Practical screening priorities for hearing impairment among children in developing countries.

Reviews/Analyses Practical screening priorities for hearing impairment among children in developing countries F.M. Gell,1 E. McC. White,2 K. Newell,3 1. Mackenzie,4 A. Smith,5 S. Thompson,6 & J. Hatcher7 Routine screening for hearing impairment in childhood is now widespread in industrial countries, although there is considerable controversy over the most efficient techniques and procedures. In most developing countries, however, routine screening programmes for hearing impairment do not currently exist. The problems involved in implementing screening programmes in developing and industrial countries are very different, and in selecting screening procedures for a particular population the fol- lowing factors have to be taken into consideration: the environmental test conditions; the availability of resources for equipment and the training of testers; the local attitudes towards disability; the level of hearing impairment that may cause handicaps; and the major types of pathology causing hear- ing impairment. We suggest that in developing countries children should be screened at school entry using a simple field audiometer and that the external ear be inspected for the presence of a discharge. There is an urgent need to develop reliable and simple screening procedures for infants and young children; where possible, all children should be screened for severe or significant hearing impairment before the age of 2 years. No screening should, however, be implemented until appropriate follow-up services are available. Introduction Childhood hearing impairment,a even when mild, may have a detrimental effect upon linguistic and I Anthropologist, Acci6n Rural Fe y Alegria, Casilla 704, Sucre, Bolivia. 2 Epidemiologist, Newcastle upon Tyne, England. 3 Deceased. Formerly Professor of Tropical Community Health, Liverpool School of Tropical Medicine, Liverpool, England. 4 Neuro-otologist, Walton Hospital, Liverpool, England. 5 Convenor, Hearing Impairment Research Group, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, England. Requests for reprints should be sent to this author. 6 Research Assistant, Hearing Impairment Research Group, Liverpool School of Tropical Medicine, Liverpool, England. 7 Biostatistician, Liverpool School of Tropical Medicine, Liver- pool, England. a The term "hearing impairment" is used here to encompass the entire continuum from a very mild impairment to profound deaf- ness. Reprint No. 5327 educational development, which can result in social and psychological problems for affected children and their families. The effects are most severe when there is a profound sensorineural hearing defect present from birth or early life, but even less severe conductive impairments that develop later in chilhood can slow down the child's linguistic and educational development (1). Poor educational per- formance has been observed in children with mild and transient conductive impairments, which are frequently associated with recurrent middle ear infection (2, 3). Since prompt appropriate therapy and supportive services can prevent many of the damaging effects of early hearing impairment, timely detection is essential to alert both parents and health care professionals to the need for appropriate action. Routine screening of infants and preschool and school-age children is now widespread in industrial countries, and a considerable range of methods has been developed to detect hearing impairment in chil- dren and to assess its extent. If a possible hearing impairment is detected in this way, otoscopy and a Bulletin of the World Health Organization, 70 (5): 645-655 (1992) © World Health Organization 1992 645 F.M. Gell et al. full audiometric examination normally follow, and referral may then be made to an appropriate special- ist. In developing countries the demands on health resources can be such that the detection of children with hearing impairments takes low priority (4), even though the prevalence of ear pathology and hearing impairment is higher than in industrial countries. Also, audiological services may be limited or nonexistent, and equipment for the routine screening of hearing impairment is rarely available. Informal enquiries suggest that where audiometers are pro- vided, they tend to be expensive nonportable diag- nostic instruments, which are rarely recalibrated. Markides has pointed out that these are often second- hand instruments donated by industrial countries which are eager to dispose of out-of-date equipment in a "useful" way (5). Breakages are common, spare parts rarely provided, and since local maintenance and repair services are seldom available the costs in the long run can be very high. In addition, the equip- ment may be unsuitable for use in hot, humid, and dusty conditions, where air-conditioned testing areas may not be available. In developing countries, audiometers are usually located in urban clinics or hospitals; rural clinics and school medical services are almost never equipped to screen for hearing impairment (4); and staff are rare- ly trained to use the instruments. Where appropriate- ly skilled personnel and quiet testing areas are avail- able, acceptable assessment and diagnostic services can be provided using relatively simple equipment: battery-powered audiometers and tympanometers can meet most of the criteria of referral. Such assess- ment, with diagnostic facilities for referral, and ser- vices for those found to have hearing difficulties, should clearly be available before the establishment of a screening programme is considered. While the facilities for referral in developing countries may be similar to those used in industrial countries, it is more difficult for developing coun- tries to adopt comparable screening procedures. Altemative methods are needed to accommodate the possibility of poor environmental test conditions, dif- ficulties in maintaining equipment, minimal training of testers, the difficulty in reaching all members of the community, and poor compliance because many parents attach a low importance to hearing impair- ments. Ideally, the planning and development of an appropriate audiology service should be preceded by epidemiological studies to determine the prevalence of hearing impairment. Such studies are feasible only when