WHO meeting on community control of stroke and hypertension: Geneva, 6-13 December 1973: one-year experience of multicentre stroke registration

' • WORLD HEALTH ORGANIZATION ORGANISATION MONDIALE DE LA SANtt MEETING OF INVESTIGATORS ON COMMUNITY CONTROL OF STROKE AND HYPERTENSION Geneva, 6-13 December 1973 CVD/S/73.10 ENGLISH ONLY INDE)(.EO ONE-YEAR EXPERIENCE OF MULTICENTRE STROKE REGISTRATION by J. Marquardsen Department of Neurology Frederiksberg Hospital, Copenhagen s. Hatano Cardiovascular Diseases Unit, WHO, Geneva K. Uemura Health Statistical Methodology Unit, WHO, Geneva The Issue of this document does not constitute formal publication. It should not be reviewed, abstracted or quoted without the agreement of the World Health Organization. Authors alone are responsible for views expressed in signed articles. Ce document ne constitue pas une publication. II ne dolt faire l'objet ct•aucun ~ompte rendu ou resume ni d'aucune citatioh sans l'autorisation de l'Organisation Mondlale de la Sante. Les opinions exprlm6es dana lea articles slgnes n'engagent que leura auteurs. CVD/S/73.10 page 2 Participating Centres and Principal Investigators * Centre No. Place Country Principal Investigator 01 Goteborg Sweden P. Harmsen 05 ~openhagen Denmark J. Marquardsen 06 Dublin Ireland A. Radie 21 Espoo Finland K. Aho 22 North Karelia Finland P. Puska and K. Salmi 28 Moscow USSR V.E. Smirnov 29 Zagreb Yugoslavia z. Poljakovic 33 Zerifin Israel L. Geltner 35 Ibadan Nigeria o.o. Akinkugbe and B.O. Osuntokun 45 Akita Japan s. Kojima 46 Saku Japan K. Isomura 47 Fukuoka Japan T. Omae 48 Osaka Japan Y. Komachi 49 Japan National Japan Y. Fukuda Railways 55 Ulan Bator Mongolia N. Dondog WHO CVD s. Hatano HSM K. Uemura * in the WHO cardiovascular Centre No. denotes code number for each centre disease register programmes \~ CVD/S/73.10 page 3 CONTENTS PREFACE ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• tit!TBODS ••••••••••••••••• •,. • • , ••••••••••• , ••••••••••••••••• • •• , •••••••• Collaborating centres ••••••••.••••••••••••••••••••••••••••••••••••• Subjects included in this Report ············~···················· Procedures used for improving comparability •••••••••••••••••••••• I. MAGNITUDE OF STROICE 1. 2 .: 3. 4. s. 6. Incidence ••••••••••••••••••••••••••• It t t et t t t t e t t t t t t e t t t t e t Incidence rate t t t t t t t a a It t It t t t t t t t t t t t t t t t t I I I t t I t t t t t t t t t t Seasonal variation • ••••••••••••••••••••••••••••••••••••••••• Stroke with hemiplegia and mild symptoms •••••••••••••••••••• Pathological type of stroke ••••••••••••••••••••••••••••••••• Reliability of ,troke type diagnosis •••••••••••••••••••••••• II. PROPILE OF STROKE PATIENTS 1. 2. 3. Age and sex distribution •••••••••••••••••••••••••••••••••••• Personal conditions prior to the stroke t t t t t • I I t t et t t t t I t t t t Previous medical history • ••••••••••••••••••••••••••••••••••• III, HOSPITALIZATION 1. 2. 3. 4. s. Admission to hospital • •••••••••••••••••••••••••••••••••••••• Patients who had stroke while in hospital for other reasons ••••••••••••••••••••••••••••••••••••••••••••• Hospitalization and severity of stroke •••••••••••••••••••••• Hospitalization and neurological symptoms ••••••••••••••••••• Hospitalization and living conditions ••••.••••••••••••••••••• IV. CLINICAL STATE AT THE TIME OF MAXIMUM IMPAIRMENT 1. 2. 3. Disturbance of consciousness •••••••••••••••••••••••••••••••• Neurological deficit •••••••••••••••••••••••••••••••••••••••• Blood pressure at acute stage • •••••••••••••••••••••••••••••• Page 5 5 6 8 8 9 9 9 10 10 11 12 13 13 14 14 16 16 16 17 17 18 18 18 18 19 CVD/S/73.10 page 4 v. VI. CLINICAL STATE AT FOLLOW-UP 1. 2. 3. 4. 5. 6. 7. a. 9. Neurological deficit t e t t t t t t I t I t I t I t t t I t t I t t I t t t t I t t t t t t I t t I t Medical consultation, laboratory or radiological examinations • , • , , • , , •••.••••• , , , •••..•.••. , • , ••• , •••••..... Complications at follow-up t t t I t t I I t t t t t I I t I t I I t t t t t t t t t t I I I I t Working conditions at three months ........................... Personal help for care of stroke patients t t t t t t t I t I t I t t t t t t I t Rehabilitation ••••••••••••••••••••••••••••••••••••••••••••••• Activity in daily life t I t t I t t t t t t t t I I t t I t t t I t t t t t t t t t I t t t t I t t Location of patients t t I I t t I I t t t I t t I t t t t t t t t I t t t t I t t I I t t t t t Patients whose diagnosis of stroke was revised ••••••••••••••• CASE FATALITY 1. 2. 3. 4. 5. lllmediate fatality rate t t t a t t t t I I t t t t t I t t t t t t t t t t I t t t I I t I t I I t Factors influencing the inanediate fatality rate t I t t t I t I I t I I t I Place of death t t I I t t I I t I I I t t I I t t t t I t I I I t t t t I t t t I t t t I t t I t t I t t t Autopsy ••••••• , ••• , .•• , , , • , ••• , •••• , •• , •.• , •• , •• , , , •• , , , , , , , • Mortality from stroke t t t t I t I I t t t t t t t I t I t I t I I t t t t I t t t t t t t I t I t • VII. SUMMARY AND IMPLICATIONS OF STROKE CONTROL 1. 2. 3. 4. 5. Annex 1: Magnitude of the problem•••••••••••••••••••••••••••••••••••••• • ••••••••••••••••••••••••••••••••••• Case fatality of patients Effort in prevention • •••••••••••••••••••••••••••••••••••••••• Hospitalization and home care . .............................. . Rehabilitation • •••••••••••••••••••••••••••••••••••••••••••••• WHO Stroke Register Study Record Forms Page 21 21 22 23 24 24 25 26 28 29 30 30 31 33 33 33 34 34 35 35 36 36 PREFACE CVD/S/73.10 page 5 Stroke occurs everywhere. Psychological and economical burdens on the family and on society are increasing: rational planning for control of stroke is urgently required. Basic information on the magnitude and nature of the problem is. as yet, largely insufficient. A study group was organized from interested participants of the WHO Semif&r on Stroke Prevention. Treatment and Rehabilitation held in Monaco in May 1970 • and methodology for uniform data collection has been developed through pre-pilot trials. The short-term goal of the stroke register is to demonstrate its feasibility and use on a limited scale, The stroke register serves for estimation of morbidity and mortality from stroke, collection of knowledge on social and clinical profile of stroke victims. assessment of preventive, diagnostic and rehabilitative measures used at community level, and the progress of surviving patients in their work and daily activities. International comparison should promote insight into the etiology and evaluation of the effect of preventive and curative services. The eventual goal is to utilize this information to promote control of stroke at the community level. The present paper reports the results of data collected during the period of 12 months in 1972-1973, from an international cooperative study of stroke registra- tion in the· community. METHODS Details of planning, procedures and necessary facilities are reported in the WHO Internal Document CVD/S/73.6, and only the essential points are described here. · Stroke is defined as rapidly developed clinical signs of focal (or globa12) disturbances of cerebral function, lasting more than 24 hours or leading to death, with no apparent cause other than vascular origin; the term refers to patients presenting clinical signs and symptoms suggestive of subarachnoid haemorrhage, intracerebral haemorrhage. or cerebral ischaemic necrosis. The stroke register is therefore based on clinically diagnosed stroke. 1 Wld Hlth Org. techn. Rep. Ser., 1971, No. 469 2 'Global' - applies to patients with subarachnoid haemorrhage and to some patients in deep coma, but does not include systemic circulatory failure. e.g. shock, Stokes-Adams syndrome, or hypertensive encephalopathy. CVD/S/73.10 page 6 All possible sources of information were requested to notify new stroke patients to the register centres. New patients of both sexes and all ages from a population in a defined study area have been registered since the initiation of the study (varies from May 1971 to April 1973) 1 except in a few centres who set an upper age limit of 60 or 65. In each case the patient wasmterviewed as soon as possible after being registered at the centre. Information obtained was recorded on the "initial record form". After three weeks, when clinical conditions were generally stabilized and various diagnostic examinations usually completed, the patient was interviewed again for the first review. The information collected was entered into the "follow-up record form". Thereafter, follow-ups were made at three months and one year after the onset of stroke. Each time, the data were entered into the "follow-up record form". When recurrent stroke occurred, an interview took place immediately and the information was recorded on the "recurrence record form", which is a simpler version of the follow-up record form. Three weeks after onset of recurrence, another follow-up was carried out in the same manner as for an initial stroke event. When a registered patient died from any cause, the "death record form" was completed. If the initial diagnosis of stroke was found to be incorrect, the patient was not followed-up further, but the three week follow-up record was completed with the correct diagnosis. If the diagnosis of stroke was changed after the three-week follow-up, then the "exclusion record form" was com- pleted and the patient was not followed up after this. The record forms mentioned above are attached to this report in Annex 1. The same record forms with standardized instructions have been used in all collaborating centres, and the completed forms sent to WH0 1 Geneva, for processing and analysis. Collaborating Centres Fifteen centres are collaborating in the WHO stroke register study; seven in Europe, one in Africa, and seven in Asia. A description of the study areas is summarized in Table 1. The study areas differ in their extent, population size, geographical location, climate and major industries. The area chosen for study is either a part of a city (Copenhagen, Dublin, Zagreb), a whole city (Goteborg) 1 selected towns and villages in a district (Akita, Fukuoka, Osaka), or a whole district (Espoo, North Karelia, Zerifin, Saku), differing in extent from 8.7 Km.2 to 21,000 Km.2. There is one occupational group - the Japanese National Railways (JNR). The size of the total population in the study area differs from 36,112 to 450,860, comprising a total of 2.04 million (Table 1). The number of men and women is almost equal in all centres, except in the older age group which comprises more women and in the occupational group which consists of male employees. No age limit was imposed on patients registered, except in three centres (JNR: age up to 60; Goteborg and Fukuoka: age up to 65). The percentage of subjects aged over 65 varies from 4.7% (Osaka) and 5.1% (Espoo) to 13.2% (Zagreb) and 21.0% (Copenhagen). A tremendous difference exists in population density, ranging from 8.6 (North Karelia) to 11 1 700 per 1Qi2 (Copenhagen). Table 1: LIST OF THE STUDY AREAS CVD/S/73.10 page 7 Centre I Country: i code , I Study Area Characteristics of the study area ·size of popu-j % Popula- lation x lOOOj tion aged (survey year) I over 65 Sweden 1 01 i .. : Goteborg Urban, industry, commerce, and services.Age up to 65. 450,860 (1970) I 9.6 Denmark: 05 Ireland 06 21 Finland I l 22 i USSR i 28 I I l Yugo- i 29 slavia' I Israel 33 Nigeria Japan ; I I Mongolia 35 45 46 47 48 49 55 . 451 Km2 Frederiksberg, Copenhagen 8.7 Km2 Mainly residential area, plus commerce and light industry. I I 101,970 (1971) Southern Dubli~ Mainly residential area, 19.3 Km2 plus some industry ! 133,681 ! (1971) Espoo and Kauniainen 315 Km2 North Karelia 21000 Km2 Developing neighbour towns I • • I to Helsinki. Commerce,serv-1 ices,industry. Population ' increase ea. 5000 annuallv. Sparsely populated, agri- culture. Large emigration ,. ...... Tulino rayon, Urban Moscow.30.2Km2 Zagreb, part. Urban 14 Km2 Ramle and Urban and rural Rehovot 1.610 km2 Ibadan Epe Akita 895.8 Km2 Saku 1.044 Km2 Fukuoka 79.3 Km2 Urban Rural A city and a village rural town Rural. Hilly district. Agriculture. light industry 2 towns, plus part of urban city. 103,508 (1970) 180,000 (1970) 209,000 (1970) 85,548 (1968) 218,000 (1970) ? 36,112 (1970) 105,185 (1970) 36,786 (1970) Osaka Developing neighbour town to 40,391 Osaka.Commuter area plus agri- (1970) culture. Population doubled in 8 years. Japan National Male employees in Tokyo Railwav reaion. Ulan Bator Urban 74,733 (1971) 261,254 (1969) 'I : i j I i I 21.0 8.4 5.1 8.6 ? 