Savings offset costs associated with CPOE: Can you afford to omit it in future strategic plans?

Since the Institute of Medicine report in November 1999, professional and lay media have given unprecedented coverage to computerized prescriber order entry (CPOE). While few could argue with the clear evidence that well-designed CPOE systems hold enormous potential to reduce errors1, this technology could require millions of dollars to implement and maintain. For example, the CPOE system at Boston's Brigham and Women's Hospital (BWH) cost about $1.4 million (in the mid-1990s) for in-house development and hardware, and at least $500,000 a year for maintenance. While this dollar outlay seems staggering at first glance, the cost savings that accompany CPOE are even more impressive - between $5 and $10 million per year at BWH.2 How are such large cost savings achieved? Just a glimpse into CPOE through the following scenario can quickly demonstrate its power to vastly improve care and reduce costs.

Before morning rounds, a physician logs on to the CPOE system to print a list of all patients on his service. He is immediately presented with an alert about a crucial dose modification for one of his patients on tobramycin with a low creatinine clearance based on today's lab values. With the click of a mouse, the physician enters the appropriate patient screen and makes the suggested dose modification. As he attempts to leave the patient profile, the system also suggests ordering appropriate follow-up creatinine levels. As rounds progress, most orders are straightforward and easily entered into the system. Other orders trigger assistance, reminders, or alerts as appropriate. For example, when the physician orders TPN, the system calculates the additives based on the patient's most current lab values, age, and weight. When prescribing an H2-blocker, a screen succinctly explains a recent formulary change and the physician readily orders the hospital-selected H2-blocker at the dose suggested. At one point, the system alerts the physician to a positive sputum culture and suggests appropriate medications while considering sensitivity information, drug interactions, and patient allergies. On another patient, the physician orders lab studies and easily requests the system to page him (via a beeper number already in the computer) as soon as the results are available. When discharging a patient, a template appears on the screen with all current drug therapy for review. After any necessary revisions, the physician prints a copy for the patient. After rounds, he prints patient-specific information sheets to give to covering physicians for reference. If a covering resident overrides a serious dose alert (e.g., chemotherapy), the order will be electronically conveyed to a senior staff physician, who must cosign the order before implementation.

Later during office hours, the physician diagnoses an elderly patient with community-acquired pneumonia and notifies the hospital of admission. He accesses the hospital CPOE system from his office and easily reviews information about prior hospital care. If a standard order set/pathway has been established for community-acquired pneumonia, the system displays an admission template with order options so he can check the parameters for each order. If there are no standard orders and he prescribes a third-generation cephalosporin, the system prompts for an indication and suggests another available choice that would reduce the risk of resistant strains. Additional prompts may suggest drug levels for certain antibiotics prescribed and low-level anticoagulation therapy if the patient is on bedrest. Each order is immediately transferred to the nursing unit and pharmacy, thus avoiding problems with delays, verbal orders, illegible handwritten orders and signatures, error-prone transcription, and time consuming order clarification. By the end of the day, the physician has spent about 27 minutes using the CPOE system, similar to time previously spent with paper order systems3.

The overall financial impact at BWH from CPOE was further broken down by specific interventions3. For example, over one year, enhanced allergy warnings and drug-drug interactions resulted in cost savings of $500,000 and $160,000, respectively. Simply displaying lab charges averted about $1 million in charges, and alerting prescribers to redundant lab orders saved another $75,000. Specific guidance when ordering human growth hormone resulted in an 85% reduction in orders and a cost saving of $177,000 in charges. Likewise, the hospital saved $500,000 after 92% of prescribers switched to an effective but less costly dosing frequency of ondansetron suggested by the system. Another $640,000 in costs was saved through suggestions to change doses based on the patient's renal function and age. These and many more examples point to real bottom-line savings when CPOE systems are fully maximized. Further, these savings relate to costs associated with extended length of stay and additional tests and treatments. It does not account for costs to the patient or health system for disability due to adverse outcomes.

To help offset the initial costs of error-reducing technology, Sens. Bob Graham (D-FL) and Olympia Snowe (R-ME) recently introduced a bill that would make $97.5 million in grants available from 2002-2011 to assist hospitals and skilled nursing homes. Sen. Schumer (D-NY) also plans to introduce similar legislation to fund CPOE. Sen. Snowe notes that every day lost in implementing new technology means more lives lost. While it's true that CPOE is very costly to implement and that vendor systems today may not perform at the precise level described above, we can no longer use financial constraints as a compelling reason to avoid such expenses in our strategic plans for the future. CPOE is a cost effective solution

References: 1) Bates DW, et al. Effect of computerized physician order entry and a team intervention on prevention of serious medication errors. JAMA 1998;280:1311-16. 2) Bates DW. Using information systems to improve performance. Swiss Medical Weekly 1999;129:1913-9. 3) Teich JM, et al. Toward cost-effective, quality care: the Brigham integrated computing system. Ed. Steen EB. Proceedings from the Second Annual Nicholas E. Davies CPR Recognition Symposium. Computer-based Patient Record Institute; May 1-2, 1996, Washington, DC