluation of groups of data using statistics.

line graph

Used to represent data over a period of time

Primary data

The original, firsthand account of the patient's treatment. In healthcare, items that are obtained directly from the patient record and which specifically identify that patient.

Data and information

Used synonymously, but not similar. Data are units of observation and information is organized data for context and meaning.

Bar graph

Used to compare categories with each other.

Population

a complete group of all of the items that are under study. Data that are studied (ex: all of the patients, salaries, doctors, etc..)

Sample

a subset of the population that has been SELECTED using a particular sampling method.

Health Care data

administrative, financial, and clinical data