13.2 6.8 5.9 10.5 7.2 4.7 0 4.4 Starting date of data in this report 1971 May 1971 July 1972 May 1972 January 1972 May 1973 January 1972 January 1972 January 1973 April 1971 November 1971 Sei,tember 1971 January 1971 November 1972 Januarv 1972 January CVD/S/73.10 page 8 This is a preliminary report which includes data from the centres which started between January 1971 and May 1972 and which have completed a full year's study. Three centres (Moscow, lbadan and Ulan Bator) have not completed a full year's reporting and are not included in this preliminary report. In Fukuoka, Osaka and the JNR the number of patients is too small (less than SO) and detailed analysis for each centre cannot be made; but they are included in the larger group of combined Japanese centres. Subjects included in this Report The number of stroke patients registered during the 12 month-period and included in this analysis is shown in three age groups in Table SB-2(i) • The number of registered patients for each centre during the year differed from 18 to 383 1 and the total of 2067 patients comprised 1047 men and 1020 women. The information given includes one complete year's registration and three-week follow- up, but later follow-up still continues. Thus the data presented concerning the course of patients beyond three weeks are preliminary. Procedures Used for Improving Comparability Measures have been taken in order to ensure a high degree of comparability. In the pre-pilot study, provisional record forms were tested in the field to ascertain whether the terms used were correctly understood by interviewers and patients, and whether valid information was obtainable. Case reports of stroke patients were circulated by WHO for testing purposes, and record forms were filled in by an interviewer in each centre: discrepancies between centres and inter- viewers, and any misunderstanding in completing the record forms, were checked. Based on the experience gained during one year of this exercise, all items were critically reviewed, and the record forms were finalized.2 The decision on whether or not to register a patient is left to each centre, and bias might be introduced by different ways of diagnosing stroke. Some case reports for which a centre had difficulty in making a decision regarding inclusion or exclusion have been distributed to all collaborating centres for testing, and their decisions according to their daily practice were reported to WHO. Many centres reported that more information than was given in these test cases is usually obtainable at their centre in order to enable a definite conclusion. Differences in interpretation of criteria and procedures were noticed as a result of this test, and have been rectified. 1 In CVD Internal Document: CVD/S/73.7 2 Report of WHO Meeting of Investigators on the Connnunity Control of Stroke and Hypertension, Goteborg, November/December 1971 (Document: CVD/72.1) CVD/S/73.10 page 9 Record forms received in WHO were routinely inspected before key-punching. Obvious errors, contradictory statements, and missing information were queried to the centres, who provided explanations or sent back the corrected information. Newsletters have been issued from time to time to facilitate communication between the coordinatina centre and collaborating centres about the progress and problema encountered. A meeting of investigators has been held at least nnce a year to resolve queries and to review the data obtained. Visits to the centres by collaborating investigators and WHO personnel have been arranged as much as possible. By these means, understanding of term.a and procedures has been continuously improved for the collection of comparable data. I. MAGNITUDE OF STROKE 1. Incidence table SB-2 (i) and (ii)1 gives, for each participating centre, the number and relative frequency of the registered patients according to age and sex. Stroke patients were younger in centres 01, 47 and 49,who imposed an upper age limit, and in centre 45; patients were older in centres 05, 33 and 46. Up to the age of 65 there were more male t~1!_-~male patients; the ratio was reversed afterwards, due to the larger number ofl,~\ation and also to the reversed incidence rate. 2. Incidence Rate The annual incidence rates,calculated by applying the above figures to the official census fiaurea for the respective populations·(table SA-2), are given in table SA-3. For male and female total populations combined the incidence rate varied by centre from 1.4°/oo to 3.2°/oo •. The incidence rate rose steeply with age. The rate was similar for each age group in various European coDlll.unities (centres 01- 29). and in Israel (centre 33), but was higher in Japanese centres (45-49). Among European centres, the Finnish ones (21, 22) showed the highest and Zagreb (29)the lowest value. Among the Japanese centres, Akita (45) revealed a particularly high incidence, with early increase in both males and females. At ages under 65 years, males generally showed a higher incidence rate than females; beyond that age, the male excess disappeared, except in two Japanese rural centres (45, 46). A tentative explanation for this might be that according to the traditional culture pattern in Japan a male patient, being the head of the family, is more likely to be referred for medical treatment than a female patient. This possibility, however, is small, as will be discussed in Section I.4. 1 this, and subsequent tables referred to in the text, are shown in the background document CVD/S/73.7. CVD/S/73.10 page 10 3. Seasonal Variation The number of onsets of stroke fluctuated by month in each centre (Table SB-13), Slightly more frequent· occurrence of stroke in the first half of the year and less in midsummer to autumn was seen. Much effort will have been made and perhaps unconscious inflation of reporting may have occurred during the initial few months of the register - i.e. around January or May. However, the low rate in December and high rate in January must be a product of delayed reporting because of holidays. A concentration of events in March in the occupational group (centre 49) must be an artificial phenomenon resulting from administrative reasons. Patients were classified by the month of onset of disease, but not by the month of registra- tion, which may well be influenced by holiday periods. Since there is no other reason for the gap in incidence between December and January, this may be taken as a warning sign for the accuracy of information gathered. 4. Stroke with Hemiplegia and Mild Symptoms Because of the lack of objectivity in the procedure for diagnosis of stroke, validity and comparability of the diagnoses in each centre need to be checked further. Hemiplegia is the most co111DOn symptom in stroke, easily recognized by the patient himself, the patient's family, and the interviewer, and for which observer bias would be minimal, whereas patients with milder neurological deficits may .not always seek medical care, and the clinical diagnosis of mild symptoms may differ by physician. This latter group therefore may possibly be underestimated and subject to more diagnostic bias. * Since a fairly uniform fraction of stroke patients (60-80%) were hemiplegic, the incidence of hemiplegic stroke did not introduce any controversy in the overall picture of stroke incidence (Table SA-4). For the total population, the incidence rate of hemiplegia varied from 0.8 to 2,5 per 1,000. A striking increase with age, male excess in younger age groups, lower figures for European and Israel centres (01-33) versus higher figures in·Japan, were all observed. The age specific incidence was again lowest in Zagreb (centre 29), and highest in Akita (centre 45). A higher incidence in males, even in the older age groups, was observed again in the two Japanese centres 45 and 46. Unless the unlikely possibility that female hemi- plegics were more often missed ai registration is accepted, male excess in these centres should be taken as real. The paragraph marked*••••••* is based on the data obtained by the end of July 1973. CVD/S/73.10 page 11 Hemiplegia, particularly hemiparesis, is however very difficult to assess when a patient is deeply comatose. The relatives may say that a hemiparesis was in fact observed initially, before the patient lost conscious- ness. The doctor may then either record "hemiplegia present" or "not assessable", according to whether he will make use also of the information from the history, or only record what he can actually demonstrate at the neurological examination. Variations in this type of coding habit may create artificial differences in the incidence of hemiplegia. The incidence rate of cases with hemiplegia and/or coma will be calculated as a core group of stroke patients. For milder strokes the numbers were too small to obtain reliable figures. However, the Japanese centres did not reveal a higher rate. From these observations, if any bias in the diagnosis of stroke was present, it is more likely that it occurred in the registration of milder stroke patients and that the Japanese centres were underestimating the incidence of milder stroke, and therefore underestimating the incidence of overall stroke rather than over- estimating it. 5. Pathological Type of Stroke Table SC-2 shows the relative frequency of ICD-rubrics 430 - 436 (with the exclusion of 435: transient cerebral ischaemia), according to the diagnoses gi~en by each centre at three weeks after the stroke. These diagnoses, being to some extent based on the results of X-ray and laboratory examinations, were considered more reliable than the preliminary diagnoses made at onset. For the patients who died before three weeks, autopsy diagnosis was taken if the brain was examined, otherwise the clinically aetermined type of stroke was taken. The strong influence of age an the distribution according to type of stroke is apparent. For example, haemorrhagic stroke, either subarachnoid or intracerebral haemorrhage (43o-431), was diagnosed in about one-half of the youngest patients (under 55 years), but in less than one-fifth of those over 65 years of age. The opposite trend appeared for presumed ischaemic cerebral necrosis (432-434), in particular for cerebral thrombosis (433). The degree to which the incidence of stroke rises with age was strongly dependent on the type of stroke, being very marked for cerebral thrombosis, less pronounced for intracerebral haemorrhage, and non-existent, or even decreased, for subarachnoid haemorrhage (SAH). The higher stroke incidence for males than for females could be found also in the individual types. In all types, however, the reverse occurred in the age group over 65 years of age, except 432, occlusion of pre-cerebral arteries, in which male excess persisted. SAH was more frequent in females, except in the younger age group of under 55 years. CVD/S/73.10 page 12 A comparison of the distribution according to diagnostic subgroups reported by individual centres showed marked variations (Table SC-2 Annex 1) 1 some of which are readily explained by the influence of age. The frequency of SAH was high in areas with relatively young study populations (centres 01 1 49) and low in areas with comparatively old populations (centres 05 1 33). It is surprising, however, that one centre (29) did not find a single instance of SAH among 124 cases of stroke; the reason for this aberration is not clear. Intracerebral haemorrhage (431) was diagnosed in a much higher proportion in the Japanese than in the European centres. This is noteworthy since strikingly high mortality rates for intracerebral haemorrhage were previously reported from Japan. It has been suggested, however, that such high rates may not be real, but merely reflections of particular diagnostic habits (Kurtzke). * The frequency of cases originally thought to be strokes, but later shown to be misdiagnosed, may be a measure of the available diagnostic facilities. The percentages were higher in the centres with lower frequency of intracerebral haemorrhage and vice versa. * Further, a comparison between the Japanese centres showed an inverse relationship between the percentage of cases diagnosed as intracerebral haemorrhage and the proportion of cases in which lumbar puncture was performed. These trends seem to throw further doubt on the validity of the high incidence of intracerebral haemorrhage in Japan. In most centres, about two-thirds of all strokes were diagnosed as some type of cerebral ischaemic necrosis, mainly as "cerebral thrombosis" (ICD 433). It is well known, however, that the differentiation between cerebral thrombosis and embolism is often extremely difficult (even at autopsy), and that the diagnosis of "pre-cerebral arterial occlusion" usually requires angiographic examination. A more detailed analysis of the percentage of patients allocated to groups 432-434 is therefore without interest. Since angiography was performed in less than 20% of all the registered cases, the present series cannot give reliable information about possible geographical differences in the incidence of carotid arterial occlusion. 