Intravenous thrombolysis for ischaemic stroke: short delays and high community-based treatment rates after organisational changes in a previously inexperienced centre
- 1Department of Neurology, Sørlandet Hospital Kristiansand, Norway
- 2Department of Neurology, Haukeland University Hospital, Bergen, Norway
- 3Institute of Clinical Medicine, University of Bergen, Bergen, Norway
- 4Hospital of Rehabilitation, Rikshospitalet University Hospital, Kristiansand, Norway
- Dr A Tveiten, Department of Neurology, Sorlandet Sykehus Kristiansand, 4604 Kristiansand, Norway;
- Accepted 30 September 2008
Aim: To evaluate hospital delays in thrombolytic treatment before and after organisational changes and community-based treatment rates in a previously inexperienced centre.
Methods: The delays before and after organisational changes made in 2006 were compared using a prospective treatment database. In a 6-month period in 2007, a community-based search was performed for all hospitalisations for ischaemic stroke. The number of patients admitted within the 0–3 h time window and the proportion treated with tissue plasminogen activator were analysed.
Results: The number of treatments increased fourfold from 2005 to 2007 with a significant reduction in mean door-to-needle time from 60 min to 38 min (p = 0.002). In the community-based series, 14/137 patients (10%) hospitalised with ischaemic stroke and 13/32 patients (41%) admitted in the 0–3 h window were treated.
Conclusions: An inexperienced stroke centre can rapidly implement the necessary logistics to deliver thrombolysis to a large proportion of patients with acute stroke with short hospital delays. Important factors are probably prenotification of a team and the initiation of thrombolytic treatment in the emergency room.
In the SITS monitoring study (SITS-MOST), treatment with intravenous tissue plasminogen activator (tPA) was safe even in centres without previous experience from randomised controlled trials.1 There are few published data on hospital delays in inexperienced centres and treatment rates in communities served by inexperienced centres.
Sorlandet Hospital Kristiansand in Vest-Agder County in Southern Norway serves a well-defined catchment area with 160 000 inhabitants, of whom 80 000 live in the city of Kristiansand and the rest in smaller cities and rural areas, the most remote within 3.5 h drive from the hospital. All patients with acute stroke are admitted regardless of age and severity. When tPA was licensed in Norway in 2003, our hospital had no experience in thrombolytic treatment of ischaemic stroke.
The aims of this study were (1) to evaluate in an inexperienced centre the treatment rates and hospital delays before and after organisational changes; and (2) to assess in a well-defined community the number of patients being hospitalised for ischaemic stroke, the proportion admitted within 3 h and the proportion treated with tPA.
All patients treated with tPA since licensing in 2003 were prospectively entered into a database. Age, gender, postal code, time of symptom onset, time of admission, time of treatment and stroke severity as assessed by the National Institute of Health Stroke Scale (NIHSS) were entered.
A community-based assessment of acute ischaemic stroke leading to hospitalisation was performed during a 6-month period from January to June 2007. We searched the electronic hospital patient files for the ICD10 codes I63.0–I63.9 and G45.9, restricted to patients in the defined catchment area. Patients from outside the catchment area referred to us for thrombolysis were excluded.
Practical routines were continuously improved, with emphasis on teaching of involved personnel. In 2006 two major organisational changes were made: in January a flow chart for the emergency medical services with screening criteria for tPA treatment was implemented and, if criteria were met, the emergency room (ER) was notified and a team response triggered. Upon arrival of the patient the neurologist was present in the ER, the CT machine was ready and blood was drawn. In June we moved the start of treatment from the stroke unit to the ER.
For each year, the number of treatments, onset-to-door time (ODT), door-to-needle time (DNT) and onset-to-needle time (ONT) were analysed. To evaluate changes in delays before and after the organisational changes in 2006, we compared ODT, DNT and ONT in 2005 with the first 6 months of 2007. We also compared the number of treated patients living more than 30 min (by car) from the hospital in the two periods.
