We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.


Using Healthcare Business Intelligence to Optimize Patient Panel Size and Shape

Originally published July 19, 2011

Large vs. Small Patient Panels?

An entire analytical application with significant clinical, financial and operational value can be built around one question: What is the ideal patient panel size? The reason is simple. Determining the size, shape and composition of patient panels involves a large number of business questions that cross an equally large number of functions within the organization.

A patient panel is the roster of patients that a primary care physician (PCP) is responsible for managing. Healthcare provider organizations push for larger panels to get the most for their physician compensation dollars. In a capitated environment, a physician would push for a smaller panel size. If the physician is also an owner, then he or she may push for a larger panel in order to maximize personal revenue. This tension between forces favoring large panels vs. small panels drives the need for detailed analysis of the factors used to determine the ideal panel size.

In addition, emerging forces are making this analysis even more critical. Quality may be sacrificed if panels are too large. In other words, too many patients means some can get lost in the shuffle. Open access makes patients happy due to its convenience, but puts more pressure on PCPs to have greater flexibility when it comes to managing their patient panels. And the advent of accountable care organizations will require PCPs to think differently about their patient populations because it essentially tips the care model on its side, putting population management and continuity of care at the forefront.

Business intelligence capabilities are essential to establish ideal patient panel size and complexity, and to effectively manage those panels once established. Setting panels is not a one-time event because the factors underlying this analysis are changing constantly. What was optimal five years ago when your organization set the panels may be sub-optimal or even maddening today because the sands have shifted under your feet.  Continuous change requires continuous updating of the model as well as continuous updating of the data going into that model.

We will take a look at the key analyses required to establish, run and support patient panels. We will also look at the types of business intelligence functionality needed by various participants in the organization and by patients who are in the panels. In addition, we will look at the business, clinical and process impact of having this information available to the organization.

Reasons to Analyze Patient Panels Frequently

Often, physician groups determine the size and makeup of patient panels and attempt to let them run themselves. The analysis is akin to a research project, but there is good reason to analyze your patient panels frequently. For various reasons, you may not reallocate patients to other panels, but frequent analysis keeps the sizes and composition of panels fresh.  In addition, there are ways to use the results to improve a number of aspects of the practice and to make more money.

Some reasons to analyze frequently using business intelligence and analytics include:

Improved Continuity of Care
Continuity of care is a concept that just makes sense. Patients are happier, and clinically they are healthier when not shifted around constantly. They are more apt to listen to and follow the advice of a doctor they know and trust. For healthcare providers, this concept also makes sense because the providers know their patients, know their history and can get through to them better. In addition, operationally it is smoother for patients to remain with the same doctor and within the same system.
 
Increased Satisfaction for Patients and Providers
Patients generally like “their” doctors and prefer to stay with a particular one versus shifting to another. Especially irritating for patients is being transferred to another provider without their voice being heard in the decision process. 

A balance exists between shuffling patients among panels too often or too seldom. For example, at one time my family had four PCP changes in one year; it was clear that the healthcare provider was using an automated algorithm for setting panels and running it quarterly. We complained, and I imagine other patients complained too. We now have had the same PCP for twelve years, changing only once when our doctor moved to another state.

Better Workload Definition
For providers, setting patient panels intelligently and reviewing them frequently helps to define their workload in a smarter fashion and identify workload problems before they become critical.

Greater Accuracy in Demand Prediction
The size and composition of a particular patient panel is a great predictor of demand. For individual doctors, they now know the outer bounds of volume because they have a defined population. In addition, they know the likelihood of classes of patients in terms of probability of visits. We will take a deeper look at the demographic factors that predict demand later, but setting panels with solid data makes it easier to see revenue and workload.

Reveals Process and Performance Issues
Wasted steps in the process of seeing patients bubble up to the top in the process of defining and analyzing patient panels. If the panels seem like they should be in balance but are out of whack, then there may be a process step that is impeding patients at one facility but not others. Perhaps patient geography was not taken into account. While patients should be visiting in a balanced flow, they may be constantly rescheduling or not showing up because the location is inconvenient. The factors described in the the next section represent a good set of clinical and service complexity variables that should cover problems such as the geography issue.

Likewise, if a particular provider is not keeping up with demand, it may signal an issue with job performance. On the other hand, if providers are not bringing in the revenue they should, perhaps they are not actively managing their patient populations for best clinical and financial results.  Most of the time, the problem is a process problem and not a provider problem, but the process of setting panels can alert management to either type of issue.

Analysis of Patient Panels

There are three basic variables that must be analyzed in order to set patient panel sizes. These are:

  1. Patients Seen per Day
    This is the measure of load on the provider. Historical information from the patient scheduling system gives the panel-setting analysts the averages they need.

  2. Days Provider is Available
    This is the provider’s capacity to handle the load. Analysis should be done on a historical basis to get an accurate picture of what already happened. Analysis should also be done on the projected working days for the upcoming year to see if any big changes took place, such as a provider taking a sabbatical, taking on new hospital duties, being involved in a process improvement project, etc.