6. Reliability of Stroke Type Diagnosis In the total of 964 deaths reported, autopsy diagnosis was available in 249 (25.8%); out of these no clinical diagnosis was given in 26 cases and clinical diagnosis other than stroke in 18 cases (Table SD-7 Annex 1). The remaining 205 cases were clinically diagnosed as acute stroke. Out of these, 5 cases had no cerebrovascular lesions and 3 had chronic or generalized lesions only. Thus the clinical diagnosis of acute stroke was confirmed by autopsy in 197 cases (96.1%) in the present series. This series is not adequate for estimating "false negative" diagnosis; according to the criteria for inclusion, patients without clinical indications The section marked*•••••* is based on the data obtained by the end of July 1973. CVD/S/73.10 page 13 of stroke were not registered, even when acute cerebrovascular lesion was found in autopsy. Nevertheless, 12 cases previously suspected as having had a stroke and registered, but.diagnosed as non-stroke at three weeks were again proved to have acute cerebrovascular lesions by autopsy. These comprised 5.8% of 209 strokes verified by autopsy. False negative rate must have been much higher if all fresh cerebrovascular lesions found in autopsy were included. Pathological type of stroke was specified in 165 cases (Table SD-7 Annex 2). The diagnosis given clinically was confirmed by autopsy in 138 cases (83.5%). When the types are grouped as haemorrhagic (ICD 430 or 431) and ischaemic (ICD 432, 433, or 434) strokes, the confirmation rate increases. In 97 cases of clinical diagnosis of haemorrhagic stroke, 89 (92%) were correct, and in 73 cases of ischaemic stroke 60 (82%) were correct. The non-specified diagnosis of ICD 436 may be counted as failure in type diagnosis. The failure in diagnosing grouped categories of haemorrhagic or ischaemic strokes was 17 and 31 respectively and accounted for 23.2% of 207 cases with type verified by autopsy. These figures were no less accurate than the so far reported series (Kagan et al.) and are supporting observations described in the preceding section I.5. Of course these figures do not secure the validity of clinical diagnosis for deceased cases without autopsy,or surviving patients. Thus, reliability of clinical diagnosis mentioned here is not applicable to the Japanese centres, where the autopsy rate was low and reliability of diagnosis remains unchecked. Much error must be expected among surviving stroke patients who had milder lesions and symptoms. II. PllOFILE or STROKE PATIENTS 1. Age and Sex Distribution The number of stroke patients in this study is shown in Tabie SB-2(i). The number varied by population size and incidence rate. The general tendency was that there were more male victims in the younger age group and more female in the older age group. The male excess continued up to the older age group in centres 29, 33, 45 and 46. The relative frequency of age groups was illustrated in Table SB-2(ii). The age distribution in centres with an upper age limit (01, 47, 49) was not comparable to others. Young patients were more frequent in Espoo (centre 21), North Karelia (centre 22) and Akita (centre 45). This is due not only to the age structure of the background population, but also to an accelerated occurrence of stroke in these centres (Table SA-3). Older patients were more c0111DOn in Copenhagen (centre 05), Zerifin (centre 33) and Saku (centre 46) in both sexes, and in females in Dublin (centre 06): these reflect the age structure of the background population. CVD/S/73.10 page 14 2. Personal Conditions Prior to the Stroke An analysis of the civil status of the stroke patients failed to show anything remarkable (Table SB-4(i)). In accordance with the advanced age of most stroke patients, the percentage of widowed persons was relatively high, for the total series. Table SB-4 (ii),illustrating the living conditions of the patients, shows marked differences in domestic structure. Thus, in spite of an almost uniform percentage of widowed persons, the proportion of patients who were living alone or in institutions was very high in centres 01, OS, 06 and 21, much lower in a rural Finnish conmunity (centre 22), Zagreb (centre 29) and Israel (centre 33), and negligible in Japan (where the family ties are known to be strong). The working conditions at the time of the stroke appear in Table SB-4 (iii). About thirty percent of the patients were still in gainful work, thou"gJi the per- centage varied considerably between centres, depending upon the proportion of elderly people: less than one-fifth were working in the centres in Copenhagen (05), Zagreb (29) and Zerifin (33). More than half of the patients were working in the Japanese centres. The distribution according to occupation, shown in Table SB-4 (iv), mainly reflects the differences between urban and rural study areas. For instance, "farmers, fishers, hun~ers" (occupation code 6) varied from 0% in Goteborg and Copenhagen to 66% in Akita. Housewives and others not doing gainful work (occupation code X) comprised nearly half in many places. This group comprised almost 90% in Israel (explanation is necessary). This was low in a Finnish developing community (centre 21) and in an agricultural coumunity in Japan (centre 45), reflecting the involvement in agricultural work of housewives, although it is possible that differences in coding practice may exist. 3. Previous Medical History Table SB-5 illustrates the past histories of the stroke patients, mainly with regard to cardiovascular diseases. The percentages may represent minimum values, since many patients - particularly those with impairment of consciousness or intellect - were unable to give a relevant history. Even in fully cooperative patients, the retrospective interpretation of past symptoms was often difficult, and may not be identical in various centres. A striking variability in the frequency of previous strokes was observed (Table SB-5 (i)): in particular, extremely high figures were given by one centre (29). Sex difference in the prevalence of past history of stroke, and increase in prevalence with age, were slight, perhaps a higher incidence of stroke in older subjects being cancelled out by high mortality. The prevalence of previous acute myocardial infarction (AMI), other heart disease, and of diabetes mellitus varied considerably, but in a roughly parallel CVD/S/73.10 page 15 way (Table SB-5 (ii)): a comparatively high frequency was reported by the European centres and Israel. whereas in Japanese centres the frequency was low. This is to some extent a reflection of prevalence of these diseases among background populations. It is remarkable that about a quarter to half of stroke patients had some kind of heart disease in the.past. Prevalence of diabetes was about 13% on average. high in Zagreb and Zerifin and low in Japan. In contrast to AMI. the prevalence of known hYpertension was lowest in Copenhagen and higher in Zagreb and Japan (Table SB-S(iii)). Treatment of hYper- tension before stroke was much lower in Japan than in Europe or Israel; altogether a little more than half, only had been treated prior to stroke. In aother centre which reported a high frequency of hypertension, i.e. Zagreb. the majority of cases were reported to be under treatment. The relative frequency of various types of stroke as diagnosed clinically (Table SC-2 Annex 1) was parallel to the frequency of known hypertension and treat- ment for it. lschaemic stroke was diagnosed in more than two-thirds in Copenhagen, whereas in Japan half of the strokes were diagnosed as haemorrhagic. The type of stroke was not specified in more than half the cases in Zagreb. The influence of sex was apparent mainly in relation to previous AMI, the frequency of which was almost twice as high in males as in females. In "other heart disease" the opposite trend was found. although much less marked. In almost all these diseases. the frequency in males exceeded that in females in the younger age groups, but frequency increased rapidly with age in females and generally reached a level equal to, or higher than, in males in the older age group over 65 years. One out of six stroke patients were dependent in self care since before the onset of stroke. Frequency in the older age group was 3 times as high as in the group under 65 (Table SB-6(ii)). *They represent patients with multiple pathology. Nearly half had stroke. and half had heart disease other than MI in the past. Forty per cent had more than one of ·the.following: stroke, AMI, other heart disease or diabetes.* The frequency of totally dependent patients was a matter of a few percent only• and partial dependence accounted for the majority. The frequency was much higher in North Karelia, Zagreb and Zerifin centres, which reported a high prevalence of stroke and/or diabetes in the past history. Disturbance in control of bladder since before stroke was seen in a few percent of young patients and in about 10% of old patients. A higher frequency was reported from Zagreb and Zerifin, followed by Dublin and Copenhagen centres, where a large proportion of the stroke patients were old. The sentence marked*•••••••* is based on data obtained by the end of July 1973. CVD/S/73.10 page 16 III. HOSPITALIZATION 1. Admission to Hospital Table SB-7 shows the admission rate of stroke patients in various centres. In the European centres and Israel, around three-quarters of stroke patients were admitted to hospital, whereas in Japan more than half the patients were treated at home. The difference was also influenced by age and sex (Table SB-4(ii)): in Japan the admission rate was much lower for elderly stroke victims than for younger ones, and much lower for female than for male victims - whereas in Western Europe and Israel the difference in rate was negligible between males and females. In the centres with high admission rates and with a high frequency of old patients, a considerable proportion of the strokes (about 10 - 20%) occurred in patients who were already staying in hospital for some other reason (included under "other" in Table SB-7(ii)). This point is discussed in the next section, III.2. When patients were actually referred to hospital, admission almost invariably took place with little delay (Table SB-7(iii)). In nearly all centres 80-90% of the hospitalized patients were admitted on the day of onset, only a negligible proportion of the patients arrived after the end of the first week. It should be emphasized that the figures given in the above paragraphs are not necessarily representative of the hospitalization trends for stroke patients in Europe as a whole. For example, in most parts of the United Kingdom the · admission rate for stroke victims is said to be only about 50% (Hewer); the same is probably true of many rural areas in other parts of Europe, including the Scandinavian countries. 2. Patients who had Stroke while in Hospital for Other Reasons A relatively large proportion of strokes which occurred in hospital in Western European and Israel centres may be explained in two ways: (a) Owing to a high morbidity, the patients had spent a considerable part of their time in hospital, thereby increasing the statistical chance of a stroke occurring during a stay in hospital. (b) Some of the strokes in this group may have been complications of the disease that had caused admission (e.g. AMI), or of the medical treatment given in hospital (e.g. excessive dehydration, surgical procedures). On the basis of the available data, only the former possibility can be assessed. Table SB-7(vii) clearly demonstrates the large numbers of previous strokes, AMI, other heart disease, and diabetes mellitus,in patients whose recent strokes occurred in hospital, as compared with the patients from the whole series. The finding of almost a quarter being diabetics in the group is impressive; in over one-third of these,case histories of at least two previous cardiovascular diseases per patient were obtained. Only one-fifth had none of these diseases. It is thus CVD/S/73.10 pa1e 17 obvious that the patients whose strokes occurred in hospital were characterized by a very hi&h cardiovascular morbidity as a whole 1 the only exception beins that arterial hypertension. whether treated or not 1 was reported with the same. or even lower. frequencies as in the total series. This is understandable. since hyper- tension per se does not cause hospitalization unless heart disease or stroke occurred. 3. Hospitalization and Severity of Stroke Some patients were not hospitalized because their condition was critical and death seemed inevitable. and others because their symptoms were mild and ambulatory care seemed to suffice. The state of consciousness is the best indicator of the severity of stroke, and admission rate in relation to the ~esree of consciousness and disturbance has been examined (Table SB-7(iv)). The 1eneral picture may be lar1ely influenced by the centr•where a large number of patients were or were not hospitalized. Therefore. individual centres need to be observed separately. In Western European centres and Israel the admission rate was very hi1h (about 95%) amon1 youn1er persons, irrespective of their state of consciousness, and a little lower in the older 1roup. Comatose older patients were least hospitalized. This may be due to the fact that they represent the most fragile part of stroke patients and may have died before they could be admitted to hospital. In the Za1reb centre1 on the contrary, hospitalization rate was hi1her in comatose patients, and in the older a1e 1roup. The most severe patients may have priority in admission in this centre, as the hospitalization rate was much lower am.ons the younger patients. In the Japanese centres where a large number of patients were taken care of at home, home care was the general rule irrespective of the condition of consciousness. These differences in admission rates may also be explained by the availability of resources, i.e. hospital and emergency services (such as mobile care), by the policy of local hospitals, and also by traditional attitudes of the family and physicians. 