HYPERTENSION IN GENERAL PRACTICE 3
Screening for hypertension: some practical problems DAVID CHRISTIE, md,fracp,mfcm Department of Medical Statistics and Epidemiology, London School of Hygiene and Tropical Medicine.
SUMMARY. In 1967/68, a screening examination carried out on 18,277 male London civil servants, of whom 488 were referred to their general practitioners with high blood pressure. After this referral, 23 per cent did not attend their doctor and among those who did, a relatively high frequency of anxiety was noted. In one third of the patients, the general practitioners were already aware of the presence of hypertension, but were not treating it. This reluctance to treat asymptomatic people continued with the man¬ agement of referred and confirmed newly diag¬ nosed hypertensive patients. The problems as¬ sociated with mass screening examinations can be overcome by careful, expensive and timeconsuming preparations. Where the purpose is to find and treat newly diagnosed hypertensive patients then screening by general practitioners is likely to be a more cost-effective approach. was
Introduction T"\ETECTION of disease before its usual clinical *^ presentation should lead to more effective inter¬ vention, with consequent reduction in morbidity and mortality. However, this presupposes that a presymptomatic diagnosis can be made, that treatment is avail¬ able, and that such treatment can influence for the better the natural history of the disease. One of the few conditions that unquestionably fulfil all of these requirements is hypertension and for this reason numerous screening programmes have been car¬ ried out to detect individuals with raised blood pressure. When such programmes are conducted on a mass basis ©
Journal of the Royal College of General Practitioners, 1979, 29, 597-601.
relatively short period, as opposed to the longscreening in general practice, then other criteria become equally important in 'finding* the disease. These include follow-up procedures designed to ensure that patients referred for treatment in fact receive it, that no adverse psychological reactions are induced, and that good liaison is maintained with the doctors re¬ sponsible for continuing care. This study describes a large screening survey carried out 10 years ago with over a
term
reference to the last factor.
Method The London Civil Service Health Survey was carried out in 1967/68, when 18,277 men over the age of 40 were screened for cardiorespiratory disease and diabetes mel¬ litus. The aims and methods of the survey have been described elsewhere (Reid et al., 1974). The parts of the examination relevant to this study were a selfadministered questionnaire and a single blood pressure measurement taken with a bias-free sphygmomanometer at the London School of Hygiene. Ninety-four per cent of patients gave permission for survey results to be given to a named general prac¬ titioner, and this was done within six weeks of the examination. Some patients gave no indication on the questionnaire that either they, or their doctors, were aware of the presence of hypertension, yet their blood pressure was greater than 200 mm systolic and/or 115 mm diastolic. The doctor was advised by letter about these patients (subsequently referred to as 'hyper¬ tensive') within two weeks and each patient also re¬ ceived a letter advising him to visit his doctor. The doctor was sent the results of the examination for the other patients, but no comment was made. At least three months after referral, follow-up questionnaires were sent to the general practitioners. Ninety-five per cent replied and the replies form the basis of this report.
Journal of the Royal College of General Practitioners, October 1979
597
Hypertension in General Practice Table 1. Comparison Qf apparent and effective yields of patients with high blood pressure detected at a screening examination. (Percentages are given in brackets.)