  3. Patient Visits per Year
    This is the demand figure. Average patient visits per year tell the panel-setting decision makers how likely the members of the patient population are to visit, which in turn drives the load on the provider.

These will allow the practice to set the appropriate size of panels, but will not be sufficient to guide management in determining the optimal composition of the panels. To determine the shape of individual panels, you will need to analyze factors such as:

  1. Patient Age
    Statistically, very young and very old patients have higher visit volume. This makes sense because these groups typically need the greatest amount of care. Analytical data can tell you how this stacks up against your overall practice’s patient demographics, and how this will affect the size and composition of each provider panel.

  2. Patient Gender
    Once again, in statistical terms, females are typically more likely to visit providers than males. This difference seems to be especially true for teenage and adult women vs. teenage and adult men. Using this information and applying it to your own system’s patient populations will allow you to balance panels more effectively.

  3. Patient Geography
    Your healthcare system may be widespread and algorithmically it may make sense to balance panels using demographic factors, but geography also plays a part. If a patient’s doctor is “way across town” or even in another town, the patient may be less likely to visit and may have to work harder to bundle visits (e.g., schedule visits for the entire family on the same day in consecutive time slots). Geography could play a part in clinical outcomes if people are less likely to visit. Geography, therefore, is another balancing factor in setting panels.

  4. Patient Condition Acuity
    Often  a physician may attribute overload to the fact that his/her patients are sicker and require more time.  This could be true and would need to be factored into setting and maintenance of his or her panels. Historical information for analysis tells management if this is truly the case or just an excuse for process or performance problems. Acuity is a key input to patient panel management.

Monitoring Effectiveness of Patient Panel Definition

There are several barometers that tell a manager that he or she needs to look at resetting the panels, either in terms of size or shape.  Once again, business intelligence applied to operational data can assist the manager in this monitoring process.

  1. Increased Phone Calls into the Practice
    If your on-call nurses and administrative staff are taking more calls or longer calls, then you probably have an imbalance in your patient panels. This situation indicates that patients are experiencing scheduling issues. In addition, if the calls involve a greater complexity of issues handled, then your panels may be seriously out of balance. The reason is that patients are doing an end-around in order to get medical advice because they cannot see their doctor. Call volumes are captured by your call management system and can be fed into your business intelligence applications for analysis alongside the other factors used to set patient panels.

  2. Increased Complaints
    Many of the reasons for increased phone calls apply to the drivers for increased complaints. Patients may not be getting what they want in terms of availability of provider time and are speaking out about it. Once again, feeding complaint data into the panel-setting analytical application can make the latter a richer information source.

  3. Increased No-Shows, Cancels and Reschedules
    Some patients give up and simply do not show up for an appointment. Others cancel because it is not convenient or they have waited too long for an appointment. If this is happening, it may signal an overloaded panel. This situation needs to be analyzed alongside call volumes, complaints and the other factors  of  panel determination decisions.

  4. Increased Walk-Ins
    Whereas some patients call to get medical advice, and others complain or cancel, still others simply show up. Perhaps their thinking is that it is harder to turn away a patient when they are physically in the office than it would have been had they called for an appointment. Walk-in volume is a good indicator to include in your panel-setting application.

  5. Increased Triage Effort
    Triage effort indicates that instead of planned responses to a particular volume level, reactive effort is needed to prioritize and rearrange patients for treatment. This is a significant indicator that something is wrong with the panel sizes and shapes as set.

  6. Increased Discontinuity Rate
    Discontinuity is measured by the number of times a patient sees a provider to whom he or she does not belong. Of all of the barometers described above, this is the clearest signal that an individual provider is overloaded, or that his or her patient panel is inappropriately set in terms of acuity. This is another key measure to include in the decision-support application used to define panels.

Business Uses of Patient Panel Analysis

If patient panels are out of balance, then management has a number of actions they can take to correct the situation. These actions include:

  1. Use Trends to Predict Balance Point Changes
    Business intelligence data and capabilities can tell you if there are trends in your patient populations or provider capacity that signal a future panel balance problem. Perhaps your patient populations are getting older. Or maybe the neighborhoods surrounding your clinics are rolling over, with elderly moving or dying and new families moving in. These will indicate potential patient panel imbalances. It could be that your populations are moving upscale or deteriorating in economic terms. This too can signal future changes.

    On the other hand, your roster of providers may be changing. You may have some pending retirements that will require provider changes. But it may also require a fresh look at all of your panels to keep them balanced and appropriately set.

  2. Eliminate Wasted Effort
    Every process has some wasted effort. Before any costly changes are made to any system or  process, it is essential to root out and get rid of wasted effort. If your providers are overloaded, chances are they are wasting some worktime. Using analytics with data from operational systems such as workflow systems, scheduling systems and financial systems can pinpoint areas of waste.

  3. Offload Non-Clinical Work
    Doctors and other providers are often surprised at how high the ratio is between clinical and non-clinical hours spent during the day. Some of this non-clinical work can be offloaded to non-clinical staff at a price that is cheaper. Before doing so, it is essential to eliminate as much waste as possible to ensure that you are not simply shifting the problem to a staffer who is already overloaded. Analytical data from workflow and other systems can help identify smart offloading decisions.