appropriate tools for the identification and assessment of hearing impairment are available, and if some referral and rehabilitative services can be provided for children identified to have such an impairment. This article first reviews the methods employed to screen for hearing impairment in industrial coun- tries. Subsequently, some of the problems that are likely to be encountered if these methods are used in developing countries are identified, and possible methods of screening for hearing impairment that might be appropriate and realistic in such settings are outlined. Standardization of techniques and classifications Despite the widespread use of screening programmes to detect hearing impairment in industrial countries, there is no standard intemationally accepted pro- cedure for identification of the condition, nor a standard criterion for "failure" and referral. Even within a country standardization of procedures is rare, and a survey of screening procedures in each state of the USA revealed "a chaotic myriad of standards, regulations, guidelines, techniques, and recom- mendations" (6). Classification of hearing impairment also varies widely, making comparisons of prevalences both within and between countries extremely difficult. In 1986, WHO prepared an intemational classification of hearing impairment, based on bilateral hearing performance tests rather than audiometric measure- ments:b comparable audiometric values were also given. Some modifications were made to this classi- fication in 1991 (see Table l).c We recommend that all countries without access to audiometric tech- nology should adopt a similar performance-based system of classification. Screening procedures for infants and preschool children Screening methods Many congenital hearing impairments are not identi- fied, even in industrial countries, until a relatively late stage. For example, a study carried out in the United Kingdom in 1977, of children bom in 1969, found that the average age when severe congenital b Prevention of deafness and hearing impairment: report of the Director-General. WHO unpublished document A39/14 (annex to document EB 79/10) c Report of the Informal Working Group on Prevention of Deaf- ness and Hearing Impairment Programme Planning. Unpublish- ed document WHO/PDH/91.1. 646 WHO Bulletin OMS. Vol 70 1992 Screening children for hearing impairments in developing countries Table 1: Grades of hearing impairmenta Grade of impairment Performanceb Recommendations Comments 0 (.25 dB)C No (or very slight) 20 dB also recommended. No impairment hearing problems. People with 15-20 dB levels Able to hear may experience hearing whispers problems. Those with unilat- eral hearing losses may experience problems even if the better ear is normal 1 (26-40 dB) Able to hear and Counselling. Some difficulty in hearing, Slight repeat words Hearing-aids may but can usually hear normal impairment spoken in normal be needed level of conversation voice at 1 metre 2 (41-60 dB) Able to hear and Hearing-aids Moderate repeat words using usually impairment raised voice at 1 metre recommended 3 (61-80 dB) Able to hear some Hearing-aids needed. Discrepancies between Severe words when shouted If no hearing aids pure tone thresholds and impairment into better ear available, lip-reading speech discrimination and use of sign score should be noted language should be taught 4 (2 81 dB) Unable to hear and Hearing aids may Spoken speech distorted, the Profound understand even a help in understanding degree depending on the impairment shouted voice words. Additional age at which hearing was including rehabilitation needed. lost deafness Lip-reading and sometimes use of sign language essential a Modified from: Report of the Informal Working Group on Prevention of Deafness and Hearing Impairment Programme Planning. Unpublished document WHO/PDH/91.1. b Performance tests for speech discrimination levels should be carried out using hearing alone, i.e., without any visual clues. c Figures in parentheses are the corresponding ISO audiometric values for the better ear, i.e., the averages of the results at 500, 1000, and 2000 Hz. Speech discrimination against a background of noise requires good high-frequency hearing. It is therefore recom-i mended that epidemiological studies include testing and reporting of the hearing threshold at 4000 Hz. hearing defects were detected was 2.5-3 years (7). Also, a study carried out in 1982 reported that about 40% of hearing-impaired children in the United Kingdom were not detected at their first screening test (8). This situation is common in industrial coun- tries and is likely to be worse in developing coun- tries, where programmes for the early detection of hearing impairment are rare. While programmes to screen the hearing of entire infant populations have met with limited suc- cess and enthusiasm (9), the importance of identify- ing those infants who are likely to develop defects as a result of exposure to perinatal or other hazards has been extensively documented (10-12). In 1982 the American Joint Committee on Infant Hearing suggested the following seven-item check-list to identify such at-risk children (13): a family history of childhood hearing impairment; congenital or peri- natal infection; anatomic malformations involving the head or neck; birth weight <1500 g; hyper- bilirubinaemia above the level which indicates the need for exchange transfusion; bacterial meningitis, especially with Haemophilus influenzae; and severe asphyxia. The value of such a register has been questioned, since 100% follow-up is impossible to achieve and a significant proportion of children with impaired hearing are not identifiable as being at risk (8, 14). However, there is general agreement that these registers remain a useful and basic guide for identification of some of the infants most at risk of hearing impairment (15). In industrial countries, it is recommended that routine screening be carried out in the first year of life, normally around 8 months of age, and again at 2.5-3 years of age, when the development of speech and articulation can also be reviewed (7). Behav- ioural response screening is used for both these tests. Children can also be examined between routine screening tests if there is any suspicion of hearing impairment (16, 17). WHO Bulletin OMS. Vol 70 1992 647 F.M. Gell et al. Many attempts have