PMC patient severity scale: derivation and validation - Patient Management Category

Health Services Research, August, 1994 by Wanda W. Young, Susan Kohler, Jeanne Kowalski

Consumers, employers, and regulators alike have increased their demands in the past few years for accurate comparisons of health care quality. Although their shared goal has been to ensure that appropriate, cost-effective care is delivered consistently to all patients, there is less consensus on how to measure whether or not we are achieving that goal. The specific task of measuring the appropriateness and outcomes associated with certain health care services is complicated by the fact that each patient may be unique in his or her combination of diseases and risk factors at any given point in time. To measure these severity of illness distinctions, this study describes the development and validation of the Patient Management Category (PMC) Severity Scale, a severity measure to improve the precision of patient outcomes research and assessments of provider performance.

The PMC Severity Scale is an adjunct to the existing Patient Management Category (PMC) Classification and Intensity Scoring System, a computerized patient classification and scoring system that is both clinically specific and linked with different levels of care (Young, Kohler, and Macioce 1992). The PMC system was originally developed with a grant from the Health Care Financing Administration to the research staff of Blue Cross of Western Pennsylvania (Young, Joyce, Schuchert, et al. 1985). The original system was made available through the National Technical Information Service, while the maintenance and dissemination of the system has been managed by The Pittsburgh Research Institute, a 501(c)(3) not-for-profit health services research organization.(1) As part of this ongoing enhancement, we sought in this project to create an independent method for quantifying the overall severity of each patient's unique combination of diseases, interaction of the diseases with effective treatment, and the resultant combined risk of morbidity and mortality. The goal was to develop a severity index associated with PMCs that could be used directly to improve the standardization of outcome measures and to complement the PMC Relative Intensity Score (PMC-RIS), which is useful for adjusting costs.

Before describing the PMC Severity Scale, the PMC Classification and Intensity Scoring System as well as some general issues in designing patient scoring systems will be reviewed. These two areas will provide a framework for assessing potential applications of the severity measure described in this article.

BACKGROUND

The original objectives of developing the Patient Management Categories and the PMC Relative Intensity Scoring System were twofold: (1) to identify, in a clinically specific way, the types of patients treated at different institutions, and (2) to measure severity and resource intensity differences among those patient types. The goal was to design a system that is both clinically meaningful and useful in predicting expected hospital resource use and costs. To accomplish these objectives, the following analytic tools were developed:

Patient Management Categories. A computerized, diagnosis-based patient classification, developed with extensive clinical input from physician panels, that incorporates severity of illness distinctions and defines comorbidity;

Relative Intensity Scores. A set of cost-based relative weights that (1) reflect the intensity of hospital services required to manage a clinically specific patient type in relation to the average hospitalized patient, and (2) are combined in a computerized methodology to yield one Relative Intensity Score (RIS) for each patient based on that patient's particular disease conditions (single disease or comorbid) and complications; and

Patient Management PATHs. Computerized, physician-specified clinical management strategies (one for each PMC), each of which consists of diagnostic and treatment services for effective care of the typical patient in that PMC.

PATIENT MANAGEMENT CATEGORIES

PMC Classification (Release 5.0; see Young, Kohler, and Macioce 1992) consists of 830 Patient Management Categories that describe all patients treated in general acute care hospitals. In addition to the categorization of general medical and surgical diseases (e.g., diabetes, pneumonia, vascular emboli, AIDS), special populations (e.g., neonates and deliveries), and psychiatric disorders, patient categories have been defined for specific complications as well (e.g., septicemia, wound dehiscence, hemorrhage).

PMCs were originally defined, by physicians, in clinical terms independent of historic patient data. Extensive physician consultation was obtained through the formation of more than 50 disease-specific panels, each of which consisted of four to six physicians (both generalists and specialists) who treated patients with the disease(s) being modeled in that panel session (Young, Joyce, Schuchert, et al. 1985). Initially, more than 125 physicians from southwestern Pennsylvania participated in these expert panels. Since its original development, however, many more physicians from the United States and abroad have contributed to the research that has been part of the ongoing modification and enhancement of the PMCs and PATHs.