| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorial |
1 Division of Laboratory Medicine, Washington University School of Medicine, Box 8118, 660 S. Euclid Ave., St. Louis, MO 63110, Fax 314-362-1461, E-mail mscott{at}labmed.wustl.edu
With the enormous popularity of audioconferences, meetings, and special inserts in nonarchived journals regarding point-of-care testing (POCT), it is clear that laboratorians are struggling with questions about implementing and managing POCT. Yet there is a paucity of serious studies in the peer-reviewed literature that address the benefits of POCT. In this issue, Nichols et al. (1) present a study that examined the impact of POCT implementation on patient waiting time for interventional cardiology procedures. This study provides an example of the proper way to address decisions regarding the implementation of POCT. The authors established an hypothesis about the impact of POCT in their setting, implemented POCT, examined a measurable outcome, and then made adjustments to patient management and workflow when simple implementation of POCT did not substantiate their hypothesis. The bottom line from this study is that despite the initial intuition that faster laboratory values would be better, it took more than just fast laboratory data to obtain the desired benefit. These authors found that simply moving the point of testing from the central laboratory to the clinical service did not improve patient waiting time until significant changes in workflow were made. The predictions of the second phase of their study [when the potential time-savings of POCT vs central laboratory turnaround time (TAT) were determined] suggested that 92% of patients requiring either renal function or coagulation tests could meet their scheduled procedure time with POCT. By contrast, at baseline, only 8% and 29% of patients met their scheduled procedure times with central laboratory testing of renal function and coagulation, respectively. However, even after the workflow changes were implemented, the improvements in patient waiting time barely reached significance, and only 33% and 42% of patients requiring renal or coagulation tests, respectively, met the goal of starting their procedure when scheduled.
That faster laboratory data alone did not lead to improved outcome should not be surprising. The available prospective, controlled studies that have examined the impact of POCT on length of stay (LOS) in emergency departments have found that institution of POCT testing for electrolytes, hemoglobin, or blood gases did not have a positive impact on patient LOS (2)(3)(4). We concluded (2) that the reason POCT did not improve LOS in our emergency department was that laboratory TAT was not the rate-limiting factor for discharge. Time spent waiting for other procedures (e.g., radiology and electrocardiography), other central laboratory tests, consults, and other factors usually overlapped or exceeded the time to obtain results for electrolyte and hematology tests from the central laboratory. Even in the well-performed randomized study by Kendall et al. (4), when the time to treat emergency department patients decreased with implementation of hematological, electrolyte, and blood gas POCT, there was no improvement in LOS or medical outcome. Nevertheless, it is important not to extrapolate these findings to all settings, as the advantages possible from faster tests can be institution-dependent. Indeed, the setting for the study by Nichols et al. (1), where pneumatic tubes were not available and the central laboratory was several floors and buildings away, would seem ideal for the hypothesis that POCT would improve the measurable outcome of patient waiting time. However, even in this setting, changes in workflow were necessary before any improvement was seen. What makes the present study unique is that it recognized and acted on the fact that "faster is better" was not the case and that other, non-laboratory, limiting factors needed to be altered in conjunction with the faster provision of test results.
What gives this study particular credibility is that it follows many of the recommendations for performing POCT outcomes studies made by Rainey in this journal 18 months ago (5). By doing so, it serves as a good example of how to perform an outcome study to assess the benefits of POCT in a particular setting without huge numbers of subjects or large expenditures. Strong points of this study included: (a) it was a real observational study; (b) it was prospective; (c) it was limited to a single service; (d) it used an easily quantifiable outcome; (e) the outcome had intrinsic financial benefits; and (f) it identified institution-specific factors. Before implementing POCT, it should be the responsibility of the laboratory to develop a hypothesis about what POCT implementation will accomplish and then to design a study to test the hypothesis. To obtain the maximum benefit from a POCT outcome study, it is necessary to define an outcome that can be easily measured. Examples are LOS (2)(3)(4), blood product usage (6)(7), therapeutic TAT (8), operating room time, and decreased time waiting for the operating room or other procedures (1). Furthermore, tests chosen for POCT should be those for which it can be anticipated that immediate action (or entry to a treatment algorithm) will be taken upon receipt of the results. Examples might include changing of anticoagulation therapy during surgery (6)(7), moving of a patient to a procedure (1), or indicating that surgical resection is complete (9).
To their credit, Nichols et al. (1) did not attempt to do a cost per result analysis before studying POCT implementation. Although this may seem negligent to some, they correctly state that cost analysis of POCT can be biased by viewpoint and the failure to include all associated costs. It should be clear that POCT costs more per reportable result than does testing in the main laboratory. POCT reagents are more costly; point-of-care coordinators must be hired; training must be done; and utilization, competency, and quality control must be continually monitored. Furthermore, it is seldom possible to fully eliminate a person or work station in the main laboratory when POCT is implemented at one or two sites, and many cost analysis studies do not account for the time a patient care technician spends performing tests. Attempts at counting "fingers and toes" to cost-justify POCT are doomed to heavy scrutiny by skeptics and cost-conscious administrators. Instead, it is best to pick a setting where the hypothesized outcome has an inherent financial value that will make POCT implementation a "no-brainer" and the laboratorian a hero! Indeed, you do not need an accountant to justify POCT if: LOS in the emergency department decreases by 15–30 min; LOS in the intensive care unit decreases by 1 day; empty operating room waiting time decreases by 30 min; blood product usage decreases by two units per patient; or if an inpatient procedure becomes an outpatient procedure.
The present study can serve as an example of how to approach these studies without a large number of subjects or large expenditure. It did not take many patients before the current authors could establish that their hypothesis would not be substantiated by simple implementation of POCT. In contrast to most other POCT studies, these authors then took the next step to modify steps in patient management and workflow and again performed a relatively small study. They determined that POCT coagulation was clearly beneficial to a small subpopulation of their patients, whereas renal POCT testing was not. I would recommend that anyone facing POCT implementation decisions reread the editorial by Rainey (5) and pick the settings and anticipated outcomes well because, as these authors have clearly shown, it is not always that simple to show that faster is better!
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  C. P Price Regular review: Point of care testing BMJ, May 26, 2001; 322(7297): 1285 - 1288. [Full Text] [PDF]
|
|
|
|
 |
|
 C. P. Price Evidence-based Laboratory Medicine: Supporting Decision-Making Clin. Chem., August 1, 2000; 46(8): 1041 - 1050. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