been made to develop behavioural methods of screening for hearing impair- ment in preschool children, but it is difficult to develop a satisfactory test that adheres to the basic principles of screening, particularly that the method should be simple, low-cost, and suitable for the rapid testing of large populations. A test that requires a voluntary response from a preschool child can be neither rapid nor simple, and is particularly difficult with 2-3-year-olds. The effectiveness of the tests depends largely on the skill of the tester in assess- ing a child's developmental progress, in selecting and performing an appropriate test, and in correctly interpreting response behaviour in terms of hearing impairment. The first-year screening test for hearing impairment includes testing the infant's hearing using the distraction technique and questioning the mother about the developmental status, communica- tion abilities, and hearing behaviour of her infant. It is important that parental suspicions of hearing prob- lems should never be disregarded, since they are valid indicators of the presence of hearing disorders (18). The distraction test involves the production of an orienting response to "meaningful sounds" of dif- ferent frequency ranges immediately after the child's attention is distracted by a play activity. The test sti- muli recommended at known decibel levels are the hum (<500 Hz), the consonant "s" (about 4000 Hz) and, if available, a high frequency rattle (>6000 Hz). The responses of the test subject are compared with those expected for a normally developed child of the same age using an auditory behaviour index (19, 20). Although the current procedure for carrying out distraction testing appears to be fairly simple, it is beset with difficulties and potential hazards. Con- siderable skill and experience are required to select the appropriate test procedure for the infant's devel- opmental age group, since general development progresses rapidly during the first year of life (20). If inaccurate stimulus intensity and frequency levels are used, distraction techniques are poor, while if clues are given about the source and timing of the stimulus, high false-negative rates can occur. McCormick has suggested that insufficient training of health visitors, poor test technique, and the lack of sound-level meters may account for the low levels of detection of hearing impairment among young children in the United Kingdom (17). In developing countries, staff training difficul- ties are likely to be considerable (21). Provision of equipment may be inadequate; quiet, distraction-free testing environments may be difficult to find; and health services that are overburdened and under- staffed may not be able to allocate trained and skilled staff for the time necessary to ensure adequate quality of testing. The 2-3-year-old age group is usually screened using a cooperative test, such as the verbal-auditory screening test in which the child is required to point to a picture representing a word that is spoken aloud (22). Children aged 3-4 years or more are screened using a performance test in which the child is trained by play-conditioning to manipulate objects in response to either speech or pure-tone stimuli. Most behavioural tests are culture-specific in that they use stimulus sounds, test materials, types of response behaviour, and procedures that are familiar and appropriate to the culture in which they were developed. Cross-cultural studies on cognition have shown that test performance depends greatly on culture, level of formal education, and familiarity of test materials (23-25). In industrial countries the sounds used in the distraction test have included a spoon scraping a china cup, a sound that would be un- familiar and therefore not "meaningful" to cultures that use gourds or calabashes as drinking utensils. Picture-pointing tests are only appropriate for children who are familiar with pictorial representa- tions; for those who have had no access to books and pictures, real and familiar objects should be used instead. At a more fundamental level, the concept of "testing" and of conditioning a behavioural response may be quite alien to some societies. The development of behavioural methods of testing hearing that are appropriate to the needs of different countries is urgently needed. While such tests are time-consuming and require an experienced, thoroughly trained and sensitive tester, they have the considerable advantage that they use inexpens- ive equipment. Community-based screening programmes Routine testing of hearing should ideally form part of general developmental screening procedures. In industrial countries, the health visitors who screen for hearing impairment in the home are normally nurses trained in developmental paediatrics; how- ever, in developing countries, health workers with sufficient training and experience to perform devel- opmental screening tests are likely to be employed in hospitals and clinics rather than in the community. Their expertise is, nevertheless, needed to test children with a suspected hearing impairment or those in a "high-risk" group who can be brought to a clinic or hospital. The routine screening of children's hearing in the community is therefore undertaken by the primary health care workers, who often receive only minimal training and have a multitude of other duties to perform. Helander et al. have produced a manual on com- munity-based rehabilitation for use by primary health WHO Bulletin OMS. Vol 70 1992648 Screening children for hearing impairments in developing countries care workers, which describes a procedure of screen- ing for hearing impairment using distraction and speech tests (26). This is the first real attempt to develop screening methods for use in developing countries, but the methods advocated are crude and involve the use of nonfrequency-specific home-made rattles and hand-clapping. Weiner has produced a similar manual with a smaller number of more accu- rate but simpler tests that can be used by families themselves (27); however, this approach indicates that distraction testing can be performed by one per- son when, for the best results, teamwork is essential. WHO recommends that primary health care workers should be trained to perform these screening tests routinely on all children in the community.d The first screening