In their provocative article concerning the debate regarding the value of applying significance testing in social research, Baghi, Noorbaloochi, and Moore take the position that the P value alone is not sufficient as an index of statistical significance. They argue that it should be coupled with other measures such as the point estimation of the effect size and a confidence interval around the effect size.
Drake, Cohen, and Cohn examine 2 aspects of pneumonia antibiotic timing as endorsed by the Centers for Medicare and Medicaid Services (CMS) Hospital Quality Improvement Demonstration Pay for Performance project. Their findings suggest a possible association between antibiotic utilization to comply with the CMS-endorsed timing standard and the use of antibiotics for conditions in which their use is inappropriate. They argue that such an association may contribute to the risk of promoting antibiotic resistance.
It is common practice to use 3-sigma limits with the normal approximation to the binomial distribution in comparing various data sources charted on a P chart. Hart, Hart, and Schmaltz advance the idea that using the normal approximation may not always be appropriate in the presence of small subgroups. For samples with small subgroups, they propose the use of the exact binomial distribution in place of the normal approximation, and they demonstrate a method to accommodate numerous subgroups without increasing the false alarm risk.
Alemi and Alemi suggest that, to the patient, the precise scientific reasons one treatment intervention rather than another is effective probably is less significant than the degree of improvement he or she experiences from that intervention. Therefore, they argue that it is not always necessary to use the full-factorial design including control groups in studying the effectiveness of each of a series of different interventions. The authors suggest that in some cases obtaining an estimate of each factor effect is redundant. They present a modification of the classic factorial design in which the number of trials can be significantly reduced.
Organizational issues in connection with facility-wide adherence to medication guidelines in nursing homes were studied by Lau, Banaszak-Hall, and Nigam. To obtain CMS certification, nursing homes must comply with CMS guidelines addressing treatment. The authors found that pharmacists and nurses held perceptions differing from those of the physicians as to the significance and authority of medication guidelines. In addition, they found that a number of external factors worked to interfere with adherence to treatment guidelines. The authors discuss the implications of the factors hindering guideline adherence on efforts to improve quality.
Intuitively, it would not be surprising if seriously ill children tended to be admitted to advanced-care hospitals that have robust information technology (IT) systems. In their study of Florida hospitals, Menachemi, Brooks, and Simpson investigated the possibility of such an association, and demonstrated a positive relationship between the presence of an advanced IT system and high levels of pediatric discharges. They also found that heavy reliance on public, Medicaid funding was negatively associated with the presence of advanced IT systems.
Organizational structure is the underlying theme in the study reported by Naessens, Culbertson, LeFante, and Campbell. They suggest that consumers' perceptions of a hospital's safety rates can be distorted as a result of misinterpretation of publicly reported data by the media and other information sources. The authors argue that, in the case of multicampus and multiunit medical centers, adhering to the patient safety reporting policies established by The Leapfrog Group has the potential for confusing the media and consumers as to the individual safety rates of the respective reporting institutions. They review the organizational modeling theories developed by Weber, Mintzberg, and Leavitt and Scott, and present a comparison system based on organizational features coupled with related criteria. They propose the formulation of criteria for determining whether a multiunit complex organization should conduct its public reporting as a single entity or on the basis of the individual units.
Being able to classify the various operational features of an action program in accordance with a conceptual framework serves to enhance the forward progress of program development and implementation. Wurster, Pearson, Sonnad, Mullen, and Kaiser reviewed 114 published studies of patient safety improvement strategies to detect and codify the structural factors that could be employed to organize effective patient safety strategies.
A team from the veterans administration conducted a multisite observational study to test a model of changing clinical behaviors, focusing on the implementation of evidence-based guidelines for the treatment of hypertension in patients with type 2 diabetes. The objective of Wyszewianski, Kowalski, Lowry, and Green was to determine to what extent implementation followed the evidence. Their findings suggest that the adoption of clinical guidelines by institutions, as well as by individual practitioners, is influenced not only by knowledge but by attitudes. One interesting finding is that the implementation strategies most commonly used by the Veterans Administration Medical centers were those likely to require the fewest resources.
This issue's "interview" is actually a panel discussion among 3 medical center professionals and executives who have led the implementation of an integrated quality management program across the many divisions within a complex organization. Nancy Lorenzi, who organized the panel, Wright Pinson, and John M. Starmer of Vanderbilt University Medical Center sit down with a tape recorder for an informal discussion of the complicated and delicate process. In their off-the-cuff comments and personal accounts, they address the practical issues that are encountered by healthcare leaders responsible for instituting major changes in complex organizations.
David C. Aron contributes an insightful review of "IOM Report: Paying for Performance in Healthcare. Cutting the Die While Tinkering With a Broken System."
Jean Gayton Carroll, PhD
Editor