4. Hospitalization and Neurological Symptoms As expected, the non-hospitalized patients had lower percentages of hemi- plegics (Table SB-7(v)). However, the most notable findins is the very high percentage (20%) of "not assessed" cases among the non-hospitalized patients. CVD/S/73.10 page 18 This group must refer mainly to comatose patients who died rapidly without being thoroughly examined. We shall never know how many of such cases were really stroke, nor the extent of their neurological deficit. The above argument is not applicable to the Japanese centres, where patients are very often treated at home. Hospitalized and non-hospitalized patients had an entirely similar make-up of neurological symptoms. No special admission policy seems to be operating here. In one Eastern European centre with a relatively low admission rate, cases with milder deficits were hospitalized less, perhaps because of low priority in admission. S. Hospitalization and Living Conditions Patients who lived alone at the time of the attack were hospitalized at the same rate as those who lived with their family (Table SB-7(vi)). In Zagreb and Japan, a large proportion of patients who lived alone were also taken care of at home, though the actual number was small. Another source of non-hospitalized patients were those in nursing homes. More than half of the stroke patients in these permanent institutions were not transferred to hospital, perhaps because the necessary nursing facilities were available and/or because the patients were already disabled and the recovery potential seemed low. IV. CLINICAL STATE AT THE TIME OF MAXIMUM IMPAIRMENT In order to make data comparable, maximum impairment within the first 24 hours after the onset of stroke, without taking account of latter deteriora- tion, was recorded. 1. Disturbance of Consciousness Disturbance of consciousness, being the most reliable indicator of the severity of the cerebral lesion,. was recorded in about one-half of the cases, and almost one-quarter of patients were completely comatose (i.e. unresponsive to painful or other stimuli) (Table SB-B(i)). The frequency of coma or semi-coma was lower in centres 05 and 29, where the relative frequency of haemorrhagic stroke was less than 20%. Coma and semi-coma were more frequent in centres 06, 21, 33, and all Japanese centres. This may mean that severe stroke, e.5, intracerebral haemorrhage, was more frequent and/or that there was less reporting of milder stroke. 2. Neurological Deficit The frequency of stroke with mild symptoms and of stroke with hemiplegia are shown separately in Table SB-8(ii). Mild symptoms comprise neither speech dis- turbance nor paralysis (to be referred to as "no" neurological deficit), speech CVD/S/73.10 page 19 disturbance only. or monoplegia. The proportion varied widely. Patients with milder symptoms without hemiplegia comprised more than 20% in the Scandinavian centres. whereas they formed leas than 10% in Zerifin (33). Saku (46) and Osaka (48). The difference in the frequency of milder strokes did not correspond to the age of the population. In the Scandinavian centres. the principal investigator is a neurologist. who may more often diagnose milder symptoms as stroke. It may also be explained that milder patients more often visit. or are referred to. a neuro- logical centre and this facilitated registration of the milder cases. Thus the differences may be partly due to completeness of case finding and partly to local diagnostic practice. Among the focal neurological signs. hemiplegia was the most prominent. being present in 60-80% of patients. inversely related to the frequency of registra- tion of milder strokes, It can be estimated that 20-40% of patients may be missed if hemiplegiaaone is adopted as a criterion for acute stroke. Hemiplegia is easily recognized. However. bias in diagnosing hemiplegia may still occur when it is very severe or very mild. In up to 31% of patients. differing by centre, neurological symptoms were not assessable; this may represent patients who died soon after the attack or who were in deep coma and whose neurological symptoms were difficult to assess during the first day of the attack. 3. Blood Pressure at Acute Stage Table SB-9(i) shows the frequency of hypertension at initial review. If the patient died before interview the first blood pressure measured after the stroke was taken from available medical records. Hypertension is defined as either systolic blood pressure equal to or over 160 11111. Hg. and/or diastolic blood pressure equal to or over 95 11111. Hg. Excessive hypertension means either systolic pressure of 210 mm Hg or more. or diastolic pressure of 115 11111. Hg or more. The average blood pressure at the acute' stage, particularly that of comatose patients. must have shifted to higher values due to reactive hypertension because of elevated intracranial pressure, but the frequency of moderate or excessive hyper- tension did not differ much between comatose plus semi-comatose patients and non- comatose patients at the initial examination. The frequency of hyper- tension varied widely between centres, without showing any systematic tendency. Damage to the vasomotor centre of the brain by stroke, resulting in a resetting of the blood pressure level. is theoretically conceivable but rarely experienced. Blood pressure measured when the acute phase is over, and by the same interviewer in the stroke register centre. would give equivalent blood pressure values prior to the stroke, in a fairly uniform way. unless the patient is not receiving anti-hypertensive therapy. Therefore. blood pressure measured at the 3-week follow-up is described here as an indicator of the baseline blood pressure. The shortcoming of the use of 3-week blood pressure measurement is the lack of information on patients who had died by that time and the equivalent blood pressure values in patients who are under antihypertensive treatment. CVD/S/73.10 page 20 Patients who are under treatment should have had hypertension when the treatment commenced. The sum of patients under treatment and hypertensive. patients without treatment would thus indicate the real frequency of hypertensive patients, which accounted for about 70% of stroke survivors at three weeks after stroke (Table SC-5(iv)). This figure was much greater than the patients who reported presence of hypertension in the past (51%) (Table SB-5(iii)). Although normotensive patients had a little higher immediate fatality than hypertensive patients in the older group (Table SD-l(v)) 1 here again we confirm the common observation that hypertension remained unrecognized in a large proportion, in stroke victims also. apuftared to be hv~ertensive but Average blood pressureltaiea to reveaL spec1a1 trends between three age groups of males and females (Table SB-9(ii)). Slightly higher average systolic blood pressure in the Japanese centres was the only finding in the centre com- parison. Blood pressure measured at 3 weeks after onset was compared with the initial blood pressure (Table SC-5(iv)). About 70% of surviving patients at 3-week follow-up were hypertensive and about 70% of these were under treatment. Two- thirds still remained hypertensive in spite of treatment. The frequency of all hypertension remained the same. This is the con- sequency of the following compensatory phenomenon. About 30% of hypertensive patients died; blood pressure was reduced to normal range in another 30%; about 15-30% of survivors with normal blood pressure at the initial examination revealed elevated blood pressure at the time of the first follow-up. The number of excessive hypertensive patients decreased from 274 at the initial interview, to 23 at the first follow-up. This was largely due to the death of excessively hypertensive patients; to the initiation of anti-hypertensive treatment; and partly due to the stabilization of blood pressure when the acute phase was over. Patients diagnosed as having intracerebral haemorrhage (ICD 431) had the highest frequency of hypertension in both age groups,and a particularly high frequency of excessive hypertension in the younger age group. This was common to all three centre groups (Table SB-9(iii)). Blood pressure measured at the initial examination is supplementary to the observations at 3 weeks, since very high blood pressure levels will have most likely been controlled by treatment at that time. Actually, a much greater part of survivors from intracerebral haemorrhage were under treatment and their blood pressure seemed more strictly controlled at 3 weeks than patients having other types of stroke (Table SC-5(ii)). The average blood pressure level of patients who had intracerebral haemorrhage without treat- ment still showed higher values than that of patients with other types of stroke. The treatment group showed varying blood pressure levels, obviously depending on the condition of the treatment and not upon other factors. The average blood pressure of this group was higher than the group without treatment; this suggests inadequate treatment, although cautious and gradual decrease of blood pressure is the rule in the treatment of hypertension of stroke victims. V. CLINICAL STATE AT FOLLOW-UP CVD/S/73.10 page 21 Data collection on follow-up beyond 3 weeks is still incomplete. Results and conclusions in this chapter must therefore be taken as preliminary. 1. Neurological Deficit Neurological deficits at the initial phase (time of maximal impairment within 24 hours from the onset) at 3 weeks and 3 months have been compared (Tables SB-S(ii) and SC-lO(i-ii)). Both the number and the percentage of patients without neurological deficit (neither paralysis of limbs nor speech disturbance) increased remarkably from 123 (6%) to 250 (19%) at the three-week follow-up, and still increased at 3 months to 321 (28%), whereas at the same time the number and percentage of hemiplegica gradually decreased: 67% to 62%, to 53%. It is evident that some hemiplegica gradually improved, thus being transferred to the category of "monoplegia" or even to the "no deficit" group; a further reduction of the number of hemiplegica was brought about by a comparatively high fatality rate in this group. The importance of the latter factor is demonstrated by the fact that centres with spectacular increase of "no-deficit" cases, or decrease of hemiplegic cases, were those with the highest three-week and three-month fatality rates. A more detailed analysis of the course of recovery in neurological deficit was illustrated in Table SC-16. The table shows the comparison between neuro- logical symptoms at initial examination and at 3-week follow-up. As is evident, the "no deficit" group was the moat stable. Speech disturbance accompanying paralysis of limb(a) always hampered recovery. The improvement of hemiplegic patients to a monoplegic state or, further, to no paralysis of limb(a) was about 40% leas if speech disturbance was associated with hem~plegia of either side. The cumulative fatality rate at 3 weeks was doubled if speech disturbance was associated with hemiplegia. Recovery from monoplegia·with or without speech disturbance did not differ greatly, but the cumulative fatality rate differed significantly. Another interesting feature was that patients with right hemiplegia had better recovery potential and less fatality than those with left hemiplegia in both groups with and without speech disturbance. The recovery rate for patients with right hemiplegia was worse than those with left hemiplegia only for recovery from speech disturbance. In 5 cases, change of side of hemiplegia was reported. This must be an error in either recording or coding, and must be corrected. Tables SC-10 (iii) and (iv) show the relative frequency of neurological deficits by type of stroke. Apparently no focal sign was manifest in a large majority of SAR (70%), other focal signs appearing only in 30%,of which hemiplegia comprised a half. The reverse was true for intracerebral haemorrhage; more than 70% presented hemiplegia. lschaemic strokes revealed hem.iplegia in more than half after 3 weeks and after 3 months. Patients with occlusion of pre-cerebral arteries had leas, and only 50% manifested hem.iplegia at 3 weeks and leas at 3 months. CVD/S/73.10 page 22 In this group, minor symptoms classified as "no deficit" or symptoms due to scattered multiple lesions classified as "multiple paralysis" were coD1110n. Milder symptoms without hemiplegia were least frequent in cerebral haemorrhage. "Not assessed" at the initial examination was as frequent as 14%, and is supposed to be due to the severity of stroke. This category decreased to a negligible fraction at 3 weeks and 3 months. When the frequency of neurological deficit at 3 weeks and 3 months was compared (Tables SC-10 (iii) and (iv)) hemiplegia decreased and "no deficit" increasedh all disease categories, particularly in cerebral thrombosis (ICD 433) and unspecified stroke (ICD 436), but less marked in haemorrhagic strokes (ICD 430, 431). This is no wonder as clinically diagnosed haemorrhagic strokes usually destroy a wider area than ischaemic strokes. 