Central triage of acute stroke patients across a distributive stroke network is safe and reduces transfer denials

Abstract

Background and Purpose:

Mechanical thrombectomy has dramatically increased patient volumes transferred to comprehensive stroke centers (CSC), resulting in transfer denials for patients who need higher level of care only available at a CSC. We hypothesized that a “distributive stroke network” (DSN), triaging low severity acute stroke patients to a primary stroke center (PSC) upon initial telestroke consultation, would safely reduce transfer denials, thereby providing additional volume to treat severe strokes at a CSC.

Methods:

In 2017, a DSN was implemented, in which mild stroke patients were centrally triaged, via telestroke consultation, to a PSC based upon a simple clinical severity algorithm, while higher acuity/severity strokes were triaged to the CSC. In an observational cohort study, data on acute ischemic stroke patients presenting to regional community hospitals were collected pre- vs. post-DSN implementation. Safety outcomes and rate of CSC transfer denials were compared pre-DSN vs. post-DSN.

Results:

The pre-DSN cohort (N=150), triaged to the CSC, had a similar rate of symptomatic intracerebral hemorrhage and discharge location compared to the post-DSN cohort (N=150), triaged to the PSC. Time-to-stroke unit admission was faster post-DSN (2h 40min) vs. pre-DSN (3h 29min), P<0.001. Transfer denials were reduced post-DSN (1.8%) vs. pre-DSN (3.8%, P=0.02), despite an increase in telestroke consultation volume over the same period (median 3 calls per day pre-DSN vs. 5 calls per day post-DSN, P=0.001). No patients who were triaged to the PSC required subsequent transfer to the CSC.

Conclusions:

A DSN, triaging mild ischemic stroke patients from community hospitals to a PSC, safely reduced transfer denials to the CSC, allowing greater capacity at the CSC to treat higher acuity stroke patients.

Background and Purpose

Stroke networks provide acute stroke expertise to remote hospitals that otherwise would have limited access to stroke care¹. Within these networks, the majority of patients are initially cared for by acute stroke-ready hospitals (ASRH) and primary stroke centers (PSC), which administer intravenous thrombolysis and follow standard quality measures for patient management². A comprehensive stroke center (CSC) additionally provides specialty care including neurosurgical/neurocritical care and mechanical thrombectomy, and participates in stroke research¹. As dedicated stroke units impart clinical benefit³, more ‘spoke’ hospitals (both ASRHs and PSCs) are joining stroke networks, resulting in network hubs (often CSCs) accepting transfers from an increased number of spokes^4–6.

The advent of guideline-based mechanical thrombectomy within extended windows has resulted in increased patient volumes transferred to CSCs^{4, 6}. Studies have proposed pre-hospital triage of patients with severe strokes to centers capable of providing mechanical thrombectomy^{3, 7}. As many transferred patients do not ultimately undergo thrombectomy^{4, 8}, experts have suggested that system protocols be put in place to determine which patients should be treated at which facilities^{5, 9}. Nonetheless, increased pressure on the bed capacity at CSCs has resulted in their inability to accommodate transfers for patients who require specialty management such as thrombectomy^{1, 4–6}.

In response to increased demand on bed capacity at our CSC and consistent bed availability at a well-established PSC, we designed a distributive stroke network (DSN) whereby stroke patients at remote, community hospitals were screened by a CSC telestroke neurologist, who transferred lower acuity patients to the PSC, thus conserving CSC beds for higher acuity patients. We hypothesized that the distribution of stroke patients from the community hospital to the appropriate level of stroke center (CSC vs. PSC) within the network, would reduce transfer denials to the CSC, without compromising outcomes for milder stroke patients.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request. The study was approved by the institutional review board with a waiver of informed consent. A STROBE checklist and flow diagram were completed for transparent reporting and available as Supplemental Figure and Table 1.

Table 1.

Distributive Stroke Network (DSN) Criteria for Central Triage to Primary Stroke Center (PSC)

Inclusion criteria * - Clinical diagnosis of ischemic stroke - < 6 hours from last known normal and NIHSS ≤ 4 (with or without tPA) OR ≥ 6 hours from last known normal and NIHSS ≤ 8

Exclusion Criteria - Acute hemorrhage on head CT - Any patient who may benefit from transfer to CSC specifically for an intervention (such as mechanical thrombectomy) - Any patient requiring Emergency Department services due to unclear diagnosis or further evaluation required - Care at CSC in recent past or patient preference for transfer to CSC - No head CT available for review by triaging telestroke neurologist - Transient ischemic attack or suspicion of stroke mimic

^*See Supplemental Methods for additional details.

Distributive Stroke Network Protocol.

Our network CSC provides 24-hour/7-day acute stroke consultation for 72 hospitals in the southern Missouri/Illinois region. The DSN protocol was developed to address the problem of increasing volume of strokes transferred to the CSC, which resulted in decreased bed capacity and transfer denials, thus, delaying care for patients eligible for acute interventions.

The DSN protocol was jointly developed by representatives of the CSC, the pilot PSC, and hospital system (Supplemental Methods). The Figure depicts patient flow through the DSN starting from: 1) patient arrival at the community hospital (spoke site), 2) central triage by the CSC telestroke neurologist, and 3) patient transfer to the PSC for low acuity/severity patients vs. transfer to the CSC for high acuity/severity patients. The DSN algorithm included clinical criteria for the telestroke neurologist to standardize the definition of low acuity patients who could be safely transferred to the PSC (Table 1).

Figure. A distributive stroke network (DSN) allows for central triage of mild acute ischemic stroke patients to a PSC, while increasing bed capacity at the CSC for thrombectomy-eligible patients.

The DSN process begins when the treating provider at a spoke site contacts the CSC “physician access line” by telephone (01). The physician access nurse connects the spoke site with the CSC telestroke neurologist. The two providers discuss the patient’s care. At this time, the telestroke neurologist can elect to see the patient via televideo consultation, though this is not required for triage. The telestroke neurologist initially determines if the patient meets criteria for acute interventions such as thrombolysis and mechanical thrombectomy. Subsequently, the telestroke neurologist follows the DSN algorithm to determine if the patient is a candidate for transfer to the PSC (Table 1) (02). If the patient is a candidate for triage to the PSC (low acuity), the spoke site physician discusses transfer options with the patient, and if the patient is agreeable, the patient is triaged to the PSC. If the patient does not meet DSN transfer criteria, the patient is triaged to the CSC (03).

Data collection.

Baseline variables, outcomes, process metrics, and transfer denials were prospectively collected. After DSN implementation, the “post-DSN” cohort included the first 150 acute stroke transfers to the PSC (enrolled September 2017 – March 2018). To obtain a control “pre-DSN” cohort, we identified the 150 most recent, consecutive CSC-transferred patients meeting DSN transfer criteria (Table 1) prior to DSN implementation (enrolled May 2015 to May 2016) (Supplemental Methods).

Statistical analysis.

Pre-DSN vs. post-DSN cohorts were compared using Welch’s t-test and Chi-squared test. Raw P-values were reported with significance at <0.05. All analyses were performed using SPSS and Minitab.

Results

The pre-DSN and post-DSN cohorts (N=300) were similar, except there were more patients with a history of stroke pre-DSN (P=0.008) (Supplemental Table 2).

First, we evaluated the clinical and safety outcomes of stroke patients transferred to the PSC after DSN implementation. Importantly, no patients transferred to the PSC via the DSN required subsequent transfer to the CSC for higher level of care. Discharge location and rate of symptomatic intracranial hemorrhage were similar pre- vs. post-DSN (Table 2). Rate of stroke mimics transferred, defined as a discharge diagnosis other than stroke, was similar pre-DSN vs. post-DSN.

Table 2.

Clinical and Safety Outcomes, CSC Call Volume, and CSC Transfer Denials Pre-DSN vs. Post-DSN.

	Pre-DSN (CSC) (N=150)	Post-DSN (PSC) (N=150)	P-value
Discharge Disposition			0.23
Home	100 (66.7%)	96 (64.0%)
Inpatient Rehabilitation	44 (29.3%)	39 (26.0%)
Nursing Facility	5 (3.3%)	13 (8.7%)
Died	1 (0.7%)	2 (1.3%)
Outcomes
Symptomatic intracranial hemorrhage (%)‡	0 (0 %)	1 (0.7%)	1.00
Time to Admission (h:mm) *§	3:49 [3:00, 5:42]	2:40 [2:11, 3:37]	<0.001
Length of Stay (days) *	2 [1, 4]	2 [1, 3]	1.00
Stroke Mimic (%)†	23 (15.3%)	20 (13.3%)	0.62
	6 Months Pre-DSN (CSC)	6 Months Post-DSN (CSC)	P-value
Transfer Denial Rate **	3.8% (19 / 501)	1.8% (16 / 894)	0.02
Transfer Call Volume (per day) ***	3 [1, 4]	5 [3, 6]	0.001

^*Data described as median [IQR: 25^th, 75^th];

^‡Defined as intracranial hemorrhage within 36 hours of last known normal accompanying decline in neurological status.

^§Defined as time from initial telestroke consultation to patient admission to stroke center

^†Defined as discharge diagnosis other than stroke.

^**Calculated as number of denied transfers divided by total number of calls (numerator=number of requested transfers that were denied because of limited bed capacity at the CSC; denominator, total community hospital calls for telestroke transfer).

^***Defined as the median number of telestroke calls for stroke neurology transfer per day. The total transfer call volume was 501 pre-DSN and 894 post-DSN.

Second, we evaluated if DSN implementation increased delays in admission to the accepting stroke center or impacted length of hospital stay. Comparing time from initial consultation to hospital admission, patients experienced faster stroke center admission after DSN implementation (3h 29min pre-DSN vs. 2h 40min post-DSN, P<0.001) with no difference in length of hospital stay between the two cohorts (Table 2).

Finally, we evaluated the impact of DSN on bed availability at the CSC by comparing percent of transfer denials (due to insufficient bed availability) for the six months before vs. six months after DSN implementation. The CSC was unable to accommodate transfers due to overcapacity for 3.8% of patient referrals pre-DSN compared to 1.8% post-DSN (P=0.02). This reduction in transfer denials was present despite an increase in daily telestroke call volume post-DSN (P=0.001) (Table 2).

Discussion

Mechanical thrombectomy has dramatically increased transfer volumes to CSCs^4–6. This has resulted in delay or denial of transfers due to insufficient bed availability, with negative consequences on treatment times for thrombectomy-eligible patients⁶. Our study provides a novel mechanism by which a stroke network can maximize its delivery of individualized care across the stroke system, ensuring optimal treatment across a range of stroke severities. We found that central triage of patients from a community hospital (undesignated or ASRH) to a PSC vs. CSC using an acuity/severity-based DSN algorithm was safe and reduced transfer denials to the CSC. Such reduction will likely improve clinical outcomes across the stroke system by maximizing capacity to treat severe strokes at the CSC. Importantly, safety outcomes, length of hospital stay, and stroke mimic rate remained stable after the practice change. While this proof-of-concept study demonstrates that central triage to a designated PSC can improve access to care, this algorithm could be expanded to allow distribution of stroke patients across a subset of hospitals within the network (including PSCs, CSCs, and thrombectomy-capable centers) based on stroke severity, acuity, inter-hospital distance, and real-time bed availability.

As national guidelines have called for comprehensive stroke systems of care^{1, 3, 10}, many have advocated prehospital stroke severity screens to bypass the PSC^{5, 9}; however, this process may result in relatively low positive predictive value for patients who ultimately undergo thrombectomy¹¹. Such practice could stress CSC bed capacity, as well as, lower PSC patient volumes, thus, challenging clinical operations in different ways. For rural patients, whose closest stroke-ready hospital is neither a PSC nor CSC, our findings suggest that central telestroke triage may avoid direct-to-CSC transfer for all patients.

Protocols distributing patients across multiple hospitals in the field of trauma and mass casualties have prioritized: (1) real-time bed management systems, (2) “centralization” of coordinated care and transfers, and (3) streamlined communication from EMS or outside hospitals^12,13. While telestroke networks inherently provide centralized care, additional inputs such as real-time bed management, combined with flexible transfer patterns incorporating patient location, stroke severity, and last known normal time, may further expedite access to regionalized stroke care.

These data were obtained prior to national guidelines recommending delayed mechanical thrombectomy up to 24 hours since last known well. In 2018, we provided education to consulting telestroke neurologists to reinforce the DSN criteria, which excluded any potential candidates for acute procedures such as thrombectomy, as reflected in Table 1. Our study excluded patients with transient symptoms due to a potentially higher rate of mimics in this population; however, the triage of TIAs who warrant hospitalization is another potential population to include in a DSN. Given the single-center and pre-post study design, study findings would ideally be replicated within prospective, multi-center cohorts comparing the DSN approach to standard care.

Conclusion

A DSN, triaging mild ischemic stroke patients from community hospitals to a PSC, safely reduced transfer denials to the CSC despite increased stroke patient volumes over time, allowing for greater capacity to treat higher acuity stroke patients at the CSC.

Non-standard Abbreviations and Acronyms:

DSN

Distributive Stroke Network

CSC

comprehensive stroke center

PSC

primary stroke center

ASRH

Acute Stroke Ready Hospital