- Evaluation of Inpatient Clinical Documentation Readiness for ICD-10-CM
- Leveraging the Cloud for Electronic Health Record Access
- Factors in Medical Student Beliefs about Electronic Health Record Use
- Justice-Involved Health Information: Policy and Practice Advances in Connecticut
- Giving Raw Data a Chance to Talk: A Demonstration of Exploratory Visual Analytics with a Pediatric Research Database Using Microsoft Live Labs Pivot to Promote Cohort Discovery, Research, and Quality Assessment
- At the Intersection of Health and Justice
- Health Information Exchange between Jails and Their Communities: A Bridge That Is Needed under Healthcare Reform
- Winter 2014 Introduction
by Dilhari R. DeAlmeida, PhD, RHIA; Valerie J. Watzlaf, PhD, RHIA, FAHIMA; Patti Anania-Firouzan, MSIS, RHIA; Otto Salguero, MPH, PhD; Elaine Rubinstein, PhD; Mervat Abdelhak, PhD, RHIA, FAHIMA; and Bambang Parmanto, PhD
This research study examined the gaps in documentation that occur when coding in ICD-10-CM. More than 4,000 diagnoses from all chapters were coded from 656 electronic documents obtained from a large integrated healthcare facility at the time the study was conducted (2012). After the documents were coded, areas for documentation improvement were identified for chapters that resulted in deficiencies in documentation, and a quick reference guide was developed.
The overall absent documentation percentage was 15.4 percent. The 10 chapters with the highest percentage of absent documentation were chapter 7 (Diseases of Eye and Adnexa), with 67.65 percent (p < .001); chapter 8 (Diseases of Ear and Mastoid Process), with 63.64 percent (p < .001); chapter 13 (Diseases of the Musculoskeletal System and Connective Tissue), with 46.05 percent (p < .001); chapter 14 (Diseases of the Genitourinary System), with 40.29 percent (p < .001); chapter 10 (Diseases of Respiratory System), with 35.52 percent (p < .001); chapter 1 (Infectious and Parasitic Diseases), with 32.88 percent (p < .001); chapter 12 (Diseases of the Skin and Subcutaneous Tissue), with 32.35 percent (p < .001); chapter 2 (Neoplasms), with 25.45 percent (p < .001); chapter 4 (Endocrine, Nutritional and Metabolic Diseases), with 14.58 percent (p < .001); and chapter 17 (Congenital Malformations, Deformations, and Chromosomal Abnormalities), with 12.50 percent.
We addressed the deficient areas in the quick reference guide developed for clinicians and technology vendors. Having complete and accurate documentation would benefit both the clinician and the patient in providing the highest quality of care.
Keywords: ICD-10-CM (International Classification of Diseases, Tenth Revision, Clinical Modification), clinical documentation improvement, reimbursement
Having accurate and up-to date documentation is vital as the US healthcare system approaches the transition from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) to the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM). ICD-10-CM has many more codes available to choose from, along with a higher degree of specificity. For coding to be accurate, the documentation needs to be in place in the medical record. Dimick (2011) identified a list of 10 areas that had documentation issues according to feedback from facilities that were in the process of transitioning to ICD-10-CM and the International Classification of Diseases, Tenth Revision, Procedure Coding System (ICD-10-PCS).1 Those areas included diabetes mellitus, injuries, drug underdosing, cerebral infarctions, acute myocardial infarction, neoplasms, musculoskeletal conditions, pregnancy, and respiratory/ventilators, and all of ICD-10-PCS. Furthermore, Moczygemba and Fenton (2012) evaluated the clinical documentation needs in specific areas (heart disease, pneumonia, and diabetes) and highlighted the importance of having accurate clinical documentation and identifying gaps along with the importance of coder training and education.2 Documentation needs must be reevaluated in every healthcare setting in preparation for the transition to the ICD-10-CM coding system in 2014.
