CMS Levels of Service
The Primary Care Decision Tree has built in logic that allows the computation of the CMS Level of Service, which is used to determine the CPT Code to use in billing. The computation of the Level of Service is a detailed and complex exercise that is documented under the CMS Documentation Guidelines. When a provider is audited by CMS, the documentation of the patient encounter must match the CPT code used in billing or fines and other penalties may be applied, including a demand for a refund. It is thus critical to a practice to make sure that the documentation matches the Level of Service billed.
This chapter will describe the implementation of the Level of Service logic that is the essential part of the Evaluation and Management (E&M) billing module. The description of the logic provided here is based on the current implementation of the CMS Documentation Guidelines as known today. Changes in the guidelines may occur which may change the way the system works as described here
Workflow
The provider will complete the appropriate forms in a decision tree, including History, Examination, Assessment and Plan. After all the steps are complete, he or she clicks on the Document button on the Decision Tree toolbar to review the final documentation. A window like the one shown below titled “Documentation” will appear. This Documentation window has check boxes on the the headings “HPI”, “ROS”, “PH”, “EXAM”. When the check boxes contain a checkmark, that section of the documentation has been completed.
Under the guidelines, there are minimum documentation requirements (such as HPI and EXAM). However, documentation of ROS (Review of Systems), and PH (Past History) is dependent upon the complexity of the case and is not always performed.
The provider is able to review the documentation at this point. If the level is agreeable, select the E&M billing button which opens the E&M Level of Service screen shown below. The screen displays the actual computation of the Level of Service is displayed.
New or Established Patient
Typically, the E&M Level of Service screen displays whether the patient is New or Established. This option may need to be manually selected for a new installation as prior history (the system’s logic for determining whether a patient is new or established) is not established at that point.
The system searches first for the presence of charges, and then for the date of the latest charge. If the latest charge is over three years old, the patient will be classified as “New”. This logic is not foolproof, however, since the CMS guidelines do allow patients not ever seen by a particular doctor (but seen by other doctors in the same practice) to be classified as a new patient. Thus, special discussions must occur in a provider group setting when patients crossover to different doctors.
Type of History
One of the major elements in the computation of the Level of Service is the determination of the type of history that has been documented, which can be classified as “Problem Focused”, “Expanded Problem Focused”, “Detailed”, or “Comprehensive”. These are displayed in the field in order from the lowest level to the highest level.
The first step in determining the level of history is done by checking the boilerplates that have been used in the documentation. Each boilerplate is checked against the file LOS.TBL under the section [LOS CATEGORY] to determine if these belong to HPI, ROS, PH, or EXAM. Only boilerplates listed in LOS.TBL are considered in the computation. The contents of LOS.TBL LOS CATEGORY section are shown below. It is thus important for a tree designer to define all boilerplates that are part of the History and Exam section in LOS.TBL.
[LOS CATEGORY]
HPPRESNT.TXT=HPI
HPTRAUMA.TXT=HPI
HPHEAD.TXT=HPI
HPMEDFUP.TXT=HPI
HPPHYSIC.TXT=HPI
HPGENRC1.TXT=HPI
HPGENRC2.TXT=HPI
HPGENRC3.TXT=HPI
HPGENRC4.TXT=HPI
HPPROSE1.TXT=HPI
HPPROSE2.TXT=HPI
HPPROSE3.TXT=HPI
HPFU1.TXT=HPI
HPFUP1.TXT=HPI
HPFUP2.TXT=HPI
HPHEAD.TXT=HPI
HPBACK.TXT=HPI
HPMEDFUP.TXT=HPI
HPPAIN.TXT=HPI
HPDIZZY.TXT=HPI
HPANXIET.TXT=HPI
HPSYNC.TXT=HPI
HPDIAB.TXT=HPI
HPJOINT.TXT=HPI
HPDIAFUP.TXT=HPI
HPHYP.TXT=HPI
HPURI.TXT=HPI
HPUTI.TXT=HPI
HPSKIN.TXT=HPI
HPFATIG.TXT=HPI
HPGYN.TXT=HPI
HPOBESE.TXT=HPI
HPABDOM.TXT=HPI
HPRHINIT.TXT=HPI
HPCHEST.TXT=HPI
HPCHOLE.TXT=HPI
HPDYSPNE.TXT=HPI
HPMODEL.TXT=HPI
HPNEW1.TXT=HPI
HPNEW2.TXT=HPI
HPNEW3.TXT=HPI
HPMEDS.TXT=HPI
HPFREE.TXT=HPI
RSBRIEFS.TXT=ROS
PHCOMP.TXT=PH
PHHMAINT.TXT=PH
PHPREGH0.TXT=PH
PHMENS.TXT=PH
FHCOMP.TXT=FH
SHCOMP.TXT=SH
PEVITAL.TXT=EXAM
PEBRIEFL.TXT=EXAM
PEBRIEF.TXT=EXAM
PEBRIEFU.TXT=EXAM
PEVITALS.TXT=EXAM
PECONST.TXT=EXAM
PESKIN.TXT=EXAM
PENECK.TXT=EXAM
