Hard AF
A research instrument that is used to gather data and information from respondents in a uniform manner through the administration of a predefined and structured set of questions and possible responses is called a(n): a. Survey b. Interview c. Process measure d. Affinity diagram
A survey is a common tool used in performance improvement to assess the level of satisfaction with a process by its customers. When designing a survey, the PI team must define the goal of the survey in clear and precise terms (Shaw and Elliott 2012, 98-101; Shaw and Carter 2014, 117, 221).
You are conducting a patient satisfaction survey in an outpatient clinic. Because you typically see about 300 people per day in the clinic, you decide to have the interviewers administer the questionnaire on every tenth patient. You are using: a. Systematic sampling b. Stratified sampling c. Cluster sampling d. Convenience sampling
A systematic random sample is a simple random sample that may be generated by selecting every fifth or every tenth member of the sampling frame. If the population includes N members and we wish to draw as sample of size n, then a systemic random sample could be selected by choosing every N/nth member of the population as the sample (White 2013, 120).
We want to compare the average weight of patients before and after a specialty diet is administered. We expect the weight to decrease after administration of the diet. Which of the following tests should be conducted to determine if the average weight for each patient is less at the conclusion of the diet? a. paired t-test b. one-sample t-test c. t-test for two independent sample means d. z-test for comparing two population proportions
Correct Answer: A A paired t-test may be used to compare a variable measured at two time points on the same subject (White 2013, 86).
In the APC system, a high-cost outlier payment is paid when which of the following occurs? a. The cost of the service is greater than the APC payment by a fixed ratio and exceeds the APC payment plus a threshold amount. b. The LOS is greater than expected. c. The charges for the services provided are greater than the expected payment. d. The total cost of all the services is greater than the sum of APC payments by a fixed ratio and exceeds the sum of APC payments plus a threshold amount.
Correct Answer: A An outlier payment is paid when the cost of the service is greater than the ambulatory payment classification (APC) payment by a fixed ratio and exceeds the APC payment plus a threshold amount (Casto and Forrestal 2013, 185).
What is a random number seed? a. The starting point for the random number generator b. The type of function used to generate random numbers c. The first random number in a series d. A value that must be in the random sample
Correct Answer: A Every random number generator has a starting point, or random seed. If the seed is designated and recorded as part of the sampling plan, then the series of random numbers can be replicated by another analyst using the same software and seed (White 2013, 117).
Mary Smith, RHIA, has been charged with the responsibility of designing a data collection form to be used on admission of patients to the acute-care hospital in which she works. What is the first resource she should use? a. UHDDS b. UACDS c. MDS d. ORYX
Correct Answer: A In 1974, the federal government adopted the Uniform Hospital Discharge Data Set (UHDDS) as the standard for collecting data for the Medicare and Medicaid programs. When the Prospective Payment Act was enacted in 1983, UHDDS definitions were incorporated into the rules and regulations for implementing diagnosis-related groups (DRGs). A key component was the incorporation of the definitions of principal diagnosis, principal procedure, and other significant procedures, into the DRG algorithms (LaTour et al. 2013, 196).
Which of the following code set(s) is utilized in the RBRVS prospective payment system? a. HCPCS/CPT b. ICD-9-CM c. Both a and b d. None of the above
Correct Answer: A In 1992, CMS implemented the resource-based relative value scale (RBRVS) system for physician's services such as office visits covered under Medicare Part B. The system reimburses physicians according to a fee schedule based on predetermined values assigned to specific services. CPT coding is the code set used to determine the fee schedule amount (LaTour et al. 2013, 434).
A sample of medical records is selected by randomly selecting all records from hospitals selected at random. What type of sampling is this? a. Simple random sampling b. Systematic sampling c. Cluster sampling d. Stratified sampling
Correct Answer: A In simple random sampling, every member of the population has an equal chance of being selected for the sample. Simple random sampling is the statistical equivalent of drawing sampling units from a hat (White 2013, 117).
We have conducted a t-test to determine if the hospital mean for DRG XXX is significantly different from the population mean for DRG XXX. The calculated t is -3.39 and the critical t is -1.96. In this case we: a. reject the null hypothesis. b. fail to reject the null hypothesis. c. conduct the test again because the results are inconclusive d. not enough information provided
Correct Answer: A The test statistic used to compare a proportion to a standard is called the one sample Z-test for proportions (White 2013, 82).