Adoption of ICD-10-CM also would facilitate international comparisons of quality of care and the sharing of best practices globally. Overall, ICD-10-CM is more effective at capturing public health diseases than ICD-9-CM. It is more specific and more fully captures nationally reportable public health diseases.3 ICD-10-CM’s increased specificity offers payers and providers the potential for considerable cost savings through more accurate trend forecasting and cost analysis. Greater detail can improve payers’ abilities to forecast healthcare needs and trends and analyze costs.4
It is important that US healthcare providers learn from the challenges that have arisen in the implementation of the International Classification of Diseases, Tenth Revision (ICD-10) in other countries. The major challenges in the transition to ICD-10-CA (the Canadian version) were the fact that the entire coding system transformed from paper to electronic, the need for coder education, and the lack of professional coders.5 A study of ICD-10 implementation in Australia (ICD-10-AM) identified similar attributes needed for a smooth implementation: education, and early preparation and planning by clinicians and workgroups.6 Both of these studies found that coders took approximately four to six months to regain their pre–ICD-10 coding productivity.
The purpose of this study is to identify the barriers related to documentation specificity, identify absent documentation across all 21 ICD-10-CM chapters, and develop a documentation improvement tool kit for providers and technology experts. The findings of the research could alert physicians and other documentation specialists as to what, if any, practices need to be changed in order to obtain accurate coding. Furthermore, the results could be used in the development of better inpatient computer-assisted coding (CAC) products. The industry needs automated solutions to allow the coding process to become more productive, efficient, accurate, and consistent. CAC in outpatient care has been well researched and studied, and several successful software products are being used, while the application of CAC to inpatient care is still minimal because it requires a more complex set of tools.
A descriptive research study using quantitative methods was conducted; the study focused on coding electronic documents across each major diagnostic category for ICD-10-CM. Each of the records was categorized into each of the ICD-10-CM chapters. Coding was performed using the 2011 version of the ICD-10-CM draft manual that was available at the time of the study. A thorough investigation of the de-identified database which included data from an integrated healthcare system identified a total of 656 electronic inpatient documents with a total of 4,791 diagnoses. To study the entire available population, the researcher decided to code the entire data set (4,791 diagnoses) (See Table 1). The electronic document consisted of multiple sections (Chief Complaint, History of Present Illness, Past Medical History/Family History/Social History, Review of Symptoms, Physical Exam, Labs and Studies, and Assessment and Plan).
These documents consisted of a combination of structured (Clinical Document Architecture [CDA] level 2) and unstructured information. For the structured sections, metadata, section headings, and subsection headings were all structured in the Extensible Markup Language (XML) format. However, clinical facts within the sections were not structured; therefore, it is possible that a medical transcriptionist could remove part of the structure because transcriptionists often follow an “as dictated” method of transcription and might remove section or subsection headings. The coding methodology involved reviewing all diagnoses listed in the last section (Assessment and Plan). The first listed diagnosis was identified as the principal diagnosis, and all remaining diagnoses were identified as secondary diagnoses. After identifying the diagnoses, the researcher went back to review all sections of the record to determine if the supporting documentation was present and/or if any documentation was absent. The term absent relates to documentation that was not present in the electronic documents that were used for coding. (For example, Ear infection is stated as a diagnosis; however, the laterality is not present in the electronic document.) An AHIMA-approved ICD-10-CM trainer and a researcher at the University of Pittsburgh were able to validate 5 percent of the total records (44 records including 737 diagnoses). We obtained feedback from physicians, coders, and technology experts in order to obtain valuable insights and suggestions in developing the quick reference guide. After developing the recommendations, we invited a physician, a coding professional, a health information management professional, and an informational technology professional responsible for developing an inpatient CAC system to review the documentation improvement tool kit and make further suggestions and recommendations for improving documentation.
After evaluating the inpatient document database and performing a thorough cleanup of the database (omission of duplicate records and erroneous records), we identified 656 patient records, each including 1 to 29 different diagnoses. A total of 4,791 diagnoses were coded for the study, and all possible diagnoses were coded to gain knowledge of the extent of the ICD-10-CM documentation requirements. The records were approximately one year old. Electronic documents reflecting only a patient’s diagnosis conditions were selected because the research involved evaluation of only ICD-10-CM (diagnosis coding).
Each of the diagnoses was categorized into the ICD-10-CM chapters, as depicted in Table 1. The distribution of the records into the ICD-10-CM chapters was uneven, with some chapters having more than 500 records (chapters 4, 9, and 18), some having fewer than 50 records (chapters 7, 8, 12, 17, 19, and 20), and two chapters containing no records (chapters 15 and 16). Overall, 736 diagnoses were identified with absent documentation, generating an overall absent documentation percentage of 15.4 percent (see Table 2).
According to the results of HCPro’s 2009 coding productivity benchmarking survey, it was estimated that the average time spent coding an inpatient record was 20 minutes.7 Therefore, the coding of 656 records at 20 minutes per record was anticipated to take the researcher (D.R.D.) approximately 219 hours. In reality, however, the coding of the entire set of records took approximately double that time.
The 10 chapters with the highest percentage of absent documentation were chapter 7 (Diseases of Eye and Adnexa), with 67.65 percent; chapter 8 (Diseases of Ear and Mastoid Process), with 63.64 percent; chapter 13 (Diseases of the Musculoskeletal System and Connective Tissue), with 46.05 percent; chapter 14 (Diseases of the Genitourinary System), with 40.29 percent; chapter 10 (Diseases of Respiratory System), with 35.52 percent; chapter 1 (Infectious and Parasitic Diseases), with 32.88 percent; chapter 12 (Diseases of the Skin and Subcutaneous Tissue), with 32.35 percent; chapter 2 (Neoplasms), with 25.45 percent; chapter 4 (Endocrine, Nutritional and Metabolic Diseases), with 14.58 percent; and chapter 17 (Congenital Malformations, Deformations, and Chromosomal Abnormalities) with 12.50 percent. We further analyzed the data by comparing the 10 individual chapters with the highest absent documentation percentages to all other chapters to evaluate if the differences observed between chapters were significant at p < .001. For example, when chapter 7 (Diseases of Eye and Adnexa), the chapter with the highest percentage of absent documentation, was compared with the rest of the chapters, we found the differences between them to be significant only for chapters 3, 21, 18, 5, 11, 9, 19, 6, 4, 2, and 10. The chapter with the next highest percentage of absent documentation was chapter 8 (Diseases of Ear and Mastoid Process). When we compared this chapter to the rest of the chapters, we found the differences to be significant for chapters 3, 21, 18, 5, 11, 9, 19, 6, and 4. (See Table 3.)
We developed a quick reference guide identifying the deficient areas in documentation to help the clinician in documentation improvement. As shown in Table 4, areas of deficiency from across all ICD-10-CM chapters were examined. The table highlights the areas of deficiency and offers recommendations for improvement. The recommendations are based on the descriptions found in the specific coding areas.
As the complexity and specificity of the documentation requirements for healthcare entities increase, it is crucial to plan for implementation of technology to capture accurate and timely documentation. This technology will greatly aid clinicians and coders in the move toward the electronic health information management arena and the ICD-10-CM/PCS coding system. Having accurate documentation will, in turn, improve the quality of the data in electronic health records and ultimately improve the quality of care that is rendered to the patients.8 Schiff and Bates (2010) highlight some of the many ways in which accurate clinical documentation could help in preventing diagnostic errors.9 The survey highlights the incorporation of checklist prompts to alert the physician to questions to evaluate prior to making a diagnosis. Having accurate documentation would enable better use of decision support software as well as CAC software. In our evaluation of the ICD-10-CM chapters for documentation specifics, we find that educating the clinician, who is the first point of contact, is vital to maintaining accurate documentation. Furthermore, frequent refresher sessions for clinicians and coders, possibly looking at ICD-10-CM chapter by chapter or at specific areas (if the facility is a specialty facility), would help in understanding the depth and specificity that is required in ICD-10-CM coding.
After reviewing the recommendations, the health information technology professionals commented that the tool kit we developed (consisting of the table of deficiencies we found and how to overcome them) was quite helpful for them in designing their interface because our recommendations directly corresponded to the initial point of contact between the physician and the patient. Being able to capture the needed data at the first point of contact saves organizational resources downstream.
Limitations of this study included the fact that the study was conducted on electronic documents and not on a complete electronic medical record. The current study could be further expanded to evaluate the documentation requirements for ICD-10-PCS, which is used for coding of procedures.