Age (years) 40to49 50to 59 60+ Total
Number Number of of patients 'new' patients Number referred to known to of men general have attended examined practitioner general practitioner 7,306 52(0.7) 97(1.3) 7,902 230(2.9) 117(1.4) 161 (5.2) 3,069 84(2.7) 18,277 488 (2.7) 253 (1.4)
1 9 7 8
Results After screening, 488 men were referred to their doctors with hypertension which had apparently been unrecognized previously. The prevalence was thus 2 7 per cent but of these men 113 (23 per cent) did not attend their doctor as advised (Table 1). Of the 375 patients who did attend 122 (33 per cent) were, according to their doctors, already known to have hypertension. None of these men had ever received treatment for the condition, and there was no evidence that the survey findings influenced the doctor's management. One measure of the yield of a screening programme is simply the number of patients found with potentially treatable diseases. A more useful measure might be the number of patients who are not known to have a disease, who actually attend their doctor for assessment and evaluation. The final column in Table 1 shows yield measured in this way and in each age group it is about half the apparent yield. These calculations of yield assume that all patients diagnosed on a single blood pressure measurement as 'hypertensive' will in fact remain so classified on repeat examinations. This is, of course, most unlikely and so each general practitioner was asked to record repeat blood pressures on the referred patients who visited him. The decision to refer in the first place was based on the rather high level of above 200 mm systolic and/or a diastolic of 115 mm or more. To assess final useful yield, a repeat blood pressure of at least 90 mm diastolic was selected as one at which most doctors would consider starting therapy (Veterans Co-operative Study Group, 1970). Of the 253 patients with apparently newly diagnosed hypertension who actually attended their doctor, 192 (76 per cent) had repeat blood pressures above 90 mm diastolic. The useful yield of the screening examination, in terms of hypertension, thus dropped to one per cent of patients examined, or 192 out of 18,277. The management of these 192 patients by their own doctor is shown in Table 2. At the end of three months, 38 per cent had no form of therapy and a further eight per cent had been given sedation or other non-specific measures. Fifty-four per cent were receiving diuretics and/or some other form of hypotensive treatment; 598
Prescribing Information Presentations 'Tagamet' Tablets PL0002/0063 each containing 200mg cimetidine. 100, £13.22; 500, £64.75. 'Tagamet' Syrup PL0002/0073 containing 200mg cimetidine per 5ml syrup. 200ml, £6.29. Indications Duodenal ulcer, benign gastric ulcer, reflux oesophagitis. Dosage Duodenal ulcer: Adults, 200mg tds with meals and 400mg at bedtime (lOg/day) for at least 4 weeks (for full instructions see Data Sheet). To prevent relapse, 400mg at bedtime or 400mg moming and evening for at least 6 months. Benign gastric ulcer: Adults, 200mg tds with meals and 400mg at bedtime (lOg/day) for at least 6 weeks (for full instructions see Data Sheet). Reflux oesophagitis: Adults, 400mg tds with meals and 400mg at bedtime (1.6g/day) for 4 to 8 weeks. Cautions Impaired renal function: reduce dosage (see Data Sheet). Potentiation of oral anticoagulants (see Data Sheet). Prolonged treatment: observe patients periodically. Malignant gastric ulcer may respond symptomatically. Avoid during pregnancy and lactation. Adverse reactions Diarrhoea, dizziness, rash, tiredness. Rarely, mild gynaecomastia, reversible liver damage, confusional states (usually in the elderly or very ill), interstitial nephritis.
References 1. Cimetidine in the treatment of active duodenal and prepyloric ulcers. (1976) Lancet, ii, 161. 2. The effect of cimetidine on duodenal ulceration. (1977) Proceedings of the Second Intemational Symposium on Histamine H2-Receptor Antagonists. Excerpta Medica, p.260. 3. Oral cimetidine in severe duodenal ulceration. (1977) Lancet, i, 4. 4. Cimetidine treatment in the management of chronic duodenal ulcer disease. (1978) Topics in Gastroenterology. (In Press). 5. Maintenance treatment of recurrent peptic ulcer by cimetidine. (1978) Lancet, i, 403. 6. Prophylactic effect of cimetidine in duodenal ulcer disease. (1978) Brit. med.J., 1,1095.
Full prescribing information is available from
SK&
a Smgttligm comipane Smith Kline & French Laboratories Limited Welwyn Garden City, Hertfordshire AL7 lEY Telephone: Welwyn Garden 25111 'Tagamet' is a trade mark C) Smith Kline & French Laboratories Limited 1979
TG: AD59
Hypertension in General Practice
surprisingly, the proportion on specific therapy increased with age. The final column of Table 2 shows that the total proportion of patients referred to hospital was 21 per cent, and that the younger the patient the more likely was such referral. A risk of mass screening is the possibility of engendering alarm and anxiety unnecessarily which is certainly not helpful. In 24 per cent of the referred patients who visited their doctor, the level of anxiety was high enough for the general practitioner to comment. In the majority of patients, the doctor thought that reassurance and explanation was enough to allay these fears; in six, more serious psychological reaction was reported. Whilst there is no real evidence that the screening examination played a causal role in these events, their occurrence underlines the care that needs to be taken by the organizers of such surveys. Many doctors expressed reservations about mass screening particularly in view of the anxiety it created.