2. Medical Consultation, Laboratory-or Radiological Examinations Tables SC-2 Annex 2 and Annex 3 illustrate the frequency of physician or specialist consultations and various diagnostic procedures applied. Almost 40% of patients failed to be examined by a physician. Presumably these were patients who died suddenly or who resided in a remote area and died before reaching medical services. The number much exceeded that of patients who died before registration and the reason for such unfortunate circumstances requires further analysis. The frequency was highest in.Fukuoka, followed by North Karelia and Zerifin: it is least in Zagreb, followed by Copenhagen, Akita and Goteborg. This may also be inversely related to the efforts made by each centre in finding suspected cases of stroke. If no scrutiny is made of death certificates, the proportion of patients without physical examination would be very small. Electrocardiogram (ECG) examination was most widely used in over half of patients, facilitated by its availability. The rate of consultation with a neuro- logist was relatively high in the ICD diagnosis groups of 430 1 432 and 434. This may mean that either neurologists were more often consulted when the stroke showed specific features, or that neurologists could distinguish and/or more often gave specific diagnosis to a stroke. ·Lumbar puncture was the most frequently applied specific examination (in one-third of patients). It plays a more decisive role in the diagnosis of haemorrhagic strokes (430 and 431), and the application rate was high in subarachnoid haemorrhage but not in intracerebral haemorrhage. Angiography was performed only in less than 10% of patients. The frequency of its use was obviously higher in the diagnosis of occlusion of precerebral arteries. It is important to decide on the surgical indication for patients with subarachnoid haemorrhage, where angiography was used more frequently but still only in one-third of cases. Electroencephalogram (EEG) examination was the second most common specific diagnostic procedure applied. It was used in about one-third of the patients in some centres (01 1 05, 29, 33), but seldom in other centres. CVD/S/73.10 page 23 Brain scan and echoencephalography are more recent developments, requiring special apparatua, and their use is not widespread. An attempt to use brain-scan more often was aeen moat in one centre (29) and to a lesser extent in centres 01 and 21: with echoencephalography this was apparent only in two centres (01 and 47). Although the therapeutic implication of elaborated examinatiom is not very high, their uae may be an indication of an active attitude towards diagnosis and therefore ratianal treatment. Frequent application of these examinations does not guarantee the accuracy of clinical diagnosis but can be a safeguard against mis- diagnosis. * The frequency of cases where at least one examination except ECG was performed was over 80% in Goteborg and over 70% in Espoo, Zagreb and Zerifin; over 50% in Copenhagen; and less in other centres. The frequency was highest in the diagnosis of subarachnoid haemorrhage (91%), occlusion of precerebral arteries (88%), and least in cerebral thrombosis and unidentified stroke (ICD 436) (46%). This clearly indicates that if we want a higher degree of diagnostic accuracy, we must advocate the more extensive use of relevant examinations.* 3. Complications at Follow-up Table SC-17 shows that at the three-week follow-up about 5% of all the surviving patients presented bed or pressure sores. There was considerable variation ~etween centres, the percentages ranging from 0.4 to 12.3%, probably becauae of differences in age distribution, in admission rates, and perhaps mostly due to differences in the quality of nursing care. After three months the pro- portion of patient• with bed sores had decreased markedly, the average then being 3%. Nearly half of the patients with bed-sores had eventually died, their fatality rate being four times higher than average, while in one-third of cases the sores had healed. In about half of patients with bed sores at 3 months, the sores developed later than three weeks after stroke. This appeared in a few percent only in European centres, but 2-3 times more frequently in other centres (Israel and Japan). As seen in Table SC-18, some degree of joint contracture was found in about 7% of the three-week survivors; the proportion at three months was stable. One- half of these died by 3 months, while in nearly a half contracture remained. In more than a half, joint contracture developped after the first follow-up. Differences between centres ranged from 5 to 10%. The frequency was stable in individual centres and the differences certainly reflect variations in criteria and thoroughneas of examination. The paragraph marked*••••* is based on data obtained by the end of July 1973. CVD/S/73.10 page 24 4. Working Conditions at Three Months Table SC-7 Annex l(i) illustrates the working condition of patients who had gainful work prior to the stroke, and who were still alive at three months. Twenty- six percent of the males and 18% of the females had returned to work. Seven percent of females belonged to the group who, although back in employment, were working less than before, as against 13% of the males. Since most of the females were house- wives and presumably part-time workers in rural areas, it is not surprising that females much less than males went back to their work. As expected, younger patients had a higher potential of recovery for work. A comparison between centres discloses only moderate variations as regards working capacity after stroke. Apparently, return to work was frequent for patients in a large occupational group (centre 49), presumably because of the provision of protective work environment and the younger ages of the patients. At attempt was made to correlate the post-stroke working conditions with the presumed type of stroke (Table SC-7 Annex l(ii)). In the group diagnosed as intracerebral haemorrhage - and thus including many severe cases - occupational rehabilitation was largely unsuccessful; only 1 out of 27 such cases in males, and none out of 17 female patients, had returned to even part-time work at three months. The difference in other groups was less significant. Five to fifteen percent returned to their previous work, and another 5 - 15 % to lesser work. Female patients recovered less than males in nearly all disease groups, perhaps reflecting less social obligation for them to return to their jobs. It should be remembered that the above figures illustrate the working condition at three months after the stroke. Data collection has not yet been completed in some centres. Some patients will resume work at a later time, but the ultimate figures are not yet available. 5. Personal Help for Stroke Patients Table SC-8 shows the type ·of services used to take care of stroke patients in each centre at three months after onset of stroke. Patients younger than 65, and aged 65 and over are illustrated separately. Centres 01, 47 and 49 studied only a younger subgroup. The number of patients who needed help with personal care significantly increased with age: more than half of the patients under 65 still required help, and about 80% of the patients aged 65 and over needed help. It is apparent that the family plays an important role in the care of stroke patients everywhere. This applied for roughly a quarter to one-third of stroke patients in Western Europe on average, slightly more in Zagreb, and much more (over AO%) in Zerifin and Japan. About one-quarter of cases were still cared for in hospital. This rate was rather uniform except in Zagreb, Osaka and JNR (low), CVD/S/73.10 page 25 and Saku (high), presumably because of planned discharge from hospital in Zagreb, because of the originally low hospitalization rate in Osaka, and because of delayed hospitalization and rehabilitation in Saku. Information was not available for about 30% of patients in Copenhagen, and analysis was impossible for this centre. The type of services differed strikingly with the age of patients. In all European countries the older patients were cared for in hospital much more than the younger patients, whereas older patients were cared for much less than younger ones in hospital in Japan. This was obviously a consequence of low admission rates, particularly among older patients, in Japan. Unpaid helpers other than the family may refer to a relative invited to help, or at times a friend or neighbour: this type of help accounted for less than 5% in any centre. Paid help other than hospital staff might be a maid, part-time housekeeper or other private services. The uee of this type of service varied from Oto 15% but was more often seen for older patients than for younger ones, imposing the cost of care either on the family and/or the community. Social and family services were thus absorbed in taking care of about 60% of surviving patients. Expensive hospital services were particularly much used by European centres, where hospitalization is a rule for medical care of patients with stroke. Detailed analysis of patients who were taken care of at home or in hospital, and their outcome, will be done when sufficient follow-up data are accumulated. 6. Rehabilitation Whether or not rehabilitation was given by a trained specialist is the criterion used to gather information on the rehabilitation service. Some patients may have recovered spontaneously, or physical impairment was so mild that they did not require rehabilitation. However, centres where mild stroke patients were registered more frequently were also giving early rehabilitation services (Table SC-15 (i)). No differences in rehabilitation policy were observed between two age groups. Rehabilitation commenced before three weeks in two-thirds or more in centres 01 1 OS, 06, 21 and 33. Early start and termination of rehabilitation was a common practice in Zerifin (centre 33). This was very infrequent in the Japanese centres, except for centre 46, and was in between these rates in other centres. Patients who have recovered do not need rehabilitation services by a specialist, and the number of recipients decreased at three months in many centres. Delayed commencement of rehabilitation after 3 weeks was rather frequent in the centres 45, 46 and 49. CVD/S/73.10 page 26 More detailed information on the initiation and duration of rehabilitation during periods before and after 3 weeks after onset of stroke was examined in relation to neurological deficit at 3 weeks and 3 months in Table SC-lO(v). Rehabilitation started early and was intensive in centres 01, 05 and 21, except for patients with "no" deficit or only mild symptoms. The Japanese centres gave least rehabilitation, but neither was rehabilitation given to a fair proportion of hemiplegic patients in centres 22 and 29. In centres 06, 33 and 46, rehabilitation was more often terminated after 3 weeks, even in patients who still had hemiplegia. The effect of rehabilitation may be assessed by comparison of neurological deficits at 3 weeks and at 3 months. In patients who had not rehabilitation, also, a definite improvement was revealed. Between 3 weeks and 3 months the "no deficit" group increased by a quarter, the mild deficit group doubled, while hemiplegics decreased only slightly. This sponuaneous improvement forms a baseline for evaluation of the effect of rehabilitation. The most successful group was that for which rehabilitation was terminated within 3 weeks. Here, the "no deficit" group almost doubled and the number of hemiplegics decreased by more than a quarter. Of course, improvement was the reason for the termination of rehabilitation, but not vice versa. Improvement in the group where rehabilitation continued was also apparent. The increase in the "no deficit" group was double that in the no rehabilitation group. However, the decrease in hemiplegics was negligible, and the increase in the mild deficit group was less. The group for which rehabilitation started after 3 weeks showed the least changes, since the reason for initiation of rehabilitation would probably have been no spontaneous improvement. The schedule for rehabilitation was examined in different ADL classes (as defined in the next section) in Table SC-15(iii). Differences between centres were great, but the general. ·tendency was for rehabilitation to be given for a longer time with the increase of disability. Altogether half the ind,pendent patients did not receive rehabilitation at all, whereas nearly half of the totally or more dependent patients received con~inued rehabilitation before and after 3 weeks. Rehabilitation was terminated in about 30% of patients who were less disabled, and in about 15% of patients who were greatly disabled, after 3 weeks. The schedule of rehabilitation given was in no way related to occupation, as shown in Table SC-15(iv). 