Medicine & Science in Sports & Exercise:
Clinical Sciences: Clinical Investigations
Usefulness and Cost Effectiveness of Cardiovascular Screening of Young Adolescents
TANAKA, YUJI1; YOSHINAGA, MASAO1; ANAN, RYUICHIRO2; TANAKA, YASUHIRO2; NOMURA, YUICHI2; OKU, SHOZO3; NISHI, SEIJI3; KAWANO, YOSHIFUMI2; TEI, CHUWA2; ARIMA, KATSURA3
1National Hospital Organization Kyusyu Cardiovascular Center, Shiroyamacho, Kagoshima City, Kagoshima, JAPAN; 2Graduate School of Medical and Dental Sciences, Kagoshima University, Sakuragaoka, Kagoshima City, Kagoshima, JAPAN; and 3Kagoshima City Medical Association, Kajiyacho 3-10, Kagoshima City, Kagoshima, JAPAN
Address for correspondence: Masao Yoshinaga, National Hospital Organization Kyusyu Cardiovascular Center, Shiroyamacho 8-1, Kagoshima City, Kagoshima, 892-0853, Japan; E-mail: email@example.com.
Submitted for publication February 2005.
Accepted for publication August 2005.
The authors thank the physicians and participants who made essential contributions to this study.
Objectives: This study was conducted to assess the usefulness of a screening system for cardiovascular disease in Kagoshima, Japan, and to compare its cost-effectiveness with that of a similar system reported in the United States.
Background: Preparticipation screening of young athletes has been implemented in many countries to prevent sudden death, but sudden death in young nonathletes remains a problem. In Japan, both athletes and nonathletes have been screened for the presence or absence of cardiovascular diseases for more than 20 yr.
Methods: From 1989 to 1997, all seventh graders in schools in Kagoshima, Japan, were screened for cardiovascular disease using a questionnaire and electrocardiogram before physical examination. They were screened again in the same way 3 yr later. One subject newly diagnosed with cardiovascular disease and recommended to limit athletic participation was defined as “high-risk.” Situations leading to cases of sudden death were verified with a report from the school in question.
Results: Of the initial study population, 99% participated in the program every year. A total of 37,807 subjects, including nine high-risk subjects, were evaluated consecutively for 6 yr. Of these nine subjects, six, including three patients with hypertrophic cardiomyopathy, were nonathletes. Three sudden deaths occurred during the study period; one student was from the high-risk group. The cost of this screening system was lower than that reported in the United States.
Conclusions: Population-based screening for heart disease in this age range is limited by various factors. To analyze the mechanisms of sudden death in adolescents, we, therefore, are in need of a nationwide registry that includes autopsies for all deadly events.
Sudden cardiac death in young adolescents is an infrequent, yet tragic and alarming event, with an estimated occurrence rate among high-school athletes of 1:100,000 to 1:300,000 individuals per year (10,14,15,25,28). Preparticipation cardiac screening of young athletes has been implemented in many countries to prevent sudden cardiac death (1,7,8,13–15,19,21,23,25,28), because sudden death associated with athletic participation is a major cause of death from cardiovascular diseases, such as hypertrophic cardiomyopathy (HCM). Three preparticipation screening methods for reducing sudden cardiac death have been recommended: specific cardiovascular history and physical examination, 12-lead electrocardiogram (ECG), and two-dimensional echocardiography. Fuller (6) reported that the 12-lead ECG is the most cost-effective preparticipation cardiovascular modality of the three.
In 1973, a national screening system for cardiovascular diseases was introduced in Japan for all 1st-, 7th-, and 10th-grade students. In this screening system, a questionnaire and electrocardiogram (ECG) are used to screen both athletes and nonathletes for the presence or absence of cardiovascular diseases. It is unclear, however, whether this strategy reduces the risk of sudden death among Japanese adolescents and whether it has an appropriate level of cost-effectiveness.
The present study was conducted, therefore, to evaluate whether this screening system prevents sudden cardiac death among athletes and nonathletes during school-related activities, and to verify its cost-effectiveness in Kagoshima, southern Japan.
