Research Activities at NCCH
The last decade has seen an increase in the number of routine data collections at the population level. Datasets that have resulted from these collections are a good source of information for health services administration and management, the monitoring of population health and disease trends, strategic health planning and policy development, as well as for research purposes. However, many researchers have raised questions regarding the quality of data contained in these datasets, one of these being the hospital morbidity dataset. Issues of concern
included data reliability and accuracy, as well as the specificity of assigned codes.
The National Centre for Classification in Health has been developing a strong research program in health classifications and terminologies now for around five years. The implicit principles of this research program are to:
- Identify and explore current and future areas of need with regards to health classification and terminology development.
- Address real-world issues of concern for the user workforce (including clinicians, clinical coders, data users).
- Involve and value the contributions and knowledge of the user workforce.
- Be responsive to the needs of our clients and funders (state and national health departments, statistical agencies, related health organisations and research bodies).
- Gather evidence from multiple sources including routine data collections, related data collections, public submissions, coder queries, user surveys and feedback, and clinical audits.
- 6. Promote the importance of health classifications and terminologies to the production of high quality data and information.
Research at NCCH
The research team at NCCH is currently
conducting the following research studies on code usage and coding quality:
PICQ and Obstetrics Coding Quality
The project aims to assess the overall quality of coded genealogical/obstetric data in the 2004/05 National Morbidity dataset using the Performance Indicators for Coding Quality (PICQ). These indicators function as a tool to assess the coding quality of hospital morbidity data, and provide an objective measure of data accuracy. Among the 31 PICQ indicators developed for the gynaecological/obstetric chapters of ICD-10-AM/ACHI Fourth and Fifth Editions, 11 indicators measure the coding accuracy of obstetric procedures and 20 test the correct assignment of obstetric diagnosis codes. Results indicate that PICQ is a useful initial screening tool for the assessment of ICD-10-AM/ACHI coding quality. The overall quality of codes assigned to gynaecological/obstetric records in the 2004/05 Australian National Morbidity dataset is of fair quality. A paper has been submitted to the Health Information Management Journal and is currently under review.
The Quality of Obstetric Coding in NSW Admitted Patient Data Collection ? A Data Linkage Study
Data Linkage utilises the concept that different data sets (usually containing data collected under different circumstances and stored in large data files) containing different information on the same group of individuals within the same population can be linked together by a statistical linkage technique. The advantage of data linkage is to maximise the value of data collected by linking information on individuals together to provide a fuller picture for clinical as well as research purposes. The newly established Centre for Health Record Linkage (CHeReL) in NSW provides the services and mechanism for achieving data linkage among various large datasets resulting from routine data collection on vital health-related statistics.
Utilising the data linkage service offered by the CHeReL, the research team is in the process of establishing a project to examine the quality of obstetric data found in the NSW Admitted Patient Data Collection (APDC) through linkage with two other related datasets: the NSW Midwives Data Collection (MDC), and the Registry of Birth, Deaths and Marriages (RBDM) birth registration data.
Code Sequencing Studies
The assignment of the correct codes in the correct order is important for many chapters in ICD-10-AM and for the interpretation of the codes. These include the obstetric chapter (Ch15), the injury chapter (Ch19), and the health services contact chapter (Ch21). Currently two studies are being conducted to examine code sequencing issues using the National Morbidity dataset.
The obstetric code sequencing study aims to assess the accuracy of the order in which codes are being assigned for obstetric cases recorded in the National Morbidity dataset. According to ACS 1521 and instructions given in ICD-10-AM Chapter 15, an additional diagnosis code from other chapters of ICD-10-AM should be assigned to identify the specific condition if a code from the range O98.0 to O98.6 (Maternal infectious and parasitic diseases classifiable elsewhere but complicating pregnancy, childbirth and the puerperium) or the range O99.0 to O99.8 (Other maternal diseases classifiable elsewhere but complicating pregnancy, childbirth and the puerperium) is assigned. The additional code should follow immediately after the O98 or O99 codes to reflect the association. Recent analysis of sequencing issues related to the O99 block by the team found that this sequencing rule is not followed very closely, thus reducing the usefulness of some of the coded data for research and reporting purposes. Further research is underway to examine the reason of this problem.
Similar to the obstetric code sequencing study, the procedural complication study aims to investigate the characteristics of codes assigned following the assignment of code I97.8 (Other postprocedural disorders of circulatory system, not elsewhere classified). The proportion of records that were assigned with the correct sequela code which further describe the condition will be analysed. The characteristics of records that were assigned an incorrect sequela code will also be described, and reasons for of this problem further explored.
Impact of Code Addition on Code Usage and Coding Quality
New codes are often added to different editions of ICD-10-AM in response to classification needs. It is a generally held belief that additional codes enable the collection of more detailed and in-depth information about the person admitted to hospital. Such additional information can be used in the subsequent care and management of the patient. However, the degree to which these additional codes function to provide additional information has not been scientifically ascertained. A project is being designed to evaluate the usage of additional codes for coding injury patients, and how the addition of these new codes impact on coding quality using National Morbidity datasets from 2002-2006. Results generated from analyses can be used to advise relevant health organisations, and contribute to future ICD-10-AM revisions
top