Delays in the presentation of stroke patients to by nuu18388


									Hong Kong Journal of Emergency Medicine

Delays in the presentation of stroke patients to hospital and possible
ways of improvement
KK Lau, KM Yeung, LH Chiu, B Sheng, KW Choi, YN Shih

       Objective: Stroke patients often came late to hospital and arrived beyond the therapeutic time window
       for thrombolytic therapy. We studied the time from stroke onset to arrival at Accident and Emergency
       (A&E) department and examined what barred them from early medical attendance. Methods: All acute
       stroke patients attending A&E between 15 March 1999 to 14 June 1999 were recruited. For those
       brought in by ambulance, their time intervals were divided into three: phase I was between stroke onset
       to call 999; phase II was between call 999 to A&E arrival; and phase III was between A&E arrival to
       being seen by doctor. For those who did not come by ambulance, they were divided into two groups:
       those who consulted other doctors and those who did not consult other doctors before coming to A&E.
       Their time lags from stroke onset to A&E consultation were compared. Results: One hundred and fifteen
       stroke patients were consecutively recruited. Sixty-five ambulance users had median time for phase I as
       151 minutes, for phase II as 32 minutes, for phase III as 17 minutes. The total median time lag was 190
       minutes. Fifty were ambulance non-users. For those who did not consult other doctors before A&E
       attendance, the median time lag was 641 minutes. For those who consulted others doctors before A&E
       attendance, their median time lag was 3,672 minutes. As a group their median time lag was 950 minutes.
       For the 65 ambulance users, we further studied the time intervals between A&E arrival and being seen by
       doctors; and the median waiting time for doctors was 17 (range 0 to 60) minutes. Conclusions: Public
       education was of paramount importance. Some common stroke signs could be widely propagated for
       recognition. Phase I should be less than 80 minutes. The median time for phase II would likely remain
       to be 32 minutes. Further shortening could be achieved in phase III. As category III & IV patients were
       most likely potential candidates for thrombolysis, they should be seen within 15 minutes. This would
       leave only 53 minutes for clinical assessment, CT brain and preparation of thrombolytic agent. These
       measures could increase the chance of providing thrombolytic treatment within the therapeutic time
       window. (Hong Kong 2003;10:76-80)

       Keywords: Cerebrovascular accident, emergency service, thrombolytic therapy, time factors

Correspondence to:
Lau Kwok Kwong, MRCP, FHKAM(Medicine)
Princess Margaret Hospital, Department of Medicine and           Recombinant tissue plasminogen activator (rtPA) can
Geriatrics, Lai Chi Kok, Kowloon, Hong Kong                      improve morbidity of acute ischaemic stroke patients.
                                                                 The success of rtPA depends on a narrow therapeutic
Yeung Kwan Mo, MRCP, FHKAM(Medicine)                             time window of less than three hours. When rtPA is
Sheng Bun, MRCP, FHKAM(Medicine)                                 given intravenously within the time window, patients
Choi Kin Wing, MRCP, FHKAM(Medicine)
                                                                 can have better functional outcome and the effect is
Princess Margaret Hospital, Accident and Emergency Department,   sustained. 1,2 Despite all evidence, thrombolytic
Lai Chi Kok, Kowloon, Hong Kong                                  therapy has not been a general practice in most
Chiu Lai Hong, FHKAM(Emergency Medicine, Medicine)
Shih Yau Ngai, MRCP, FHKAM(Emergency Medicine)                   hospitals. The cautious attitude of some clinicians
Lau et al./Delays in presentation of stroke patients                                                                           77

may cause some restriction, but the real limiting                   frame was divided into three phases after stroke onset.
factor is time. Patients usually attend late after stroke           Phase I was between stroke onset to call 999, phase
and most of them come beyond the therapeutic time                   II was between call 999 to arrival at A&E, and phase
window.3 To examine the problem of delay, we studied                III was between arrival at A&E to medical attendance.
the time from stroke onset to arrival at Accident and               For the ambulance non-users, the time frame from
Emergency (A&E) department. We looked for factors                   stroke onset to A&E arrival was recorded. In
that barred them from early attendance and proposed                 particular they were asked whether they had visited
some possible solutions. 4 All these measures aimed                 general practitioners before they came to A&E.
at making thrombolytic therapy feasible as the                      Exclusion criteria were stroke patients who had
treatment had been proved to reduce morbidity and                   attended other hospitals for the same episode, or
mortality.                                                          stroke occurring during hospitalization. The stroke
                                                                    onset time was specifically asked. If a patient was
                                                                    found to have stroke upon awakening, then the time
Methods                                                             patient went to bed was counted as the possible stroke
                                                                    onset time. Time records were up to the minute,
This prospective observational study was carried out                which were checked with patients and their relatives
at the Accident and Emergency Department of                         and also with the time records in the Ambulance
Princess Margaret Hospital (PMH) from 15 March                      Service.
1999 to 14 June 1999 inclusively. We consecutively
recruited all acute stroke patients attending A&E                   By the time of this study, it had already been a
PMH within this period. 5,6 The inclusion criteria                  practice that an experienced nurse would triage most
included all acute stroke patients. They were divided               (more than 90% of all) A&E attendance within ten
into two groups by whether they used 999 emergency                  minutes after registration. Patients were prioritized
ambulance or not. For the ambulance users, the time                 according to a triage system. (Table 1) Patients with

