Summer 2004 - Copyright 2004 by S. I. Inc.

It gives me great pleasure to announce that S. I. Inc. will be presenting a new seminar series being given by Allegiant Systems, Inc. These seminars will be of interest to owners and managers of small to medium sized businesses in the suburbs of the Boston metropolitan area. We will begin with three topics during August and September and follow it up with more during the fall. Each of these sessions will begin at 6:00 PM and last roughly an hour. Following the seminar there will be plenty of time to meet my associates and continue the discussion. All of the events will take place in the Allegiant offices in Newton. Light refreshments will be served.

The first program is entitled Customer Relationship Management (CRM), "it´s about people not computers." It will run on Thursday, August 5, 2004. Robert Schuldenfrei will be the principal speaker. We will discuss what CRM is and why it is critically important for smaller enterprises. As the tag line indicates, the focus is on management and not systems, although we will be explaining five programs that help you realize the benefits of CRM. S. I. Inc. has had 30 years of experience delivering value from computer based systems to managements large and small. As readers of the WORKS know we have done a number of CRM engagements.

Three weeks later Michael Kibler, President of Allegiant Systems, will present the second program: Information Technology (IT) Security, "the threats are real and the consequences can be costly." Viruses, worms, spyware, hackers, spam, phishing, browser exploits - the threats are constantly evolving. We will discuss current security issues, how you can mitigate your exposure, how you can limit damage when your defenses are compromised (I intentionally said when, not if), and some steps to take to recover if your computers or networks become seriously impaired. Allegiant has had years of experience preventing disasters and bringing computer systems back from the dead. We can bring the benefits of that experience to you.

For the final program in our first series we will bring Bob back again to discuss your corporate web site. In a program called Web Site Development and Maintenance, "everyone has them, few are good," we will explain how your site can deliver on the promise you had when you first developed the site. This will be held on Tuesday, September 14, 2004. In the course of the hour Bob will explain what your site should do for your firm. We will address when you need professional help and when you can do it yourself. In the latter case, Bob will explain how you can use Microsoft Word, an no other software product, to manage the web site. Yes, if you can type a letter using Word, you can manage your company´s site!

Excel Analysis with SQL Data

Historic data is often a rich source of information that can be productively mined and analyzed by management. There are, however, issues with this approach such that organizations fail to obtain the benefits that might accrue from using historic data. This is a shame because this data is usually free for the taking as a by-product of the firm´s accounting system. Since the founding of S. I. Inc. in 1981, we have been working with companies to help them produce information from existing data sources. In this brief article we will give the reader an overview of how we do it.

The principal reason why management fails to do the analysis is because the accounting data is in a form that most people find very "user hostile." Often this is in a database such as Oracle or SQL Server. The Microsoft spreadsheet, Excel, has some great analytical tools, but if you can not get at the data all effort goes for naught. What follows is an example of what might be called the SAE approach. SAE is the first initials of SQL Server, Access, and Excel. Here is the broad-brush of a problem drawn from Marketing Research. Suppose there is a set of observations locked way in a SQL Server database on attitudes about a town. This dataset also contains data about length of residence in the town and the importance the respondent attaches to weather. Our goal is to build a forecasting model that claims the longer you live in the town and the more importance weather is to you the more favorable your attitude will be towards the town.

The approach will be to use Access to "dip into" the SQL database and pass that information on to Excel. Then we will use Excel to perform a multiple regression (a statistical tool that is built into Excel, but hidden) to construct our model. Since Excel wants to see rows and columns, we need to form that data in Access so as to "please" Excel. This whole process could be automated in Excel (using VBA programming), but for ease of use and explanation, we are going to move the data from package to package on the Windows clipboard. This is the familiar copy and paste operations.

In order to use a SQL Server database, instead of the built-in Jet database native to Access, you create an Access "project." This is practically effortless since all of the tables have been defined when the SQL database was created. The New Project from the Existing Data menu item is executed. The user explains to Access where the SQL database is located and Access does the rest. When the process ends you have tables and relationships that appear to be lodged within Access, although you are directly using them in SQL Server. Often the data you desire is in a single table with the observations being the rows and the variables formed into the columns. If this is not the case you will use an Access query to form the array of rows and columns. Select all of the rows and columns in either the table or the query and copy them to the clipboard.