PECHEST.TXT=EXAM
PELUNGS.TXT=EXAM
PEHEART.TXT=EXAM
PEABDOM.TXT=EXAM
PEMALE.TXT=EXAM
PEFEMALE.TXT=EXAM
PERECTAL.TXT=EXAM
PEBACK.TXT=EXAM
PEUPEX.TXT=EXAM
PELOEX.TXT=EXAM
PENEURO.TXT=EXAM
PEMENTAL.TXT=EXAM
PEIMMUNE.TXT=EXAM
PELYMPH.TXT=EXAM
PECOMP.TXT=EXAM
PEHEAD.TXT=EXAM
PEEYES.TXT=EXAM
PEENT.TXT=EXAM
PEOB1.TXT=EXAM
PEOBP.TXT=EXAM
PEFUNDUS.TXT=EXAM
PHGENE.TXT=PH
PHINFECT.TXT=PH
PHMENSES.TXT=PH
PHOB.TXT=PH
PHPREG1.TXT=PH
PHPREGH.TXT=PH
Scoring History of Present Illness
HPI is computed by checking how many variables are filled in on the QA form. This is easily checked using a count of VARCODES in CLDICTIONARY where the section code starts with the code “HP”. The VARCODES used in the HPI are hardcoded. Thus a tree designer must use only the variable names indicated here in order for the Level of Service to be computed. A score is computed based on the number of variables filled in from the following list:
LOCATION
DURATION
ONSET
QUALITY
TIMING
ASSOCIATEDSS
CONTEXT
SEVERITY
AGGRAVATING or RELIEVING (counted as one)
A score of 9/9 would be the maximum given. A minimum of 1/9 is required in order for Level of Service computation to occur. In reality, the maximum credit is already achieved when there are 4/9 entries or higher.
If designing a custom HPI QA form that does not use the hardcoded variables above, it is possible to allow the system to recognize additional varcodes by making the following override entries in LOS.TBL.
[HPI VARIABLES]
LOCATION2=1
DURATION2=1
ONSET2=1
QUALITY2=1
TIMING2=1
ASSOCIATEDSS2=1
Items defined here will be counted as 1
If one does not use HPI QA’s in their Decision Tree, it is possible to allow the system to provide a default credit for HPI’s. It will then be the responsibility of the provider to alter the final history credit depending on the actual amount of HPI done using some other documentation form (such as dictation). To override, make the following entry in LOS.TBL:
[Override]
HPI Elements=3
This example always gives a credit of 3 HPI elements even though an actual HPI QA form was not filled in.
Scoring Review of Systems
Scoring of Review of Systems is similar to that done for HPI. It is a count of unique varcodes of today’s entries in CLDICTIONARY where the section code is “ROS” is made. However, any varcodes titles may be used here as it is not hardcoded like the HPI. The logic of scoring is based on 3 possible levels which are:
Zero (0/3)
Greater than Zero Varcodes (1/3)
Greater than Two Varcodes (2/3)
Greater than Nine Varcodes (3/3)
Scoring Past History
Past History can now be computed as a category. There are tree types of Past History: Social History, Family History, and Past Medical History. The system determines the extent of past history by checking how many of these options are filled in. 3/3 is the highest score and 0/3 is the lowest score. A score of 2/3 or 3/3 is the maximum. In other words, the presence of at least two history items gives the maximum allowable points. Additionally, users have the ability to entry in a Diagnosis Note when making a Past History entry. If a user enters a value in that isn’t correct, there is an ‘Entry Error’ status that can be applied to it.
Scoring of All of History
After HPI, ROS and Past History are individually scored, the total score will now be applied to the entire History section.
Problem Focused - HPI Score greater than 0
Expanded Problem Focused - HPI Score greater than 1 and ROS Score greater than 0
Detailed - HPI Score greater than 3 and ROS Score greater than 2 and Past History Score greater than 0
Comprehensive - HPI Score greater than 3 and ROS Score = 3 and Past History Score greater than 1
Scoring of Examination
Examination is scored by counting both the number of rows (called elements) and the number of unique varcodes (categories) that are found in CLDICTIONARY in today’s records. The table of scores is:
Problem Focused - Greater than Zero elements and greater than Zero categories
Expanded Problem Focused - Greater than or equal to Six Elements (1/3)
Detailed - Greater than or equal to 12 elements and greater than or equal to two categories (2/3)
Comprehensive - Greater than or equal to 18 elements and greater than or equal to nine categories (3/3)
In a QA Form, an element is a single selection in a choice box. Thus if multiple selections are made in a choice box, each selection counts as an element. It is therefore crucial – to be in full compliance with the CMS DG guidelines – to ensure that the choices are valid elements. This means, do not split up items as multiple choices in a choice box if CMS considers the selected items as only a single item or they may be given more points than is allowed.
