It would be hard to overestimate the impact of the diagnosis-related group (DRG) system on rationalizing health care cost structures and in providing an indispensable organizing tool for use in health care quality and utilization research. Norbert I. Goldfield, one of the earliest and most influential researchers and proponents of DRGs, traces the development and refinement of the system over the past quarter century. It may be hard now to picture the murkiness that enveloped such concepts as health care quality and how to measure, count, and evaluate episodes or units of care and their clinical outcomes before 1980. Many expressed misgivings about the perceived association of "cookbook" methodology with a scholarly professional activity like providing medical care. As an involved participant in countless discussions of the subject leading up to Medicare's adoption of DRGs, I can report that the arguments often grew heated. After reviewing the system's development over the years, Dr Goldfield goes on to discuss the more recently developed variations designed to reflect clinical severity. These are the Centers for Medicare & Medicaid Services' (CMS') severity-adjusted version of DRGs, the Medical-Severity DRGs (MS-DRGs), and the 3-M Corporation system's All-Patient-Refined DRGs (APR-DRGs).
In DRG designations and quality evaluation, POA (Present on Admission) plays a vital part in evaluating the quality of care, affecting the reimbursement level, and, in many cases, the Joint Commission or CMS evaluation and accreditation or certification of a hospital or skilled nursing facility. Since October 2008, CMS has, in effect, penalized hospitals where discharge diagnoses show the presence of hospital-acquired conditions not documented as POA by reducing payments, lending its hefty weight to the perception that "quality pays." Robert McNutt and colleagues present a method for estimating the proportion of cases in which the MS-DRG assignment changes when hospital-acquired diagnoses are removed from the reimbursement calculation; subsequent changes in hospital reimbursements; the attenuation in changes in MS-DRG assignment when possible POA diagnoses have been factored in; and the effect of the numbers of ICD-9 diagnosis codes on MS-DRG assignment. The team studied nearly 185 000 discharges with at least 1 hospital-acquired condition. Although their sample was limited to academic medical centers, their findings point to the need for more study of broader hospital populations and of the effect of variations in coding practices.
Randomization is usually a goal in constructing a sample for study of the effects of a clinical intervention. However, in the case of a managed care organization analyzing the effects of a population-level intervention such as the introduction of a new medication, randomization may not be possible. Nonrandomized observational studies can serve well in these circumstances. However, it is difficult and often impossible to derive valid conclusions about causal relationships between the intervention and the observed outcomes from a nonrandomized study. It becomes necessary to take measures to reduce the effect of bias. One such measure is propensity scoring, a measure of the probability that the subject received a specified intervention. Mohamed Hussein and colleagues present an example of the use of propensity scoring coupled with statistical matching to mitigate the possible effect of bias.
As Pavani Rangachari points out, complex systems theory supports the notion that organizational learning is essential to organizational improvement. She proposes using network theory to develop a framework to facilitate understanding how learning and improvement take place in the context of hospital infection prevention initiatives. In the course of her studies of reported research, the author draws upon reports involving 2 types of systematic improvement strategy: the Toyota Production System (TPS) and the Positive Deviance (PD) model.
Using an extension of Donabedian's structure, process, and outcome model, Haiyan Qu, Richard M. Shewchuck, Yu-ying Chen, and J. Scott Richards studied the relationship between the functional outcomes of patients being treated for recent spinal cord injury and the structural and process elements of their care. The authors report that their findings in this study suggest that the final outcome of therapy does not depend entirely on the volume of process or therapeutic measures employed or on the number of therapist hours expended. They conclude that more attention should be paid systematically to the profiles of patients that reflect their individual therapeutic needs, capacities, and potential for improvement.
The possibility of online counseling and treatment of patients living with chronic conditions such as diabetes and weight problems arouses widespread enthusiasm among both practitioners and patients. Farroukh Alemi et al studied the impact of online counseling on drug use in a population of underserved at-risk individuals. Although they comment that their study showed promise for using technology in treating hard to reach patients or clients, they acknowledge that there are logistical problems in dealing with these patient populations. The high attrition rate they encountered was one such problem.
Citing the need for more study of clinical outcomes in behavioral health settings, Alok Madan and colleagues suggest that various issues in collecting and processing behavioral health outcomes data may play a role in limiting the scope and significance of such studies. They discuss the advantages that might accrue to patient self-report of treatment outcomes in mental health. The authors present a Web-based system for collecting patients' self-report data on the construct called "quality of life" in an outpatient setting. They acknowledge that design constraints imposed by this approach and the extreme brevity of the 4-item quality of life measure preclude drawing inferences about causality and making conclusions on the basis of the self-reports.
In their "Discussion" section, Hamid Khour, Patricia Perreault, and Denise Herzog put it plainly. Their initial hypothesis was not confirmed. It's refreshing to see scientific investigators make such an admission. They predicted that low satisfaction levels reported on a patient questionnaire would point to\break individual problems experienced by patients in the course of outpatient digestive tract endoscopy. They report that although the rate of global satisfaction with a pediatric endoscopy service was high, patients and their parents reported far lower satisfaction levels in connection with a range of individual components of the outpatient endoscopy experience. Social psychology could add a few insights to explain the observed phenomenon.
Administrative and clinical staff of health care provider organizations, from large teaching hospitals to group practices to physicians' offices, are constantly bombarded with marketing materials and sales calls by representatives of proprietary electronic medical record and order entry systems. In their tutorial, Marie H. Federowicz, Mila N. Grossman, Bryant J. Hayes, and Joseph Riggs present a step-by-step guide for using activity-based costing (ABC) to estimate the true cost of implementing an electronic health record (EHR). They point to the fact that although ABC's concepts and techniques are widely used in manufacturing, ABC is not commonly used in health care. They assert that the traditional costing methods do not work well when applied to studies of the costs of EHR systems. In the interest of demonstrating how ABC can be used in a health care setting, they present a case study of its testing and adoption in a 3-physician clinic in the Midwest.
Efficient utilization of a hospital's patient care resources is a factor in the overall quality of care. Keeping room turnaround time at a level that serves the objective of facilitating necessary admissions and prompt medical and nursing attention to a patient's needs qualifies as a contributing element in good care. At the same time, as authors Evelyn C. Brown and John Kros acknowledge, it is important not to allow the need for promptness in room turnaround to interfere with infection control. Employing process mapping and heuristic measures, the authors analyze and evaluate the process used to reduce room turnaround time while maintaining optimal infection control in a regional hospital.
Jean Gayton Carroll, PhD
Editor