If a 95 percent confidence interval for a facility's mortality rate is 5% +/-1% or (4%, 6%), would a 90 percent confidence interval for the same sample be: a. Wider b. Narrower c. The same width d. Not enough information
Correct Answer: A The width of the confidence interval is a measure of the precision of the point estimate of the proportion and is sometimes referred to as the margin of error. A narrow confidence interval is more precise than a wider confidence interval. Notice from the formula that the width of the confidence interval is the function of three variables: confidence level, sample size, and sample proportion (White 2013, 71).
Data on the geometric length of stay for various long-term care DRGs is used in determining ______ adjustments. a. High cost outlier b. Short stay outlier c. Interrupted stay d. Area wage index
Correct Answer: B A short-stay outlier is an adjustment to the payment rate for stays that are considerably shorter than the average length of stay (ALOS) for a particular MS-LTC-DRG. A case would qualify for short-stay outlier status when the LOS is between one day and up to and including five-sixths of the ALOS for the MS-LTC-DRG (LaTour et al. 2013, 440).
In terms of grouping and reimbursement, how are the MS-LTC-DRGs and acute care MS-DRGs similar? a. Relative weights b. Based on principal diagnosis c. Categorization of low-volume groups into quintiles d. Classification of short-stay outliers
Correct Answer: B Both the MS-LTC-DRGs and the acute care MS-DRGs are based on the principal diagnosis in terms of grouping and reimbursement (Casto and Forrestal 2013, 223).
By querying the healthcare entity data, you find that patients admitted on a weekend have a mean length of stay that is 1.3 days longer than patients who are admitted Monday through Friday. This method of finding information is called: a. Structuring query language b. Data mining c. Multidimensional data structuring d. Satisficing
Correct Answer: B Data mining is the process of sorting through the organization's data to identify unusual patterns or to apply analytical models that will assist in predicting future events. Current applications of data-mining activities in healthcare include models to support fraud detection, utilization review, and clinical pathways (LaTour et al. 2013, 86; Shaw and Carter 2014, 19, 169).
The descriptive statistic most often used with nominal and ordinal data is: a. Median b. Frequency distribution c. Z-test d. Confidence interval
Correct Answer: B Frequency charts are an excellent tool for displaying the distribution of categorical variables (White 2013, 56).
In which form of database would data mining to support complex data analysis most effectively take place? a. Clinical data repository b. Clinical data warehouse c. Database management system d. Electronic health record
Correct Answer: B Generally, only two kinds of operations occur in a data warehouse: data warehousing and data mining (LaTour et al. 2013, 93).
You are conducting a one-sample t-test; the sample contains 30 observations. The number of degrees of freedom for the one-sample t-test is: a. 30 b. 29 c. 28 d. 31
Correct Answer: B The degrees of freedom for a one-sample t-test are one less than the sample size, n (White 2013, 83).
If an analyst wishes to predict future ancillary charges for hip replacement patients based on the age of the patient, which of the following is a correct statement? a. Age is the dependent variable; ancillary charges is the independent variable b. Age is the independent variable; ancillary charges is the dependent variable c. The average ancillary charge is the best estimator d. The two variables cannot be related
Correct Answer: B The next step in exploring the relationship between two variables is to analyze the ability of the value of one variable to predict an outcome or value of a second variable. In this scenario, the variable that is used to predict is called the independent variable. The outcome or variable to be predicted is called the dependent variable. An easy way to recall which variable is dependent or independent is to remember that the dependent variable depends on the value of the independent variable (White 2013, 107).
We want to determine if Critical Care Hospital's ALOS (µ1) for DRG XXX is significantly less than the national ALOS (µ2) for DRG XXX. The alternative hypothesis would be stated as: a. µ1 ≤ µ2 b. µ1 < µ2 c. µ1 > µ2 d. µ1 = µ2
Correct Answer: B The null hypothesis is the status quo (White 2013, 60).
Which information found on the chargemaster describes payment codes for services or items, and is usually a three or four digit number? a. CPT/HCPCS code b. Revenue code c. Ledger number d. Activity code
Correct Answer: B The revenue code is a three-digit code that describes a classification of a product or service provided to the patient. These revenue codes are required by CMS for reporting services (LaTour et al. 2013, 449; CMS 2013).
Which of the follow attributes is required for a sample to be statistically valid? a. It must include at least 100 units b. It must be reproducible c. It must include at least one of each type of unit d. It must not include more than 500 units
Correct Answer: B The term "statistically valid sample" is used often in practice. In order for a sample to be statistically valid, it must be large enough to provide information with sufficient precision to meet the goals of the analysis. Guidelines for selecting a sample size are discussed later in this chapter. A statistically valid sample is typically a probability sample where each item in the population has an equal chance of being selected. Finally, a statistically valid sample is reproducible. A reader or end user should be able to recreate the sample with the documentation provided by the analyst who selected the sample (White 2013, 116).