  4. Keep Providers Informed
    One of the best ways to head off an impending operational problem such as over-paneled providers is to let the people who will be affected see the analytical information and the pending issue. They will have thoughts on how to solve it before it happens, and will be happier to have their opinions heard.

  5. Increase Patient Education Effectiveness
    Patients themselves can help solve panel overloads by taking on some of the care process themselves. Patient education is key to helping patients help themselves, and to then see the physician for matters that are too complex for self-care. A great deal of information is now available from a variety of sources with varying degrees of accuracy, quality and effectiveness. A way to use analytics to drive patient self-reliance is to show patients simple statistical information on what has worked for your organization’s patient populations. This puts some of the control into your hands by effectively giving a particular treatment or information source a “stamp of approval.” Patients will take it from there and help you to rebalance your patient panels.

    The actions above may get or keep patient panels completely in balance. Other actions are more reactive in nature. Some actions that may need to be taken as last resorts either cap potential revenue, increase practice costs or increase the risk of patient dissatisfaction.

In these cases, business analytics can still help. Examples include:

  1. Let Natural Patient Attrition Reduce Overloaded Panels
    Patients naturally leave practices because they move, pass away or simply change in terms of their medical conditions. One strategy is to let panels balance themselves over time. This is a passive approach and may take quite a bit of time. As such, it is not as powerful as the actions described previously. Nevertheless, business intelligence can be useful in monitoring the situation as it progresses, if for no other reason than to indicate that stronger measures are needed.

  2. Close a Doctor’s Practice to New Patients
    A reactive measure that is commonly used when a panel becomes overloaded is to close it to new patients. The potential loss of income is great. It signals that the practice group has not seen this coming and missed an opportunity to increase revenue capacity just ahead of demand. Even in this situation, analytics can be useful to support panel-setting decisions if the number and even names of potential new patients are tracked.

  3. Add Support Staff
    Support staff can be added to take on clinical or non-clinical activities from an overloaded doctor. Either way, costs increase as a result of not seeing the situation coming in time to eliminate unnecessary work or to offload non-clinical work, as described above. Predictive analytics of trends and simulation capabilities can help a practice get ahead of the curve, and potentially add doctor capacity without adding staff.

  4. Move Patients Away from an Individual Panel
    This may be necessary, but likely signals a severe failure of proactive measures to set appropriate panels or to increase doctor capacity. In some cases, a group of patients may require a higher level of doctor ability, or different set of skills or personality traits. In this case, it is not a deficiency in the ability to predict demand, but a missed factor that could have prevented moving patients to a different panel. This newly discovered factor should be added to the analytical mix. The reason is simple. Most patients are happy with their doctor and are dissatisfied, confused and even angry when they are shifted to another provider (even if they like the new one!).

    Even if the panels are in balance, management can take some of these same actions to continuously improve the practice as a whole, and to improve the working lives of individual providers.  For example, elimination of non-value-added work is a best practice regardless of whether thing are going poorly or well. Trend analysis is another good example. As management gets better at predicting demand, this may allow them to foresee and avoid closing a practice to new patients.

Summary

In the end, providers (doctors, physician assistants, nurse practitioners, registered nurses, et al) pretty much want two things. That is, they want enough time to really care for their patients, and they want to make enough money to support their practices. This is an ever-present tug-of-war, as the first favors smaller panels and the second favors larger panels. Thoughtful analysis must go into setting the appropriate size and shape of panels for individual providers, and for the mix of panels for a group’s practice. Frequent analysis of panel size and complexity is essential to keep pace with changing demographics and with changing business models. This change is becoming more rapid and more disruptive, with such efforts as healthcare reform, accountable care, pay-for-performance, etc. Business intelligence can help by making this frequent analysis possible, and by making the results more meaningful.

Get started now by using analytics to a greater extent in determining, monitoring and continuously refreshing your patient panels.

Thanks for reading!

References:

  1. Murray M., Davies M., Boushon B. Panel Size: How Many Patients Can One Doctor Manage?, American Academy of Family Physicians April 2007. http://www.aafp.org/fpm/2007/0400/p44.html

  2. Bodenheimer T. Primary Care – Will it Survive? New England Journal of Medicine Aug 2006  http://www.nejm.org/doi/full/10.1056/NEJMp068155

  3. Wanless S. Answers Clinical Quality Analytics Application. Resource Management Professionals Jul 2011. More information available from author.
  • Scott WanlessScott Wanless

    Scott is the Healthcare Analytics Director for Cipe Consulting Group. He has more than 30 years of experience in business intelligence strategic planning, analytics application development and business analysis across numerous industries including hospitals, physician groups, healthcare payers, laboratory research, insurance, lending, manufacturing, retail and state government. Scott can be reached at scott.wanless@cipeconsulting.com.

    Editor's note: More healthcare articles, resources, news and events are available in the BeyeNETWORK's Healthcare Channel featuring Scott Wanless and Laura Madsen.

Recent articles by Scott Wanless

 

Comments

Want to post a comment? Login or become a member today!

Be the first to comment!