SUBJECTS AND METHODS
Kagoshima, a city located in southern Japan, has a population of 600,000. Our subjects included young adolescents who entered junior high school (at the seventh-grade level) between 1989 and 1997; we chose not to include elementary school children because the prevalence of sudden death among this population is low. A follow-up was conducted 3 yr later, between 1992 and 2000, when the subjects entered senior high school (at the 10th-grade level). The same methods of examination were used in both the initial and follow-up screenings. Because the aim of this study was to elucidate whether the screening system in Kagoshima is useful in preventing sudden cardiac death, students previously diagnosed with cardiovascular disease were excluded. Permission to perform this study was obtained from the committee for ethics of Kagoshima City Medical Association, under the condition that confidentiality regarding all personal data would be maintained. Written, informed consent was obtained from all patients.
Screening for heart diseases in Kagoshima.
During initial and follow-up screenings, which were conducted within schools, all enrolled 7th and 10th graders were examined using 12-lead ECG at a paper speed of 25 mm · s−1 and asked to give a questionnaire to their parents for completion. The questionnaire consisted of items related to any history of heart murmurs, cardiac diseases (including congenital heart defects, and Kawasaki disease), and cardiac symptoms (syncope, chest pain, an irregular heart beat or palpitations on exertion, and shortness of breathe), as well as any family history of sudden death at <40 yr of age. Subjects with abnormal ECG findings or a history of cardiac disease were examined further by physical examination and, if needed, chest x-ray, exercise stress tests (Master's two-step test or the treadmill exercise stress test), and echocardiography. High-risk subjects were defined as those newly diagnosed with cardiac disease and recommended to limit their athletic participation, whereas low-risk subjects were defined as those newly diagnosed with cardiac disease but not placed under athletic limitations.
Hypertrophic cardiomyopathy was suspected at the primary screening if a relevant family history was indicated or abnormal ECG findings were observed. Definitive diagnosis was made based on left ventricular hypertrophy (wall thickness ≥ 13 mm) in the absence of other cardiac diseases capable of producing this magnitude of hypertrophy and a nondilated left ventricle as observed by echocardiography (11,22). Left ventricular dilatation was diagnosed by determining the left ventricular end-diastolic dimension relative to the body surface area (5).
Sudden death during school-related activities among high school students.
Sudden death during school-related activities was also evaluated among the enrolled subjects from 1989 to 2002. Sudden death was defined as unexpected death as a result of natural causes in which loss of all functions occurred instantaneously or within 6 h of the onset of symptoms or collapse (1). Situations leading to cases of sudden death were verified with a report from the school in question.
Cost-effectiveness of screening programs.
To evaluate the cost-effectiveness of the screening program in the present study, we followed the method of Fuller (6), who examined cost parameters during screening of 70 high-risk subjects from 700,000 high-school athletes in the United States. The report assumes that 10% of high-school athletes determined as “at risk” will live an additional 40 yr, whereas 90% will live an additional 20 yr. The estimated cost per year of life saved was then calculated. The costs of the primary screening program in Kagoshima is financed by Kagoshima city ($16.7 per examination), whereas the costs of secondary screenings, which might include a chest x-ray, Master's two-step test, treadmill exercise stress test, and ECG, are fixed at $14, $27, $58, and $83, respectively, and covered by national health insurance. In Japan, parents, therefore, must pay 30% of all medical expenses incurred.
Final subject population.
A total of 69,033 students who entered junior high school in Kagoshima City between 1989 and 1997 were enrolled in this study. Approximately 99% of the parent 7th- and 10th-grade population participated annually in the primary screening program, and overall, a total of 68,503 students (99.2% of 69,033) had primary screening. During the study period, 30,696 students moved out of the area in their 10th grade, leaving a total of 37,807 students who were followed up for the following 3 yr; that is, 37,807 students were evaluated for six consecutive years. According to the questionnaire, 632 students were previously diagnosed with cardiovascular disease by their family doctor and, of all 7th- and 10th-grade students, 975 and 901 showed abnormal findings in the primary screening, respectively. Thus, a total of 1876 students (2.7% of 37,807) participated in secondary screening by physical examination, exercise tests, or echocardiography (Fig. 1).
FIGURE 1Cardiovascul...Image Tools
High- and low-risk subjects identified in secondary screenings.