test should be carried out on chil- dren before 2 years of age and the second just before school age. Children with discharging ears or with the clinical signs and symptoms of hearing problems (e.g., earache, vertigo, or tinnitus) and infants or children considered to be "at-risk" (e.g., where a parent or sibling is deaf, or following certain severe infections) should also be tested. In most developing countries the number of home births is high and "at-risk" children are seldom identified soon after birth. In addition, it is rarely possible for infant development to be monitored by skilled health professionals. Parents and primary health care workers must therefore be responsible for identifying, as early as possible, children with impaired hearing. McCormick's simple checklist of behavioural indicators of hearing impairment in the first year of life (17) is a useful guide for use by both parents and health visitors in the United Kingdom and (apart from reference to the noise of a vacuum cleaner, which could be easily modified to a more appropriate sound) could also be of value in developing countries (see Table 2). It is stressed that routine screening of the hearing of infants and young children should not be implemented unless a referral programme and realistic and appropriate methods of assessment are available, and until adequate follow-up care and support can be provided. Children who fail the first screening test, infants who are considered to be "at-risk", and any child with a suspected hearing impairment should be referred to an appropriate health centre for examination and further assessment. Further assessment of infants and young children In the assessment of hearing impairment in infants and young children in industrial countries, distrac- d See footnote b, p. 646. tion, cooperative, and performance tests are used in much the same way as screening tests, but the threshold of hearing is sought rather than a simple pass/fail result. Children aged 2-3 years are often assessed using play audiometry. Because of the complexities of testing this age group, it is desirable that diagnostic audiometric tests for preschool children should be carried out only by trained audi- ologists or audiological technicians; also, quiet test- ing environments are essential. Currently there is no simple, accurate and low- cost alternative to these tests for the assessment of hearing impairment in infants and young children. Although McCormick has shown that rigorous train- ing of health visitors in the use of the distraction test can considerably increase the number of hearing- impaired infants identified (17), such training may not be available in all countries. WHO has stressed the urgent need for research into the development of more accurate and reliable techniques for use in both industrial and developing countries as no single method of testing infant hearing has proved to be ideal (28, 29).e In industrialized countries various techniques of evoked response audiometry, such as the auditory brain stem response (30, 31) and other complex tests are used to test at-risk and difficult-to-test children. Much controversy surrounds the use of such tests as screening techniques because of their technical com- plexity, the requirement for expensive equipment, and the need for highly trained clinicians. Even as a diagnostic tool, their cost-effectiveness in the United Kingdom has been questioned (32). For most developing countries such high technol- ogy equipment would be inappropriate and financial- ly inaccessible. Scarce resources would be better spent on screening and assessing older infants and school-age children for whom treatment and rehabili- tation would be easier to provide. Screening procedures for school- age children Even where preschool screening for hearing impair- ment is performed, routine screening of all school- age children is advisable. In addition, those children with the most profound hearing impairments are unlikely to attend school, and hence attempts should be made to reach all children in the community (33). Children with a less severe hearing impairment that has existed from birth or the first few years of life are often not identified prior to school entry, particu- larly if the impairment is progressive. e See footnote b, p. 646. WHO Bulletin OMS. Vol 70 1992 649 F.M. Gell et al. Table 2: Can your baby hear you?a Here is a checklist of some of the general signs you can look for in your baby's first year: Shortly after birth Your baby should be startled by a sudden loud noise such as a hand clap or a door slamming and should blink or open his eyes widely to such sounds. By 1 month Your baby should be beginning to notice sudden prolonged sounds like the noise of a vacuum cleaner and he should pause and listen to them when they begin. By 4 months Please tick if response present li He should quieten or smile to the sound of your voice even when he cannot see I you. He may also turn his head or eyes towards you if you come up from behind and speak to him from the side. By 7 months He should turn immediately to your voice across the room or to very quiet noises made on each side if he is not too occupied with other things. By 9 months He should listen attentively to familiar everyday sounds and search for very quiet sounds made out of sight. He should also show pleasure in babbling loudly and tunefully. By 12 months He should show some response to his own name and to other familiar words. He may also respond when you say 'no' and 'bye bye' even when he cannot see any accompanying gesture. Your health visitor will perform a routine hearing screening test on your baby between seven and nine months of age and will be able to help and advise you at any time before or after this test if you are concerned about your baby and his development. If you suspect that your baby is not hearing normally, either because you cannot answer yes to the items above or for some other reason, then seek advice from your health visitor. a Reprinted with permission from: McCormick, B. Hearing screening by health visitors: a critical appraisal of the distraction test. Health visitor, 56: 449-451 (1983). Otitis media in its various forms is common in children under 7 years of age and may cause both transient and permanent hearing impairment. Chron- ic suppurative otitis media (CSOM) is often ignored by