Table 1. Triage Guideline for Accident & Emergency Departments, Hospital Authority 1999 Edition and Results
Triage Category        I                     II                   III                     IV                   V
Vital signs            Unstable vital        Borderline vital     Stable vital            Stable vital         Stable vital
                       signs                 signs                signs                   signs                signs
Terminology            Critical              Emergency            Urgent                  Semi-urgent          Non-urgent
Definition             • Life threatening    • Potential life     • Suffer from major • Suffer from            • Suffer from a
                          condition             threatening          condition with          an acute but         minor and stable
                       • With unstable          condition            relatively stable       stable condition     condition
                          condition          • Require emergency     signs                   with stable vital
                       • Require immediate      treatment and     • Require treatment        signs
                          resuscitation         close, immediate,    as early as possible
                                                and continuous
Neurological deficit • Inadequate            • Respond to verbal • Acute onset of         • Bell's palsy
                          respiratory effort    command              focal neurological
                       • Respond to pain                             deficit e.g. limb
                          or below                                   weakness,
                                                                     numbness or
                                                                     slurring of speech
Number of patients 3                         3                    56                      3                    0            65
Time interval          0                     4                    19                      16                   -            17
(median) between
arrival to being seen
by A&E doctor (min)
Time range (min)       0                     2~7                  0~60                    9~20                 -            0~60
78                                                                          Hong Kong j. emerg. med.       Vol. 10(2)    Apr 2003

unstable vital signs would be given a higher priority.              median time interval between call 999 to arrival at A&E
Under this system, patients in category I, II, III and              Department (phase II) was 32 minutes. In A&E, patients
IV would expect to be seen by A&E doctors within                    were triaged into different categories. Three patients
the time frame of immediate, less than 15 minutes,                  were in category I, and A&E doctors attended them
less than 30 minutes and less than 90 minutes                       immediately. Three patients were in category II, and
respectively. We measured the time intervals between                A&E doctors attended them after a median time of
registration and medical attendance.                                four minutes (range 2 to 7 minutes). Fifty-six patients
                                                                    were in category III, and A&E doctors attended them
                                                                    after a median time of 19 minutes (range 0 to 60
Results                                                             minutes). Three patients were in category IV, and A&E
                                                                    doctors attended them after a median time of 16
One hundred and fifteen patients were recruited, 65                 minutes (range 9 to 20 minutes). No patient was in
used ambulance and 50 arrived by other means of                     category V. As a whole group, A&E doctors attended
transportation. Two stroke patients who had attended                these patients after a median time of 17 minutes (range
other hospitals for the same episode, and two other                 0 to 60 minutes).
patients who had stroke during their hospitalization
were excluded. Patients' demography and CT results                  Fifty patients did not use ambulance to attend A&E
were recorded in Table 2. For the 65 ambulance users,               and 36 of them did not consult private practitioners
their stroke onset time and time lag were recorded                  before their A&E attendance. Their median time lag
in Table 3.                                                         between stroke onset to A&E consultation was 641
                                                                    minutes. Four teen of them consulted private
For the 65 patients who used ambulance, the median                  practitioners before attending A&E. Their median
time interval between stroke onset to call 999 (phase I)            time lag was 3,672 minutes. As a whole group, their
was 151 minutes (ranged 0 to 10,751 minutes). The                   median time lag was 950 minutes. (Table 4)

Table 2. Patients' demography and CT results
Male: Female                                                        61: 54
Age (mean; median; range) (yr)                                      69; 70; 35~94
CT results:           Haemorrhage (%)                               22 (19.1%)
                      Ischaemia (%)                                 88 (76.5%)
                      Subarachnoid haemorrhage (%)                  3 (2.6%)
                      Undefined (no imaging)                        2 (1.7%)

