 |
|
Evaluation and Management Services
A Comparison of Medical Record Documentation With Actual Billing in Community Family Practice
George E. Kikano, MD;
Meredith A. Goodwin, MS;
Kurt C. Stange, MD, PhD
Arch Fam Med. 2000;9:68-71.
ABSTRACT
| |
Objective To compare the concordance of family physicians' billing for evaluation and management services with medical record documentation.
Design Multimethod, cross-sectional observation study.
Setting Eighty-four family practices in northeast Ohio.
Participants Four thousand fifty-four outpatients visiting 138 family physicians.
Main Outcome Measure The degree of concordance between evaluation and management Current Procedural Terminology codes billed by physicians, with those codes assigned by trained research nurses using American Medical Association criteria to code medical records for the same visits.
Results Discrepancies between the multifactorial nature of family practice outpatient visits and the Current Procedural Terminology coding criteria, which dictate overcoding for depth rather than breadth, made coding difficult (multiple-rater  statistic between research nurses = 0.36). Among 4137 outpatient visits with complete billing information, 57% of the Current Procedural Terminology codes generated by medical record review were concordant with the actual billing code assigned by physicians. Undercoding and overcoding occurred at a similar frequency (21% and 19%, respectively) and differed by more than 1 code in fewer than 4% of visits. Visits by new patients were more likely to be inaccurately coded than visits by established patients.
Conclusions Record documentation by community family physicians largely reflects the level of services billed using evaluation and management codes. Undercoding is as common as overcoding. Efforts from regulatory agencies should be redirected from penalizing physicians for overcoding to focusing on the development of coding criteria that reflect the multifactorial nature of outpatient primary care practice.
INTRODUCTION
HISTORICALLY, physicians have been accustomed to billing for their clinical services based on the time spent with the patient.1 Prior to 1992, there were no uniform guidelines for appropriate billing levels for physicians' professional services. In 1992, the American Medical Association (AMA) and the Health Care Financing Administration (HCFA) jointly developed coding criteria for evaluation and management (E&M) services.2 The purpose of these guidelines was to standardize claims for different levels of service provided by physicians in various settings. These criteria for E&M services were revised in 1994.3-4 To make the E&M guidelines more explicit and to reflect the content of services provided across different specialties, a second revision was jointly developed by the AMA, medical society representatives, and HCFA in May 1997.5 Implementation of these revised guidelines was initially delayed from January to July 1998, and under pressure from the AMA and other physician groups, exclusive implementation was suspended pending further discussions.
The HCFA is planning to spend considerable resources in monitoring compliance with E&M guidelines by auditing physicians' billing practices and reviewing record documentation (Medicare newsletter. Nationwide Insurance Enterprise; Columbus, Ohio. March 1998). This subject has generated considerable concern among practicing physicians. According to AMA Past President Percy Wootton, MD, the new E&M guidelines are "one of the biggest issues on the minds of practicing physicians at the moment."6 Audits for compliance with the guidelines are based on reviews of medical records for outpatient visits and comparisons of the level of documentation with claims submitted to local Medicare carriers. Documentation in medical records is therefore the standard by which appropriate E&M coding is judged. While physicians agree that proper documentation is essential for good patient care and well-maintained records facilitate communication between physicians, the new guidelines are widely seen as cumbersome and distracting from patient care (Neil H. Brooks, MD, written communication, May 1998).
In the 1994 guidelines, which were in effect during our study, E&M code selection included 7 components (history, examination, medical decision making, counseling, coordination of care, nature of presenting problems, and time). The first 3 components were the essential factors that determine the level of billing for E&M services. The guidelines did not provide explicit criteria on how to code for the common primary care outpatient visit scenario in which multiple problems are addressed to a limited degree. Time spent on counseling only affected coding if it accounted for more than 50% of the visit.4 The proposed 1997 documentation guidelines for E&M services still described 7 components for the level of services. In the revised guidelines, examination is defined by both single-organ system and multisystem examination.5, 7
Our previous work identified patient and visit characteristics that are associated with discordance between billed Current Procedural Terminology (CPT) codes and those assigned by direct observation of outpatient visits by established patients.8 Since HCFA's reimbursement and compliance monitoring is based on record review, we now extend this line of inquiry by comparing family physicians' billing for outpatient E&M services with documentation in the medical records. The objective of this analysis is to compare the concordance of the level of billed E&M services with actual chart documentation for the same visits. Our hypothesis is that family physicians properly document the level of services they bill for in most visits.
