Extending Patient Safety Beyond the OR and PACU

Michael W. Russell, MD; C. William Hanson, III, MD

Anesthesiologists continue to improve patient safety beyond the doors of the operating rooms and post-anesthesia care units. A team at the University of Pennsylvania, Philadelphia, PA, led by anesthesiologist-intensivist, Michael W. Russell, MD, is addressing the issue of postoperative patients admitted to general care wards requiring subsequent, unanticipated ICU admission.

Review of medical records revealed that respiratory events account for the majority of these unscheduled ICU admissions among postop patients. This observation suggests that routine monitoring with EKG telemetry for patients cared for on general hospital wards is poorly suited for early detection of respiratory events (early enough to allow appropriate diagnosis and treatment before any EKG change, a late development in the critical incident, would be detected). Therefore, new technology and equipment was installed which permits continuous monitoring (both bedside and centrally displayed at the nurses’ station) of arterial oxygen saturation. The system, known as “Oxinet,” developed originally by Nellcor, (which is now part of Mallinckrodt, Inc.), provides continuous pulse oximetry with centralized data and alarm displays (similar to current EKG telemetry systems). The centralized alarm display can also be controlled at the bedside, allowing patients and caregivers to minimize false alarms due to movement and artifacts arising during routine nursing or medical care.

To test this new technology, 250 postoperative cardiothoracic patients admitted to a general ward were randomized prospectively to receive either routine EKG monitoring, or EKG monitoring plus continuous pulse oximetry. Data is downloaded in 24-hour epochs from this system. Preliminary analysis suggests a reduction in primary admissions, as well as return admissions to higher acuity care areas, of approximately 25%1. In the group admitted to the ICU with a diagnosis of “respiratory distress,” the reduction was 3-fold (2.8% vs. 0.9%). These data support prior suggestions of the utility of this type of technology.2 Careful coordination and nursing education is instrumental in assuring the success and acceptance of the new system by patients and staff. Any new monitors for general ward patients must be integrated in a way to avoid severe and impractical restrictions of normal patient and provider activities. Thus, patients are monitored only while in bed and when ambulating as part of standard post-operative care.

Encouraged by these findings, the anesthesiologists at the University of Pennsylvania are currently enrolling patients in a prospective, randomized study of approximately 2,000 patients. In addition, the new study will address the costs of routinely implementing a new technology for care of ward patients. Therefore, a full cost-benefit analysis is underway by J. Sanford Schwartz, MD, Professor of Medicine and Health Care Management and Economics of the Wharton Business School. Comparison of cost savings in the form of reduced ICU transfers and potential reductions in length of hospital stay will be balanced against the potential additional costs of routine oximetry monitoring which includes capitalization costs, laboratory tests, respiratory therapy services. This large, prospective study should definitively answer whether a centralized system of oxygen saturation monitoring for postop patients on general care nursing units will reduce ICU admissions of at-risk patients in a cost-effective manner.

Dr. Russel and Dr. Hanson (who is current President of the American Society of Critical Care Anesthesiologists) are from the Department of Anesthesia, University of Pennsylvania, Philadelphia.

References

  1. Russel MW, Hanson CW. Continuous pulse oximetry monitoring and surgical ICU admission rates. Anesthesiology 1998; 89:B4. [An .ASCCA abstract]
  2. Eichhorn JH, Cohen MS, Darvish AH. Postanesthetic hypoxemia detected by continuous telemeteric pulse oximetry. Anesthesia and Analgesia 1992;74:S83.