Abstract
It is a pleasure to provide the guest editorial for this Special Issue of the Health Information Management Journal (HIMJ). The Journal has had a long and interesting history that can be traced back to the 1970s with humble beginnings, being manually produced with a stencil and duplicating machine by volunteers (Watson, 2019). The fact that we are now publishing a special issue through SAGE Publishing speaks volumes about how far the Journal has come and the title of this Special Issue, Clinical Coding and the Quality and Integrity of Health Data, speaks volumes about how important clinical coding has become to the management of health information in Australia and around the world. In Australia, as the Health Information Management Association of Australia celebrates its 70th anniversary, clinical coders are facing challenges on many fronts. Clinical coded data influences diverse aspects of our health systems, from quality and safety monitoring and funding models to health service planning and infrastructure development. In addition, we have technological developments that will change the clinical coders’ roles substantially over the next 5–10 years.
The articles published in this Special Issue reflect these challenges and illustrate the far-reaching consequences of data that lack integrity and are of poor quality. Campbell and Giadresco (2020), through a literature review, investigated the effect of computer assisted coding on the accuracy and quality of clinical coding and its impact on clinical coding professionals. The articles, dissertations and case studies they reviewed demonstrated value in improving clinical coding accuracy and quality through computer assisted coding. Campbell and Giadresco concluded that clinical coders should view computer assisted coding as an opportunity to develop new skills, particularly in monitoring and auditing coding outputs, and that sound change management strategies are needed to ensure a successful transition of the clinical coding workforce to new roles. Improved clinical coding accuracy will benefit our health system enormously but it would be naïve to think that computer assisted coding is the complete answer. Clinical coders will be needed in different roles to help realise the benefits of computer assisted coding.
To that end, Hay et al. (2020) discussed the role of documentation improvement specialists and how they can ensure adequate documentation that can be translated into clinical codes. This is a potential role for clinical coders who understand both the clinical documentation and the needs of the end users of the coded data. Hay et al. (2020) also outlined the work of the Australian Commission on Safety and Quality in Health Care, which has promoted improved documentation through its National Safety and Quality Health Service Standards and the use of coded data for monitoring patient safety through its hospital-based outcome indicators. The development of the hospital-based outcome indicators has further elevated the need for high-quality clinical coding.
However, barriers exist to achieving quality clinical coding outcomes. Canadian authors, Doktorchik et al. (2020), discussed these barriers in their article ‘A Qualitative Evaluation of Clinically Coded Data Quality From Health Information Manager Perspectives’. Their interviews with health information managers and clinical coding managers revealed that expectations were increasing for high-quality data collection but without additional resources to support this endeavour. They also found that incomplete and disorganised clinical documentation and lack of good communication with clinicians impacted on the quality of clinical coding. These same issues exist in Australia, and I am sure in many other countries around the world.
The integrity of clinical coding depends fundamentally on the quality of the patient record. The Portuguese study by Alonso et al. (2020), ‘Health records as the basis of clinical coding: Is the quality adequate? A qualitative study of medical coders' perceptions’, highlights that clinical records are not just for patient treatment but that the data derived from them are stored in administrative databases and used for many downstream purposes. To that end, the authors conducted focus groups to elicit from clinical coders the problems they face in the health records that influence the quality of the coded data. They identified several issues including missing or incomplete discharge and/or surgical notes, the use of abbreviations, variability in documentation between specialties and lack of specificity in diagnosis descriptions. They also identified that in spite of electronic health records solving illegibility problems, they have created problems of their own, notably the copy and paste facility that results in errors being repeated throughout the record and very large volumes of notes to be perused by clinical coders. Importantly, they also found that no solutions are being found for these issues. Australian clinical coders would sympathise with these comments, as would many others.
Three recent articles, two in this Special Issue, have focused on the congruence between the clinical codes assigned to the case and the clinical documentation in the medical record. Given that clinical coders are governed by guidelines and standards that limit the assignment of codes in certain circumstances, very important questions are raised by these papers. Australian authors, Nguyen et al. (2019), studied the level of agreement between documentation in the medical records and ICD-10-AM coding of mental health, alcohol and drug conditions in trauma patients. These authors concluded that despite documentation in the medical record, these conditions are not always coded, rendering incomplete the administrative databases on which epidemiologists and other researchers rely. Sveticic et al. (2020) from Queensland, Australia, conducted a medical record review to assess the validity of data on suicide and self-harm. They concluded that suicide and self-harm are under-enumerated in the administrative data and issued a warning that the data should be used with caution. In the third paper, the UK authors Handley and Emsley (2020) studied medical records that had been identified by the allocation of specific codes for intracranial venous thrombosis (ICVT). They concluded that ‘the coded data reported a higher incidence of ICVT than previously thought’. This goes to the question of specificity of clinical codes in the international classification of diseases and its various modified forms around the world.
The problem of classification keeping up with current clinical definitions was outlined by Phillips et al. (2020) in their article from the United States, ‘Malnutrition Definitions in Clinical Practice: To be E43 or Not to Be?’ When the definitions used by clinical coders are out of step with the latest clinical definitions, the integrity of the coded data is compromised. Should the classification be updated more regularly, or does that compromise the stability that many processes need?
Perhaps the most high-profile use of the clinical coded data is as a foundation for diagnosis-related groups (DRGs), which are used by many funding models. Two articles in this Special Issue have raised issues associated with the coding of co-morbidities, which are important for determining complexity splits in DRGs. In a study based on Portuguese data, Souza et al. (2020) argued that all co-morbidities, pre-existing or newly diagnosed, should be coded in order to achieve optimum severity splits in the all patient refined diagnosis related groups (APR-DRGs). Following the publication of an Australian report that stated the complexity model in Australian DRGs did not correlate with cost, Kim et al. (2020) undertook a study of the complexity model in Korean DRGs concluding that ‘if highly accurate coding data and cost data become available the performance of secondary diagnosis as a variable to reflect the case complexity should be re-evaluated’ (p. 6).
This Special Issue of HIMJ will help to raise awareness of how important the clinical coding function is to the quality and integrity of our health data. Across the spectrum of documentation improvement, clinical code assignment and end uses of the data, such as for funding models based on DRGs, the articles in this issue challenge us all to find solutions that will improve the quality of coded data, protect its integrity and support the clinical coding workforce.