Table 3. Time interval for 65 ambulance users from stroke onset to A&E arrival
Stroke onset time     mn-        3a.m.-    6 a.m.-        9 a.m.-   12 noon-     3 p.m.-   6 p.m.-        9 p.m.-       Total
                      3 a.m.     6 a.m.    9 a.m.         12 noon   3 p.m.       6 p.m.    9 p.m.         12 mn
Number of patients    5          4         15             11        7            9         7              7             65

Ambulance call time   mn-        3a.m.-    6 a.m.-        9 a.m.-   12 noon-     3 p.m.-   6 p.m.-        9 p.m.-       Total
                      3 a.m.     6 a.m.    9 a.m.         12 noon   3 p.m.       6 p.m.    9 p.m.         12 mn
Number of calls       3          5         8              15        10           13        8              3             65
Time lapse from       <1         <2       <3         <6        <9      <12        <15      <18       <21        <24           
                                                                                                                                 ≧ 24
stroke, from onset
to ambulance
call (hr)
Cumulative no. of     23         28       35         45        51      52         54       56        58         59           65
(%)                   35         43       54         69        78      80         83       86        89         91           100
Lau et al./Delays in presentation of stroke patients                                                                 79

Table 4. PMH A&E department arrival time for stroke patients
                             Ambulance users
All strokes                  median time lag = 190 min                              Not visit GP before PMH
(n=115)                                                                             (n=36)
                             Ambulance non-users                                    median time lag = 641 min
                             median time lag = 950 min                              Visit GP before PMH
                                                                                    median time lag = 3,672 min

Discussion                                                     stroke and they all died within the first seven days and
                                                               30 days respectively. Among the ambulance users, 44
Why did patients wait for more than two and a half             out of 65 patients (67.7%) had ischaemic stroke. They
hours before they decided to call 999 for emergency            belonged to category III and IV, and they were seen
ambulance? We believed that they could not recognize           after a median time of 19 minutes and 16 minutes
the signs and symptoms of acute strokes. They were             respectively. It is important to note that they are
not aware of the possibility that the patient had an acute     potential candidates for rtPA treatment and they belong
medical problem; hence they did not call 999. The lack         to category III and IV. We therefore suggest that
of knowledge was also seen in patients who did not use         suspected stroke patients should be upgraded to a status
emergency ambulance. The 36 ambulance non-users                of category II, and be seen by doctor within 15 minutes.9
who did not visit their general practitioners had been         Even with these arrangements, the doctor can only have
waiting for more than ten hours before they came to            about 53 minutes to take the history, perform
A&E. Presumably the hemiplegia did not improve, or             examination, arrange CT brain and prepare the
had further deteriorated while they were waiting. The          thrombolytic agent. The success of stroke treatment
14 ambulance non-users who visited their general               depends on streamline arrangement as well as providing
practitioners had waited even longer. They waited for          effective treatment within the therapeutic window.
two and a half days before they attended A&E. Our
first important message was to emphasize the                   PMH is a general hospital with 1,130 acute beds that
importance of public education to recognize the signs          serves about 800,000 people. The 999 emergency
of stroke and to encourage the use of emergency                ambulance service has provided a high standard of
ambulance service properly. While the ambulance                service with no charge. In 1999 more than 93% of
service will most likely remain as efficient as before,        all 999 emergency ambulance requests could be met
the interval between stroke onset to call 999 (phase I)        within 12 minutes.10
should be within 80 minutes if not shorter. The reason
is that triage, physical examination, CT brain and             With the success of rtPA treatment in acute ischaemic
preparation of thrombolytic agent will require another         stroke, reasons for late admission were carefully
60 minutes. We conclude that patient delay is the main         searched. Possibilities included: low awareness, lack
obstacle in providing early therapeutic treatment to           of knowledge, lack of supporting relatives,
stroke patients and changes can be done in phase I.            geographical reason, means of transportation, time
                                                               of onset, age, availability of insurance cover and
The triage system in our A&E department was the same           transportation.8,11 In our study, the lack of knowledge
as the standard practice for all A&E departments in            and prompt responses were the essential hurdles. We
Hong Kong and also in many hospitals overseas. Patients        put much emphasis on stroke onset time. It had been
with worse symptoms and poor outcomes tended to                arbitrarily recorded as the midpoint between going
come early.7,8 Nevertheless when we examined the result,       to bed and waking up. 8 Our study used the earliest
all category I and II stroke patients had haemorrhagic         possible time because the actual time of stroke onset
80                                                                   Hong Kong j. emerg. med.        Vol. 10(2)   Apr 2003

