Boston-based researchers find outpatient adverse events are numerous, serious and understudied

Adverse events are “relatively common and often serious” in outpatient settings, according to authors of a recent retrospective review of patient records in Massachusetts. This assessment comes from the same well-regarded research group that produced the SafeCare Study of harm in inpatient care last year. 

Key findings from “The Safety of Outpatient Health Care” include:

  • 7% of patients experienced at least one adverse event 
  • 17.4% of adverse events identified were considered serious in nature; 2.1% were life-threatening; none were fatal
  • 10.4% of Black or African American patients and 13.1% of patients over the age of 85 experienced an adverse event
  • 64% of adverse events identified were medication-related  
  • Variation in adverse event rates across the 11 study sites ranged from 1.8% to 23.6% of patients who received care
  • Study likely undercounted the true number of adverse events

“Our findings are quite worrisome,” says lead author David M. Levine, M.D., M.P.H., general internist and Clinical Director of Research and Development, Healthcare at Home, Mass General Brigham. “We found that harm was common in the outpatient setting and often preventable.”

Levine David

David M. Levine, M.D., M.P.H.

Records for the outpatient study came from health systems that participated in the earlier inpatient research and represented care delivered in 2018. A total of 3,103 patient records were reviewed from outpatient sites with between 74 and 2,670 physicians and 1 to 31 locations. Most were associated with teaching hospitals in the Boston area.

“Much more is known about inpatient safety, although patients spend nearly all their lives outside the hospital, so this is an important contribution,” notes David W. Bates, M.D., Chief of the Division of General Medicine and Primary Care, Brigham and Women’s Hospital, and Medical Director of Clinical and Quality Analysis for Information Systems at Mass General Brigham. Drs. Bates and Levine were also involved in the earlier SafeCare Study, of which Dr. Bates was the lead author. 

David Bates highres

David W. Bates, M.D.

Dr. Bates observes that “medications were by far the most frequent cause of adverse events.” In 2011, a large national survey of ambulatory care estimated that adverse drug events were responsible for 4.5 million ambulatory visits. Urmimala Sarkar, M.D., M.P.H., lead author of that study, later commented, “Just to give a sense of proportion, that's more visits than for strep throats.” 

Adverse drug events are relatively easy to detect in record reviews, which may help explain why they accounted for 64% of the adverse events identified. Diagnostic errors, on the other hand, are difficult to ascertain and are almost certainly underrepresented in this and other studies.

Other adverse events identified in the Massachusetts study included healthcare-associated infections (14.8%), surgical/procedural adverse events (14.2%), patient care adverse events (8.3%) and perinatal/maternal adverse events (0.7%)

Wide variation poses a challenge to analysis of outpatient safety

The variety of settings included in the study — primary or specialty care, radiology, phlebotomy, ambulatory surgery, “or any other outpatient service cataloged in the EHR” — demonstrates the challenge researchers face in analyzing adverse events in outpatient care.

Emergency department encounters were included as outpatient visits when the patient was not admitted or placed under observation for over two midnights. 

Researchers recorded the setting where each adverse event was understood to have occurred. An ambulatory patient seeking treatment for an apparent adverse event may have been harmed elsewhere, including in self-care at home. Conversely, patients harmed by outpatient care in one setting may go to another for treatment and, therefore, not appear in the study’s results. The authors note that the two circumstances offset each other to some unknown degree, adding uncertainty to the data.

Older adults were more likely to have experienced an adverse event than younger adults, with 13.1% of people aged 85 or older and 10.6% of those between 65 and 84 affected. 

Results also show some demographic differences in harm. Among patients who are Black or African American, 10.4% experienced an adverse event versus 4.1% of patients who are Asian. Differences among other racial groups were not significant. 

Dr. Sakar, who authored the 2011 study on outpatient adverse events, wrote an editorial that accompanies the new outpatient study. She notes that patients and family members often play a more active and consequential role in managing their outpatient vs. inpatient care, a dynamic that should be considered in safety studies.

Urmimala PSB

Urmimala Sarkar, M.D., M.P.H.

Traditional methods are inefficient and may undercount adverse events 

For the outpatient study, the selected patient records were distributed among seven trained nurse reviewers who examined them for “triggers” that indicate a possible adverse event. Reviewers documented and summarized those results and forwarded them with the records to six physician adjudicators for further analysis.

This approach likely undercounts adverse events. It sees only data recorded in each health system’s electronic record. The small size of some samples limits comparison, and despite training and expert knowledge, reviewers and adjudicators are subject to human error. This method, which has long been used to estimate the incidence of adverse events, is also time- and labor-intensive.

The study's authors note that automating the identification of adverse events could produce more complete and reliable information. Dr. Bates observes, “The rates of some types of harm, like adverse drug events, have been detected primarily through spontaneous reporting, which results in large underestimates of the true rate of the events.”

Dr. Bates, who serves on a task force currently advising the Betsy Lehman Center on a statewide voluntary pilot initiative to bring automated safety surveillance tools to Massachusetts hospitals, thinks that automated surveillance will “become the norm and that this will give us more accurate estimates of safety than we have today, although this approach is much less well developed for outpatients than inpatients.”

Inpatient safety lacks proven solutions

The authors also point to a lack of proven methods for reducing harm in ambulatory settings. They cite a Cochrane systemic review of methods used to reduce medication errors in primary care, which they note found that "interventions probably made little or no difference.” Other initiatives, such as clinician decision support programs and indication-based medication labeling, have shown promising if not conclusive results. 

In addition to increased efforts to measure and understand adverse events in ambulatory care, Dr. Levine says, “Interventions for improved patient safety in the outpatient setting need to be a primary focus nationwide.”


We want to hear from you!

Email us your feedback and comments: