Good is not Good Enough: The Benchmark Stroke Door-to-Needle Time Should be 30 Minutes

The importance of treating ischemic stroke patients quickly has long been recognized, and the mantra “Time is brain”, is now ubiquitous. ^{1 , 2} Unfortunately, the thinking “We still have time in the treatment window…” is occurring too often during the acute stroke code. The treatment window from time of onset is 4.5 hours ³ in most guidelines, ^{4 , 5} yet there is declining benefit as time elapses. A 1997 National Institute of Neurological Disorders and Stroke (NINDS) Symposium and the subsequent Brain Attack Coalition ⁶ set the standard of 60 minute door-to-needle time. ⁷ This door-to-needle time was arbitrary but designed to provide a useful metric. It has now been incorporated into both national guidelines and accreditation standards ^{8 – 10} but has been treated more like a guide or a range rather than a hard target. Parkinson’s law - “The job expands to fit the time available” - is as true in stroke care as it is in economics. ¹¹ We argue that to change this mentality we must revise our target downward to a 30-minute median door-to-needle time.

Today we know that: (1) “Time is brain” in the most urgent way; (2) with attention to systems change, median door-to-needle times of 30 minutes can be achieved; (3) we fail to meet a 60-minute door-to-needle time for most patients in most centres. The sooner blood flow is restored in ischemic stroke, the better the outcomes. ^{12 , 13} For each 15 minute reduction in delay, there is an estimated 4% improvement in good clinical outcomes. ¹³ The number of neurons that are lost in a middle cerebral artery stroke has been quantified to be 1.9 million per minute or 114 million neurons every hour, which translates to 3.6 years in accelerated aging. ² While this is based upon an average rate of infarct progression and individual patients may vary substantially, the damage is staggering. Acute ischemic stroke is a medical emergency that is as, or more, time sensitive than myocardial infarction ¹⁴ and trauma. ¹⁵ We need to treat it as such. Using conservative calculations of 800 000 strokes per year in North America, where 40 000 are treated with intravenous tissue plasminogen activator (tPA), and 16 000 (40%) of these have a good outcome, an additional 3 000 patients will have good outcomes based on a decrease of 30 minutes in time to treatment. ¹⁶ In addition, this calculation does not allow for a projected increase in the number of patients treated. The optimal door-to-needle time is the fastest possible time that preserves safety and appropriateness.

When hospitals and stroke teams use a systematic quality improvement approach to stroke thrombolysis, significantly lower door-to-needle times can be realized. In Helsinki, an initial median door-to-needle time of 1 hour 45 minutes in 1998, was reduced to 50 minutes by 2004, ¹⁷ and then further reduced to 20 minutes in 2011 (excluding basilar artery occlusion). ¹⁸ This model was replicated at the Royal Melbourne Hospital, where median door-to-needle times were reduced from 61 minutes to 25 minutes. ¹⁹ Improvements in the United States (US) have been reported by the Target-Stroke initiative. ^{20 , 21} Importantly, these improvements in door-to-needle treatment time have not resulted in an increase in complications. A key concept in fast treatment is the paradigm of parallel (rather than serial) diagnostic evaluation, assessment and treatment. The Emergency literature discusses the concept of “swarming” for acute trauma and acute myocardial infarction. The same must apply to acute stroke.

In Canada, comprehensive stroke centres have been able to achieve similar reduction in door-to-needle time. In Calgary, as a result of a six-sigma quality improvement process, the stroke team’s median door-to-needle time was reduced to 32 minutes (range 11–121 min) from 52 minutes (range 6–193 min). Comparable magnitude improvements have been seen in Montreal (median 54 min reduced to 30 min), Ottawa (median 74 min reduced to 37 min) and at other Canadian stroke centres. However, both the US and Canada have globally failed to meet the 60-minute door-to-needle target. Results from the American Heart Association “Get With the Guidelines” data show that only 26.6% of patients received tPA within 60 minutes of arriving at a hospital. ²² In Canada, results from the 2011 national stroke audit on quality of stroke care showed that only 34% of patients had a door-to-needle time equal to or less than 60 minutes. ²³

To improve we can learn from success. The process changes that were employed by the groups in Helsinki, Melbourne and our groups in Canada are not novel. [Table] These groups have simply recognized the urgency in hyper-acute stroke care of ischemic stroke. They developed processes to ensure this mindset permeates to Emergency Medical Services (EMS), Emergency Department, and stroke team. They have worked in parallel as a team. Based on results from these groups, we ask: How long does it take to administer tPA? Is a median 60-minute door-to-needle time acceptable? Internationally, other professional groups have recognized that, globally, we are not treating stroke fast enough. ^{22 , 24}

Table 30 changes to get to 30 minutes A list of process changes, used at centres across Canada, that may expedite the treatment of acute ischemic stroke with intravenous alteplase

EMS=emergency medical services; IV=intravenous; PACS=picture archive and retrieval system; ED=Emergency department; CT=computed tomography; tPA=tissue plasminogen activator; STAT=“immediately” (used in medical contexts from latin “statim”); Code=requiring a team of providers to rush to the specific location and begin immediate treatment; MD=Medical Doctor; RN=Registered Nurse; DTN=Door-to-Needle

We are calling for an aggressive update of our targets as a necessary step to better patient care. Today we should set a door-to-needle benchmark at a 30-minute median (60-minute 95^th percentile). To achieve a system where 95% of patients get treated within 60 minutes, where patients are treated with the fastest possible door-to-needle time, we need a median door-to-needle time of 30 minutes. Systems that aim for and achieve this target will have engineered their processes to allow for predictable and unavoidable variance due to patient factors that can necessarily delay treatment. Centres that can achieve these results will see more of their patients walking out of hospital and will have met an essential criterion for being named a top stroke centre.

This target is not easy but it is demanded by the immutable biology of the disease. These efforts will require ongoing efforts to maintain success. Key implications of such a change could include centralization of acute stroke care to specialized centers where this makes sense geographically, increased use of telemedicine, ²⁵ a strong emphasis on training to avoid treatment of stroke mimics, and more careful and discriminative use of diagnostic imaging. Success will be achieved in different ways at different centres. For our patients, let us commit to reducing delays in stroke thrombolysis and set a new standard for ischemic stroke treatment.