could be as early as the beginning of sleep. In future         References
when rtPA is contemplated as the therapeutic agent,
the earliest possible onset time must be taken.1 In other      1.  Tissue plasminogen activator for acute ischemic stroke.
studies, poor outcome was associated with late admission           The National Institute of Neurological Disorders and
of the day and also nocturnal onset of stroke.8 We could           Stroke rt-PA Stroke Study Group. N Engl J Med 1995;
not demonstrate these phenomena in our study.                  2. Kwiatkowski TG, Libman RB, Frankel M, et al. Effects
                                                                   of tissue plasminogen activator for acute ischemic stroke
We did not have a standard questionnaire for why                   at one year. National Institute of Neurological Disorders
patients were late to hospital. A close examination of             and Stroke Recombinant Tissue Plasminogen Activator
                                                                   Stroke Study Group. N Engl J Med 1999;340(23):
the possible reasons would require another study. But
when some patients or their relatives were interviewed         3. A system approach to immediate evaluation and
later, they often expressed their inability in recognizing         management of hyperacute stroke. Experience at eight
early signs and symptoms of stroke. We had to reiterate            centers and implications for community practice and
the importance of public education as clinical features            patient care. The National Institute of Neurological
                                                                   Disorders and Stroke (NINDS) rt-PA Stroke Study
such as sudden onset of hemiplegia, hemiparesis,                   Group. Stroke 1997;28(8):1530-40.
dysphasia, weakness of facial muscles, ataxia, double          4. Adams HP Jr, Brott TG, Furlan AJ, et al. Guidelines
vision, stupor, confusion or coma could be the first               for Thrombolytic Therapy for Acute Stroke: a
symptoms of stroke, no matter how simple and straight              Supplement to the Guidelines for the Management of
                                                                   Patients with Acute Ischemic Stroke. A statement for
forward they might seem to be. Unfortunately their
                                                                   healthcare professionals from a Special Writing Group
knowledge, attitude and subsequent responses were not              of the Stroke Council, American Heart Association.
systematically studied in this study. Probably another             Stroke 1996;27(9):1711-8.
study would be necessary to identify the significance of       5. Huang CY, Chan FL, Yu YL, Woo E, Chin D.
these factors.7,10,12                                              Cerebrovascular disease in Hong Kong Chinese. Stroke
                                                               6. Kay R, Woo J, Kreel L, Wong HY, Teoh R, Nicholls
The main obstacle in providing timely treatment to acute           MG. Stroke subtypes among Chinese living in Hong
stroke patients is low public awareness of the symptoms            Kong: the Shatin Stroke Registry. Neurology 1992;
and signs of stroke. Patients and their relatives often            42(5):985-7.
wait for several hours, if not days, to recognize that it is   7. Alberts MJ, Perry A, Dawson DV, Bertels C. Effects of
                                                                   public and professional education on reducing the
a medical emergency, and then decide to call 999. Some             delay in presentation and referral of stroke patients.
would even visit their family doctors before they were             Stroke 1992;23(3):352-6.
referred to A&E. This practice has been proved to cause        8. Harper GD, Haigh RA, Potter JF, Castleden CM.
further delay.9 Our suggestions are: educate the public            Factors delaying hospital admission after stroke in
                                                                   Leicestershire. Stroke 1992;23(6):835-8.
to recognize the symptoms and signs of acute stroke
                                                               9. Harraf F, Sharma AK, Brown MM, Lees KR, Vass RI,
and then come to A&E by ambulance. In A&E triage,                  Kalra L. A multicentre observational study of
stroke patients should be upgraded to category II even             presentation and early assessment of acute stroke. BMJ
if they appear stable, because these groups of patients            2002;324(7354):17.
can benefit from thrombolytic therapy in the narrow            10. Lo CB, Lai KK, Mak KP. Prehospital care in Hong
                                                                   Kong. Hong Kong Med J 2000;6(3):283-7.
time window.                                                   11. Esagunde RU, Navarro JC, Conde BL, Marasigan SM,
                                                                   Javier RS. Reasons for the delay in hospital admissions
                                                                   among stroke patients seen at JRRMMC [Abstract].
Acknowledgement                                                    4th Biennial Convention of the ASEAN Neurological
                                                                   Association (ASNA) 2001:91.
                                                               12. Jorgensen HS, Nakayama H, Reith J, Raaschou HO,
We would like to acknowledge the Emergency                         Olsen TS. Factors delaying hospital admission in acute
Ambulance Service of Fire Services Department for                  stroke: the Copenhagen Stroke Study. Neurology
their contribution in this study.                                  1996;47(2):383-7.

To top