Medical Billing Forum

General Category => General Questions => Topic started by: rmk123 on January 16, 2013, 04:45:39 PM

Title: Performance Standards/Workload
Post by: rmk123 on January 16, 2013, 04:45:39 PM
Revising workload performance standards for billers who handle the process from insurance verification through resolution. Tasks are not divided by person but one person performs all tasks. I have located industry standards by task which is helpful. I am challenged, though, on how to create a hybrid of this for our model. 

I have the industry standards for measuring success on the AR management/collection side. This will be coupled with the workload goals.

What are your suggestions for an effective system?

Title: Re: Performance Standards/Workload
Post by: RichardP on January 16, 2013, 11:17:43 PM
Are you in any position to be able to do time and motion studies?  That would be the only reliable way to collect your data (as opposed to somebody else's data).  What specialties do you handle: chiropractor; social worker; psychiatrist; heart surgeon?  How much of your practice (practices?) is Medicare? Medicaid? Worker's Comp.? HMO?  Does your doctor have a Physician Office Lab that you have to bill for?  Hopefully you get the point.  The billing work involved is going to vary based upon how busy your doctor(s) is/are, what their specialty is, what insurance carriers are involved, etc.  There is no list of performance standards out there that accurately reflects your unique set of circumstances.

Having said that, a rule of thumb is that one person who knows what they are doing should be able to handle two doctors from start to finish.  That runs more true than not for my business, and the majority of my clients have Physician Office Labs that we bill for.  We employ roughly one person for every two clients we have.  Some times are busier than others, so there is variation in the workload at times.  But we use slack times to catch up on things we maybe put off during busier times.
Title: Re: Performance Standards/Workload
Post by: PMRNC on January 17, 2013, 09:36:35 AM
Are you in any position to be able to do time and motion studies?

Yes, I call these work flow studies. More and more companies are very successful using this method. It's one of the things I do in my consulting for larger offices. I go in and cross train all of the admin department, this way each person knows what the other person is doing, it kills a few birds with one stone in the long run if someone is out for extended period of time, work doesn't stop. It also of course increases efficiency as well. The staff really likes it too, as it gives each of them a break from their normal routine. :)
Title: Re: Performance Standards/Workload
Post by: RichardP on January 17, 2013, 03:46:11 PM
Just to be clear - a work-flow study is different from a time and motion study.

A work-flow study follows the work on its entire path from the time it comes into the office until the time it leaves - looking for any bottle-necks or any ways the path might be made more efficient.

A time and motion study looks at repetitive tasks and times them.  It then looks at the motions required to carry out the task and analyzes whether anything could be changed to cut down on the motions required to carry out the task.  An example of this might be having your clerk type in three weeks worth of diagnosis and proceedure codes for a specific doctor, and time how long that takes.  Then divide the time it takes by the number of fee slips to get a time-per-fee-slip figure.  Another example would be timing how long it takes someone to highlight, cut, and then paste something using the mouse (looking at range of motions).  Then time them doing the same thing using key-strokes on the keyboard (usually faster than reaching for the mouse).  When you arrive at a time-per-unit figure, you can multiply that figure by the number of instances of that task to get a ballpark idea of how long it should take to get the job done.  Folks who consistantly exceed the norm can get paid more.  Folks who consistantly fall below the norm can be replaced by those who can at least match the norm.

As I pointed out in my previous response - the norms vary, based on all the things I listed in my previous response, and more - such as specialty of doctor, type of insurance, whether the doctor does his own labs, etc.  But, if you have people working for you, and you have not established these types of norms, it is difficult to tell whether you are getting a fair level of work from your employees for a fair level of pay.
Title: Re: Performance Standards/Workload
Post by: QueenAlicia on January 18, 2013, 12:09:42 PM
How do you implement these studies?  Are there a bunch of charts and surveys?
Title: Re: Performance Standards/Workload
Post by: RichardP on January 19, 2013, 06:18:25 PM
If you are a large corporation, you hire a consulting firm.  They send over lots of people who sit and stare at the workers with stop watches in hand.  They also have the workers fill out surveys.  At the end, when they present the results of their studies to top management, they will have lots of pretty charts.

Us, not so much.  The following discussion is useful only for situations where you have employees.  And it is going to be boring unless you are really interested in this stuff.  If you do all the work yourself, there is not much point in getting involved in the detail described below.

A person who has been billing for a while should have a good idea of all tasks involved in working a client's billing cycle from A to Z (whatever that might be for your client).  If you don't have a listing of all of your tasks for your client, creating such a list would be the place to start.  Identify every single thing you do from start to finish (however you define start, and wherever your finish is).  This listing should contain everything you do to get your client paid for seeing a patient - starting with wherever you start (taking the phone call asking for an appointment?).  And it should end with the tasks involved in working accounts receivable all the way through sending someone to collections and writing off their balance.