Which of the following does not influence the level of type II error? a. Sample size b. Selection of test statistic c. Null hypothesis value d. All of these influence type II error
Correct Answer: C A Type II error depends on the statistical test used to test the hypothesis and the sample size (White 2013, 60).
Once all data has been posted to patient's account, the claim can be reviewed for accuracy and completeness. Many facilities have an internal auditing system that runs each claim through a set of edits. This internal auditing system is known as a: a. Chargemaster b. Superbill c. Scrubber d. Grouper
Correct Answer: C A claim scrubber is used by facilities as an internal auditing system to limit the number of denied claims (Casto and Forrestal 2013, 262).
Which of the following does not describe capitation? a. Provides all contracted healthcare services needed by an individual b. Individual enrollee or third party payer pays a fixed premium for the covered group c. Payment for each medical service is usually based on the actual charges of the provider d. Contract stipulates exactly which healthcare services are covered and which ones are not
Correct Answer: C Capitation is based on per person premiums or membership fees rather than on itemized per-procedure or per service charges. The capitated managed care plan negotiates a contract with an employer or a government agency representing a specific group of individuals. According to the contract, the managed care organization agrees to provide all the contracted healthcare services that the covered individuals need over a specified period of time (usually one year). In exchange, the individual enrollee or third-party payer agrees to pay a fixed premium for the covered group (LaTour et al. 2013, 430).
An analyst wishes to determine the amount of time coders require to code inpatient records. She samples all records from 10 randomly selected dates. This is an example of a. Simple random sampling b. Systematic sampling c. Cluster sampling d. Stratified sampling
Correct Answer: C Cluster sampling may be performed as single-stage or two-stage versions. In single-stage cluster sampling, clusters are selected at random and all units in that cluster are included in the sample. An example of single-stage cluster sampling is to randomly select a day in the month and select all cases coded on that day as the sample (White 2013, 122).
If a researcher calculated a Pearson's r of 0.76, what is the coefficient of determination? a. 0.24 b. 0.42 c. 0.58 d. 1.52
Correct Answer: C Coefficient of determination: A statistic that measures the amount of variance in a dependent variable explained by one or more independent variables. If there is one independent variable, then this value is the Pearson Correlation Coefficient squared (White 2013, 240).
Survey responses collected using a 7 point Likert scale are an example of: a. Interval scale data b. Ratio scale data c. Ordinal scale data d. Nominal scale data
Correct Answer: C Ordinal data is categorical data where the categories are mutually exclusive and they do have a natural order (White 2013, 5).
To ensure quality of data, the cancer committee reviews the abstracting done by the cancer registry personnel. This type of reliability check is called: a. Precision b. Recheck c. Interrater d. Construct
Correct Answer: C Reliability is frequently checked by having more than one person abstract data for the same case. The results are then compared to identify any discrepancies. This is called an interrater reliability method of checking. Several different people may be used to do the checking. In the cancer registry, physician members of the cancer committee are called on to check the reliability of the data (LaTour et al. 2013, 383).
A researcher would like to test the null hypothesis that the mean wait time at four urgent care sites is equal. What statistical test should the researcher use in this situation? a. Chi square b. T-test for means c. ANOVA d. Z-test for proportions
Correct Answer: C The statistical tool used to compare more than two populations is called an analysis of variance (ANOVA). The null hypothesis for an ANOVA is that the population means to be compared are equal (White 2013, 92).
Binary variables are useful in calculating a. Rates b. Proportions c. Percentages d. All of these
Correct Answer: D A special case of a categorical variable is a binary variable. A binary variable has two values: yes or no. Most rates used in healthcare are based on binary variables (White 2013, 55).
The CPT codes for Emergency Department visits are: 99281 - Level 1 Emergency Department Visit 99282 - Level 2 Emergency Department Visit 99283 - Level 3 Emergency Department Visit 99284 - Level 4 Emergency Department Visit 99285 - Level 5 Emergency Department Visit This set of CPT codes in an example of: a. Nominal data b. Interval data c. Ratio data d. Ordinal data
Correct Answer: D Categorical data where the categories are mutually exclusive and they do have a natural order (White 2013, 5).