parents, although high prevalences of this condi- tion have frequently been reported among popula- tions of American Indians, Aborigines, and Maoris (34, 35), as well as in developing countries (36-38). This contrasts with the situation in industrial coun- tries, where noninfective, allergic and acute infective diseases of the middle ear are the predominant causes of aural pathology (39). Children with chronic otitis media may either be unable to express the nature of the problem or be unaware of the associa- ted mild-to-moderate hearing impairment. Conse- quently their condition often goes unnoticed by both health care professionals and parents. Children who have bilateral chronic otitis media may suffer 1-2 years' educational retardation with significant delays in speech and language acquisi- tion, even though their hearing impairments are rare- ly >45 dB (40). Paradise has concluded that early episodes of otitis media cause developmental impair- ments only if residual hearing impairment results (41), but others workers have documented language learning problems caused by even transient hearing impairments at an early age (42, 43). There is evi- dence to indicate that a mild hearing impairment (10-15 dB) may be sufficient to impair the acquisi- tion of language skills by young children and may lead to educational retardation (44). In industrial countries, it is now considered important to identify such mild and transient hearing impairments as early as possible. There is considerable controversy in industrial countries over the most appropriate and cost- effective methods of screening school-age children for hearing impairment. Screening audiometry, using either pure-tone or speech stimuli, is the most widely used method of identifying individuals with a possible hearing impairment, but the screening level generally used (around 25 dB HTL) does not reliably identify the mild hearing impairments associated with serous otitis media. Over the last 20 years WHO Bulletin OMS. Vol 70 1992650 Screening children for hearing impairments in developing countries increasing concern about the effects of middle ear disease, particularly in North America, has focused interest on the detection of fluid in the middle ear using tympanometry (45-47). Tympanometry Tympanometry is a quick, objective, and reliable means of identifying the patency of the tympanic membrane or the presence of fluid in the middle ear; it requires minimal cooperation from the child. However, tympanometry may result in over-referrals, and consequently unnecessary follow-up costs (48), while the currently available equipment is neither rugged nor low cost, and the results can be difficult to interpret. The technique has been used as a screening pro- cedure for populations with a high prevalence of serous otitis media (otitis media with effusion). Where CSOM is a greater problem than serous otitis media, as appears to be the case in many developing countries, accurate tympanometry may not be pos- sible since it can be difficult to obtain a good seal between the ear canal and the measuring probe if the ear is full of discharge. CSOM can best be diagnosed by examining for the presence of a purulent dis- charge, using otoscopy to identify a perforation of the tympanic membrane. We suggest that for populations where CSOM is likely be a major problem, inspection of the ear for a purulent discharge is the most effective form of screening. Where the priority is the identification of moderate-to-severe impairments, pure-tone audi- ometry is undoubtedly the most suitable screening tool. Audiometric screening The most commonly used form of audiometric screening for school-age children in industrial coun- tries is pure-tone air conduction audiometry. The American Speech and Hearing Association recom- mend use of a limited frequency test, manually administered to individuals, as a rapid and efficient means of screening children as young as 3 years of age (49). Simple, portable, pure-tone screening audiometers are generally used for this purpose. Opinion differs widely about which frequencies should be used in audiometric screening. The most commonly used are probably 500, 1000, 2000, 4000, and 6000 Hz. Although 500 Hz is an important speech frequency, it is often omitted since it is easily masked by ambient noise, while 6000 Hz tends to produce a high false-positive rate (50). Ballantyne & Martin recommend that 500, 1000, 2000, and 4000 Hz are probably enough for testing primary schoolchildren (51). The pass/fail criterion for screening must be selected with regard to both availability of treatment and rehabilitation services, as well as to the degree of impairment considered to be a handicap. Consider- ation should be given to whether the objective is to identify only "disabling" impairment or all impair- ment over a specified intensity and frequency range, and whether the criterion for failure is based on the audiometrically better or worse ear. Haggard et al. state that in the predominantly sensorineural impair- ments of adults the disability is generally determined by the better ear, while speech testing indicates that overall disability in children may be more closely related to the situation in the worse ear (52). A range of intensity levels are currently used for screening in industrial countries; most commonly, 20 dB HTL at 1000 Hz and 2000 Hz, and 25 dB HTL at 4000 Hz to compensate for the 4000 Hz "dip" that can be caused by noise-induced hearing impairment. Hearing impairment of this extent may be of little practical significance in a developing country, and for many individuals who are identified treatment may not be available (3). Screening at 30 dB HTL may be more appropriate for practical reasons because of the higher levels of background noise and because children with this level of hearing im- pairment need special attention at school, even if they do not necessarily require a hearing aid. Alterna- tively, screening at 50 dB HTL would exclude such children, and the smallest hearing impairments detected at this level would indicate the need for a hearing aid in the absence of chronic discharge (D. N. Brooks, personal communication, 1985). Knight's review of currently available speech testing materials in non-European languages cites work on the development of such materials in Arabic, Cantonese, and several Indian and African