7. Activity in Daily Life (ADL) Activity in daily life is divided into four classes by degree of independence in self-care and walking, as follows: (Table SC-19) Class 1: Completely independent independent in self-care and walks without help from another person Clasa II: Partly dependent (grade 1) CVD/S/73 .10 page 27 partly dependent in self-care but walks without help of other person, or independent in self-care but walks with help of other person or does not walk at all although need not be carried in transfer Class III: Partly dependent (grade 2) partly dependent in self-care and walks aided or does not walk at all Class IV: Totally dependent totally dependent in self-care Unknown: Information about self-care and/or walking not available. At 3 weeks, about one-third of the patients were independent, one-third were completely dependent, and one-third partly dependent (Table SC-ll(i)). At three months (Table SC-ll(ii)) 1 about half became independent, and complete dependence decreased to one out of six survivors. Partial dependence decreased as a whole and among these cases the degree of dependence lessened (Class III to Clasa II). The relationship between age and ADL was clear; a steady decrease of independence and increase of complete dependence with age was observed. At 3 months about three-quarters had become independent and only 6% were completely dependent in patients younger than 55, whereas at the ages of 65 to 74 nearly half were completely independent and 20% were completely dependent. Female patients seem slightly more dependent than men, particularly those aged over 75, due to the mu~h older age composition of female patients in this age group. Table SC-14 shows the relation of ADL classes at 3 weeks and 3 months, by three age groups. In an additional table (Table SC-14 a) increase in class number is defined as "worsening" and decrease in class number as "improvement". More than half the patients did not change their ADL status during this observation period, irrespective of age groups. One-third to one-quarter improved, depending on age. The frequency of deterioration was negligible. Death accounted for 5 to 15%, increasing with the age of the patients. The potential for improvement was greater in ADL classes II and III, rather than Class IV. Patients who were independent at 3 weeks were the most stable; more than 90% of those aged under 55 did not change, but the possibi- lity of both worsening and of death increased slightly with age. Remarkable progress in ADL was observed everywhere, but the degree of independence differed by centre (Table SC-ll(iii) and (iv)). The relative frequency of completely independent patients varied from one- to two-thirds among survivors at CVD/S/73.10 page 28 three months. Complete dependency decreased at various rates and remained in about 10% to one-third at three months. A high frequency of independence and low frequency of complete dependence was observed in the centres 01 1 21, 48, whereas the reverse was true for centres 29, 45 and 46. Of course the centres with higher frequency of mild stroke and/or with high frequency of younger patients, experienced a higher rate of recovery. The reason for a high recovery rate in centre 48 may also be partly explained by the high initial fatality in males. Bed- or chair-fast patients accounted for 40% at 3 weeks and decreased to about 20% at 3 months with less centre difference than at 3 weeks (Table SC-12(iii)). Bladder control is a neurological function, but if impaired the patient requires help of another person (Table SC-12(iv)). A quarter of stroke survivors still had trouble with bladder control at three weeks, but the frequency decreased to 16% at three months. The frequency varied widely by centre at three weeks, but to a lesser extent at 3 months. 8. Location of Patients The location of patients is more related to social factors 1 e.g. the delivery system of medical care 1 available facilities, the economic condition, traditional attitude of the family, than to the clinical condition of the patient. The observation periods are too short for evaluation of these different environ- ments, and only the general situation is described (Tables SC-6, SC-13). At three weeks (Table SC-6) 1 a larger part of hospitalized patients were still in hospital and about one-fifth had been discharged. At three months (Table SC-13(i)) 1 more than half had been discharged, and only one-fifth of male patients, one-quarter of yqunger female and one-third of older female hospitalized patients were still in hospital. The rate of discharge was high in European centres, headed by Zagreb (nearly 90%), but was low in Japanese centres, particularly in Saku (30%). ADL was a determinant factor for discharge (Table SC-13(ii)). The rate of continuous hospitalization at 3 months was less than 10% in ADL Class I, about 20% in Class 11 1 and 50-60% in Classes III and IV. The rate differed by centre: in the Japanese centres, where the hospitalization rate was low, a longer stay in hospital was common; here, the discharge rate for ADL Class I was the same as that for Class II in other centres, and that for Class II was the same as that for Classes III and IV. The low discharge rate may be related to insufficient pro- vision of services after discharge and may also indicate low efficiency in use of hospital beds. * In Zagreb, more than half were discharged by 3 weeks,* and no hospitalization continued after 3 months for patients of ADL Classes I to III. The phrase marked* ••• * is based on data obtained by the end of July 1973. CVD/S/73.10 page 29 The majority of patients of ADL Classes III and IV were moved to nursing care facilities, and only a few of them. remained in hospital. Rather strict discharge criteria seem to have been operating here. Transfer to another hospital was reported in about one-fifth of patients at 3 weeks. This happened more frequently in young patients and hence perhaps for the purpose of rehabilitation.. In ADL Claes I and in centres 22 and 33, one-fifth of patients at 3 months were placed :h the category E ("Others" - Table SC-13(ii)). What this category really represented needs to be clarified. A majority of the patients who, at the time of initial examination, were being taken care of at home, continued to stay at home. By 3 weeks 15% of them had been hospitalized. At 3 months, putting together these patients with those discharged from hospital, about 60% of independent patients (ADL Classes I and II) and one-third of patients who were more dependent (ADL Classes III and IV) were at home. Patients who were in nursing homes at 3 months accounted for 8% of females aged over 65 and 5% of other age/sex groups. This frequency depends upon available facilities and was high in centres 05, 22 and 33. Living condition. may have some influence upon the subsequent location of patients (Table SC-13(iii)). The rate of discharge from hospital and the rate of patients at home were lower in patients who lived alone at the time of stroke than those who had been with their family. This may be explained by a preponderance among patients who lived alone of older females, whose recovery is usually slow and imperfect. The fact that more than half of stroke survivors who lived alone were discharged and were staying at home at 3 months after the onset of stroke means that continuous care and services for home care need to be provided for these people. 9. Patients whose Diagnosis of Stroke was Revised In about 5% of all the cases the original diagnosis of "stroke" was revised during the weeks following onset. These patients were followed for a period of three weeks and then withdrawn from the register. One-half of this group were patients whose neurological deficit disappeared within 24 hours, according to the definition used in the present work, such cases are not stroke but are transient cerebral ischaemia. Ten percent were generalized cerebral lesions, which had to be excluded from the (acute) stroke register. In the remaining cases the revised diagnosis varied. Table SE-2 Annex lists the large number of patho- logical conditions which were initially registered as stroke and were later changed. CVD/S/73.10 page 30 It is noteworthy that while there were five cases of primary or secondary intra- cranial neoplasm, no case of subdural haematoma appears in the list. In some of the published stroke series, autopsy revealed a number of misdiagnosed cases of subdural haematoma. In view of the low autopsy rate in the present series, it cannot be excluded that some of the dead patients might actually have had undiagnosed subdural haematoma. The number of revised diagnoses, particularly transient cerebral ischaemia, was high for those centres (OS, 06) in which registration was carried out soon after the onset of the disease. It is surprising that no revised diagnosis was experienced in any Japanese centre; this may be due to several factors, for example low hospitalization rate, more frequent referral from practitioners with less application of laboratory examinations, postponement of referral until a reasonably correct diagnosis is established, or low autopsy rate, etc. The background information for these patients was sought in order to ascertain a reason for the misdiagnosis (Table SE-2). The past medical history, disability before the event, and hospitalization rate, were essentially the same as that of the remaining stroke patients. Hypertension and dependency in self- care were less frequently found. The presenting neurological symptoms were milder, but fatality rate at 3 weeks was of a similar magnitude. Under these conditions, it seems possible that a falsely high incidence of stroke may have been introduced in the Japanese centres. VI. CASE FATALITY The reporting of deaths may not be complete, particularly for deaths occurring after 3 weeks from onset of stroke. Nevertheless a tentative analysis has been made to illustrate special features. 1. Immediate Fatality Rate In Table SD-1 the fatality rate is shown according to the interval between stroke and death. As in other published series, there was an accumulation of deaths during the first day of the disease. Altogether, 13% of the registered patients died on the day of onset; after three days, 18%; after one week, 25%; and after three weeks, 33%. The fatality rate for males and females at the same age was almost identical (Table SD-1 add.1). The high fatality rate of young patients on the first day should be a result of the high frequency of severe stroke, i.e. intracerebral haemorrhage~ After a week, the fatality rate of older patients overtook that of young patients. The fatality slows down in two to three weeks in young patients, the cumulative rate being about 30% at 3 weeks, and 35% at three months. In old patients the high fatality rate persisted, and the cumulative fatality rate was about 40% at three weeks and 50% at three months. The influence of age on the cumulative fatality rate therefore increased with the time that passed since the stroke. CVD/S/73.10 page 31 When individual centres are reviewed separately (Table SD-2, add.2 1 3) 1 higher fatality rates were recorded in the centres where frequency of coma was high. In the young age group, the Saku 1 Fukuoka and JNI. centres, and in the old age group Espoo, Zerifin and Osaka centres showed a high rate, whereas the fatality rate was generally low in Goteborg 1 Copenhagen and Zagreb centres where there was a lower frequency. of comatose patients. However. the interpretation of these data was difficult, because of the fact that the highest percentages of early death were found in the Japanese centres where the admission rate for stroke patients was particularly low. The different proportions of early deaths may therefore merely indicate that emergency treatment in hospital did in fact delay the time of death in severe stroke cases. The fatality rate for young patients in Akita and Osaka appeared to be ezceptionally low, and the reason for this requires further analysis. 2. Factors Influencing the Inmediate Fatality Rate (a) Previous diseases: It appears from Table SD-l(iv) that for patients with a history of previous stroke the immediate prognosis for life was only slightly worse than for those without previous episodes. At ages under 65 years, the three-week fatality rate was 33% in the former group, and 27% in the latter. while at ages over 65 the correspond- ing difference does not exceed 2%. The influence of previous stroke was thus so mild that it may only reflect the fact that patients with previous stroke were, on average, older than those who had never had stroke previously. *Manifestation of pre-existing cardiovascular disease was expected to be an adverse factor in the survival after stroke. Early fatalities, i.e. those occurring within 1-3 days of the stroke 1 were less frequent in patients with a history of either AMI or diabetes mellitus than in those without such histories. This may mean that persons with arteriosclerotic heart disease and/or diabetes mellitus were more likely to have ischaemic strokes than haemorrhagic ones. Death rates after the first three weeks became higher in patients with history of cardiovascular disease.* A history of hypertension, on the other hand, appeared to have no significant influence on the time at which death occurred after stroke (Table SD-l(v)). In fact, the proportions of very early deaths (within three days) were almost identical for patients with known hypertension, and for those with a negative history. It reflects a well-known fact that a negative history of hyper- tension does not preclude the presence of hypertension which is very often present without symptoms. * A slightly higher i11111ediate fatality (within 3 days) was observed in hypertensive patients not under treatment than in treated hypertensive patients. This may be due to untreated hypertensive patients being more prone to intracerebral haemorrhage.