Discussion The examination of apparently well people, especially men, for high blood pressure is generally regarded as a worthwhile exercise in preventive medicine. As Tudor Hart (1975) points out, there are two approaches to such screening. The first is examination by the general practitioner of all patients presenting, for whatever reason, and the second, usually organized by a university department or other agency, consists of the examination of large numbers of people in a relatively short time. In view of the problems inherent in the latter approach, I suggest that screening in general practice, for high blood pressure at least, may be a more effective way of using limited resources. The most useful definition of hypertension is "that level of blood pressure above which there is evidence that treatment does more good than harm". In 1967 the available evidence (Hamilton et al., 1964; Veterans Co-operative Study Group, 1967) indicated that the appropriate level of diastolic pressure was 115 mm and above. The subsequent report from the Veterans Group (1970) lowered this to 105 mm diastolic; the current trial
Table 2. Management by general practitioners of confirmed 'new hypertensives' referred from a screening survey. (Percentages are given in brackets.) Percentage referred NonNo treatment Hypotensive specific Total to hospital Age 43 21 7 30 15 40 to 49 21 87 5 36 46 50 to 59 3 62 16 23 36 60+ 21 15 192 74 103 Total (54) (8) (100) (38)
600
by the Medical Research Council (MRC) may demonstrate a benefit from treatment in those people with a diastolic of 90 mm or above. This successive definition and re-definition of 'disease', dependent as it is upon cumulative scientific knowledge, may be regarded as a model of the way scientific medicine should advance. Doctors responsible for continuing care should be aware of such advances in preventive medicine, and be willing to implement them; it is equally necessary for patients to accept such intervention. Ten years ago, from the data collected in this study, there was considerable reluctance on the part of general practitioners to treat asymptomatic hypertensives, even those patients whom they 'already knew' to have high blood pressure. There is little evidence that the situation has changed (Heller, 1976) and this is a strong argument against the mass screening of populations without previous consultation with, and perhaps education of, the general practitoners in the area. If there is little point in referring hypertensives to doctors who have no intention of treating them, there is even less point in advising people to see their doctor when 23 per cent decide, of their own volition, not to do so. This is not always the case: in a recent Australian survey (Christie et al., 1976) hardly any referred patients failed to visit their doctor as advised. The difference, once more, lies in securing not only consent but the full co-operation of general practitioners in the area. Lessons can be learned from the past and the MRC pilot trial (1977) provides a model of how a mass screening survey for hypertension should be carried out. Of particular interest is the attention given to possible psychological sequelae of screening. In contrast to the rather high incidence of such sequelae. in the Civil Service Survey, the careful procedures of the MRC trial appear to have virtually eliminated the problem. Almost all the difficulties encountered in screening for hypertension can be circumvented by timeconsuming preparation including persuasion and education of both patients and their doctors. Certain types of investigation, such as those used in the MRC trial, do require such preparation but this is not the case where the purpose is simply to find and treat presymptomatic hypertension. Under these circumstances, a doctor working in his own practice has a degree of control that is impossible to achieve even in the best mass surveys. The development and spread of case finding for hypertension by general practitioners would be a major step forward in prevention of stroke and possibly of coronary heart disease.
References Christie, D., McPherson, L. & Vivian, V. (1976). The Queenscliff Study-a community screening programme for hypertension. Medical Jornal of Australia, 2, 678-680. Hamilton, M., Thompson, E. N. & Wisniewski, T. K. M. (1964). The role of blood pressure control in preventing complications of hypertension. Lancet, 1, 235-238.
Journal of the Royal College of General Practitioners, October 1979
Hypertension in General Practice Hart, J. Tudor (1975). Screening in primary care. In Screening in General Practice (ed. Hart, C. R.). Edinburgh, London and New York: Churchill Livingstone. Heller, R. F. (1976). Detection and treatment of hypertension in an Inner London community. British Journal of Preventive and Social Medicine, 1, 1437-1440. Medical Research Council Working Party (1977). Randomized controlled trial of treatment for mild hypertension: design and pilot trial, on mild to moderate hypertension. British Medical Journal, 1, 1437-1440. Reid, D. D., Brett, G. Z., Hamilton, P. J. S., Jarrett, R. J., Keen, H. & Rose, G. A. (1974). Cardio-respiratory disease and diabetes among middle-aged male civil servants. Lancet, 1,469-473. Veterans Administration Co-operative Study Group on Antihypertensive Agents (1967). Effects of treatment on morbidity in hypertension: results in patients with diastolic blood pressure averaging 115 through 129 mm Hg. Journal of the American Medical Association, 202, 1143-1152. Veterans Administration Co-operative Study Group on Antihypertensive Agents (1970). Effects of treatment on morbidity in hypertension II: results in patients with diastolic blood pressure averaging 90 through 114 mm Hg. Journal of the A merican Medical Association, 202, 1028-1034.
Addendum Dr Christie is now First Assistant in Clinical Epidemiology, Department of Community Health, University of Melbourne, Australia.
Middle ear effusion We observed the frequent occurrence of persistent middle ear effusion in children with acute otitis media and followed them according to standard procedures for otologic diagnosis. We performed a life-table analysis to identify risk factors for such persistent disease. Sixtytwo patients were free of middle ear effusion at one or more clinic visits two to 13 weeks after presentation and were considered cured; 45 had effusion at all clinic visits during this period and were defined as having persistent effusion. The life-table analysis showed that the relative risk for persistence was 3 * 8 times higher in children less than 24 months of age compared with children 24 months of age or older (p
Screening for hypertension: some practical problems DAVID CHRISTIE, md,fracp,mfcm Department of Medical Statistics and Epidemiology, London School of Hygiene and Tropical Medicine.
SUMMARY. In 1967/68, a screening examination carried out on 18,277 male London civil servants, of whom 488 were referred to their general practitioners with high blood pressure. After this referral, 23 per cent did not attend their doctor and among those who did, a relatively high frequency of anxiety was noted. In one third of the patients, the general practitioners were already aware of the presence of hypertension, but were not treating it. This reluctance to treat asymptomatic people continued with the man¬ agement of referred and confirmed newly diag¬ nosed hypertensive patients. The problems as¬ sociated with mass screening examinations can be overcome by careful, expensive and timeconsuming preparations. Where the purpose is to find and treat newly diagnosed hypertensive patients then screening by general practitioners is likely to be a more cost-effective approach. was
Introduction T"\ETECTION of disease before its usual clinical *^ presentation should lead to more effective inter¬ vention, with consequent reduction in morbidity and mortality. However, this presupposes that a presymptomatic diagnosis can be made, that treatment is avail¬ able, and that such treatment can influence for the better the natural history of the disease. One of the few conditions that unquestionably fulfil all of these requirements is hypertension and for this reason numerous screening programmes have been car¬ ried out to detect individuals with raised blood pressure. When such programmes are conducted on a mass basis ©
Journal of the Royal College of General Practitioners, 1979, 29, 597-601.
relatively short period, as opposed to the longscreening in general practice, then other criteria become equally important in 'finding* the disease. These include follow-up procedures designed to ensure that patients referred for treatment in fact receive it, that no adverse psychological reactions are induced, and that good liaison is maintained with the doctors re¬ sponsible for continuing care. This study describes a large screening survey carried out 10 years ago with over a
term
reference to the last factor.