languages (53). However, while speech audiometry has been used in chilhood screening programmes (44), it is not generally employed for this purpose because of its complexity and the need for highly skil- led personnel. The variation in the overall intensity of spoken words and their frequency-specific inten- sity patterns make them less easy to calibrate than pure-tone stimuli (52). Furthermore, since all speech frequencies give overtones of several frequencies, a child with a sharp high-frequency "dip" may not be identified if screened with speech stimuli (54). Speech audiometry is therefore more appropriate and useful for assessing the extent of hearing disability, particularly with preschool children who will not respond to pure tones. Combined screening methods Various combined screening methods have been sug- gested, particularly for situations where a high level WHO Bulletin OMS. Vol 70 1992 651 F.M. Gell et al. of ambient noise precludes the use of low frequency tones in audiometry. Suggestions have been made about combining tympanometry with a pure-tone test using five frequencies (55), with a single high fre- quency tone (56), or with a single warble tone at 3000 Hz (57). However, use of a combination of techniques increases the complexity and time taken to perform the tests. Other methods which have been combined with audiometry, such as otoscopy or tuning-fork testing, require the skill and insight of an experienced exami- ner and these methods may not be effective for mass screening by less experienced individuals (58). WHO recommends a minimum of 3 years' formal training in otoscopy for community health workers and health professionals! Screening audiometry in developing countries In most developing countries, routine audiometric screening is not performed in schools since staff and the appropriate equipment are not available. Clinics in major towns may be equipped with expensive, complicated, diagnostic audiometers, but it may be difficult to obtain spare parts or to service them, they require highly trained operators, and they are in any case too bulky and delicate to be transported to schools. Screening surveys in developing countries have therefore generally used simpler, portable screening audiometers (59, 60); nevertheless, even these instruments are relatively costly, and in field studies in India and Swaziland we have found them unsuitable for use in rough terrain if much travelling is involved. Simple, low-cost, robust screening audiometers are needed for routine use in schools, clinics and field surveys, and in 1985 WHO requested that research be carried out into developing such an audiometer. In response, the Liverpool Field Audio- meter was produced.9 This is a hand-held, battery- powered field audiometer with a single earphone that emits warble tones of 500, 1000, 2000, and 4000 Hz at intensities of 30, 50, or 80 dB HTL (Fig. 1). It has been used successfully in the field in the United Kingdom, India, and Swaziland (Fig. 2).h The results of screening tests using the field audiometer f See footnote b, p. 646. 9 The Liverpool Field Audiometer was designed and manufac- tured by MEG Instrumentation, Sheffield, England, to a specifi- cation developed in conjunction with the Liverpool Hearing Impairment Research Group. h Hearing Impairment Research Project. The national sample study of the prevalence of hearing impairment in children in Swaziland, 1987: full report. University of Liverpool, 1988. Fig. 1. The Liverpool Field Audiometer. '.,: :..:... :; ......::.i:.. .i:. :.: ..::.:.: :-:: :. .,....... ',''..*: :: ...:... : ...- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.::.i:.:i:i.::':!:.::.::..:...:: ................... ...... .. ..... :10. ko* .ii i- E E | | ~~~~~~~iiL _ Fig. 2. Testing a child's hearing in Swaziland using the Liverpool Field Audiometer. agree well with those obtained using full audiometric testing in trials with children in the Nottingham Children's Hearing Assessment Centre.' Hoare, N.W. et al. Acoustical and operational characteristics of a miniature hand-held screening audiometer. Unpublished results. 652 WHO Bulletin OMS. Vol 70 1992 Screening children for hearing impairments in developing countries The field audiometer is suitable for use not only as a screening tool in developing countries but can also perform several useful functions in industrial countries. For example, it could be useful in supple- menting impedance testing with a high-frequency tone, in the testing of difficult children who cannot be examined using conventional headphones, and for use by non-specialists, such as general practitioners, speech therapists or teachers, where a simple check of hearing is needed to justify referral to an audiolo- gist or otologist. Further assessment of school-age children School-age children who fail a screening test should be referred for otoscopic examination and a full audiometric assessment to determine the type and extent of their hearing impairment. In developing countries such referrals are normally made to hospi- tals or special hearing clinics. In district hospitals, there should ideally be small sound-proofed booths for audiological assessment, but with ingenuity a satisfactorily quiet area can be found in most hospi- tals. Conclusions If intemationally standardized procedures for the identification and classification of hearing impair- ment are to be developed, an exchange of ideas and cooperation between countries is needed. Agreement on such procedures would facilitate both the plan- ning and development of audiological services and comparisons of the prevalence of hearing impairment in different populations. Routine screening programmes should only be implemented where there are adequate referral services. These services include diagnosis and assessment, production and fitting of ear moulds and hearing-aids, with follow-up and maintenance ser- vices and rehabilitation programmes, including educa- tional support. In such situations, routine screening for hearing impairment should be performed at school entry and, where possible, before the age of 2 years. If it is not feasible to screen whole popula- tions of young children, attempts should be made to screen those children considered to be "at-risk" of hearing impairment. The problems that arise in the behavioural test- ing of infants and young children in industrial