New world of data collection revolves around patient count
It’s no longer all about visit numbers
Home health agencies traditionally made it a priority to collect data on the number of visits made per month, quarter, and year. The results drove hiring, services expansion, and other growth decisions.
But it’s a whole different data collection world now. The number of visits is far less important under the interim payment system (IPS) and the prospective payment system (PPS) than is the patient count. This means quality managers should brush up their data collection and decision-making processes if they are still mired in the old philosophy of counting visits.
"We paid little attention to the patient number in past years, and now that’s the critical number," says Celeste Roberson, BSN, MPH, senior partner with Roberson, Muck & Associates in Wilmington, NC. Roberson has spoken at home care conferences about how to make data collection more efficient.
"We used to have agencies say, We have 40% growth now,’ and it wasn’t new patients causing the growth. It was more visits for every patient," she adds.
Since IPS put pressure on home care agencies to reduce the number of visits per patient, then all new growth is related to increasing the number of new patients. While most home care directors and quality managers understand this shift in focus, they may not have made all the data collection changes that will support the philosophical change. And this is where Roberson has some advice.
Suppose you want to increase your patient count by courting referral sources. Traditionally, that might mean basing a marketing campaign on reaching new potential referral sources and reminding current sources that your services are available. But under IPS, is this really the best and most efficient way to increase your patient count?
Before you answer, "Yes," consider this scenario: Doctors A and B habitually refer patients to you with chronic illnesses requiring long lengths of stay (LOS). However, Doctors C and D refer patients to you who can be treated and released within eight weeks. Under IPS, you might desire to encourage Doctors C and D, and not spend as much time courting Doctors A and B, whose patients probably cost you more money than they bring in.
But you wouldn’t even know this unless you have revamped your data collection system to include this type of information.
"We have all kinds of data coming out of our information systems, but have very little data that is analyzed or summarized for management," Roberson explains, "and that’s where people are hampered. They can’t develop programs to make for good management accounting."
Know what you seek
Here’s a quick guide to solving this problem:
1. Define your data use.
You need to know what you’re looking for and choose computer selection fields that reflect that.
This is trickier than it sounds. For instance, the selection field for LOS could mean different things, depending on your data collection system. In some systems, this means patient count. In others, this is based on the number of discharged patients in a particular period.
Here are some questions to find answers to before using a particular data relating to patient count:
• Do you measure the number of patients receiving services in a particular period, or do you measure the number of patients discharged?
• How long of a period do you want to measure, and will your system measure for that length of time?
"A lot of times providers never question a report if it says what it is," Roberson says. "But you should check a few, and if you want data for three months year-to-date, and the report says 120 days, which is longer than three months, then it’s not giving you the information you want."
2. Get the bugs out.
One common mistake is an agency will include non-patients in patient count data. For whatever reasons, a patient who has received no visits during that year is continuing to pop up in the computer file that relates to unduplicated patient counts.
"You should take a look to see if there are any patients on the data sheet who have zeros across the visit column," Roberson advises.
Also, quality managers should be skeptical of any patients who are listed with only one nursing visit. If an agency has a significant number of these one-visit patients, it could skew the agency’s LOS figures. "I encourage people to not spend hours, but to take a quick look and get a colored pen and highlight these patients," Roberson says.
3. Ask your computer vendor to help with data collection problems.
"People often buy a big computer package and don’t feel they can approach the vendor for help," Roberson notes.
However, computer vendors usually want to improve and update their software packages, and they need input from consumers in order to do so. "It’s not the vendor’s intention to print data that you don’t need," Roberson says. "Vendors can do patches and corrections on reports."
Approach the vendor in a non-antagonistic manner, she advises. Simply state how you are using the reports and what you need to have corrected.
4. Limit LOS data to a fiscal year of 365 days.
There are two ways to determine a patient’s length of stay. One way is to count the number of full days for each patient, by taking the patient’s start of care date and subtract it from the discharge date.
Or, to obtain a more accurate figure for a particular fiscal year, take that same figure and limit it to 365 days of the fiscal year. So, if a patient is admitted on Dec. 1, 1998, but is not discharged until Feb. 1, 2000, then the total number of days would exclude any days between December 1998 and January 2000.
The easiest way to adjust your data collection to reflect the number of actual days a patient is served within a particular year is to create a start of care date of Jan. 1 for all patients carried over from the previous year, and create a discharge date of Dec. 31 for all patients who will continue to receive services in the next year.
It is important to know a patient’s annual LOS because this is the number used in calculating an agency’s average LOS under IPS. However, quality managers may want to run reports on patients’ total LOS as well, because this information more clearly tells them how many visits the agency is taking for patients of different diagnoses.
One of the more common and least useful LOS data fields is when a computer program divides LOS into categories of 30 to 60 days and 61 to 90 days, etc., Roberson says.
"It looks like an accounts receivable report, but you look at that and say, Who knows what it means?’ and you don’t know what your average is."
5. Learn your unduplicated patient count.
"The unduplicated patient count is more important than it’s ever been," Roberson says. "You need it to gauge new patient growth and to see if that growth is coming from the referral sources you want, and to compare it with related cost data."
For example, suppose a quality manager wants to collect numbers for stroke patients. First, the manager will need an unduplicated patient count of people who are stroke patients, and then the manager will divide this number into the total cost of stroke patients for that period of time. If the cost per case turns out to be $5,000, it will give managers an important piece of information.
They can decide perhaps that a $5,000 per-case cost for stroke patients is reasonable. But if the data showed that the agency’s cost per congestive heart failure (CHF) case was also $5,000, administrators might decide the cost for a CHF diagnosis is excessive, and a quality improvement project should be started in that area.
By knowing an agency’s unduplicated patient count, quality managers can better decide which projects to tackle next. "Look at practice patterns. Look at what the numbers tell you, including what kind of LOS the agency has and what kind of patient mix," Roberson says.