Once you have your list assembled, and the separate events listed in some logical sequence of occurance - you need to break the list up into a.) those events where you have absolute control over the time you spend on them (e.g., how fast you can type the codes from a fee slip into your computer); b.) those events where you have absolutely no control over the time you spend on them (e.g., on hold with the insurance carrier as they check the status of a claim), and; c.) all the events that fall in between these two extremes.

A.  Analyze your list of events where you have no control over how much time you spend on them.  Is there another way of doing this task that would give you (some? total?) control over how much time you spend on it?  For example, could you log on to the carrier's website to check the status of a claim rather than calling and being put on hold (time each method for a month, and do whichever has the fastest average time)?  Can you afford software, or a service, that would automatically call and remind patients that payment is overdue?

B.  Analyze your list of tasks where you have some, but not total, control over the time you spend on them.  Are there other ways of doing these tasks that would give you more control over the time you spend on them?  Maybe there is no way to do them that would give you more control over the time you spend on them.  But without first identifying which tasks these are, and really looking at how you are doing them versus how you could be doing them differently, you won't really know.

C.  Analyze your list of tasks where you have total control over the time you spend on them.  This will be the area where you do most of your time and motion study.  The simplest example for this category is determining how long it should take to type into the computer the diagnosis and procedure codes off of a fee slip.  I know that an EMR can pass these codes directly into a Practice Management system; I'm just using this as an example.  Take this example and apply the principles to whatever is actually happening in your situation.

We have for clients Ob/Gyn, Heart specialists, Lung specialists, General Practitioners, Proctologists, and Pediatricians.  Each of these specialties code differently, both in types of codes and modifiers required, as well as the total number of codes used.  Most of our clients have physician office labs, and so those codes are also on our fee slips - with the number of labs varying among specialties.

I create an Excel spreadsheet for each employee I wish to track.  On a given spreadsheet, each doctor has his own heading.  Under each doctor heading I have a column for Date, Start Time, Stop Time (entered by employee), Total Time (calculated by formula), Number of Fee Slips (entered by me, not the employee), and Time per Fee Slip (calculated by formula).

                                   Doctor 1

             Start      Stop     Total    Number of    Time per
Date      Time      Time     Time     Fee Slips      Fee Slip

xxxx       xxx        xxxx      xxx         xxxx           xxxx
xxxx       xxx        xxxx      xxx         xxxx           xxxx
xxxx       xxx        xxxx      xxx         xxxx           xxxx
xxxx       xxx        xxxx      xxx         xxxx           xxxx
xxxx       xxx        xxxx      xxx         xxxx           xxxx

  Total:                             xxx         xxxx           xxxx

When I'm doing these tests, I collect a weeks worth of fee slips for the target doctor and give them to the person I'm testing.  There are 30-40 fee slips per day (average = 35) times 5 days = about 175 total.  I give instructions to type at a sustainable pace that they can keep up all day, and then I work to keep their distractions to a minimum.  We then type the relevant figures into the spreadsheet - which gives us the time per fee slip for that day, for that employee, for that doctor.  Over the course of three or four months I will do this with each employee, testing them on all of our doctors, five or six times per doctor.  Since the number of codes per fee slip varies from doctor to doctor, I expect that the average time per fee slip will be different for each doctor.  By testing all employees, with all doctors, and doing this a few times a year, we can calculate an average time per fee slip for each doctor.

Looking at the table provided above, you can see that I calculate across for any given day.  But I can also calculate down to get a total in each column.  So - for a given employee and a given doctor - I can get say five separate data points:  175 fee slips on each of 5 separate testing dates = 875 fee slips total.  X number of minutes on each of 5 separate testing dates = xx number of minutes total.  By re-testing over time, I can weed out the effect of an employee feeling particularly slow or particularly energetic on any given testing day.  This also weeds out the effect of a fee slip having more or less codes on it than normal.  The average I end up with is a good number for that employee.  I collect this number for each employee.  The average of all of them gives me a benchmark number of how long it should take, on average, to process one fee slip for a particular doctor.  When I hand out work, I know how many fee slips I have assigned, times the amount of time it should take to do one fee slip, equals how long it should take the employee to complete the task.  There will be variations on any given day.  But at least I have a legitimate number in my head to use as a benchmark over time - that I can use to identify those performing below standard, and those performing above standard.

I'm a pretty fast typer and I run these same tests on myself.  I know how long on average it takes me to type the fee slip codes into the computer for a given doctor.  When someone's average time differs from mine significantly, I can either counsel them to pick up the pace, or maybe find different work for them to do, or maybe replace them altogether.

That's probably more than enough for now.  Come back with any questions.
Title: Re: Performance Standards/Workload
Post by: rmk123 on January 22, 2013, 12:33:17 PM
Thanks for the comments, points, and discussion!