Beth Reid

Evaluation of the performance of diagnosis-related groups and similar casemix systems: methodological issues

With the increasing recognition and application of casemix for managing and financing healthcare resources, the evaluation of alternative versions of systems such as diagnosis-related groups (DRGs) has been afforded high priority by governments and researchers in many countries. Outside the United States, an important issue has been the perceived need to produce local versions, and to establish whether or not these perform more effectively than the US-based classifications. A discussion of casemix evaluation criteria highlights the large number of measures that may be used, the rationale and assumptions underlying each measure, and the problems in interpreting the results. A review of recent evaluation studies from a number of countries indicates that considerable emphasis has been placed on the predictive validity criterion, as measured by the R 2 statistic. However, the interpretation of the findings has been affected greatly by the methods used, especially the treatment and definition of outlier cases. Furthermore, the extent to which other evaluation criteria have been addressed has varied widely. In the absence of minimum evaluation standards, it is not possible to draw clear-cut conclusions about the superiority of one version of a casemix system over another, the need for a local adaptation, or the further development of an existing version. Without the evidence provided by properly designed studies, policy-makers and managers may place undue reliance on subjective judgements and the views of the most influential, but not necessarily best informed, healthcare interest groups.

Comparing diagnosis-related group systems to identify design improvements

The objective of this research was to compare the casemix systems used in the United Kingdom (UK), Australia and the United States of America (USA) to identify possible improvements in the design of the UK Healthcare Resource Groups. The data consisted of over 12 million inpatient and day case discharge records from 574 National Health Service acute hospitals in England for 2001-2002. These data were grouped into four casemix systems, namely Versions 3.1 and 3.5 of Healthcare Resource Groups, the United States-based All Patient Diagnosis Related Groups, and the Australian Refined Diagnosis Related Groups. The statistical performance of the groups was measured using the reduction in variance (RIV) statistic. The Australian Refined Diagnosis Related Groups produced the best RIV overall but this grouper had the advantage of more groups than the others. The comparison of the performance of the chapters within each grouper showed that each had some chapters with a better RIV than the other groupers. Comparing the performance of these groupers was successful in identifying changes to the Healthcare Resource Groups that improved its performance. Further revision of the Healthcare Resource Groups should be focused on the chapters with the best potential for improved performance.

Clinical coder training initiatives in Ireland

The Hospital In-Patient Enquiry and National Perinatal Reporting System (HIPE & NPRS) Unit of the Economic and Social Research Institute in Ireland requested a review of its coder training programs and data quality initiatives, primarily because of the decision to implement a major change in Irelands morbidity classification in January 2005. In August 2004, a formative evaluation using qualitative methods was conducted to assess the Units programs and initiatives. A number of opportunities for building on the solid frameworks the Unit has implemented were identified. In this paper, we focus on the Units coder training programs. The Units data quality initiatives will be discussed in a subsequent paper (Bramley & Reid 2005).

Under-coding in Australia limits the performance of DRG groupers.

The Diagnosis Related Group (DRG) system is now used extensively in Australia to classify acute inpatients for many applications, including payments to hospitals. The quality of the inpatient separation data affects the performance of the DRG version, especially its predictive validity. Data from the State of Maryland, in the United States, contain more secondary diagnosis and procedure codes than Australian data. A comparison of the performance of DRG versions using data from Australia and Maryland allowed us to answer the following research question: What impact did these additional codes have on the performance of the DRGs? The best performance in predictive validity (R2) was obtained using the Maryland data no matter which DRG version was used. Casemix-adjusted code counts showed that more diagnoses were coded in Maryland. The most plausible reason for this was that conditions were not being recorded comprehensively by doctors in the medical record in Australia.

Morbidity data quality initiatives in Ireland

In 2004, the Hospital In-Patient Enquiry and National Perinatal Reporting System (HIPE & NPRS) Unit of the Economic and Social Research Institute in Ireland requested a review of its coder training programs and data quality initiatives, primarily because of the decision to implement a major change in the morbidity classification in January 2005. In August 2004, the authors conducted a formative evaluation using qualitative methods to assess the Units programs. A number of opportunities for building on the solid framework the Unit has implemented were identified. The preceding paper focused on the Units coder training programs (Bramley & Reid 2005). In this paper, the Units data quality initiatives are examined.

Evaluation Standards for Clinical Coder Training Programs

This paper reports on an evaluation of clinical coder training programs, recently carried out in Ireland. In building an evaluation framework, the literature was reviewed to identify best practice standards, current practice, and professional opinion against which a sound judgment could be made. The literature was variable but nevertheless useful for the identification of evaluation standards. These standards are reproduced here in order to add to the literature. We also discuss the areas that would benefit from further research, thus contributing to the discourse on best practice in evaluating clinical coder training programs.

Best practice in the management of clinical coding services: Insights from a project in the Republic of Ireland, Part 1:

Introduction: This article presents some of the results of a year-long project in the Republic of Ireland to review the quality of the hospital inpatient enquiry data for its use in activity-based funding (ABF). This is the first of two papers regarding best practice in the management of clinical coding services. Methods: Four methods were used to address this aspect of the project, namely a literature review, a workshop, an assessment of the coding services in 12 Irish hospitals by structured interviews of the clinical coding managers, and a medical record audit of the clinical codes in 10 hospitals. Results: The results included here are those relating to the quality of the medical records, coding work allocation and supervision processes, data quality control measures, communication with clinicians, and the visibility of clinical coders, their managers, and the coding service. Conclusion: The project found instances of best practice in the study hospitals but also found several areas needing improvement. These included improving the structure and content of the medical record, clinician engagement with the clinical coding teams and the ABF process, and the use of data quality control measures.

Regarding Questions and Principles

I am hoping to use the letters page of the HIM Journal to bring two issues to the attention of your readers. Neither is very important in the greater scheme of things but they both concern the misuse of words. Precision in language makes for better communication and if we make an error then some will say that reflects a poor education. As an educator I am keen to ensure that this accusation does not apply to health information managers. First is the use of the phrase “begs (or begging) the question”. This phrase is sometimes used in general parlance as if it means that the discussion raises an interesting (or overlooked) question. This is a misinterpretation of the phrase. Begging the question, also known as petitio principii, occurs where the person assumes as a premiss for an argument the very conclusion that they were intending to prove. Copie (1978) gave this example: “The Governor must be a good friend to the farmers of this state because he told them so in a speech last night, and no one would lie to his friends” (p. 124). Of course, this example is so brief that no one would be taken in by such a false argument. Begging the question usually occurs in a much longer line of argument where it is easy for the participants to lose sight of the fact that the argument is circular. So, next time you are tempted to say that something or someone “begs the question” think for a moment and make sure you don’t mean “raises the question”. Second is the confusion between the words principle and principal. These words sound the same but have different meanings that the readers can check for themselves in a dictionary. The two meanings are illustrated in the following sentence: “I told the person who revealed my principal diagnosis to the newspapers that I thought they had no principles.” I have seen many cases where students and health information managers have used these words in the wrong context. This may be a problem of poor proofreading, because the spell checker in a word processor will not detect where a word is spelled correctly but used incorrectly. Whatever the cause, it makes anyone look poorly educated when they make this error. Both words occur frequently in the health information management field, so let’s be sure to use them correctly.