Method The London Civil Service Health Survey was carried out in 1967/68, when 18,277 men over the age of 40 were screened for cardiorespiratory disease and diabetes mel¬ litus. The aims and methods of the survey have been described elsewhere (Reid et al., 1974). The parts of the examination relevant to this study were a selfadministered questionnaire and a single blood pressure measurement taken with a bias-free sphygmomanometer at the London School of Hygiene. Ninety-four per cent of patients gave permission for survey results to be given to a named general prac¬ titioner, and this was done within six weeks of the examination. Some patients gave no indication on the questionnaire that either they, or their doctors, were aware of the presence of hypertension, yet their blood pressure was greater than 200 mm systolic and/or 115 mm diastolic. The doctor was advised by letter about these patients (subsequently referred to as 'hyper¬ tensive') within two weeks and each patient also re¬ ceived a letter advising him to visit his doctor. The doctor was sent the results of the examination for the other patients, but no comment was made. At least three months after referral, follow-up questionnaires were sent to the general practitioners. Ninety-five per cent replied and the replies form the basis of this report.
Journal of the Royal College of General Practitioners, October 1979
597
Hypertension in General Practice Table 1. Comparison Qf apparent and effective yields of patients with high blood pressure detected at a screening examination. (Percentages are given in brackets.)
Age (years) 40to49 50to 59 60+ Total
Number Number of of patients 'new' patients Number referred to known to of men general have attended examined practitioner general practitioner 7,306 52(0.7) 97(1.3) 7,902 230(2.9) 117(1.4) 161 (5.2) 3,069 84(2.7) 18,277 488 (2.7) 253 (1.4)
1 9 7 8
Results After screening, 488 men were referred to their doctors with hypertension which had apparently been unrecognized previously. The prevalence was thus 2 7 per cent but of these men 113 (23 per cent) did not attend their doctor as advised (Table 1). Of the 375 patients who did attend 122 (33 per cent) were, according to their doctors, already known to have hypertension. None of these men had ever received treatment for the condition, and there was no evidence that the survey findings influenced the doctor's management. One measure of the yield of a screening programme is simply the number of patients found with potentially treatable diseases. A more useful measure might be the number of patients who are not known to have a disease, who actually attend their doctor for assessment and evaluation. The final column in Table 1 shows yield measured in this way and in each age group it is about half the apparent yield. These calculations of yield assume that all patients diagnosed on a single blood pressure measurement as 'hypertensive' will in fact remain so classified on repeat examinations. This is, of course, most unlikely and so each general practitioner was asked to record repeat blood pressures on the referred patients who visited him. The decision to refer in the first place was based on the rather high level of above 200 mm systolic and/or a diastolic of 115 mm or more. To assess final useful yield, a repeat blood pressure of at least 90 mm diastolic was selected as one at which most doctors would consider starting therapy (Veterans Co-operative Study Group, 1970). Of the 253 patients with apparently newly diagnosed hypertension who actually attended their doctor, 192 (76 per cent) had repeat blood pressures above 90 mm diastolic. The useful yield of the screening examination, in terms of hypertension, thus dropped to one per cent of patients examined, or 192 out of 18,277. The management of these 192 patients by their own doctor is shown in Table 2. At the end of three months, 38 per cent had no form of therapy and a further eight per cent had been given sedation or other non-specific measures. Fifty-four per cent were receiving diuretics and/or some other form of hypotensive treatment; 598
Prescribing Information Presentations 'Tagamet' Tablets PL0002/0063 each containing 200mg cimetidine. 100, £13.22; 500, £64.75. 'Tagamet' Syrup PL0002/0073 containing 200mg cimetidine per 5ml syrup. 200ml, £6.29. Indications Duodenal ulcer, benign gastric ulcer, reflux oesophagitis. Dosage Duodenal ulcer: Adults, 200mg tds with meals and 400mg at bedtime (lOg/day) for at least 4 weeks (for full instructions see Data Sheet). To prevent relapse, 400mg at bedtime or 400mg moming and evening for at least 6 months. Benign gastric ulcer: Adults, 200mg tds with meals and 400mg at bedtime (lOg/day) for at least 6 weeks (for full instructions see Data Sheet). Reflux oesophagitis: Adults, 400mg tds with meals and 400mg at bedtime (1.6g/day) for 4 to 8 weeks. Cautions Impaired renal function: reduce dosage (see Data Sheet). Potentiation of oral anticoagulants (see Data Sheet). Prolonged treatment: observe patients periodically. Malignant gastric ulcer may respond symptomatically. Avoid during pregnancy and lactation. Adverse reactions Diarrhoea, dizziness, rash, tiredness. Rarely, mild gynaecomastia, reversible liver damage, confusional states (usually in the elderly or very ill), interstitial nephritis.