coun- tries are likely to be more pronounced in developing countries, where fewer resources for equipment and trained personnel are available, and where environ- mental testing conditions may be less favourable. Also, the traditionally used procedures for the behavioural screening of hearing impairment have been developed in industrial countries and may be inappropriate for some non-Westem cultures. Re- search into the development of simple, low-cost and reliable techniques for the early identification of hearing impairment is urgently needed. Chronic suppurative otitis media appears to be the most common cause of mild-to-moderate hearing impairment in children in developing countries. Tympanometry is not the most effective technique of identifying this condition, which is best diagnosed by the presence of a purulent discharge, using oto- scopy if possible. For the detection of moderate-to- severe hearing impairment, audiometry is still the most suitable screening procedure: a simple, robust and low-cost field audiometer has recently been developed for this purpose. Acknowledgement This review was made possible by grants from the Gatsby Charitable Foundation (1985-88 and 1989-92). Resume Priorit6s pratiques pour le depistage des deficits auditifs chez l'enfant dans les pays en developpement Un deficit auditif chez l'enfant peut compromettre son developpement linguistique et intellectuel et etre a l'origine de problemes sociaux et psycholo- giques. Une detection precoce est donc essentiel- le pour avertir les parents et les professionnels de la sante de la n6cessit6 d'instituer un traitement et de mettre en place des services de soutien capables d'6viter bien des effets nWfastes d'un tel d6ficit. Des services d'orientation ad6quats devraient etre etablis avant la mise en place d'un programme de depistage, ces services pouvant etre analogues a ceux qui existent dans les pays industrialis6s; toutefois, d'autres m6thodes de depistage devront etre adoptees dans les pays en developpement en raison des circonstances et des conditions propres a ces pays. Dans les pays industrialises, le depistage des deficits auditifs chez les nourrissons et les enfants d'age prescolaire s'appuie sur 1'exploitation de registres d'enfants "a risque" et sur les methodes suivantes: 6preuve de distraction pour les bebes de moins d'un an; audiometrie vocale pour le groupe des 2 a 3 ans; et tests de performance entre 3 et 4 ans. Toutes ces methodes deman- dent du temps et un personnel experiment6 et competent. La mise au point de tests adaptes aux WHO Bulletin OMS. Vol 70 1992 653 F.M. Gell et al. diff6rentes cultures et aux diff6rents environne- ments des pays en d6veloppement doit etre une priorit6. Ces tests doivent pouvoir etre utilises par les parents et les agents de soins de sante pri- maires. Un depistage syst6matique chez tous les enfants d'age scolaire est 6galement souhaitable. Dans les pays industrialises, on utilise a cette fin les techniques suivantes, individuellement ou en association: audiom6trie tonale en conduction aerienne; audiom6trie vocale; tympanometrie; emploi du diapason. Etant donne que l'otite moyenne chronique suppur6e est beaucoup plus fr6quente dans les pays en d6veloppement que l'otite moyenne s6reuse, la tympanom6trie n'est pas particulierement indiqu6e et il est pr6f6rable de rechercher la pr6sence d'un 6coulement puru- lent ou de pratiquer une otoscopie pour verifier s'il n'y a pas perforation du tympan. Les examens audiom6triques de routine sont rarement pratiqu6s dans les pays en developpement en raison du manque de mat6riel et de personnel specialise. Un audiometre simple serait 6galement tres utile pour les tests de d6pistage et le Liverpool Field Audiometer a 6t6 mis au point pour r6pondre a ce besoin. Cet audiometre de terrain portatif fonction- nant sur piles 6met des sons modules a 500, 1000, 2000 et 4000 Hz et a des intensit6s variables (30, 50 et 80 dB HTL). 11 a et6 utilise avec succes au Royaume-Uni, en Inde et au Swaziland. Des recherches doivent etre entreprises d'urgence sur les methodes de d6pistage appli- cable aux b6b6s et aux enfants d'age prescolaire dans les pays en d6veloppement. Chez les enfants d'age scolaire, la plupart des d6ficits audi- tifs l6gers a moder6s que l'on observe dans ces pays sont dus a une otite moyenne chronique suppur6e pour laquelle le meilleur test de d6pista- ge consiste a v6rifier la pr6sence d'un 6coulement purulent et, si possible, a pratiquer en meme temps une otoscopie. La recherche des d6ficits auditifs mod6r6s a graves doit etre effectuee par audiom6trie et le Liverpool Field Audiometer a ete conqu a cet effet. References 1. Penniceard, R.M. & Rabinowitz, A. Educational retardation and constructive hearing problems, 2nd ed. London, ILEA Schools Psychological Services, 1981, pp. 3-18. 2. Hamilton, P. & Owrid, H.L. Comparisons of hearing impairment and socio-cultural disadvantage in rela- tion to verbal retardation. British journal of audiol- ogy, 8: 27-32 (1974). 3. Lewis, N. Otitis media and linguistic incompetence. Archives of otolaryngology, 102: 387-390 (1976). 4. Miles, M. Deafness in rural Asia. Volta review, 86: 274-281 (1984). 5. Markides, A. Educational provision for hearing- impaired children in Ghana. Teacher of the deaf, 74: 19-27 (1976). 6. Rosenberg, P.E. & Swogger-Rosenberg, J. Hearing screening. In: Lass, N.J. et al., ed. Speech, language and hearing. Philadelphia, W.B. Saunders, 1982, pp. 1000-1014. 7. Final report of the subcommittee appointed to con- sider services for hearing impaired children. Lon- don, Department of Health and Social Security, 1981. 8. Tweedie, J. Children's hearing problems: their sig- nificance, detection and management. Bristol, IOP Publishing Ltd, 1987. 9. Greville, A. & Keith, W.J. Detection of hearing loss in infancy. New Zealand medical journal, 88: 277-280 (1978). 10. Sheridan, M.D. Infants at risk of handicapping conditions. Monthly bulletin of the Ministry of Health and the Public Health Laboratory Services, 21: 238-245 (1962). 11. Feinmesser, M. & Tell, L. Neonatal screening for detection of deafness. Archives of otolaryngology, 102: 297-299 (1976). 12. Gerkin, K.P. The high risk register for deafness. ASHA, 26 (3): 17-23 (1984). 13. American Speech and Hearing Association Joint Committee on Infant Hearing Position Statement. ASHA, 24: 1017-1018 (1982). 