METHODS
This analysis is part of the larger Direct Observation of Primary Care (DOPC) study, which examined 4454 outpatient visits to family physicians in northeast Ohio. The methods of the DOPC have been described in detail previously.9-10 Briefly, family physician members of the Ohio Academy of Family Physicians practicing in northeast Ohio were invited to participate in a study of the content of family practice. Of the 531 physicians contacted, 138 volunteered to participate. Consecutive outpatients seen by the participating physicians over 2 different days of observation by research nurses were enrolled if the patient gave verbal informed consent.
For the larger study, research nurses collected data on the content and context of patient visits using multiple methods that included direct observation of the physician-patient encounter, patient exit questionnaire, medical record review, and a practice environment checklist. Data on the participating physicians were collected using a questionnaire that physicians completed following their participation in the study. Information on billing CPT codes for the observed visit was obtained from office staff subsequent to the observation days. Patients were classified as new patients if they had not been seen in the practice within the past 3 years, and as established patients if they had at least 1 visit during the past 3 years.3
Based on medical record review, a CPT code was assigned for each visit by the research nurse using established guidelines.3 Research nurses were trained on E&M coding using the Physicians' CPT: CPT '94, Clinical Examples Supplement4 to enhance validity and interrater reliability of the nurse-assigned CPT codes. The nurses were blinded to the actual billing for each visit at the time of the medical record review.
The main outcome measure for the study was the degree of concordance between the actual billing CPT code and the nurse-assigned CPT code based on medical record review. The E&M codes, ranging from 99211 to 99215 for visits by new patients and from 99201 to 99205 for established patient visits, were rank ordered from 1 to 5, reflecting the level of complexity of the visit. A difference score was created by subtracting the E&M code assigned by the research nurse based on medical record review from the actual E&M code billed for the visit. A positive difference score indicated overcoding, meaning that the CPT code billed for the visit was higher than the nurse-assigned CPT code based on medical record review of the visit. A negative difference score was indicative of undercoding (the actual CPT code that was billed was lower than the nurse-assigned CPT code). Additional stratified comparisons were performed for visits by new and established patients, and the difference in their distributions was tested using the  2 statistic.
RESULTS
Characteristics of the participating physicians (N = 138) have been described in detail elsewhere.10 The physicians in the DOPC study were demographically similar in age and percentage in rural and solo practice to active practicing members of the American Academy of Family Physicians in mean age and number of patients seen per week.11 They differed by having a higher percentage of female and residency-trained physicians, which reflects recent trends in the characteristics of family physicians.
A total of 4454 patient visits were observed for the DOPC study. Of these, 238 were missing billing data, 60 did not have medical records available, and 13 were missing both data elements. Of the remaining 4137 visits, 355 (9%) were made by new patients and 3782 (91%) were made by established patients who had seen their physician on average 4.3 times in the past year.10
As previously described,9-10 the mean age of patients in the sample was 41 years; 88% were white and 62% were female. These characteristics are similar to national samples of patients seeing family physicians.12 On a 5-point Likert-type scale, the mean self-reported health status was 3.8, indicating that individuals on average reported being in very good health. Nearly all of the patients in the sample (93%) had some form of health insurance, with the most common types being managed care (36%), Medicare (23%), and fee-for-service (20%) insurance. At the time of this study, managed care insurance represented predominantly preferred provider organizations in this market.13 More than half (58%) of the visits were for the treatment of an acute illness, 23% were for the treatment of a chronic illness, 12% were for well care, and 7% were for other reasons.
The interrater reliability for CPT codes assigned by 8 research nurses based on medical record review was fair ( = 0.36).9 As shown in Figure 1, discordance between the CPT code actually billed and the nurse-assigned CPT code based on medical record review ranged from -4 to +3, and was evenly distributed between undercoding and overcoding. In more than half of the visits (57%), the actual CPT code and the nurse-assigned code were concordant (difference score = 0). Billing differed by more than 1 code in fewer than 4% of visits. For the 355 visits by new patients, billing and nurse-assigned coding were concordant for 46% of visits, with 24% undercoded and 30% overcoded. For the 3782 visits by established patients, billing and nurse-assigned coding were concordant for 58% of visits, with 22% undercoded and 20% overcoded. The distributions of undercoding and overcoding were significantly different between visits by new and established patients (P<.001).
|
|
|
|
Difference between actual Current Procedural Terminology code billed and Current Procedural Terminology code based on medical record review (N = 4454) of 4137 outpatient visits with complete billing information.
|
|
|
COMMENT
Our findings show that family physicians' medical record documentation is correspondent with their billing practices for outpatient services in more than half of the visits reviewed. Compared with billing codes assigned by standardized medical record review, undercoding and overcoding occur with similar frequency, and differences are usually limited to one level of complexity different from the expected CPT code. Undercoding is more common than overcoding for visits by established patients, who constitute most outpatient visits to family physicians. For visits by new patients, both undercoding and overcoding are more common than for established patients. In addition, overcoding is more common than undercoding for new patient visits. Despite the higher fees for new patient visits, this may reflect physicians overcoding for the additional administrative burden of caring for new patients, which is not fully reflected in the medical record for the visit.