Which of the following are used to report information about mortality and morbidity at local, state, and national levels? a. Rates, populations, and ratios b. Ratios, proportions, and continuous variables c. Proportions, populations, and continuous variables d. Proportions, ratios, and rates
Correct Answer: D Many healthcare statistics are reported in the form of a ratio, proportion, or rate. These measures are used to report morbidity (illness), mortality (death), and natality (birthrate) at the local, state, and national levels (LaTour et al. 2013, 484; Shaw and Carter 2014, 119, 222).
Under what circumstances is it acceptable to use a non-probability sample for an audit? a. When the variable of interest is rare b. When a large probe sample was previously selected c. When the variable of interest is readily observable d. When not generalizing to the population
Correct Answer: D Non-probability sampling is that in which members of a sample are deliberately selected for a specified purpose". If the goal of the analysis is to gain an understanding of a process or exploratory data analysis, then a non probability sample may be used. The goal of the study should be determined prior to the collection of any data. If the goal is to generalize the results from the sample to the population then a probability sample should be used (Osborn 2006, 139; White 2013, 116).
A researcher mined the Medicare Provider Analysis Review (MEDPAR) file. The analysis revealed trends in lengths of stay for rural hospitals. What type of investigation was the researcher conducting? a. Content analysis b. Effect size review c. Psychometric assay d. Secondary analysis
Correct Answer: D Secondary analysis is the analysis of the original work of others. In secondary analysis, researchers reanalyze original data by combining data sets to answer new questions or by using more sophisticated statistical techniques. The work of others created the MEDPAR file (LaTour et al. 2013, 569).
What statistic is appropriate for measuring the association between patient severity categories (1-4) and number of laboratory tests performed during an admission? a. Pearson's correlation coefficient b. Median c. Least squares regression d. Spearman's Rho
Correct Answer: D Spearman's Rho: A statistic that measures the strength of the linear relationship between two ordinal variables or one ordinal and one continuous variable. The statistic can range from −1 to +1 (White 2013, 245).
After collecting data for a sample of patients an analyst determines that the patient's BMI explains 31 percent of the variance in the post-operative length of stay for heart bypass patients. What is the correlation between patient BMI and post-op length of stay? a. 1.00 b. 0.31 or -0.31 c. 0 d. 0.56 or -0.56
Correct Answer: D The regression statistics portion of the Excel summary output includes two important statistics: multiple R and R-square. Multiple R in the case of simple linear regression is the Pearson's correlation coefficient between the dependent and independent variable. The R-square is sometimes referred to as the coefficient of determination. The coefficient of determination measures the amount of variance in the dependent variable that is explained by the independent variable (White 2013, 109).
What statistic should the researchers in #3 use to measure the linear relationship between age and length of stay? a. Contingency table b. T-test c. ANOVA d. Pearson's r
Pearson's r Correlation Coefficient (r) measures the strength of the linear relationship between two continuous variables. The statistic can range from −1 to +1 (White 2013, 101).
Population-based registries assist public health agencies in identifying the ______of a disease. a. Prevalence b. Incidence c. Characteristics d. Demographics
a. Prevalence - The data from facility-based registries are used to provide information for the improved understanding of cancer, including its causes and methods of diagnosis and treatment. The data collected also may provide comparisons in survival rates and quality of life for patients with different treatments and at different stages of cancer at the time of diagnosis. In population-based registries, emphasis is on identifying trends and changes in the incidence (new cases) of cancer within the area covered by the registry (LaTour et al. 2013, 370).
An analyst was testing the hypothesis that there is an association between physician gender (M/F) and the willingness to use/not use CPOE. The value of the chi-square test statistic is 7.23. If the acceptable level of type I error was set to 0.01, what is the conclusion of the hypothesis test? Chi-squared with One Degree of Freedom Critical Values Alpha level X2 Critical Value (df = 1) 0.1 2.706 0.05 3.841 0.025 5.024 0.01 6.635 0.005 7.879 a. Reject the null hypothesis b. Do not reject the null hypothesis c. Accept the null hypothesis d. Accept the alternative hypothesis
a. Reject the null hypothesis - An extreme value for a test statistic is indicative of a statistically significant result. For the chi-squared test, the value of the test statistic should be compared to the chi-squared distribution with one degree of freedom to determine the probability of incorrectly rejecting the null hypothesis is less than our pre set alpha level (White 2013, 62).
Pull payer name for the payer with plan code 'M01' from a table called 'payer' that contains the following data elements: plan_code, payer_name, contract_begin_date, contract_end_date. a. Select payer_name from payer where plan_code = 'M01' b. Select plan_code = 'M01' from payer c. Select * from payer d. Join payer_name, plan_code Select 'M01'
a. Select payer_name from payer where plan_code = 'M01' - A simple query statement using the DML has three parts. The "select" statement is always first. This determines the label, or field name, of the data that is being retrieved. The next statement is always the "from" statement, telling the database which table is to be used. The last statement is the "where" statement, which lists the conditions that must be satisfied for the data to be included (White 2013, 45).