Symptomatic intracerebral haemorrhage (SICH) and functional outcome 3 months after treatment were assessed in all patients treated up to 30 September 2007. SICH was defined as any haemorrhage combined with any worsening. Functional status 3 months after treatment assessed with the modified Rankin Scale (mRS) was recorded by a study nurse or stroke neurologist, either by outpatient examination or structured telephone interview. A good outcome was defined as mRS 0–2.
Descriptive statistics were used to summarise the baseline characteristics age, gender, pretreatment NIHSS and time of admittance. A t test was used to compare DNT, ONT and the number of treated patients living more than 30 min (by car) from the hospital in 2005 and 2007. p Values <0.05 were considered statistically significant. Statistical analysis was performed with SPSS Version 16.
From 2005 to 2007 the number of patients treated increased fourfold with a statistically significant reduction in DNT. Median DNT in 2007 was 38 min. The number of treatments, mean ODT, DNT and ONT for each year are shown in table 1.
Of the treated patients, 10/18 in 2007 compared with only 1/8 in 2005 lived more than 30 min by car from the hospital (p = 0.042).
By 30 September 2007, 71 patients had been treated with tPA. All patients had complete 3-month follow-up data. The median age was 69 years (range 33–88) and 47 (66%) were men. The median pretreatment NIHSS was 10 (range 2–29). SICH was seen in two patients (2.8%). A good outcome (defined as mRS 0–2) was seen in 44/71 patients (62%).
The community-based series of hospitalisations for ischaemic stroke in the first 6 months of 2007 is shown in table 2. We identified 137 cases, equalling an annual incidence of 170 hospitalisations for ischaemic stroke per 100 000. Fourteen patients (10%) were treated with tPA, 32 (23%) arrived at hospital within the 0–3 h window and 13 (41%) of these were treated.
Our study shows that thrombolytic treatment may be effectively implemented in a routine stroke service in a small inexperienced centre. Treatment can be available with short hospital delays to a large proportion of patients with acute stroke. Following organisational changes in 2006, the number of treatments increased and median DNT fell significantly to 38 min.
Some improvement in DNT was seen before 2006. This indicates that factors other than the structural changes may have contributed to the improvement, such as increasing general knowledge of thrombolysis in ischaemic stroke and the growing experience of the centre.
However, our findings are in accordance with a reported marked reduction in hospital delays after reorganisation of the ER and a DNT of 36 min in Helsinki. Although Helsinki is a large centre with the highest reported number of treatments in Europe, logistic solutions are similar to ours with prenotification and treatment in the ER.2 In comparison, mean DNT in SITS-MOST was 68 min.1
The average ONT was unchanged in our series. This may be partly explained by an increasing number of treated patients from more remote parts of the catchment area.
The median pretreatment NIHSS in our series was 10 compared with 12 in SITS-MOST. This reflects the fact that we have treated more mild strokes. Safety was good with 2.8% symptomatic haemorrhages.
The proportion arriving in the 0–3 h window in our study is similar to the findings of others.3 4 Approximately 10% of all patients with ischaemic stroke and approximately 40% of those admitted within 3 h were treated. Some large centres have reported a similar high proportion of treated patients.2 5 6
Our study now shows the availability of treatment in a community with service provided by a single previously inexperienced centre.
In Norway we do not have ER physicians. Patients are triaged by ER nurses. Doctors from one or several specialties are alerted according to flow charts. In our centre the neurologist is responsible for thrombolysis 24/7.
A strength of our study is the prospective community-based design in a well-defined catchment area served by only one hospital. We therefore believe that this study provides a true estimate of acute strokes in our region. The study period was limited to 6 months.
This study shows that a small previously inexperienced stroke centre can rapidly implement the necessary logistics to deliver thrombolysis with short hospital delays to a large proportion of patients with acute stroke.
The authors thank study nurses Siv Pettersen and Karen Johanne Olsen for data collection.
Competing interests: AT and LT have received travel grants and honoraria from Boehringer Ingelheim GmbH.