Vila Health ® ActivityPresenting Statistical Results for Decision MakingIntroductionYour OfficeStakeholder InterviewsEmail ResponseConclusionIntroductionAt any enterprise, statistical results can point the way to better decisions. This is especially true in health care, where trends or changes in patient care or results can have powerful effects on health care organizations — clinical effects, and financial effects. Knowing the facts about trends and changes is crucial to navigating them, which means that in many ways, statistical results are powerful tools.But they’re less powerful if nobody understands them. More often than not, stakeholders in health care organizations are not well versed in statistics or data analysis. Especially for these stakeholders, it’s critical to make careful choices about which numbers to include — and how to talk about those numbers. Doing so helps you to present statistical results in a way that’s understandable and actionable even for those with less data literacy.In this activity, you’ll have an opportunity to analyze a health trend at a large urban hospital, and to decide which results of the analysis are most important to communicate to the stakeholders concerned with it.Your OfficeYou are an analyst in the Quality Assurance department at St. Anthony Medical Center, a large hospital in Minneapolis, Minnesota. You have an email from David Brooks, the manager of Quality Assurance, in which he explains your next assignment.EmailFrom: David BrooksSubject: HACsHello! So I’m aware that you’re relatively new around here, but I’ve got to pull you in on an important project. Here’s the short version, and I can answer questions later if this isn’t enough background.Since you’re in health care, you must know that hospital-acquired conditions (HACs) are a big deal. Hospitals are designed from the ground up to minimize them, but the fact is that hospitals always have a lot of bugs floating around and that means they’re a constant danger. And while we try to minimize mistakes, health care professionals are human, and sometimes mistakes cause real problems.The hospital’s board is getting worked up about them again, because we’ve had some high-profile cases where a patient went home and had to be readmitted because of a mistake or an infection they picked up while they were here. They’re worried about PR, of course, but they’re worried about patient outcomes, and — of course — financial reimbursement and penalties, too.What I need you to do is start working on a presentation about HACs here, specifically as they relate to staffing levels and skill mix. That presentation should explain our recommendations to the board. You’ll want to include some statistics to bolster our case, so take a look at the AHRQ National Scorecard on Hospital-Acquired Conditions and the CMS Hospital-Acquired Conditions Reduction Program. Look at published research about the relationship between HACs and staffing levels, too.Let’s get this right; this is our chance to get the board to understand that HACs will go down if they do what we recommend.Go talk to Rick Susskind first; he’s our senior data analyst and he’s already been working on this. He has completed a data analysis on this, and he can explain what the variables are. You’ll have to talk to a few other people too, to get everything you need.DavidEmailFrom: Rick SuskindSubject: HAC DataHi! I got your message about the HAC question. I’ve attached a spreadsheet of the data analysis we’ve got on that.Here are a few variables that might be unclear:HAC_Rate: That’s the number of HACs per 1,000 discharges. You could say it’s the measure for the extent of hospital errors.Nursing_HPPD: That’s the number of nursing staff hours per patient day. It’s the hospital’s nurse staffing level.Skill_Mix: That’s obviously about skill mix, but specifically, it’s the percentage of nursing staff hours that are provided by registered nurses, as opposed to other professionals.ALOS: That stands for average length of stay; it’s the number of inpatient days per hospital discharge. It’s a measure of how efficient the hospital is.Let me know if you have other questions! Oh, you’ll need to find out how many discharges we have per year. I don’t know that, but Troy Holland, our vice president of patient services, will know.Attachment: HAC AnalysisStakeholder InterviewsTroy HollandVice President of Patient ServicesIt’s nice to meet you! Yep, I’ve got those discharge stats. I’m assuming you’re confining your analysis to the last full year? If so, the number of discharges for last year was 10,000. That includes both adult and pediatric discharges.What I don’t have is information about nurse staffing or costs. For those, you’ll want to talk to Jackie, the chief nursing officer.Jackie SandovalChief Nursing OfficerHello. I understand you need some data about our staff costs related to nursing. I’ve got three numbers for you.Our cost per nurse is $72,000. Now, that’s an average of nurses and registered nurses.Our cost per registered nurse is $85,000.Per licensed practical nurse, it’s $52,000. Those numbers include salary, benefits, and overhead.You had asked about some other costs, but I don’t have those. You’ll want to talk to the CFO, Owen Welch.Owen WelchChief Financial OfficerHey, I’ve only got a couple of minutes but here’s what you asked about: Our cost per inpatient day is $2,600, and the penalty per HAC is $5,700.Email ResponseIt looks like David has followed up with you about the data analysis project you’re working on. Review David's latest message to you and send a reply with your initial thoughts.EmailFrom: David BrooksSubject: HACsThanks for doing all that running around. Now that you’ve got the data analysis you need, I’d like you to do some thinking about what results you’re going to include in the presentation. Remember, it’s the board, not a bunch of number geeks. You’ll want to give them the statistical results they need to make a decision, not just every bit of data we have.Write down a list of numbers you’re thinking about including in your presentation. We really need to make a data-driven recommendation to the board, and I’d really appreciate your expertise. Thanks!ReplySubject: RE: HACsThis message has not been composed yet.ConclusionIn this activity, you saw that while complicated analysis of data is an important step in guiding decisions at health care organizations, it’s not the only step. Decision makers need to understand how the data were collected, how variables are measured, and whether the analysis tools and techniques were appropriate.You also gathered necessary data for an analysis of HACs, nurse staffing levels, and skill mix, and did some pre-work on a presentation of the data analysis to the hospital’s board.

The use of statistics is often employed to highlight international, national, and state trends in crime and drug abuse.1. Explain the following terms in your own words.1. Box-and-whiskers plot2. Standard deviation3. Normal distribution (9 points)2. A BAC describes the amount of alcohol in a person's blood, expressed as weight of alcohol per unit of volume of blood. For example, 0.08 percent BAC indicates 80 mg of alcohol per 100 ml of blood." (source: http://www.austintexas.gov/faq/what-does-blood-alcohol-concentration-bac-measure (Links to an external site.); please note, you do not require to access the source to answer the question) For the data set below calculate the mean, median, mode and standard deviation. Then, for Blood alcohol levels only, plot: a. Frequency histogram b. Probability (i.e. cumulative frequency) plot What do you conclude about the distribution of the sample?From the data set presented here, would you feel comfortable to say that the age group between 30-39 years exhibits higher BAC counts than other age groups? (16 points)AgeBlood alcohol level (mg/dL)2113456532346521348218362489284543102487852563711248105619533685841319036982710425324754356764625078389236845926418134852854241094296193433276418439139743410141954088In an experiment using three tosses of the same coin, state and apply the general equation that provides the total number of possible outcomes for the experiment and list the outcome string (3 points). Then, for each of the following three scenarios, list the possible outcome string and calculate the probability of the scenario as a decimal number (to three places) showing all working (9 points).The probability of getting three heads when flipping a coin 3 times.The probability of at least getting one head and one tail when flipping a coin 3 times.The probability of getting a head in the first flip, a tail in the second, and a head again in the third flip of a coin.Below is an incomplete contingency table displaying two levels for each the gender and the political party association (13 points).Complete the table with all marginal totals and cell counts.Calculate the following probabilities.For a male to be a Republican.For a voter to be female.For a voter to be either a democrat or a female.For a person to not be a female but voting democratic.Political partyTotalGenderRepublicanDemocratMale8773?Female61??Total??300