Nine students (0.024% of 37,807) were identified as high-risk subjects: five with HCM, one with left ventricular dilatation, one with Wolff–Parkinson–White syndrome with tachycardic episodes, one with primary pulmonary hypertension (PPH), and one with long QT syndrome with torsade de pointes (Table 1). Simultaneously, 497 students were identified as low-risk subjects, whereas the remaining 1370 were diagnosed as normal. None of the HCM patients had any family history of the disease. Left ventricular dilatation was diagnosed in a student with incomplete right bundle block by echocardiography, which showed dilated left ventricular dimension; PPH was diagnosed in a 12-yr-old girl with right ventricular hypertrophy, and another asymptomatic girl was diagnosed with long QT syndrome because of a corrected QT interval of 0.53 s. Of the high-risk subjects, two boys with HCM and one with left ventricular dilatation were athletes (tennis, kendo, and baseball players, respectively) whereas the remaining six were nonathletes.
Sudden death during school-related activities.
Both high- and low-risk subjects were followed for 6 yr after the first screening program and, therefore, the occurrence of sudden death was investigated consecutively from 1989 to 2002. Of 37,807 subjects, three sudden deaths occurred at high-school age. Each student was followed consecutively without exception and, therefore, the risk of sudden death in this population of high-school students averaged 1.32 in 100,000 per year (3 of 37,807 subjects multiplied by 6 yr). All sudden death subjects were boys with no history of syncopal attacks and no family history of sudden cardiac death. One 14-yr-old boy was identified with HCM during screening and died while jogging at 11:00 a.m.; he was a kendo player, but after screening was disqualified from participating in competitive sports. The remaining two sudden death subjects, aged 13 and 16 yr, also had normal ECG findings and both died during extracurricular athletic activities (handball and basketball, respectively) in the afternoon (Table 2). No autopsies were conducted in the latter two cases. No low-risk subjects died suddenly during the study period. The prevalence of sudden death among high-risk subjects (1 of 9) was significantly higher than that among low-risk and normal subjects (2 of 37,798) (P= 0.0007).
In Kagoshima, the cost of primary screening at the 7th- and 10th-grade levels is $16.7 per examination. The cost of screening 37,807 7th and 10th graders during our first screening was $1.26 million. The cost of secondary screenings, which can include a chest x-ray, echocardiogram, and exercise stress ECG, is, on average, $156 at the 7th- and 10th-grade levels. In this study, secondary screening of 1876 students cost $290,000. The total cost to screen 37,807 students from 1989 to 2000 was therefore $1.55 million. Nine high-risk subjects were identified in the present study, one of whom died. According to Fuller's method (6), we assumed that 10% of the remaining eight high-risk students would live an additional 40 yr, or a total of 32 yr for the group, whereas 90% of the remaining eight students would live an additional 20 yr, or a total of 144 yr for the group. Thus, the total life gained from the screening program was 176 yr, giving an estimated cost of $8800 per year of life saved by the screening program. This was lower than that reported by Fuller ($44,000–$200,000).
The sudden cardiac death screening system implemented in Kagoshima, southern Japan, was shown to be useful in identifying high-risk subjects among both athletes and nonathletes and, therefore, its cost-effectiveness was considered appropriate. Preparticipation cardiovascular evaluation of young athletes by 12-lead ECG is also warranted on the basis of available evidence showing that athletes affected by HCM are successfully identified and athletic field fatalities reduced (3).
Hypertrophic cardiomyopathy is a major cause of sudden death among competitive athletes (3,4,21), and sudden death associated with exercise is a common cause of death among patients with HCM. An appropriate screening system might therefore help reduce mortality through detection of HCM in athletes and subsequent disqualification from competitive sports (1). Sudden death, however, can also occur during sedentary periods or periods of mild activity (3,9,18), and in this study, of the five HCM subjects identified, three were nonathletes. This screening system might therefore help prevent nonathletes from participating in competitive exercise in the future. Screening also identified symptomatic subjects with arrhythmias and, therefore, at risk of sudden cardiac death; for example, those with long QT syndrome and Wolff–Parkinson–White syndrome. These findings, therefore, clearly indicate the importance of identifying high-risk subjects in both athlete and nonathlete populations to decrease sudden cardiac death in young subjects.
