Many countries have adopted early hearing detection programs over the past 30 years; with several UNHS protocols such as single OAE and ABR screening, OAE followed by ABR, or AABR (Auto ABR) with a follow-up ABR, etc. being employed . In China, most hospitals follow the OAE screening followed by a final diagnosis using the ABR test, as recommended by the US National Institute of Health (NIH) in 1993 . This two-stage protocol is effective and has a low failure rate .
In the current investigation we modified this UNHS protocol to better address the needs of newborn babies of internal migrants, who have steadily increased in numbers over the last decade in Beijing and often miss out on this important screening service due to the geographic instability resulting from their parents’ way of life. In particular postpartum, families are often unable to accept regular hospital appointments due to their frequent moves, and are reluctant to commit themselves to rescreening tests at the same hospital where the baby was delivered, despite availability of free screening/rescreening tests. Indeed, a relatively lower rescreening rate of 56.87% in the current study was attributable to three major factors; namely i) some babies could not be rescreened because they had been sent back to their parent’s home provinces after a few days, ii) some babies were rescreened in hospitals closer to the their new rented accommodation and relevant data were not available to the investigators at Shangdi Hospital, and iii) some parents were reluctant to accept re-testing because their new rental location was far away. These findings thus emphasize the importance of revising currently recommended hearing screening protocols for newborn babies of frequently migrating parents. Based on previous research , our revised protocol calls for two inpatient OAE tests and two outpatient OAE tests instead of one each as recommended currently. These extra tests provide more scheduling flexibility and thus greater opportunity for newborns of migrant workers to be tested.
Using this modified protocol, we have demonstrated that there was increased coverage rate in the target population, the basic measure of screening efficiency. In a hospital such as Shangdi Hospital, where the majority of pregnant women are poor migrants without any form of health insurance and therefore eager to leave as soon as possible to avoid the expense of hospital stay, we shortened the inpatient screening time from the normal 72 hours to a maximum 48 hours by adding an earlier OAE screening test at 24-48 hours to ensure higher coverage/acceptance rates for hearing screening. Indeed, on reviewing our database we found that 2169 infants would have missed the inpatient hearing screening based on the recommended protocol. Moreover, as 492 of these infants did not pass the inpatient hearing screening at 24–48 hours, under the recommended protocol these infants with suspected hearing impairment would have been missed if they had left the hospital by about 48 hours. Using this modified protocol, the coverage rate was found to be high at 98.91%, meeting the recommended coverage by the Joint Committee of Infant Hearing (JCIH) , and was similar to the coverage rate observed for an obstetric hospital downtown (98.85%), where the majority of expectant mothers were residents . The positive rate of 27.22% for hearing anomalies observed after the first hearing screening in the present study was much higher than that reported previously in several studies [21–24]. One study indicated that OAE testing had a 15.6% false positive rate in the first 24 hours of life, but this fell to 4% by 72 hours . It is possible that the comparatively high rate for detection of hearing anomalies at the first screen in the current study was at least partly due to a high false positive rate, as most of the infants received their first OAE screening before they were 72 hours old; which has been suggested to be the ideal screening time .
In the modified UNHS protocol we kept the recommended inpatient OAE test at 48-72hours for referred infants staying in the hospital beyond 48 hours, to decrease the false positive rates. Although the present study indicated that the positive rate decreased significantly from 27.22% to 18.36% after the second inpatient OAE test 48–72 hours after birth, this was still much higher than the rate of 6.39% observed at the obstetric hospital downtown . This disparity may be a result of not all infants being rescreened at the same time point of 72 hours, and suggests that selection of an appropriate screening/rescreening time following birth may be an important factor in minimising false positive rates for hearing anomalies in newborn babies.
The referral rate is another important measure of a screening program’s efficiency and effectiveness and, according to the JCIH 2007 position paper, can be minimised by effective rescreening . Indeed, many well-infant screening protocols incorporate an outpatient rescreening within 1 month of hospital discharge to minimize the number of infants referred for follow-up audiological and medical evaluation. Some studies have demonstrated that following auditory neurological maturity and transient middle ear effusion absorption, some infants referred for outpatient OAE screening at 1 month passed the OAE test at 2 months [26, 27]. In this context we felt that a single outpatient rescreening test after 1-month of birth was not appropriate for a migrating population for reasons discussed above, and therefore modified the UNHS protocol to include a second outpatient rescreening test at 2 months to increase the possibility for a greater number of migrants being able to accept at least one OAE test before further audiological assessment. Using the modified protocol, we found that the final referral rate for ABR testing was significantly reduced from 2.08% to 1.73%.
Of the 10983 migrant newborn babies that had the first hearing screening test, only 0.32% (35/10983) was found to have SNHL. The proportion of infants with identified unilateral (n = 23) to bilateral (n = 12) hearing loss in our study was similar to that previously reported in developing countries; approximately 1:3–4 [23, 28, 29]. However, if we include the missing subjects, we estimate that the total incidence of congenital SNHL would be 1.06%. It is important to note, that our study missed all infants with neural hearing loss (auditory neuropathy) because the current screening program relies on OAE, which assesses pre-neural functioning. The prevalence of sensory and neural hearing losses would therefore exceed the current estimated. Further, over time, late-onset and progressive hearing losses will increase the number of children who would benefit from intervention .