References 1. Cimetidine in the treatment of active duodenal and prepyloric ulcers. (1976) Lancet, ii, 161. 2. The effect of cimetidine on duodenal ulceration. (1977) Proceedings of the Second Intemational Symposium on Histamine H2-Receptor Antagonists. Excerpta Medica, p.260. 3. Oral cimetidine in severe duodenal ulceration. (1977) Lancet, i, 4. 4. Cimetidine treatment in the management of chronic duodenal ulcer disease. (1978) Topics in Gastroenterology. (In Press). 5. Maintenance treatment of recurrent peptic ulcer by cimetidine. (1978) Lancet, i, 403. 6. Prophylactic effect of cimetidine in duodenal ulcer disease. (1978) Brit. med.J., 1,1095.
Full prescribing information is available from
SK&
a Smgttligm comipane Smith Kline & French Laboratories Limited Welwyn Garden City, Hertfordshire AL7 lEY Telephone: Welwyn Garden 25111 'Tagamet' is a trade mark C) Smith Kline & French Laboratories Limited 1979
TG: AD59
Hypertension in General Practice
surprisingly, the proportion on specific therapy increased with age. The final column of Table 2 shows that the total proportion of patients referred to hospital was 21 per cent, and that the younger the patient the more likely was such referral. A risk of mass screening is the possibility of engendering alarm and anxiety unnecessarily which is certainly not helpful. In 24 per cent of the referred patients who visited their doctor, the level of anxiety was high enough for the general practitioner to comment. In the majority of patients, the doctor thought that reassurance and explanation was enough to allay these fears; in six, more serious psychological reaction was reported. Whilst there is no real evidence that the screening examination played a causal role in these events, their occurrence underlines the care that needs to be taken by the organizers of such surveys. Many doctors expressed reservations about mass screening particularly in view of the anxiety it created.
Discussion The examination of apparently well people, especially men, for high blood pressure is generally regarded as a worthwhile exercise in preventive medicine. As Tudor Hart (1975) points out, there are two approaches to such screening. The first is examination by the general practitioner of all patients presenting, for whatever reason, and the second, usually organized by a university department or other agency, consists of the examination of large numbers of people in a relatively short time. In view of the problems inherent in the latter approach, I suggest that screening in general practice, for high blood pressure at least, may be a more effective way of using limited resources. The most useful definition of hypertension is "that level of blood pressure above which there is evidence that treatment does more good than harm". In 1967 the available evidence (Hamilton et al., 1964; Veterans Co-operative Study Group, 1967) indicated that the appropriate level of diastolic pressure was 115 mm and above. The subsequent report from the Veterans Group (1970) lowered this to 105 mm diastolic; the current trial
Table 2. Management by general practitioners of confirmed 'new hypertensives' referred from a screening survey. (Percentages are given in brackets.) Percentage referred NonNo treatment Hypotensive specific Total to hospital Age 43 21 7 30 15 40 to 49 21 87 5 36 46 50 to 59 3 62 16 23 36 60+ 21 15 192 74 103 Total (54) (8) (100) (38)
600
by the Medical Research Council (MRC) may demonstrate a benefit from treatment in those people with a diastolic of 90 mm or above. This successive definition and re-definition of 'disease', dependent as it is upon cumulative scientific knowledge, may be regarded as a model of the way scientific medicine should advance. Doctors responsible for continuing care should be aware of such advances in preventive medicine, and be willing to implement them; it is equally necessary for patients to accept such intervention. Ten years ago, from the data collected in this study, there was considerable reluctance on the part of general practitioners to treat asymptomatic hypertensives, even those patients whom they 'already knew' to have high blood pressure. There is little evidence that the situation has changed (Heller, 1976) and this is a strong argument against the mass screening of populations