Shortsighted JC standard lacks understanding of pharmacist's role in prospective order review

From the January 23, 2002 issue

ISMP and ASHP are urging healthcare practitioners to comment immediately on the Joint Commission on Accreditation of Healthcare Organizations' (JC) revised standards on medication use in hospitals, ambulatory care, and long-term care. Of particular concern are revisions that would make exceptions to the requirement that a pharmacist reviews each medication order BEFORE the drug is dispensed or administered. The new exceptions apply when physicians enter medications into a computerized prescriber order entry (CPOE) system that has an active physician alert mechanism and a working, approved set of appropriate checks and balances. A review by a pharmacist AFTER administration of the drug is sufficient. Furthermore, it's implied in several standards that pharmacy services may not be required in acute care settings. While there are many reasons why these proposed revisions are troubling, an error reported just last week sheds some light on the issue.

A cardiologist called the pharmacy to order INTEGRILIN (eptifibatide) on a patient in the cardiac catheterization unit. To ensure that a pharmacist would always prospectively check the patient's serum creatinine and prescribed dose, the drug was not in floor stock. The patient had been transferred from another hospital and had no lab studies drawn at this hospital. Upon finding no serum creatinine levels in the computer system, the pharmacist asked the physician if he knew the patient's most recent level. In records sent from the transferring hospital, the physician found that the patient's creatinine level was 10.7 mg/dL. Upon overhearing this, another staff member commented that a high creatinine level was no problem since the patient was on hemodialysis. The physician and staff member were unaware that Integrilin is contraindicated in dialysis-dependent patients, and that dose adjustments are necessary in patients with a serum creatinine between 2-4 mg/dl. There is no clinical information about use in patients with a serum creatinine over 4 mg/dl. The pharmacist suggested ReoPro as a safer alternative for this patient.

While one could argue about the degree of risk for major bleeding in this patient with severe renal impairment, that's hardly the point. What this error clearly demonstrates is that, even if the physician had prescribed Integrilin via CPOE with a functional alert system, no warning would have been triggered due to lack of information in the system about the patient's creatinine level or hemodialysis status. At best, the system could have prompted the physician to check the patient's creatinine, but would that have made a difference if he was unaware that the drug was contraindicated in patients on dialysis?

Unfortunately, errors like this are not isolated because there's a strong possibility that crucial information (e.g., new allergies, comorbid diagnoses, lab values from external sources, recent weights) may be missing in computer systems, which makes alert systems ineffective. While we fully support the immense value of CPOE with functional alert systems, it can't replace the pharmacist. A functional CPOE system offers a crucial double check that otherwise may be overlooked by reliance on people, while pharmacists' prospective screening of medication orders offers a crucial double check that otherwise may be overlooked by reliance on computers. Patient safety depends on both. We have to agree with the person who reported this error when he eloquently stated: "I can't support an operational framework which removes the pharmacist from being able to provide an up-front, supportive double check, assuring at least that drugs are not used when clearly contraindicated." If you feel the same way, let JC know by January 28, 2002.