Accurate Patient Census measures

 

I’m often surprised at how questions posed by hospital administrators and staff that appear to be simple on the surface are in fact much more complex then they appear. I realize this is a common complaint in general regarding software design, but this one in particular is asked often and impacting effective resource planning. For example, it’s common for a provider, operations manager or financial analyst to ask, “How many patients were in the hospital last month”. This information would drive decisions such as, “should we increase the available nursing pool” or “do we need to increase/decrease services in a certain specialty of the hospital.” Unfortunately, accurately answering this question depends on many implied assumptions by the user. A responsible requirements analysis would ferret out the complexities of such a request quickly, but it can be difficult to convince the user that the devil is in the details for such a question. Thankfully, dimensional analysis gives us powerful tools for exhibiting the need for extra time to clearly identify the question.

 

Lets first consider this question from the perspective of the operations manager. Their primary concern might be to have enough materials and personnel to accommodate demand for services. When they ask “how many patients were in the hospital last month?”, they’re really asking the inferred question of how much resources were expended last month so they can plan for future months resource needs. They want to know how many people were in beds in the hospital consuming resources. Further, they want to know what type of resources and from where these resources were being consumed. A patient who has been admitted as an inpatient will typically consume more resources as compared to an outpatient or observation patient. A patient in the ICU will likely consume more provider time then someone in oncology. Using dimensional analysis, we can construct a measure that focuses this question and provides flexibility to the user to extrapolate additional questions to improve their ability to predict future resource needs.

 

Next, lets consider the question of “how many patients were in the hospital last month” from the perspective of the financial analyst. Their primary question might be how much money is being spent and how are we doing with respect to re-imbursements. Again, when they ask, “how many patients were in the hospital last month?”, they’re really asking the inferred question of how much money was spent or how much money is owed. This helps them communicate with the various payors like Medicare, blue-cross, GroupHealth, etc. to create an aggregate billing profile for the hospital. Typically, Medicare sets these rules for hospitals and only patients that are admitted as inpatient are to be included. Many other variables come into play when identifying how this profile is created, but we’ll describe the impact of this single detail as it impacts accurate patient census calculations.

 

Now that we’ve clarified the perspectives of “patient days” for the financial analyst and the operations manager, lets outline the key differences and how they can cause confusion. Often a patient is admitted as observation to the hospital for up to a day and then is changed to an inpatient. From the financial perspective, this patient “arrived” to the hospital the moment they were changed to inpatient status. From the Operational perspective the person is considered an inpatient retrospectively back to the moment they occupied the bed. This makes sense if you consider that often a patient comes into the hospital and it takes time to determine what is wrong with them. During this “interim” period while diagnosis is evolving, the future inpatient is likely consuming inpatient resources. Nurses are attending to them, IV’s are being placed, a bed and room are occupied. Also, Medicare rules dictate that until the patient is admitted as an inpatient and a charge is posted to them while in this status, they cannot be included in “patient day” calculations. Further necessary beurocratic processes also delay the conversion of an observation patient to an inpatient such as delays inputting patient data and overworked doctors that might not get to their paperwork in a timely fashion. One fundamental truth needs to be addressed at this point. The Medicare/Financial definition of “patient days” is not a count of patients in the hospital. Rather, it’s a audit of inpatient charges. Unfortunatly, most hospitals use the medicare definition for creating census measures. Operational managers are often trying to match up resources with a measure that floats between close to accurate and completely wrong. On the flip side, financial analysts have difficulty reconciling resources consumed to what they should charge for those resources.

 

It’s a subtle but critical distinction of separating the measure from the goal. Always attempt to identify the goal of the measure so to construct it properly within the context of the goal. Why the question is being asked is critical to the understanding of how to answer it. I realize this can sound pedantic, but experience has taught me it’s better to be careful when constructing measures used in large organizations and to document them carefully less they are used as cornerstones in decisions that significantly impacts patient care.

 

In the past, I’ve resolved this issue by creating two census definitions. One for financial goals called “financial census” and one for operational objectives called “operational census”. When these numbers are compared, it becomes obvious the operational census is greater then the financial census. At first this can be confusing to users, but once the differences are explained, they begin to understand how complex this question of “How many patients were in the hospital last month” actually is.