In the United States, the estimated cost per year of life saved by taking specific cardiovascular histories into account and conducting physical examinations during primary screening is $84,000; a 12-lead ECG screening costs $44,000 and screening by two-dimensional echocardiography costs $200,000 (6). In the present study, the estimated cost per year of life saved was found to be $8800. Fuller found that ECG screening produced abnormal responses in 15.7% of cases, necessitating further evaluation by echocardiography and resulting in a secondary screening cost of $365 per person (6). In the present study, the rate of abnormal findings during primary screening was <3% and the cost of secondary screening was $156 per person. According to the Organization for Economic Development (OECD), health spending per capita is $1796 in Japan and $4358 in the United States (26). Allowing for this, the cost of further evaluation in Japan, fixed by national health insurance, is inexpensive compared with the United States. Accurate comparisons between these two countries are difficult, however, because of the different adopted health systems. Nevertheless, the greater cost-effectiveness observed here might have resulted from the use of ECG and questionnaires during the first screenings.
A number of limitations with the present study exist that need addressed. First, is the small number of patients with HCM who were screened. We identified five patients with HCM among 37,807 subjects (0.013%); however, in the United States, HCM was shown to be present in approximately 0.17% of young adults screened by two-dimensional echocardiography, and the prevalence is higher among black subjects (0.24%) than in white subjects (0.10%) (16). In this study, the prevalence of HCM among adolescents in the second screening was very low, possibly because of the lower prevalence of HCM in the Japanese population (23). Another reason might be that screening evaluation of young adolescents aged 12–15 yr is too early, with low sensitivity in detecting cardiovascular diseases related to a risk of sudden death, which usually develop at a later age (12,17,20,27). Other limitations are the absence of autopsies and limitations of ECG screening. Two deaths occurred suddenly in two boys aged 13 and 16 yr with normal ECG findings. Both occurred during high levels of athletic activity, suggesting the possibility of sudden cardiac death; however, autopsies are needed to confirm this. Recently, a prospective population-based study in the Veneto region of Italy reported a sudden death incidence of 2.3 per 100,000 athletes per year from all causes, and 2.1 per 100,000 from cardiovascular diseases (2). Because the two above-mentioned sudden deaths occurred in normal patients during high levels of athletic activity and within a few hours of symptom onset, we suggested the possibility of sudden cardiac death. ECG of these subjects showed a slight right axis deviation, but the limb and precordial leads showed no abnormal findings. Preparticipation screening systems that use ECG might therefore be limited in their potential to detect cardiovascular abnormalities capable of causing sudden death (24). To reduce the limitations of ECG screening, further studies are needed to reduce sudden cardiac death in subjects with normal ECG. Improved on-site resuscitation should also be a critical goal and the promotion of an effective resuscitation program for school personnel, family members, and adolescents is essential.
In conclusion, this study shows that the screening system in Kagoshima is useful for identifying high-risk subjects among both athletes and nonathletes, and its cost-effectiveness, therefore, is considered appropriate. Population-based screening for heart disease in this age range is limited by various factors, however, because the incidence of cardiovascular disease and sudden death is very low in adolescent populations. To analyze the mechanisms of sudden death in adolescents, we therefore need a nationwide registry that includes autopsies for all deadly events.
1. Corrado, D., C. Basso, M. Schiavon, and G. Thiene. Screening for hypertrophic cardiomyopathy inyoung athletes. N. Engl. J. Med.
2. Corrado, D., C. Basso, G. Rizzoli, M. Schiavon, and G. Thinen. Does sports activity enhance the risk of sudden death in adolescents and young adults? J. Am. Coll. Cardiol.
3. Corrado, D., A. Pelliccia, H. H. Bjornstad, et al. Cardiovascular pre-participation screening of young competitive athletes for prevention of sudden death: proposal for a common European protocol. Eur. Heart J.
4. Elliott, P. M., J. Poloniecki, S. Dickie, et al. Sudden death in hypertrophic cardiomyopathy: identification of high risk patients. J. Am. Coll. Cardiol.
5. Feigenbaum, H. Echocardiographic measurements and normal values. In: Echocardiography
, Feigenbaum H. (Ed.). 5th ed. Malvern, PA: Lea & Febiger, 1993, pp. 658–683.
6. Fuller, C. M. Cost effectiveness analysis of screening of high school athletes for risk of sudden cardiac death. Med. Sci. Sports Exerc.
7. Fuller, C. M., C. M. McNulty, D. A. Spring, et al. Prospective screening of 5,615 high school athletes for risk of sudden cardiac death. Med. Sci. Sports Exerc.