To successfully implement ICD-10-CM, clinicians need to be aware of the detailed documentation requirements. Provider organizations that are able to capture the data required in ICD-10-CM can be confident in obtaining the appropriate reimbursement in a timely manner.
This study located some documentation deficiency areas in all ICD-10-CM chapters (except the pregnancy and newborn chapters) and suggested recommendations to overcome the deficiencies and produce accurate documentation.
The authors would like to thank M*Modal for providing access to the electronic documents. The study was supported in part by the SHRS Research and Development Fund of the School of Health and Rehabilitation Sciences at the University of Pittsburgh. The Institutional Review Board at the University of Pittsburgh approved this project.
Dilhari R. DeAlmeida, PhD, RHIA, is an assistant professor in the Department of Health Information Management in the School of Health and Rehabilitation Sciences at the University of Pittsburgh in Pittsburgh, PA.
Valerie J. Watzlaf, PhD, RHIA, FAHIMA, is an associate professor in the Department of Health Information Management in the School of Health and Rehabilitation Sciences at the University of Pittsburgh in Pittsburgh, PA.
Patti Anania-Firouzan, MSIS, RHIA, is an assistant professor and clinical education coordinator in the Department of Health Information Management in the School of Health and Rehabilitation Sciences at the University of Pittsburgh in Pittsburgh, PA.
Otto Salguero, MPH, PhD, is the director of the Health Systems Engineering Initiative in the Industrial Engineering Department at the University of Pittsburgh in Pittsburgh, PA.
Elaine Rubinstein, PhD, is a senior service fellow in the Health Communications, Surveillance, and Research Support Branch of the National Institute for Occupational Safety and Health in Pittsburgh, PA.
Mervat Abdelhak, PhD, RHIA, FAHIMA, is the department chair and an associate professor in the Department of Health Information Management in the School of Health and Rehabilitation Sciences at the University of Pittsburgh in Pittsburgh, PA.
Bambang Parmanto, PhD, is a professor in the Department of Health Information Management in the School of Health and Rehabilitation Sciences at the University of Pittsburgh in Pittsburgh, PA.
- Dimick, C. “Top Documentation Issues for ICD-10.” April 18, 2011. AHIMA blog post, AHIMA Journal website. Available at http://journal.ahima.org/2011/04/18/top-documentation-issues-for-icd-10/.
- Moczygemba, J., and S. Fenton. “Lessons Learned from an ICD-10-CM Clinical Documentation Pilot Study.” Perspectives in Health Information Management (Winter 2012).
- Watzlaf, J. M., J. H. Garvin, S. Moeini, and P. Firouzan. “The Effectiveness of ICD-10-CM in Capturing Public Health Diseases.” Perspectives in Health Information Management (Summer 2007).
- Libicki, M., and I. Brahmakulam. The Costs and Benefits of Moving to the ICD-10 Code Sets (Technical Report No. TR-132-DHHS). Santa Monica, CA: RAND, March 2004.
- Roop, E. “Canada’s Slant on Smooth ICD-10 Strategies.” For the Record 20, no. 25 (2008): 20.
- Innes, K., K. Peasley, and R. Roberts. “Ten Down Under: Implementing ICD-10 in Australia.” Journal of AHIMA 71, no. 1 (2000): 52–56.
- HCPro: http://www.hcpro.com/content/238552.pdf.
- Miller, R. H., and I. Sim. “Physicians’ Use of Electronic Medical Records: Barriers and Solutions.” Health Affairs 23, no. 2 (2004): 116–26.
- Schiff, G. D., and D. W. Bates. “Can Electronic Clinical Documentation Help Prevent Diagnostic Errors?” New England Journal of Medicine 362, no. 12 (2010): 1066–69.
Dilhari R. DeAlmeida, PhD, RHIA; Valerie J. Watzlaf, PhD, RHIA, FAHIMA; Patti Anania-Firouzan, MSIS, RHIA; Otto Salguero, MPH, PhD; Elaine Rubinstein, PhD; Mervat Abdelhak, PhD, RHIA, FAHIMA; and Bambang Parmanto, PhD. “Evaluation of Inpatient Clinical Documentation Readiness for ICD-10-CM.” Perspectives in Health Information Management (Winter 2014): 1-16.