Which of the following should be considered when selecting an alpha level for an hypothesis test? a. The cost of incorrectly rejecting the null hypothesis b. Alpha level is 0.05 based on statistical standards c. The sample size d. The proportion of successes observed in the data.
a. The cost of incorrectly rejecting the null hypothesis - Type I error is set prior to performing a hypothesis test and is called the alpha level or simply the level of the test for short. The alpha level of a test should be set based on the risk or cost inherent in rejecting the null hypothesis when it is true (White 2013, 60).
During an influenza outbreak, a nursing home reports 25 new cases of influenza in a given month. These 25 cases represent 30 percent of the nursing home's population. This rate represents the: a. Prevalence b. Incidence c. Frequency d. Distribution
b. Incidence - An incidence rate is used to compare the frequency of disease in populations. Populations are compared using rates instead of raw numbers because rates adjust for differences in population size. The incidence rate is the probability or risk of illness in a population over a period of time (LaTour et al. 2013, 506; Shaw and Carter 2014, 170).
Which test statistic should be used to test the hypothesis that there is an association between patient gender and readmission rate? a. One-sample z-test for proportions b. Two-sample z-test for proportions c. ANOVA d. T-test
b. Two-sample z-test for proportions - Correct Answer: B An analyst may be asked to compare the rates in two populations such as the mortality rates at two hospitals or the surgical site infection rate in two different units of a hospital. Hypotheses regarding two proportions may be tested using the two-sample Z-test for proportions (White 2013, 72).
What is a binary variable?
c. A variable that can take on only two value - If the data used to compile a rate is thought of as a binary variable as having two values for each subject: one if the attribute is true and zero if it is not, then a rate is actually the average of those values (White 2013, 65)
Which of the following is true of the median?
c. It is based on the whole distribution. - The median offers the following three advantages: relatively easy to calculate; based on the whole distribution and not just a portion of it, as is the case with the mode; and unlike the mean, it is not influenced by extreme values or unusual outliers in the frequency distribution (LaTour et al. 2013, 519; Shaw and Carter 2014, 126, 226)
If we change sample size from 10 to 100 and all other statistics remain the same, then how is the confidence interval changed? a. The interval shifts 5% to the left. b. The interval will be wider. c. The interval will be narrower. d. The interval will not change.
c. The interval will be narrower. - A wider confidence interval will require a smaller sample size. A narrower confidence interval will require a larger sample size (White 2013, 126).
If the least squares regression line modeling wait time as a function of the number of surgeries scheduled is Y = 10 + 5 x X. Where Y is wait time and X is the number of surgeries. If the actual wait time for a patient scheduled on a day with 4 surgeries is 34 minutes, what is the residual for that patient?
d. 4 - A representation of the least squares regression line is: Y = bX + a + e, where b is the slope and a is the intercept, estimate from the least squares regression line and e is the error term or residual (White 2013, 108
A sample of 10 patients from General Hospital was selected to determine the rate that prescriptions were filled after discharge from the facility. 8 of the 10 patients did fill their prescriptions. If the target for this rate is 95%, what hypothesis test should be completed to determine if GH meets the standard? a. Chi square b. Two-sample Z-test for proportions c. T-test d. One -sample Z-test for proportions
d. One -sample Z-test for proportions - The test statistic used to compare a proportion to a standard is called the one sample Z-test for proportions (White 2013, 68).
What statistic should the researchers in #3 use to measure the linear relationship between age and length of stay? a. Contingency table b. T-test c. ANOVA d. Pearson's r
d. Pearson's r - Pearson's r Correlation Coefficient (r) measures the strength of the linear relationship between two continuous variables. The statistic can range from −1 to +1 (White 2013, 101)
5. The distribution in this curve is: -^__________
d. Skewed right - A skewed distribution is asymmetrical. Skewness is the horizontal stretching of a frequency distribution to one side or the other so that one tail is longer than the other. The longer tail has more observations. Because the mean is sensitive to extreme observations, it moves in the direction of the long tail when a distribution is skewed. When the direction of the tail is off to the right, the distribution is positively skewed, or skewed to the right. When the direction of the tail is off to the left, the distribution is negatively skewed, or skewed to the left (LaTour et al. 2013, 521; Shaw and Carter 2014, 119, 222).
