Most adverse drug events unreported: Take these steps

Use data to identify safety issues

Adverse drug events (ADEs) occur in about 3.1% of all hospital stays, according to a report from the Agency for Healthcare Research and Quality (AHRQ). Here are key findings:

  • Most ADEs (90%) were due to side effects from a medication that was properly administered.
  • Only 8.6% of ADEs were due to drug poisoning: accidental overdose, wrong drugs given or taken, or drugs taken inadvertently.
  • Older patients were more likely to experience adverse effects from properly administered drugs.
  • Drugs most commonly associated with ADEs were corticosteroids, anticoagulants, anti-cancer and immunosuppressant agents, opiates, and analgesics and fever reducers.
  • Average total hospital costs for patients who experienced drug side effects or other ADEs were $10,100 compared with an average cost of $7,600 for patients who didn't experience ADEs.

However, these results are underreported for medication errors leading to ADEs, because the report used coding data, according to Allen J. Vaida, PharmD, FASHP, executive vice president of the Huntingdon Valley, PA-based Institute for Safe Medication Practices (ISMP).

"Very few errors are going to be picked up with coding data," he says. In fact, many of the "adverse effects" coded may have actually been due to errors, which practitioners and coders wrongly attributed to "adverse effects of drugs properly administered," notes Vaida.

"We frequently see that adverse effects — better called adverse reactions — are attributed to the side effects of the drugs, rather than errors in prescribing, dispensing, or administration, which oftentimes caused the effect," he adds.

To ensure that ADEs caused by errors are tracked accurately, perform focused reviews on certain medications, recommends Vaida. For example, track blood glucose levels and episodes of hypoglycemia or hyperglycemia in patients receiving insulin or certain oral medications to treat diabetes. "Direct observation of medication administration is another proven method that has been used to quantify errors," he says.

The drug categories most commonly associated with ADEs are no surprise, says Vaida, and have been identified for years. He points to the ISMP's list of "high alert drugs" that includes opiates, anticoagulants, and other medications identified in this report. "Quality professionals must follow the safety literature, and implement known safety recommendations for this class of drugs," says Vaida.

Use data to support change

Make sure that recommendations for error prevention are implemented at your organization and track ADEs, Vaida advises. He recommends looking at elevated international normalized ratios and partial thromboplastin times, which could signal an ADE from some anticoagulants, and monitoring the use of reversal agents, such as naloxone or flumazenil, which are signals of excessive sedation from dosing or administration errors from opiates and benzodiazepines.

Use internal and external error reporting data to justify technology investments, such as electronic prescribing with decision support or bedside bar coding, says Vaida. You also can use these data to gain support for implementing low-cost measures such as independent double checks, standardized order sets, checklists, and other safety measures.

Use the external data, such as errors in the literature, to conduct a short, focused look at how your organization is performing in these areas. "If you have internal data, you can compare; if you don't have good internal data, you can gain support from committees to conduct the internal review by using the external data to justify it," says Vaida.

If the same safety issues are identified after the internal review is done, you then can use both the external and internal data to support change, says Vaida.

Your internal data should include information from your current medication error reporting system or incident reporting system, information from risk management on serious events that are in litigation or are being reviewed, and use of focused reviews.

This information doesn't necessarily give an accurate rate of errors or incidents, but it can help you see if the same occurrences from the external information are occurring in your hospital. "If you have a poor reporting system, perform the focused internal review for each of the areas identified in the safety publications," says Vaida.

Implement these practices

ADEs cannot be reliably predicted and not every ADE is preventable, says Marybeth Farquhar, director of AHRQ's quality indicators activities. "ADEs occur throughout the many steps of the medication administration process in hospitals," she says.

However, there are several practices that may minimize ADEs and their adverse effects on patients. For example, a thorough medical history should be taken from the patient and/or caregiver. This should include the names of all medications the patient is currently taking, including vitamins and herbal supplements, doses, and frequency.

"Some hospitals go so far as to ask the patient to bring all their medication bottles with them to the hospital, so that they can be recorded during admission procedures," says Farquhar.

Allergies to medications or foods and prior reactions to medications should be clearly noted in the patient record and visible to clinicians via patient identification bands, says Farquhar. Prescribing clinicians should consider the patient's age, weight, underlying condition, and renal function when ordering medications.

Good communication among the health care team is essential throughout the medication administration process, says Farquhar. "Patient monitoring is important to detecting ADEs. However, computerized systems can help hospital staff identify ADEs and minimize their effects," she says.

For example, a computer system can alert hospital staff if any signals of possible ADEs occur, such as certain lab test results, high or low levels of certain medications, and pharmacy orders for medications generally used to treat allergic reactions.

Also, by integrating the computer system to link pharmacy, lab, and other hospital information about the patient, pharmacists can identify and notify physicians about drug allergies, drug-drug interactions, drug-food and drug-condition contraindications, says Farquhar. She also recommends the following to prevent ADEs:

  • using the Food and Drug Administration's MedWatch program to report serious drug reactions;
  • improving hospital incident reporting systems;
  • creating a culture of safety that encourages reporting ADEs;
  • relying more on pharmacists to advise prescribing clinicians;
  • promoting clinician education on medications and possible side effects;
  • improving nurse medication administration and monitoring systems.

There should be more reporting of ADEs, says Jeffery Laux, clinical pharmacist at Covenant HealthCare in Saginaw, MI. "Most probably go unreported and therefore cannot be tracked."

Most events are picked up in retrospective reports, often by coders as they examine medical records, says Laux. "It makes it very hard to find complete information when the patient has already left the hospital," he says. "It is also helpful to interview personnel who witnessed the event."

If the events are reported, these must be followed up with some detail — this is the only way to determine why the event happened. "If you cannot figure out why the events happen, it is impossible to try to correct them," Laux says.

At Winston-Salem, NC-based Novant Health, a successful program to track ADEs was implemented, which included all eight hospitals in the system, says Mike Gum, director of pharmacy for Presbyterian Healthcare in Charlotte, NC.

For one week of every month, a clinical pharmacist reviewed all administrations of the following: vitamin K, naloxone, Kayexalate, Romazicon, and laboratory values of an international normalized ratio (INR) greater than 3.5 or elevated digoxin levels.

"After phase one, it was decided that elevated INRs was the biggest issue" says Gum. "We then concentrated on that. We were able to decrease the incidence of overall elevated INRs by 47%." The following steps were taken:

  • A warfarin dosing program was developed.
  • Routine laboratory monitoring of patients receiving warfarin was standardized to be done on admission and at least every three days.
  • Clinical pharmacists were given the ability to order an INR if deemed appropriate.
  • Physicians were routinely educated. "This was accomplished by letters to office-based physicians, because many of the elevated INRs were on admission, as well as routine updates of hospitalists," says Gum.

(The full report, "Adverse Drug Events in U.S. Hospitals, 2004," can be accessed at

[For more information, contact: Mike Gum, Director of Pharmacy, Presbyterian Healthcare, 200 Hawthorne Lane, Charlotte, NC 28204. Phone: (704) 384-4000. E-mail:
Jeffery Laux, Clinical Pharmacist, Covenant HealthCare, 1447 N. Harrison Street, Saginaw, MI 48602. Phone: (989) 583-4478. E-mail:]