* The general state of the patient before the stroke - as measured by the degree of dependence in ADL - was a more reliable prognostic indicator than the presence of any specified disease in the history (Table SD-l(iii)). At ages under The sections marked* ••• * are based on data received by the end of July 1973. CVD/S/73.10 page 32 65, very early and early fatality rates up to three weeks were thus about twice as high in patients who prior to the stroke were dependent in self-care than in those who were independent. In elderly patients the difference was smaller, but still of considerable size. Almost half of the patients who had been disabled since before the stroke died by three weeks; thus the proportion of patients with previous dependency in self-care decreased from 17.3% to 12.9% at 3 weeks, and to 10.8% at three months. (b) Clinical state at the time of maximal impairment: Table SD-l(i) clearly confirms the fact that the state of consciousness is the single most important factor influencing the imnediate prognosis after stroke. For each age group, at any time, the fatality rate rose steeply with increasing degrees of impaired consciousness, the difference being particularly remarkable at the earlier stage. The three-week fatality rate was no less than 80% in comatose patients. Thus, in spite of modern improvements in emergency care, the chances of a comatose ~troke~patient surviving the acute phase were still less than one in five. Even a mild disturbance of consciousness was important, particularly among the elderly: at ages over 65, the three-week fatality rate was 39% for somnolent patients, but only 10% for those who were fully conscious. It is clear also that the presence of coma is a harbinger, not only of death, but of rapid death: almost half the fatalities observed in comatose patients during the first three months occurred on the day of the stroke; in contrast, only 6% of deaths occurred on the day of the attack in patients who were alert. * The fatality rate was much higher among "excessively hypertensive" patients, at any time of the follow-up up to 3 months in young patients and up to one week in older patients, particuarly at the very early stage (Table SD-l(vi.a)). A higher blood pressure level at the initial examination can be associated either with intracerebral haemorrhage, high risk of recurrent bleeding of ruptured aneurysm or consequence of elevated intracranial pressure due to wide-spread brain lesions resulting in brain oedema, and denotes a worse prognosis. Difference in fatality rates at the acute stage in the older age group was seen only between "excessively hypertensive" and the other two groups combined (normotensive and moderately hypertensive). The adopted cutting point for "moderate" hypertension was perhaps too low for old age patients - and "excessively hypertensive" only could separate definitely high blood pressure and hence high fatality groups. The long-term prognosis seemed worse in "normo"-tensive old patients, since the blood pressure level in these patients was in fact in the hypotensive range and reduced perfusion pressure in patients with stroke having advanced atherosclerosis in the cerebral artery may be a potential risk against long-term su~ival. * The paragraphs marked* ••• * are based on data received by end of July 1973. 3. Place of Death CVD/S/73.10 page 33 As illustrated in Table SD-2, a few patients died at home and the majority died in hospital in the Western European centres, whereas the majority died at home in the Japanese centres. In one Eastern European centre and the Israel centre the situation was in between. Some patients died in nursing homes in Copenhagen and North Karelia, less in Dublin and Zerifin, and only one patient in Espoo and in Saku. No deaths occurred in nursing homes in the other centres, reflecting the usage of nursing homes in each study area. 4. Autopsy Table SD-7 summarizes the autopsy rate for each centre. ~he rate for the whole series was 26.1%, It was particularly high in Goteborg and Copenhagen, both of which centres performed autopsy examination in over three-quarters of those who died. Autopsy was carried out for almost half the deaths in Espoo, and 20% in Dublin and Fukuoka. The rate was around 10% in other centres, and nil in Osaka. Comparison between clinical and autopsy diagnosis is reported in section I.6: Reliability of stroke type diagnosis. The differences in autopsy rate according to place of death is illustrated in Table SD-6. The rate was highest in hospital deaths, varying by centre from 82% in Copenhagen, 75% in Goteborg, followed by Espoo, Saku, Akita, etc. It was only about 10% in patients who died at home. Twelve out of 13 deaths at home were autopsied in Goteborg, but the rate was much lower in some other centres. Although the number was small, deaths occurring in places. other than at home or in hospital (presumably medico-legal cases) were included in the centres in Goteborg, Espoo and North Karelia. 5. Mortality The number of deaths by different type of stroke for each centre is summarized in Table SD-4. As mentioned before, data collection of death records has not been completed, and the results are thus preliminary. The distribution of stroke deaths for various types of stroke differed widely by centre, but not greatly by sex. Highly fatal stroke accounts for a larger proportion of mortality, introducing a higher relative frequency of intra- cerebral haemorrhage and a lower relative frequency of ischaemic strokes. SAR and unspecified stroke did not change in their share of mortality and morbidity. In centres 06, 29 and 33 the type of stroke was not specified for more than half stroke deaths. A very high frequency of deaths due to SAH was reported from Goteborg and the JNR centre, where the population under study was much younger. No cases of SAH death were reported from Zagreb, Fukuoka and Osaka: however, this seems to be improbable. SAH may have been included under intracerebral haemorrhage. Deatblcaused by combined haemorrhagic strokes (ICD 430 + 431) formed a major part of stroke deaths in Goteborg and in all the Japanese centres. Ischaemic strokes accounted for a major part of stroke deaths ;fl centres 05, 21, 22 and perhaps also 33. This general situation was similar to that for mo~bidity data. CVD/S/73.10 page 34 There were more female deaths in the Western European centres and Israel, whereas the reverse was found for the Eastern European and Japanese centres. Death from SAH and intracerebral haemorrhage followed the same sex pattern in these centres. The sex ratio was more variable for deaths from ischaemic stroke. VII. SUMMARY, AND IMPLICATIONS OF STROKE CONTROL 1. Magnitude of the Problem Accurate figures on the magnitude of the stroke problem in the comm.unity are emerging from the study. The average incidence rate for total population did not differ very much by centre. The annual incidence rate ranged roughly from 1.5 to 3 per thousand population. Let us take as an example a country of moderate size, with a population of ten million. Twenty thousand people become victims of stroke each year. About a quarter to one-tenth of them are younger than 55, differing with the age structure of the background population. On the whole, a little more than half are fetnale. Men are more frequent in the younger population groups of less than 65 years of age. Out of all the stroke cases, a quarter die within a week, more than one-third within three weeks, and a little more than 40% at three months, leaving 12,000 survivors. Thirty percent, or 6,000 patients, were engaged in some gainful work prior to stroke. At three months, 21500 of them have died; about 75% of male and 80% of female survivors are not working at all 1 accounting for a loss of productivity of 4,600 people at three months after stroke. Twelve percent of survivors (400) are working less than before, and 13% (450) only are working as much as they were before the stroke. Full working capacity is regained in less than a quarter. This aspect has various implications on the loss of productive power of a country. The loss is more important in Japan where mo~e than half the victims were previously in gainful occupation, and perhaps even more so in countries where the stroke victims are much younger. Independence in self-care is generally the goal of stroke rehabilitation. About half of the survivors do not need personal assistance in self-care and walking at 3 months. About one-third require some help, and 17% are completely dependent. Help is required anew every year for 91000 patients at 3 weeks after onset. One- third1 i.e. 21 800 at 3 weeks, and 21000 at 3 months, are totally dependent and require special long-term care in an institution. For partially dependent persons, a protective environment is necessary. These facilities are often lacking and do not always operate in an organized way to help the smooth transfer of patients. Training of physiotherapists, and support and guidance for patients at home and for their families is an urgent task. Basic information derived from the study on the magnitude of the problem should help in determining the scope and scale of the planning of such services for patients. 2. Case Fatality of Patients CVD/S/73.10 page 35 At the early stage young patients have a higher fatality rate, probably due to a high frequency of intracerebral haemorrhage. Young patients have a better life expectancy when the acute stage is over. Difference by age becomes more marked later; one-third of younger patients die at three months, whereas half of the older patients die during the same period. Prognosis is largely dependent upon the level of consciousness, i.e. the extent of damage to brain function. Current treatment is not potent enough to ameliorate these high fatality rates. Hospitalization seems helpful in postponing very early death, but does not seem to improve life expectancy in the long run. It therefore seems more reasonable to put more emphasis on prevention, which is more effective and perhaps more economical. 3. Effort in Prevention A single factor considered as the most effective target for prevention of stroke is the control of hypertension (HT). Blood pressure (BP) measured at three weeks after on1et of stroke would indicate the BP level prior to stroke, as the acute phase is over at this time. Patients who are receiving antihypertensive treatment at thi1 time obviously 1hould have been hypertensive before treatment; patients who are not treated and whose BP is at a hypertensive level form. another group of hypertensive patients. By putting together these two groups, one finds that of the study population 70% of stroke survivors were hypertensive. Fifty- one percent of the stroke patients knew that they had hypertension. Out of those who knew of the presence of RT, 57% were treated prior to stroke. Because of the retrospective nature of the study, nothing can be said about 35% of the stroke patients who died within three weeks, and no information is available on how many of the patients treated for hypertension were really under effective control. The success rate of effective treatment would be much lower among patients who had stroke, as the occurrence of stroke itself means failure of hypertension control in the co11111Unity. At three weeks, half of. the patients received antihypertensive therapy. It is interesting to note that their average BP was higher than in patients without treatment. This indicates two conditions. Firstly, that their blood pressure level was much higher before treatment and still remained high even after some reducation; and secondly, that the control of hypertension was insufficient, in spite of antihypertensive therapy. We still do not know whether the treatment of RT in patients who already experienced stroke is really effective in the prevention of stroke and other complications, or in reducing case fatality. Although study of the effectiveness of HT control in patients with stroke is not the intention'of this particular study, some pointers on this will be obtained from further follow- up. It should also be noted that a quarter of surv1v1ng stroke victims were not hypertensive at three weeks after the stroke. For these patients, factors other than hypertension were causative of stroke, and research is needed for their identification and possible prevention. CVD/S/73.10 page 36 4. Hospitalization and Home Care It is too premature to try to evaluate the effectiveness of hospitalization in stroke cases. Only a sketchy forecast is given here. Three different attitudes towards hospitalization can be distinguished between the centres. Western European centres hospitalized all stroke patients, except for a few of the most severe cases who died very soon after onset, and for a few mild patients. In the Zagreb centre, severe, older patients were hospitalized more frequently, perhaps due to a preference for patients who needed more help. In the Japanese centres, the majority of patients were taken care of at home, irrespect- ive of their condition. In some Japanese centres, immediate case fatality was very high, as was the frequency of complications. Comparison of hospitalized and non- hospitalized patients in Japanese centres, where hospitalization does not seem to be related to the severity of stroke, showed a remarkable difference in case fatality in both the young and old age groups. Other related problems, e.g. occupancy of hospital beds, and the cost of taking care of stroke patients, remain to be studied. Early discharge is fairly common in Zagreb, Zerifin, Goteborg and North Karelia. Efficient use of hospital beds is enhanced by early discharge. Comparison of the profile of patients who remain in hospital, and the outcome re recovery of patients at home, in ho'spital and in nursing homes, will be made at the next stage. 