without previous consultation with, and perhaps education of, the general practitoners in the area. If there is little point in referring hypertensives to doctors who have no intention of treating them, there is even less point in advising people to see their doctor when 23 per cent decide, of their own volition, not to do so. This is not always the case: in a recent Australian survey (Christie et al., 1976) hardly any referred patients failed to visit their doctor as advised. The difference, once more, lies in securing not only consent but the full co-operation of general practitioners in the area. Lessons can be learned from the past and the MRC pilot trial (1977) provides a model of how a mass screening survey for hypertension should be carried out. Of particular interest is the attention given to possible psychological sequelae of screening. In contrast to the rather high incidence of such sequelae. in the Civil Service Survey, the careful procedures of the MRC trial appear to have virtually eliminated the problem. Almost all the difficulties encountered in screening for hypertension can be circumvented by timeconsuming preparation including persuasion and education of both patients and their doctors. Certain types of investigation, such as those used in the MRC trial, do require such preparation but this is not the case where the purpose is simply to find and treat presymptomatic hypertension. Under these circumstances, a doctor working in his own practice has a degree of control that is impossible to achieve even in the best mass surveys. The development and spread of case finding for hypertension by general practitioners would be a major step forward in prevention of stroke and possibly of coronary heart disease.
References Christie, D., McPherson, L. & Vivian, V. (1976). The Queenscliff Study-a community screening programme for hypertension. Medical Jornal of Australia, 2, 678-680. Hamilton, M., Thompson, E. N. & Wisniewski, T. K. M. (1964). The role of blood pressure control in preventing complications of hypertension. Lancet, 1, 235-238.
Journal of the Royal College of General Practitioners, October 1979
Hypertension in General Practice Hart, J. Tudor (1975). Screening in primary care. In Screening in General Practice (ed. Hart, C. R.). Edinburgh, London and New York: Churchill Livingstone. Heller, R. F. (1976). Detection and treatment of hypertension in an Inner London community. British Journal of Preventive and Social Medicine, 1, 1437-1440. Medical Research Council Working Party (1977). Randomized controlled trial of treatment for mild hypertension: design and pilot trial, on mild to moderate hypertension. British Medical Journal, 1, 1437-1440. Reid, D. D., Brett, G. Z., Hamilton, P. J. S., Jarrett, R. J., Keen, H. & Rose, G. A. (1974). Cardio-respiratory disease and diabetes among middle-aged male civil servants. Lancet, 1,469-473. Veterans Administration Co-operative Study Group on Antihypertensive Agents (1967). Effects of treatment on morbidity in hypertension: results in patients with diastolic blood pressure averaging 115 through 129 mm Hg. Journal of the American Medical Association, 202, 1143-1152. Veterans Administration Co-operative Study Group on Antihypertensive Agents (1970). Effects of treatment on morbidity in hypertension II: results in patients with diastolic blood pressure averaging 90 through 114 mm Hg. Journal of the A merican Medical Association, 202, 1028-1034.
Addendum Dr Christie is now First Assistant in Clinical Epidemiology, Department of Community Health, University of Melbourne, Australia.
Middle ear effusion We observed the frequent occurrence of persistent middle ear effusion in children with acute otitis media and followed them according to standard procedures for otologic diagnosis. We performed a life-table analysis to identify risk factors for such persistent disease. Sixtytwo patients were free of middle ear effusion at one or more clinic visits two to 13 weeks after presentation and were considered cured; 45 had effusion at all clinic visits during this period and were defined as having persistent effusion. The life-table analysis showed that the relative risk for persistence was 3 * 8 times higher in children less than 24 months of age compared with children 24 months of age or older (p