Now we are ready for the Excel work. Bring up the Excel program as usual. Starting with a blank worksheet, paste the array on to the page. The data will appear on the sheet with the variable names appearing in row one. If Excel was installed with the "Typical" options, the necessary Analysis ToolPak will not be on your system. In that case use the original Office CD (or have someone in Information Technology do it) to load this optional feature. Once loaded Selecting Tools | Data Analysis | Regression will take you to the regression statistical package. Without going into the statistical detail, here is what you do. You give the package the column you want to forecast. That is the dependent variable, or Attitude in our example. Then you point regression to the columns that purport to explain this behavior. These are the independent variables, Length of Residence and Weather Importance. Once you hit the OK button Excel gives the desired results.

Multiple regression is hardly the only sophisticated tool found in Excel. The package is "loaded" with powerful tools for analysis. I have been doing marketing research for years and I have not found a tool I wanted to use that was missing from Excel. With just a little study, you too can drill into your data mine and retrieve high assay gold.

Hospital Quality Control Using Excel

By Patricia Schuldenfrei.

The American public is hearing a great deal about hospital errors and the importance of safe practices in healthcare. As patients who are hospitalized are older, sicker and more fragile than in the past, the chance of a patient having a fall while hospitalized also grows. Using simple statistical software can assist health quality professionals to see the patterns and trends among patient care events and use the data to plan and implement better procedures.

Falls are the leading cause of injury death in persons over 65 years of age and the second leading cause of spinal cord and brain injury in the older population. Many older people become hospitalized as the result of a fall - for example the fall that causes a fractured hip. However, persons hospitalized for any reason may experience a fall during the hospitalization. These falls sometimes cause injury, further complicating the medical condition, causing further pain and suffering and increasing length of stay.

Knowing the risks and dangers involved, hospitals collect data on each patient fall. Such data is usually gathered via an "incident report" or "patient safety occurrence report". Immediately after the fall, the nurse submits this report with such vital data as the patient name and room number, time and date, and details concerning the fall. Any immediately apparent injury is noted and direct quotations from the patient are included when possible.

The safety occurrence reports from each unit are forwarded to a quality professional who aggregates data from the entire hospital, using the data for analysis of risk and to help plan strategies to reduce falls. In rehabilitative facilities there are often 7 to 12 falls per 1000 patient days. In other words, in a 100 bed facility, it is not unusual to see 30 patient falls per month. In a short period of time, the quality professional has access to volumes of data. The challenge is turning that data into useful information that can be used to implement changes that positively affect patient outcomes.

A commercially available program such as Microsoft Excel can be an invaluable tool for such data analysis. The quality manager can tap a database or enter information about falls into an ongoing Excel spreadsheet. Sorting the data by room number allows a tally of the fall rate per unit. Nursing units with high numbers of falls receive priority on falls prevention education. Sorting by patient name reveals those patients who are repeat fallers. The care team and the patient/family meet to plan specific care to avoid future injury. The pharmacist receives ongoing data about repeat fallers. Review of the medication profile may reveal drug combinations that are placing the patient at higher risk. In these cases, pharmacy contacts the physician to suggest alternative strategies. Sorting the data by time of day sometimes shows that falls are more frequent at change of shift. One current quality initiative at a local hospital is to have nurse aides make rounds in every room during nurse shift change. This practice keeps a trained caregiver visibly present to patients, allows patients to ask for assistance with toileting or bed to chair transfers, and prevents falls.

The combination of valid data, timely data analysis, and interdisciplinary quality review has had positive outcomes. Using this triple tiered approach at a local hospital has led to changes that have reduced the fall rate per 1000 patient days from 7 to 5. There is a reduction in both the frequency and the severity of falls. Patient safety is positively affected and the hospital gains in quality while reducing risk.

For additional information:

     bob@s-i-inc.com
     Phone: (781) 329-4828

     Or send USPS mail to:
     S. I. Inc.
     32 Ridley Road
     Dedham, MA 02026