In a meta-analysis of more than 200,000 patients with TIA or mIS, risk for subsequent stroke within 90 days was 2.1% for those treated in a TIA clinic vs 2.8% for patients treated in inpatient settings, which was not significantly different. The risk for patients treated in an emergency department (ED) was higher, at 3.5%.
“The message is that if you do the correct risk stratification and then triage patients based on their risk profile, you can safely discharge and have a timely follow-up for the patients who have low risk for a subsequent event,” co-investigator Ramin Zand, MD, vascular neurologist and stroke attending physician at Geisinger Health System, Danville, Pennsylvania, told Medscape Medical News.
The findings were published online January 5 in JAMA Network Open.
Higher Risk in EDs
There is currently no consensus on the care protocol for patients with TIA or mIS, and the rate at which these patients are hospitalized varies by region, hospital, and practitioner, the investigators note.
Previous studies have indicated that outpatient management of certain individuals with TIA can be safe and cost-effective.
The current researchers searched for retrospective and prospective studies of adult patients that provided information about ischemic stroke after TIA or mIS. Studies that used time- and tissue-based definitions of TIA were included, as well as studies that used various definitions of mIS.
The investigators examined care provided at TIA clinics, inpatient settings (such as medical–surgical units, stroke units, or observation units), EDs, and unspecified settings. Their main aim was to compare outcomes between TIA clinics and inpatient settings.
In all, 226,683 patients (recruited between 1981 and 2018) from 71 studies were included in the meta-analysis. The studies examined 101 cohorts, 24 of which were studied prospectively. Among the 5636 patients who received care in TIA clinics, the mean age was 65.7 years, and 50.8% of this group were men. Among the 130,139 inpatients, the mean age was 78.3 years, and 61.6% of the group were women.
Results showed no significant difference in risk for subsequent stroke between patients treated in the inpatient and outpatient settings.
Among patients treated in a TIA clinic, risk for subsequent stroke following a TIA or mIS was 0.3% within 2 days, 1.0% within 7 days, 1.3% within 30 days, and 2.1% within 90 days. Among those treated as inpatients, risk for subsequent stroke was 0.5% within 2 days, 1.2% within 7 days, 1.6% within 30 days, and 2.8% within 90 days.
Risk for subsequent stroke was higher among patients treated in the ED and in unspecified settings. At the EDs, the risk was 1.9% within 2 days, 3.4% within 7 days, 3.5% within 30 days, and 3.5% within 90 days. Among those treated in unspecified settings, the risk was 2.2% within 2 days, 3.4% within 7 days, 4.2% within 30 days, and 6.0% within 90 days.
Patients treated in the ED also had a significantly higher risk for subsequent stroke at 2 and 7 days compared with those treated in inpatient settings and a significantly higher risk for subsequent stroke at 2, 7, and 90 days compared with those treated in TIA clinics.
“Most Comprehensive Look”
“This is the most comprehensive look at all the studies to try and answer this research question,” said Zand. The results were similar to what was expected, he added.
The infrastructure and resources differed among the sites at which the various studies were conducted, and the investigators adjusted for these differences as much as possible, Zand noted. A certain amount of selection bias may remain, but it does not affect the overall conclusion, he added.
“Timely outpatient care among low-risk TIA patients is both feasible and safe,” he said.
Zand noted the findings have implications not only for patient management but also for the management of the health system.
“It’s not feasible nor desirable to admit all the TIA patients, especially with the lessons that we learned from COVID, the burden on the health systems, and the fact that many hospitals are operating at full capacity right now,” he said.
The recommendation is to hospitalize high-risk patients and provide outpatient evaluation and workup to low-risk patients, he added. “This is exactly what we saw in this study,” Zand said.
Commenting for Medscape Medical News, Louis R. Caplan, MD, professor of neurology at Harvard Medical School, Boston, Massachusetts, noted that evaluation of patients with TIA or mIS “can be done very well as an outpatient” if clinicians have experienced personnel, the outpatient facilities to do the studies necessary, and criteria in place for deciding who to admit or not admit.
However, the decision on whether to choose an inpatient or outpatient approach for a particular patient is complicated, said Caplan, who was not involved with the research.
Clinicians must consider factors such as whether the patient is mobile, has a car, or has a significant other. The patient’s symptoms and past illnesses also influence the decision, he added.
Caplan noted that in the meta-analysis, far fewer patients were seen in the TIA clinics than were seen in the inpatient setting. In addition, none of the studies used uniform criteria to determine which patients should undergo workup as outpatients and which as inpatients. “There was a lot of selection bias that may have had nothing to do with how sick the person was,” Caplan said.
In addition, few hospitals in the United States have an outpatient TIA clinic, he noted. Most of the studies of TIA clinics that the researchers examined were conducted in Europe. “It’s easier to do [that] in Europe because of their socialized medicine,” said Caplan.
But TIA clinics should be more widespread in the US, he added. “Insurance companies should be willing to pay for comparable facilities, inpatient and outpatient,” he said.
The study was conducted without external funding. Zand reported no relevant financial relationships. Caplan was an investigator for TIAregistry.org, which analyzed the outcomes of treatment in TIA clinics in Europe.
JAMA Netw Open. Published online January 5, 2022. Full text
Follow Erik Greb on Twitter: @MedscapeErik.