It is possible to design a decision tree with fixed credit being given to Examination. If this is done, then it will be the provider’s responsibility to use the correct examination option during E&M CPT Computation. To perform this override, you may enter the following settings in LOS.TBL
[Override]
Exam Elements=18
Exam Levels= 6
You need to use only one of the above since the highest credited value between the two will be used all the time
Note: The computations for CMS Level of Service has been modified. Previously, the counting of scores for achieving a “Comprehensive Exam” was based upon the rule “Organ Systems >= 9 and Elements >= 18”. In 1997 this rule was refined further but the enforcement was left in the selection of choices in the choice box as created by the tree designer rather than by the system. The updated rule is “Organ Systems with at least 2 elements >= 9 and Elements >= 18”. From the use point of view, this means that even if all the organ systems in a Brief Exam QA are filled in, you will receive a comprehensive level exam score only if at least 9 of the items have 2 or more entries. (For testing purposes, on the standard primary care tree, the following items have only one entry upon hitting normals: Constitutional, Rectal, Immunologic). This new behavior may be turned off (thus reverting to 1995 rules) by using the following setting in LOS.TBL in the Outline Folder:
[Override]
1995=1
Decision Making
A major contributor to the Level of Service is the complexity of decision-making. This category is based on the highest level of decision making score of 2 out of 3 of the factors below:
Amount of Complexity
Risk Factors
Management Options
Amount and Complexity of Data
The amount and complexity of data is difficult to approximate. However, one can arrive at a conservative yet appropriate level by assuming that the amount of data to be reviewed is directly proportional to the number of diagnostic reports to be examined. A fixed proportional relationship is assumed between the number of tests ordered from departments that do diagnostic work. The department codes that are responsible for diagnostic test reports are defined in LOS.TBL as follows:
[%Data Complexity]
Count Once="'PATH','NURSE','PROC','PROCS','LAB','RAD','ULTRA',
'CT','MRI','REQ'"
Count Per Order="'TECH','SCHEDP'"
The system will score each diagnostic department based on the above settings. If the department is placed in the “Count Once” category”, any number of tests ordered today from the same department will be counted as one. This is particularly obvious in Lab tests where one lab test contains multiple observations.
If tests are considered especially complex, these should be placed in the Count Per Order category. Remember that this should be done only if the department produces multiple tests that requires the review of different types of data. In the example above, the department “TECH” stands for Ophthalmic technicians who will perform several different types of tests such as Visual Fields, Corneal Topography, Flourescein Angiogram, A-Scan, B-Scan which are all complex and separate processes. Although the nurse may perform several types of tests, generally, in a Primary Care environment these do not result in complex or detailed reports and thus should not be given extra points for multiple occurrences.
Data complexity is scored based upon a count of departments as indicated above and the table below:
Departments <= 1 is a Data Complexity Score of 1/4
Departments = 2 is a Data Complexity Score of 2/4
Departments = 3 is a Data Complexity Score of 3/4
Departments >= 4 is a Data Complexity Score of 4/4
Risk Factors
There are several ways to gauge the patient risks in a given evaluation and management encounter. Some red flags are used. Although this may not entirely describe the complexity of the decision making process, it is a conservative approach and works most of the time. These flags are based on the following:
Prescription Management
Referral to Specialist
Surgical Referral
Management Options (Number and Status of DX)
Prescription Management
Scoring of the risks related to prescription management are as follows:
No Drugs Managed (Risk Level Score = 1)
OTC drugs or given by other provider (Risk Level Score = 2)
Drugs Prescribed by Provider (Risk Level Score = 3)
Referral to Specialist
No Referral (Risk Level Score =1)
Referral Made (Risk Level Score =3)
Surgical Referral or Plan
No Surgical Referral (Risk Level Score =1)
Surgical Referral (Risk level Score = 4)
Final Scoring of Risk Factors
The total score under risk factors is a value of 1/4 to 4/4 and is based on the highest score attained based on the occurrence of any of the items in this category. In order to precisely compute risk, the system makes the judgment based on orders made to specific departments. These departments are coded by risk level as defined in LOS.TBL as follows:
[%Risk]
Moderate Risk="'REF','REFD','IV','INJ'"
High Risk="'SURG'"
Thus, it is important to design trees that use the applicable department code in Orders to apply to appropriate amount of risk. For example, one needs to distinguish between minor surgical procedures vs. major surgical procedures. “Wart Removal”, being a minor procedure should not use a Department Code that belongs to either Moderate or High Risk as defined above. Typically, such a minor surgery is designed as “PROC” department (Procedure). In contrast, an Appendectomy is completely invasive and is a high-risk procedure, which is usually true of all surgeries requiring general anesthesia.
Another area of difference can be in the provider’s actual involvement in a risky decision. The provider who refers a patient to a surgeon for possible appendicitis should probably only get a “Referral” credit (moderate risk). On the other hand, the surgeon, or someone with the skill to determine the actual need to perform surgery would be the one to make the final decision on moving forward with surgery and thus that provider should get the “High Risk” credit. Please note that the provider making the decision to perform the surgery need not perform the surgery itself but only decide to do so to get the risk factor credit. Thus, a surgical attending provider may make the decision, while the actual surgery may be performed by a resident.
Management Options (Number/Status of DX)
First the number and status of active diagnoses in the problem list are computed.
Number of New Diagnosis
Number of Worsening Diagnosis
Number of Improving/Stable Diagnosis
If the number of New Diagnosis is greater than zero, +3 is added to the points.
If the number of Worsening Diagnosis is greater than 0, +2 is added to the points.
If the number of Improving/Stable Diagnosis is greater than 1, +2 is added to the points.
Total Management options score is based on the following table:
1 /4 score = Up to 1 point
2 /4 score = 2 or more points
3 /4 score = 3 or more points
4 /4 score = 4 or more points
Final Scoring of Decision Making
Final scoring of the decision making is based upon the highest two of three scores in the following categories:
Amount of Complexity
Risk Factors
Management Options
If the highest scoring category is by itself, the points should be moved down until 2 of 3 categories are at the same level.
The final level of Decision Making is based on the following table:
Straightforward (Score 1 /4)
Low Complexity (score 2 /4)
Medium Complexity (Score 3 /4)
High Complexity (Score 4 /4)
Computation of Level of Service and Final CPT Code
The Level of Service computation is used to determine the actual CPT code to be used. This formula varies depending on the patient’s status as a new or established patient
New Patients – Office Visit
If Type of History is Problem Focused and Type of Exam is Problem Focused and Decision Making is Straightforward, then the CPT code to use is “99201”.
If Type of History is Expanded Problem Focused and Type of Exam is Expanded Problem Focused and Decision Making is Low Complexity, then the CPT code to use is “99202”.
If Type of History is Comprehensive and Type of Exam is Comprehensive and Decision Making is Moderate Complexity, then the CPT code to use is “99204”.
If Type of History is Comprehensive and Type of Exam is Comprehensive and Decision Making is High Complexity, then the CPT code to use is “99205”.
Established Patients – Office Visit
If two out of three of the following categories is true:
Type of History is Problem Focused and Type of Exam is Problem Focused and Decision Making is Straightforward, then the CPT code to use is “99212”.
If two out of three of the following categories is true:
Type of History is Expanded Problem Focused and Type of Exam is Expanded Problem Focused and Decision Making is Low Complexity, then the CPT code to use is “99213”.
If two out of three of the following categories is true:
Type of History is Detailed and Type of Exam is Detailed and Decision Making is Low Complexity, then the CPT code to use is “99214”.
If two out of three of the following categories is true:
If Type of History is Comprehensive and Type of Exam is Comprehensive and Decision Making is Moderate Complexity, then the CPT code to use is “99215”.
Override Due to Counseling Time
The Level of Service, as computed by the complex formula described above can be overridden if face to face counseling time with the patient exceeds a certain level. The rule is as follows:
New Patients
First the counseling time spent face to face with the patient must be more than 50% of the total visit time. If this is the case, the following table can be used to determine Level of Service:
99201 - Level 1 – Less than 20 minutes of counseling
99202 - Level 2 – 20 – 29 minutes of counseling
99203 - Level 3 – 30 – 44 minutes of counseling
99204 - Level 4 – 45 – 59 minutes of counseling
99205 - Level 5 – 60 or more minutes of counseling
Established Patients
The counseling time spent face to face with the patient must be more than 50% of the total visit time. If this is the case, the following table can be used to determine Level of Service:
99211 - Level 1 – Less than 10 minutes of counseling
99212 - Level 2 – 10 – 14 minutes of counseling
99213 - Level 3 – 15 – 24 minutes of counseling
99214 - Level 4 – 25 – 40 minutes of counseling
99215 – Level 5 – 40 or more minutes of counseling