The moderately low interreliability rate in the assignment of CPT codes by highly trained research nurses based on medical record review is lower than the reliability of the same nurses using direct observation of outpatient visits.14 This shows the difficulty of coding the visit based only on medical record review. In addition, the 1994 criteria for higher CPT coding were based largely on the provision of greater depth of services for a single problem. These criteria accurately reflect the provision of specialty services and some primary care visits in which the focus is a single problem. However, the criteria are inadequate for most family practice visits in which a higher level of service often involves multiple problems and integrating the delivery of a wide range of services.10 Family practice care of patients often creates depth of service by addressing multiple problems to a limited degree during each visit and by seeing patients and families for multiple visits over time.14-15
The high level of concordance in our study occurred during the period in which E&M guidelines from 1994 were being used. These guidelines were vague as to what was expected for physicians' documentation when compared with the revised 1997 guidelines. In October 1997, HCFA started a prepayment audit program for E&M services with a random sample of all claims submitted. The initial results showed that 40% of claims were accurate, 40% were denied owing to insufficient documentation, and 20% were paid at the lower complexity level.16 These reviews by HCFA used both the old (1994) and the new (1997) guidelines to assign E&M codes.
Physician groups have been concerned with the effect of the complex new guidelines and their lack of fit with the practices and documentation needed for good patient care.6 More time can potentially be spent worrying about proper chart documentation to comply with regulations at the expense of time usually spent with patients. This shift coupled with more administrative tasks required by Medicare and managed care organizations can potentially compromise the physician-patient relationship.17-18 The AMA has been actively working with HCFA and other groups to assure proper application of the billing guidelines. At their February 1998 board meeting, the AMA Board of Trustees approved an action plan consisting of (1) advocacy to prevent physicians from undergoing penalties for inadvertent errors, (2) correcting problems by ensuring that the guidelines reflect actual clinical practice, and (3) physician education to understand the new E&M guidelines (Medicare newsletter. Columbus, Ohio: Nationwide Insurance Enterprise; March 1998). The American Academy of Family Physicians declared the new guidelines "too flawed to be fixed" and lobbied for delaying exclusive implementation. The efforts of the AMA in collaboration with other medical societies led to HCFA suspending implementation of the revised guidelines. Based on input from physicians and medical organizations, a new group has been established to evaluate and modify the guidelines before they are implemented. Pilot testing before implementation is being considered (Neil H. Brooks, MD, written communication, May 1998).
The findings of this analysis are limited by studying family physicians from only 1 geographic area. However, the new E&M guidelines are universal for the entire country and for all physician groups, and the findings are likely to be generalizable to other practicing family physicians. The research nurses were trained only to abstract data for E&M codes, excluding codes for other services such as preventive services and procedures. Therefore, our findings apply only to E&M coding.
Our findings show that even in the absence of systematic education regarding E&M guidelines, physicians are generally accurate in billing for their services. Additional factors that add to the complexity of the services provided are difficult to capture from chart reviews. Such factors include physicians' prior knowledge of their established patients' medical and family histories. In addition, during 18% of visits, family physicians provide incidental care to family members, which is not reflected in billing for the identified patient.19
With the high concordance of billing and CPT codes assigned by chart review and the lack of a systematic overcoding or undercoding bias, it might be more fruitful to redirect HCFA's efforts toward physician education. In addition, those undertaking future revisions of billing E&M guidelines should consider the unique aspects of primary care and family practice, in which greater visit complexity and work are often represented by breadth of care, as well as depth of care for narrowly defined diseases.
AUTHOR INFORMATION
Accepted for publication June 28, 1999.
This study was supported by grant 1R01 CA60682 from the National Cancer Institute, Bethesda, Md, and by a Robert Wood Johnson Foundation, Princeton, NJ, Generalist Physician Faculty Scholar Award (Dr Stange).
We are grateful to the physicians and office staff members of the Research Association of Practicing Physicians and their patients, without whose participation this study would not have been possible.
| Editor's Note: Just get a group of family physicians together and ask them to code written notes from visits and you will quickly know how poorly reproducible the coding system is! I agree with the American Academy of Family Physicians that the new guidelines are "too flawed to be fixed" and take time away from what we want to be doing: spending time with the patient.—Marjorie A. Bowman, MD, MPA
|
|
Corresponding author: George E. Kikano, MD, Department of Family Medicine, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH 44106-4950 (e-mail: gek{at}po.cwru.edu).