14. Richards, I.D. & Roberts, C.J. The "at-risk" infant. Lancet, 2: 711 (1967). 15. Swigart, E.T., ed. Neonatal hearing screening. Lon- don, Taylor & Francis, 1986. 16. Caufield, C. A developmental approach to hearing screening in children. Pediatric nursing, 4 (2): 39-42 (1978). 17. McCormick, B. Hearing screening by health visitors: a critical appraisal of the distraction test. Health visitor, 56: 449-451 (1983). 18. Watkin, P.M. et al. Parental suspicion and identifi- cation of hearing impairment. Archives of disease in childhood, 65: 846-850 (1990). 19. Murphy, K.P. A developmental approach to pedia- tric audiometry. Hearing aid journal, 6-7: 32-34 (1979). 20. Northern, J.L. & Downs, M.P. Hearing in children, 3rd ed. London, Williams & Wilkins, 1984. 21. Okunade, A.O. Screening for handicaps in children: are Nigerian nurses equipped? International journal of nursing studies, 17: 181-187 (1980). 22. Griffing, T.S. et al. Verbal auditory screening for pre-school children. Transactions of the American Academy of Ophthalmology and Otolaryngology, 71: 105-111 (1967). 23. Cole, M. et al. The cultural context of learning and thinking: an exploration in experimental anthropol- ogy. London, Methuen, 1971. 24. Cole, M. & Scribner, S. Culture and thought: a psy- chological introduction. New York, John Wiley, 1974. 654 WHO Bulletin OMS. Vol 70 1992 Screening children for hearing impairments in developing countries 25. Bruner, J.S. Culture and cognitive growth. In: Bruner, J.S., ed. Relevance of education. London, Penguin, 1976. 26. Helander, E. et al. Training in the community for people with disabilities. Geneva, World Health Organization, 1989. 27. Werner, D.B. Disabled village children: a guide for community health workers, rehabilitation workers, and families. California, Hesperian Foundation, 1987. 28. Jerger, J. & Hayes, D. The cross-check principle in pediatric audiometry. Archives of otolaryngology, 102: 614-620 (1976). 29. Klein, L.A. Audiological assessment of the young infant. Journal of the Speech and Hearing Associa- tion of Virginia, 26: 37-41 (1985). 30. Jacobson, J.T. & Hyde, M.L. The auditory brain- stem response in neonatal hearing screening. In: Swigart, E.T., ed. Neonatal hearing screening. London, Taylor & Francis, 1986. 31. Galambos, R. Use of the auditory brainstem res- ponse (ABR) in infant hearing testing. In: Gerber, S.E. & Mencher, G.T., ed. Early diagnosis of hearing loss. London, Grune & Stratton, 1978. 32. Brooks, D.N. Resource allocation in audiology. British journal of audiology, 19: 171-173 (1985). 33. McPherson, B.D. & Holborow, C.A. School screen- ing for hearing loss in developing countries. Scandi- navian audiology, 28 (suppl.): 103-110 (1988). 34. Manning, P. Purulent otitis media: differences be- tween populations in different environments. Pedi- atrics, 53: 135-136 (1974). 35. Lewis, N. Otitis media and linguistic incompetence. Archives of otolaryngology, 102: 387-390 (1976). 36. Okafor, B.C. The chronic discharging ear in Nigeria. Journal of laryngology and otology, 98: 113-119 (1984). 37. Holborow, C.A. Prevention of deafness in rural tropical areas. Tropical doctor, 15: 39-41 (1985). 38. Prasansuk, S. Audiology in Thailand and other Asian countries. International Association of Physi- cians in Audiology bulletin, 6: 3-10 (1987). 39. Gupta, R.K. et al. A study on chronic suppurative otitis media in rural area of PHC Chaka district, Allahabad. Indian journal of preventive and social medicine, 16: 49-53 (1985). 40. McCandless, G.A. Screening for middle ear dis- ease on the Wind River Indian Reservation. Hear- ing instruments, 26 (4): 19-20 (1975). 41. Paradise, J.L. Otitis media during early life. How hazardous to development? A critical review of the evidence. Pediatrics, 68: 869-873 (1981). 42. Holm, V.A. & Kunze, L.H. Effects of chronic otitis media on language and speech development. Pediatrics, 43: 833-839 (1969). 43. Paradise, J.L. On tympanostomy tubes: rationale, results, reservations and recommendations. Pedia- trics, 60: 86-90 (1977). 44. Gerwin, K.S. & Glorig, A. Detection of hearing loss and ear disease in children. Springfield, IL, Charles C. Thomas, 1974. 45. Brooks, D. The use of the electro-acoustic im- pedance bridge in the assessment of middle ear function. International audiology, 8: 563-569 (1969). 46. Alberti, P. & Kristensen, R. The clinical application of impedance audiometry. Laryngoscope, 80: 735-746 (1970). 47. Jerger, J. Clinical experience with impedance audiometry. Archives of otolaryngology, 92: 311-324 (1970). 48. Paradise, J.L. & Smith, C.G. Impedance screening for preschool children: state of the art. Annals of otolaryngology, 88: 56-65 (1979). 49. American Speech and Hearing Association Guide- lines for Identification Audiometry. ASHA, 17: 94-99 (1975). 50. Melnick, W. et al. Evaluation of a recommended programme of identification audiometry with school- age children. Journal of speech and hearing disor- ders, 29: 3-13 (1964). 51. Ballantyne, J. & Martin, J.A.M. Deafness, 4th ed. London, Churchill Livingstone, 1984. 52. Haggard, M.P. et al. Speech, admittance and tone tests in school screening: reconciling economics with pathology and disability perspectives. British journal of audiology, 18: 133-153 (1984). 53. Knight, J.J. Some aspects of speech tests in non- European languages. In: Martin, M., ed. Speech audiometry. London, Taylor & Francis, 1987. 54. Hood, R.B. & Lamb, L.E. Identification audiometry. In: Gerwin, K.S. & Glorig, A., ed. Detection of hearing loss and ear disease in children. Springfield, IL, Charles C. Thomas, 1974. 55. Downs, M.P. Auditory screening. Otolaryngologic clinics of North America, 11: 611-626 (1978). 56. Cooper, J.C. et al. An abbreviated impedance bridge technique for school screening. Journal of speech and hearing disorders, 40: 260-269 (1975). 57. Brooks, D.N. Impedance in screening. In: Jerger, J. & Northern, J.L., ed.Clinical impedance audiometry. Acton, MA, American Electromedics Corp., 1980. 58. Hildyard, V.H. et al. Tuning fork tests as aid to screening audiometry. Archives of otolaryngology, 78: 151-154 (1963). 59. David, J.B. et al. Adamarobe - a "deaf" village. Sound, 5: 70-72 (1971). 60. McPherson, B. & Holborow, C.A. A study of deaf- ness in West Africa: the Gambian hearing health project. International journal of pediatric otorhino- laryngology, 10: 115-135 (1985). WHO Bulletin OMS. Vol 70 1992 655