8. Glover, D. W., and B. J. Maron. Profile of preparticipation cardiovascular screening for high school athletes. JAMA
9. Maron, B. J., B. R. Chaitman, M. J. Ackerman, et al. Recommendations for physical activity and recreational sports participation for young patients with genetic cardiovascular diseases. Circulation
10. Maron, B. J. Sudden death in young athletes. N. Engl. J. Med.
11. Maron, B. J. Hypertrophic cardiomyopathy. A systematic review. JAMA
12. Maron, B. J., and P. Spirito. Implications of left ventricular remodeling in hypertrophic cardiomyopathy. Am. J. Cardiol.
13. Maron, B. J., T. E. Gohman, and D. Aeppli. Prevalence of sudden cardiac death during competitive sports activities in Minnesota high school athletes. J. Am. Coll. Cardiol.
14. Maron, B. J., J. Shirani, L. C. Poliac, R. Mathenge, W. C. Roberts, and F. O. Mueller. Sudden death in young competitive athletes. Clinical, demographic, and pathological profiles. JAMA
15. Maron, B. J., P. D. Thompson, J. C. Puffer, et al. Cardiovascular preparticipation screening of competitive athletes. Circulation
16. Maron, B. J., J. M. Gardin, and J. M. Flack, et al. Prevalence of hypertrophic cardiomyopathy in a general population of young adults. Echocardiographic analysis of 4111 subjects in the CARDIA Study. Coronary Artery Risk Development in (Young) Adults. Circulation
17. Maron, B. J., J. M. Isner, and W. J. McKenna. Hypertrophic cardiomyopathy, myocarditis and other myopericardial disease, and mitral valve prolapse. In: 26th Bethesda Conference. Recommendations for determining eligibility for competition in athletes with cardiovascular abnormalities. J. Am. Coll. Cardiol.
18. Maron, B. J., J. Kogan, M. A. Proschan, G. M. Hecht, and W. C. Roberts. Circadian variability in the occurrence of sudden cardiac death in patients with hypertrophic cardiomyopathy. J. Am. Coll. Cardiol.
19. Maron, B. J., S. A. Bodison, Y. E. Wesley, E. Tucker, and K. J. Green. Results of screening a large group of intercollegiate competitive athletes for cardiovascular disease. J. Am. Coll. Cardiol.
20. Maron, B. J., P. Spirito, Y. Wesley, and J. Arce. Development and progression of left ventricular hypertrophy in children with hypertrophic cardiomyopathy. N. Engl. J. Med.
21. Maron, B. J., W. C. Roberts, McAllister, H. A. Rosing, and D. R., Epstein, S. E. Sudden death in young athletes. Circulation
22. Maron, B. J., and S. E. Epstein. Hypertrophic cardiomyopathy: a discussion of nomenclature. Am. J. Cardiol.
23. Matsumori, A., Y. Furukawa, K. Hasegawa, et al. Epidemiologic and clinical characteristics of cardiomyopathies in Japan. Results from nationwide surveys. Circ. J.
24. Pfister, G. C., J. C. Puffer, and B. J. Maron. Preparticipation cardiovascular screening for US collegiate student-athletes. JAMA
25. Priori, S. G., E. Aliot, C. Blomstrom-Lundqvist, et al. Task force on sudden cardiac death of the European society of cardiology. Eur. Heart J.
26. Reinhardt, U. E. and P. S. Hussey. Cross-national comparisons of health systems using OECD data, 1999. Health Aff.
27. Spirito, P., and B. J. Maron. Absence of progression of left ventricular hypertrophy in adult patients with hypertrophic cardiomyopathy. J. Am. Coll. Cardiol.
28. Zipes, D. P., and H. J. Wellens. Sudden cardiac death. Circulation
This article has been cited 28 time(s).