From the Departments of Family Medicine (Drs Kikano and Stange and Ms Goodwin) and Epidemiology and Biostatistics (Dr Stange) and the Ireland Cancer Center (Drs Kikano and Stange), Case Western Reserve University, University Hospitals of Cleveland, Cleveland, Ohio.
REFERENCES
| |
1. Hirschl N. Eternal/mystery or essential/mastery: evaluation and management coding for physician services. J Am Health Information Manage Assoc. 1995;66:14-15, 18.
2. Kirschner CG, Edwards NK, May DN, et al. CPT: Physicians' Current Procedural Terminology. Chicago, Ill: American Medical Association; 1991.
3. Kirschner CG, Burkett RC, Coy JA, et al. Physicians' Current Procedural Terminology: CPT '94. Chicago, Ill: American Medical Association; 1994.
4. Kirschner CG, Burkett RC, Coy JA, et al. Physicians' Current Procedural Terminology: CPT '94, Clinical Examples Supplement. Chicago, Ill: American Medical Association; 1994.
5. Revised documentation guidelines for evaluation and management services. CPT Assistant. 1998;7:1-23.
6. Getting the facts about E&M. American Medical News. March 23/30, 1998;41(special issue):1A-4A.
7. Moore KJ, Henry LA. Exam documentation just got harder. Fam Pract Manage. 1997;475-82, 85.
8. Chao J, Gillanders WG, Flocke SA, et al. Billing for physician's services: a comparison of actual billing with CPT codes assigned by direct observation. J Fam Pract. 1998;47:28-32.
ISI
| PUBMED
9. Stange KC, Zyzanski SJ, Smith TF, et al. How valid are medical records and patient questionnaires for physician profiling and health service research? a comparison with direct observation of patient visits. Med Care. 1998;36:851-867.
FULL TEXT
|
ISI
| PUBMED
10. Stange KC, Zyzanski SJ, Flocke SA, et al. Illuminating the black box: a description of 4454 patient visits to 138 family physicians. J Fam Pract. 1998;46:377-384.
ISI
| PUBMED
11. American Academy of Family Physicians. Facts About Family Practice. Kansas City, Mo: American Academy of Family Physicians; 1996.
12. Schappert SM. National ambulatory medical care survey, 1994 summary: advance data from vital and health statistics [abstract]. Adv Data. 1996;273:1.
13. Baxter RJ, Kohn LT, Omata RK, Williams C. Health System Change in Cleveland, Ohio: A Case Study. Washington, DC: Center for Studying Health System Change; 1997.
14. Medalie JH, Zyzanski SJ, Langa DM, Stange KC. The family in family practice: is it a reality? J Fam Pract. 1998;46:390-396.
PUBMED
15. Stange KC, Jaén CR, Flocke SA, Miller WL, Crabtree BF, Zyzanski SJ. The value of a family physician. J Fam Pract. 1998;46:363-368.
ISI
| PUBMED
16. Martin S. Pay denied in E&M audits. American Medical News. April 6, 1998;41:1, 31, 34.
17. Emanuel EJ, Dubler NN. Preserving the physician-patient relationship in the era of managed care. JAMA. 1995;273:323-329.
FREE FULL TEXT
18. Emanuel EJ, Brett AS. Managed competition and the patient-physician relationship. N Engl J Med. 1993;329:879-882.
FREE FULL TEXT
19. Flocke SA, Goodwin MA, Stange KC. The effects of a secondary patient on the family practice visit. J Fam Pract. 1998;46:429-434.
ISI
| PUBMED
RELATED ARTICLE
The Archives of Family Medicine Continuing Medical Education Program
Arch Fam Med. 2000;9(1):79-80.
FULL TEXT
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
Inadequate Training in Billing and Coding as Perceived by Recent Pediatric Graduates
Andreae et al.
CLIN PEDIATR 2009;48:939-944.
ABSTRACT
Template-Guided Versus Undirected Written Medical Documentation: A Prospective, Randomized Trial in a Family Medicine Residency Clinic
Mulvehill et al.
J Am Board Fam Med 2005;18:464-469.
ABSTRACT
| FULL TEXT
Medical Illness and the Risk of Suicide in the Elderly
Juurlink et al.
Arch Intern Med 2004;164:1179-1184.
ABSTRACT
| FULL TEXT
Expert Agreement in Current Procedural Terminology Evaluation and Management Coding
King et al.
Arch Intern Med 2002;162:316-320.
ABSTRACT
| FULL TEXT
Is Diabetes Treated as an Acute or Chronic Illness in Community Family Practice?
Yawn et al.
Diabetes Care 2001;24:1390-1396.
ABSTRACT
| FULL TEXT
|