5. Rehabilitation Rehabilitation in this study means physical, speech or vocational training given by any qualified medical or paramedical person who had special training in rehabilitation. Early initiation and continuation of rehabilitation is expected to increase the working capacity, and independence in self-care. Initiation of rehabilitative care within three weeks after stroke was the general practice in many European centres and in the Israel centre, where more than two-thirds of the patients rece~v~d.rehabilitative care.within three weeks. In.Japanese centresl rehabilitatiqn was 1n1t1ated after 3 weeks 1n more than 10% of surviving patients, gufriu!ftSaraifajid/ other centres. Rehabilitation as defined was seldom applied in two Japanese centres (47 and 48). Bed-sores were more frequently reported in Zerifin, Akita and Saku, and joint contractures in Zagreb, Akita and Saku. The interpretation of the role of rehabilitation is difficult, since mild cases did not receive rehabilitation, and disabled patients often received rehabilitation continuously. Spontaneous improve- ment, without rehabilitation, was the general rule in slightly disabled patients. CVD/S/73.10 page 37 At three months, 52% of the patients were taken care of at home. Centres where there were relatively few rehabilitation specialists were naturally offering less rehabilitative services, and the family was involved more often in taking care of patients, even at the early stage. The training of rehabilitation specialists and the building-up of systematic services both for hospitalized and non-hospitalized patients seem to be priority problems in many places. CVD/S/73.10 Annex 1 page· 38 WHO STROKE REGISTER - Initial Record Form Study[!EEJ Identification and Registtation Registration number: Centre code: I I I I I I [D 4-8 9-10 Date of completion of ma [Dmonth rn 19 11-16 record form N ame of interviewer: N ame of patient: _____________________ _ Present home address: ----------------- ___ Tel. no. _______ _ Date of birth: ~ month rT"Tl LJ._J [I] lu.:..LJ 17-23 Code for data processing G24 Sex: l•male, 2=female 025 026 State of patient at time of notification: l•alive, 2-dead Preliminary diagnosis by notifying source (IC D code): I I I l I I 27-31 Personal C onditi.ons Prior to Stroke C i vii status: !•single, 2=married0 3cdivorced, 4-widowed. 5•separated, 0 6•other, 9-unknown 32 Living conditions: l•in private household, alone 2-in private household, not alone a-permanently living in institution 9=unknown Working conditions: l•in gainful work 2-retired because of ill health a-retired because of age 4-housewife S•other 9-unknown Occupation (2-digit ILO code): X 4=no gainful work X9aunknown , ......... ,..,.._ r"' ~I, "", .. -.. 0]35·36 Previous Medical H istorv l=no, 2=yes, 9=unknown (3=optional code) Stroke: A cute myocardial infarction: 038 A MI Registration No. , if any: I I I I I 139-43 A ny other heart diseases: D 44 A nerial hypertension: O 45 Hypertension Register No., if any:! ... _l...._ ..... l...._ ..... l.___l...._ ..... l 46·50 Treatment for raised blood pressure immediately prior to stroke: Diabetes mellitus: Patient's Capability of Self Care Before Stroke self care: !=independent 2•partially dependent 3=totally dependent 9=unknown Control of bladder: !'"not disturbed, 2=disturbed 9=unknown Early Stages of Present Attack day D ate of onset or detection: [I] Admission to hospital: l=no, 2=yes 3sstroke occurred in hospital, 9=unknown Date of admission to hospital: Clinical state at time of Maximum Impairment Level of consciousness: !=fully conscious, 2•somnolent, 3=semi-coma, 4 .. coma, 9=unknown R 051 052 month 0]55-58 month [060-63 L Motor defi.cit of limbs: l•no deficit 2=weakness/paral ysis 3=not assessable Arm Leg 0 0 O 65-66 0 67-68 Speech disturbances: l=no, 2=yes, 3=not assessable Blood pressure: mmHg (lst measurement after onset) Systolic Diastolic 069 I I I 170-72 I I I 173-75 CVD/S/73.10 naoe 39 WHO STROKE REGISTER - Follow-up Record Form Identification Initial tegl.Jtration number: l I I I I I 4-8 [I] Centre code: day month year 9-10 ... Date of Cl>D:J.Dletion of recora torm [[] [TI1ll11-16 N ame of interviewer: N .1me of patient: Present home address: Tel. no. Date of birth: rnrn Type of follow-up Optional 1 = regi,lar, 3 weeks rrn ll.__L_l__J17-23 2 • regular, 3 months . 6 ·.;. regulai; l.Syears D 3 * regular, 1 year 7 = tegular, 2 years · 24 4 • at recuirence 5 = recuirence, 3 weeks To be recorded at 3 weeks follow-up only Diagnosi.s by Registry (causing entry into Register) lCD code l I I l I Were any of the following examinations performed? J. "' no, 2 • yes, 9 • unknown F.xamination by physician F.xamination by neurologist Dao 031 Lumbar pwicture A ngiography Electtoencephalogram Brain scan Echo-encephalography Electrocardiogram Blood pressure at Systolic this interview: mm Hg Diastolic WHO/CVD 5165 - 2/72 - 2000 page l I I I I I I 138-40 ~1-43 Page 1 for Reg. No. Transfer of the Patient since previous interview lnltf.tution code: 1 • hospital, 2 • nursi.ng home 3 • home, 4 • other, 9 • unknown. Where was the patient at time of previous interview? Traasferred to:(if not transferred, write •no transfer• and leave boxes blank) 1) ___________________ Code D 45 ~ month r"f--, _ _ _ _ _ _ _ _ date LLJ [I] 19l_LJ 46-51 2) ________ _ Code date [I] 3) • _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Code· ---------- date [I] [IJ19[IJ 6<J·65 4) _____________________ code 066. ___ ----- __ date [I] [IJ19[IJ 67-72 Personal Conditions at Time of Interview Current living conditions: 1 • in private household, alone 2 • in private household,not alone 3 = in hospital or nursing home 9 • unknown Current working conditions: If in gainfu 1 work before the stroke, 1 • w orking as much as before 2 "' working less than before 3 = not working 9= pnknown If not m gainful work before the stroke, 4 = domestic activities as before 5 .. domestic activities less than before 6 .. no domestic activities 9 • unknown If help is needed in personal care, help is given by: l • family 2 = others (unpaid) 3 • paid domestic help 4 = staff in hospital or nursing home 5 = no help available 6 = no help needed 9 = unknown Card No. G]ao (T o continue to page 2) CVD/S/73.10 page 40 WHO STROKE REGISTER - Follow-up Record Form Medic.11 History Since Previous Regular Interview Recurrent stroke(s) (more than 3 weeks after onset): 1 • no, 2 = yes, 9 = unknown (3 = opt1011al coae) Acute myocardial infarction: 1 • no, 2 • yes, 9 • unknown (3 = optional coile) 026 If •yes•, AM I Registration No: (leave blank if not applicable) I I I I I I 27-31 Antihypertensive treatment: 1 • no, 2 • yes, 9 • unknown (3 = optional code) 032 Hypertension Registration No: (leave blank if not applicable) I I I I I 133-37 Rehabilitation given by qualified person: 1 • no, 2 • yea, 9 = unknown (3 = optional code) Clinical State (at time of interview) Level of consciousness: 1 • fully conscious 2 = consciousness dirturbed 9 = unknown Motor deficit of limbs: 1 • no deficit 2 • weakness/paralysis 3 = not assessable Speech disturbances: 1 • no 2"' yes 3 • not assessable Bed or pressure sores: 1 • no 2"' yes 9 • unknown Joint contracture: 1 = no 2 • yes !l"' unknown R ArrnO Leg D L 040-41 042-43 044 Page 2 for Reg. No. Activity of Dally Living Bed-fast/chair-fast and must be carried? 1 = no 2 • yes 9 • unknown Control of bladder: 1 • not disturbed 2 • disturbed 9 • unknown Self care: 1 s: independam 2 • partially dependant 3 • totally dependant 9= unknown Walking: 1 • without help of other person 2 • with help of other penon 3 • not at all 9 = unknown Dso Card No. 0 80 CVD/S/73.10 i>aae 41 WHO STROKE REGISTER - Death Record Form Study ! P! 1 !3 ! Initial registration number: I I I I I I []] Centre code: day month year [IJ [I]19[IJ Date of death: Name of data collector: Name of patient: ____________________ _ Home adclrea: ---- -------- _______ Tel.no,---- __ Date of birth: 17-23 Code for data processing ~ 24 Place of death: 1-llome, 2'1bospital, 3~home, o 25 "8other1 ~wn (boxes 26·36 are reserved for local option) Autopsy performed?a l11D.01 21-ym, in bolp.l.tal Ssya. medf.co-legal 9'ounlmown 037 . WHO/CVD 5166 • 12/72 Cause of death: ICD code ______________ __I I I I I _________________ I I I 1 I _________________ I I I I I 138-42 143-47 148-52 __________________ ! I I I I ___________________ _.I I I J I ! s3-57 ! sa-62 _________________ _I I I l I 163-67 If death occwred within 3 weeks of moke, were any of the following examinations performed! lsno, 2-ym, S,,,U.nknown Examination by physician Examination by neurologist Lumbar puncture Angiography Electroencephalogram Brain scan Echo-encephalogram Electmcardlogram Brain autopsy performed? 1 .. no. 2 s yes. 9 = unknown 068 069 070 0 71 072 073 074 075 076 CVD/S/73.10 page 42 Idendftcatlon Initial regillrmm lllllllben WHO STROKE REGISTER • Recurrence Record Form Study IP j 1 j , I ,-s Cllalcal scare at dme al mazlmwn impairment within 24 hours Level of CODICiOIIIDell: 1 • fully comclou1 Centre code: I I I I I I [I] 9-10 2 • level of comcioum.u reduced 9 • unknown 02s Da:c! ;:;:;letlon of ~ [I]month 19 r!r-, ll•lS Morar deficit of limba UJ UJ 1 • no deficit 2 = weaknea/paralyds Name of 3 • nor a.e.able interviewer: ________________________________________ _ Name of patimt1 ••••••••••••••• _ •••••••••• · •••••••••••••••••• Preaenc home adclreaa ............................................ . Date of blldl: ,,._ reL no. . ------·-~---------- ~ manm· rT"TI · UJ [I] ~LL.LJ1'1·23 Speech dldurbancea: 1 • no 2 • yes 3 • not ameaable Eltly eses of prelellt recurrence Date of omet or detecdon: 1 • no 2 •yea 3 • lb'Oke'occurred Ill hCllpkal 9 • ullkllown Date of ldmildoll to hospiUl: ICD code R L Anno 026-27 Leg D 02s-29 Dao day month [I] [I] 31-34 day month [[] [I]a6-39 WHO STROKE REGISTER • Exclullon Record Form CVD/S/73.10 page 43 (To be uaed wben tbe excludon occurred later tban 3 weekl after onaet of attack) Study IP I R I 2 I I I I I I I ,-9 Celllre code: [D 9·10 Date of first follow-up rn rn rn 19 L.l_l ll-16 N amc ·)f interviewer=--------------------------------------- Na ne of patient:-----------------------------------------------··-·· TeL no •••••••••••••••••••• Date of compledan of the forms 65·'10 lCD code I I I 1 I I '11·'15 GJ 80 lfll0/C9D 5195 12/72-500 CVD/S/73.10 page 44 WHO STROKE REGISTER Initial Record form: 1-3 4-8 9·10 p 1 Registration number Centre Date of completion code 25 26 · 27-31 Diagnosis 51 Follow-up Record Form: Transfer Sheet ICD Examinations Syst. BP Diast. BP Date 52 53-58 59 60-65 Inst Date nst Date t Date ansfer 2 ansfer 3 T aQJfer 4 I I I I I I t I 1 I I I I I I I I I t I [ - .!-.3_. _I __ r- _ i::.8 ••••• I _9:JQ I_ -- __ lJ.:lJ>_,. _____ L_, ___ • -.11:aa ___ • 1 _ J~J p I l I 2 I I I I I I I I I I I I I I I I I I I I I I - ·--· -·--·--· --·--·--·--· - •--J.- , __ . _ . __ , __ 1 __ • __ • -' - '--· _,_ ·--· 25 26 27-31 32 33-37 38 39 40-41 42-43 44 46 47 48 49 50 AMI Reg. No. Hypt Reg. No. QmRLRL Arm Leg Death Record Form: g 25 26 51 27-31 32-36 37 38-42 43-47 48-52 ,4 .... ;., ,0 ~ 0 > tJ 0 :c 3' ICD 53-57 58-62 ICD, autopsy ICD, autopsy ICD Aty IC D, clinical ICD, clinical ICD, clinical 63-67 68-75 76 ICD, autopsy Examinations ra 80 GJ Ord No. 80 ~ Card No. , ::-, ~VD/S/73.10 page 45 WHO STROKE REGISTER - Transfer Sheet Rerurrence Record Form 1 · 3 4-8 9-10 11-16 17-23 24 P 1 4 4 Initial Centre Date of completion of the form Date of birth od R,.g, st rat ion numbn rode 25 26-27 28·29 30 31-34 35 36-39 40-44 Con R L Arm R L Spc Date of onset Hsp Date of hosp. Prelim.Diagnos," Leg or detect 1 on Follow-up Record Form: ICD Examinations 52 53-58 59 60-65 nst Date Inst Date Transfer 2 Transfer 3 Syst. BP Diast. BP Pre Inst Date 66 67-72 Date Transfer 4 Inst Transfer 1 73 74 75 Liv Wk Co on [ ... T!:~ .. ---T ··· r· ··T 1:-~ .. ----. ··· !. -~";~-+-·-..-···,· !!;!~-r-- ·r··· .i.··,··· , ... t~=~~ · · ·r · · .•.... J~fJ P: 1: 2: : ! : : : : : : : : : : : : : : ' ; ; : : I I • 1 I I I I I I I 1 • I I I ' 1 I 1 1 1 ' ' - ... • L • • • .L ..... .&. ............ .&. .. • • .L ........ &. ....... J. ....... t. • • •+ • • • ...... • ....... ••"'"• • .. .&. "''"' .. ,L ._ • .t. .. • • .a. ...... &. ...... J.. ....... .&. ...... • 1. .. .... .1. ...... ..L ...... .a 25 26 27-31 32 33-37 38 39 40-41 42·43 44 45 46 47 48 49 Re M AMI Reg. No. Hy Hypt Reg. No. Re Con R L R L trk I rt Arm Leg Exclusion Record Form 1·3 4-8 9-10 11·16 24 p -R 2 1 Registration number Centre Date of I st follow-up od code 65-70 71-75 80 2 Date of completion Diagnosis by od of the form Registry 50 80 . 0 Card No. 80 0 Card No.