Journal of the American College of CardiologyThe True Incremental Cost of ECG Screening: The Price Is Not Right, But the Cost Appears EffectiveJournal of the American College of Cardiology
Journal of the American College of CardiologyThe True Incremental Cost of ECG Screening: The Price Is Not Right, But the Cost Appears Effective ReplyJournal of the American College of Cardiology
Internal MedicineSudden cardiac death in young athletesInternal Medicine
British Journal of Sports MedicineObservations on the "Lausanne Recommendations'' on sudden cardiovascular death in sportBritish Journal of Sports Medicine
CirculationElectrocardiograms should be included in preparticipation screening of athletesCirculation
Annals of MedicineSudden cardiac death: Prevalence, pathogenesis, and preventionAnnals of Medicine
Current Problems in CardiologyAdding an Electrocardiogram to the Pre-participation Examination in Competitive Athletes: A Systematic ReviewCurrent Problems in Cardiology
Paediatrics & Child Health
Cardiac risk assessment before the use of stimulant medications in children and youth A joint position statement by the Canadian Paediatric Society, the Canadian Cardiovascular Society, and the Canadian Academy of Child and Adolescent Psychiatry
Paediatrics & Child Health, 14(9):
Canadian Journal of Cardiology
Cardiac risk assessment before the use of stimulant medications in children and youth: A joint position statement by the Canadian Paediatric Society, the Canadian Cardiovascular Society, and the Canadian Academy of Child and Adolescent Psychiatry
Canadian Journal of Cardiology, 25():
CirculationShould an electrocardiogram be included in routine preparticipation screening of young athletes?Circulation
CardiologyPhysical examination and ECG screening in relation to echocardiography findings in young healthy adultsCardiology
Archives of Pediatrics & Adolescent Medicine
Heart Disease Screening of Schoolchildren in Taiwan
Archives of Pediatrics & Adolescent Medicine, 163(3):
British Journal of Sports MedicineThe Lausanne recommendations: a Dutch experienceBritish Journal of Sports Medicine
Sudden Death in Young Athletes
Scandinavian Journal of Medicine & Science in SportsDiagnostic value of different electrocardiographic voltage criteria for hypertrophic cardiomyopathy in young peopleScandinavian Journal of Medicine & Science in Sports
Asia-Pacific Journal of Public HealthThe Economic Evaluation of a Health Screening Program on Congenital Heart Disease for School Children in Taichung, TaiwanAsia-Pacific Journal of Public Health
CirculationRecommendations and considerations related to preparticipation screening for cardiovascular abnormalities in competitive athletes: 2007 update a scientific statement from the American Heart Association Council on Nutrition, Physical Activity, and Metabolism - Endorsed by the American College of Cardiology FoundationCirculation
Scandinavian Journal of Medicine & Science in SportsSudden cardiac death in young athletes: time for a Nordic approach in screening?Scandinavian Journal of Medicine & Science in Sports
Journal of the American College of CardiologyPre-Participation Screening of Young Competitive Athletes for Prevention of Sudden Cardiac DeathJournal of the American College of Cardiology
Pace-Pacing and Clinical Electrophysiology
The Role of ECG Screening in the Evaluation of Risk of Sudden Cardiac Arrest in the Young
Pace-Pacing and Clinical Electrophysiology, 32():
Circulation-Arrhythmia and ElectrophysiologySudden Death in the Young What Do We Know About It and How to Prevent?Circulation-Arrhythmia and Electrophysiology
Netherlands Heart Journal
Periodical cardiovascular screening is mandatory for elite athletes - Pro screening
Netherlands Heart Journal, 15(6):
Arquivos Brasileiros De Cardiologia
Competitive Sports for Children and Adolescents: Should an Electrocardiogram Be Required in the Pre-Participation Physical Examination?
Arquivos Brasileiros De Cardiologia, 93(2):
British Journal of Sports MedicineReturn to play? Practical considerations for young athletes with cardiovascular diseaseBritish Journal of Sports Medicine
Annals of Internal Medicine
Cost-Effectiveness of Preparticipation Screening for Prevention of Sudden Cardiac Death in Young Athletes
Annals of Internal Medicine, 152(5):
Journal of Interventional Cardiac ElectrophysiologyPrediction and prevention of sudden death in young populations: the role of ECG screeningJournal of Interventional Cardiac Electrophysiology
Clinical Journal of Sport MedicineHeat Injury Prevention Practices in High School FootballClinical Journal of Sport Medicine
Medicine & Science in Sports & ExerciseElectrocardiograms in Athletes: Interpretation and Diagnostic AccuracyMedicine & Science in Sports & Exercise
SUDDEN DEATH; NONATHLETE; HYPERTROPHIC CARDIOMYOPATHY; PREPARTICIPATION
©2006The American College of Sports Medicine
Highlight selected keywords in the article text.