The Dynamics of Health Care Resource Utilization

Michael E. Klepser, Pharm.D., Donald G. Klepser, M.B.A.


Pharmacotherapy. 2002;22(2s) 

In This Article

Abstract and Introduction

Many pharmacoeconomic studies suffer from issues relating to the sensitivity of data and ability to accurately determine treatment-related costs. In addition, cost data generally are treated as static or constant variables. Our goal was to develop an approach to cost analysis that would allow us to appreciate the dynamic nature of health care resource utilization. By collecting detailed cost data on a daily basis and focusing on a narrow interval encompassing the infectious process, we were able to detect changes in resource utilization associated with management of the infectious process. This method also allowed us to assess and compare the dynamics of resource utilization among various patient groups.

The development of cost consciousness in health care has been a complex evolution. As clinicians in the health care system and as stewards of resources, we have been thrust into the battle between maintaining high standards of patient care and containing associated costs. We are asked to advance medical practice through the implementation and utilization of new techniques, devices, and pharmacologic agents, but to accomplish this goal with little or no additional financial resources. Complicating this charge is the fact that many new therapies are specific with respect to their actions, whereas patients' conditions are increasingly complex. It then becomes difficult to assess the economic ramifications of therapies for which traditional outcomes such as length of hospitalization or mortality may not be directly affected.

A common approach used to assess inpatient costs associated with disease management is to examine the total hospital costs over the admission. Subsequently, groups of interest are created and mean or median costs incurred over the hospitalization compared. Optionally, average daily costs for the inpatient period may be determined and evaluated. These analytic approaches have become increasingly criticized because of two primary flawed assumptions. First, this approach assumes that costs incurred during the hospitalization are related to the condition or treatment of interest. For example, no distinctions may be made between a patient admitted for the treatment of a staphylococcal infection and a patient admitted secondary to an automobile accident who subsequently developed an infection caused by a Staphylococcus species. The second flaw pertains to the assumption that hospital costs can be divided by the length of hospitalization to obtain an average daily inpatient cost. This approach implies that treatment costs are fixed and constant over time. Intuitively, we know this assumption to be false. When a patient is admitted or started on therapy, clinical status and utilization of health care resources are at specific levels. As the patient progresses through hospitalization and clinical status changes, utilization of health care resources also fluctuates. If the patient's status worsens, more tests will be performed, nursing service level will increase, and more drugs may be administered. As a result, the cost of therapy will increase over time. Conversely, if the patient responds to therapy and improves, health care resource utilization will decrease and costs will diminish. Therefore, it should be evident that resource utilization, cost of therapy, and clinical status are linked and are dynamic rather than static. As a result, an analysis that treats cost as a static or fixed variable may not appropriately describe this outcome measure.

Infectious processes pose several challenges to accurate examination of costs associated with inpatient management of these events. First, identifying a beginning and an end to an infectious episode may be difficult. A patient may exhibit signs and symptoms days or weeks before diagnosis and start of treatment. In addition, when should an infectious episode be considered to be resolved? Is a patient cured when antibiotics are stopped, at discharge, or when signs and symptoms completely resolve? Failure to accurately and consistently identify an infection window within a patient's hospital-ization may significantly bias observations. Second, many patients are not admitted to the hospital specifically for the treatment of an infection. As a result, management of the infectious process may not directly affect the patient's length of stay or the type of room and nursing services they require. Therefore, if an effort is not made to dissect the infectious episode from the rest of the hospitalization, the sensitivity of the analysis to detect difference with respect to the management of the infection will suffer.

Other factors that warrant consideration when conducting an analysis of infection-related treatment costs include site of infection, patient's location (inpatient vs outpatient) when the infection was acquired, and determination if the organism was truly pathogenic. Many investigators retrospectively identify patients for study inclusion through an evaluation of microbiology records and identify individuals with the organism of interest. Failure to gather additional information regarding the patient and the pathogen may severely impair the investigator's ability to draw meaningful conclusions from the data.

Potential pitfalls associated with commonly used methods for conducting pharmacoeconomic analyses include the following: the infectious episode may last for only a short time and be unrelated to the length of stay, thus introducing noise into an analysis of total hospital costs; identifying the beginning and end of the infectious episode generally is not attempted; limited information regarding costs in various categories (e.g., room, nursing, laboratory, pharmacy) is given; adequate clinical data are not collected; and no sense of the dynamics of resource utilization is provided.

Recognizing what we perceived to be limitations of previously published economic studies of inpatient management of infectious episodes, we attempted to develop a new strategy for the collection and evaluation of these data. Our goals were to better describe health care resource utilization associated with an infectious process, describe the dynamics of resource utilization, and determine what types of resources were being used (Figure 1).

Methodological goals describing health care resource utilization associated with an infectious process, describing the dynamics of resource utilization, and depicting the proportions of resources used.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: