Implementation and evaluation of a rural community-based pediatric hearing screening program integrating in-person and tele-diagnostic auditory brainstem response (ABR)

Training village health workers to screen hearing using distortion product Oto acoustic emissions³ (DPOAE) and assist in ABR testing

Seven Village Health Workers (VHWs) were recruited for training through the non-profit, Rural Women’s Social Education Centre (RUWSEC) located in the community. All VHWs were women with minimum five years of field work experience, having minimum eighth grade education, and demonstrated good communication skills.

Evaluation

Performance in the post training evaluation determined recruitment of VHWs in the screening program. Follow-up evaluations (6 months and 1.5 years post-training) were also conducted to assess retention of knowledge and skill. Knowledge was evaluated at baseline and post-training using 15 multiple-choice questions (Additional file 1). The questions pertained to age of screening, risk factors of hearing loss, methods of hearing screening, interpretation of screening results and consequences of hearing loss. Skill was evaluated post-training using an objective structured clinical examination (OSCE) format, in which, VHWs performed the screening process and assisted in ABR on one adult each. Agreement in DPOAE screening result between audiologist and VHW was assessed on 10 infants and 20 adult ears. Images from the training program are shown in Fig. 2.

Implementation of community-based hearing screening program

Location of screening program

Fifty-one villages and hamlets with an estimated population of 32,560 in Thirukazhukunram block (Group A) and 43 villages and hamlets with an estimated population of 33,642 in Madhurantagam block (Group B) were selected for the program. The location of the screening program is shown in Fig. 3. Villages were approximately equal in distance from the tertiary care hospital. The population of Group A and Group B were nearly equal. VHWs proximity to prospective villages was a factor in village selection.

Tele-diagnostic testing

Tele-diagnostic ABR testing was conducted using a mobile telemedicine van with satellite connectivity in rural areas that lacked internet penetration or otherwise with broadband internet. This testing was conducted in real time by an audiologist at the tertiary care hospital by remotely accessed equipment. The trained VHWs prepared the child for testing (electrode placement, positioning child and ensuring that the child was asleep throughout testing). A tele-technician set up the equipment and established satellite / broadband internet connectivity. Detailed description of the mobile telemedicine van with satellite connectivity and validation of tele-ABR protocol is published in Ramkumar et al. [19]. A schematic representation of tele-ABR diagnostic testing is shown in Fig. 4.

Evaluation of knowledge

A benchmark criteria of 80% scores in knowledge assessment was set for recruiting VHWs for the screening program. On pre-training evaluations, VHWs scored between 35 and 70%. Figure 5 reflects that all VHWs achieved between 80 and 86% scores in the immediate post evaluation.

VHWs were encouraged to review the manual and videos to refresh their knowledge and maintain quality of service. Six months post-training, a secondary evaluation was conducted where all VHWs obtained a minimum of 80% scores. In the evaluation conducted one and a half years post training, two VHWs obtained less than 80%. In such cases, information provided during the training were recapitulated once again.

On the Freidman test of repeated measure, a significant difference (at α = 0.01) was obtained between baseline scores and immediate post, six months post and one and half years post training (F = 19.507, df (3,15), P = 0.00002). Prior to training, VHWs were aware of some of the risk factors of hearing loss such as trauma to the ear, perforated ear due to use of hair pins or other sharp objects, excessive noise exposure and consangnous marriage was reported as a risk factor for all disabilities. The post-training evaluations showed considerable improvement in the scores obtained by VHWs on questions related to importance of age of identification of hearing loss and effective methods for screening children for hearing loss. VHWs answered questions related to DPOAE screening such as adequate conditions for testing, what information could be obtained through DPOAE screenings, and how to make inferences based off of screening results and messages on DPOAE screeners related to noise and probe fit.

Evaluation of skill

Agreement of results in DPOAE screening between audiologist and VHW was assessed on 10 infant and 20 adult ears. Four VHWs had 90% agreement on both adults and infants. Two VHWs achieved less than 80% agreement (Fig. 6).

All VHWs except VHW S and VHW L obtained above 80% in knowledge, skill and inter-tester reliability. VHW L obtained less than 80% in reliability alone, hence was provided additional training and performance was monitored for two months, before accepting her in the program. VHW S was excluded from the program due to unsatisfactory skill levels despite retraining.

Follow-up rate for 2nd screening

The median follow-up rate for 2nd screening was 85%. Mothers remained in their maternal homes often between three to six months after delivery but relocated to their husband’s home after this period. Hence, follow-up rate was poorer for infants between zero to three months despite VHWs conducting door-to-door screening. In the four to five year age group, there were children of nomadic tribes who moved from one location to another and could not be followed up.

Refer rate

Follow-up for in-person versus tele-ABR diagnostic testing

Five of seventeen (30%) who underwent tele-ABR were identified with hearing loss, and all three children who underwent in-person ABR (100%) had hearing loss. As a result of this program, two infants and six young children with hearing loss were identified. All children with hearing loss exhibited one or more risk factors. Of the eight children with hearing loss, three were born to parents with second degree consanguineous marriage, four had a family history of hearing loss and three had more than one risk factor including low birth weight, jaundice, preterm birth and/or a family history of hearing loss.

Of the eight infants and young children identified with hearing loss, one child was under three months of age, and was subsequently followed up to rule out neuromaturational delay. Two children with bilateral mild hearing loss were identified at one and two years of age respectively, and were counseled regarding communication strategies and periodic follow-up; one three-year-old child with unilateral mild conductive hearing loss was referred for otolaryngological evaluation at the nearest government hospital. Two children between three and four years of age with bilateral severe to profound hearing loss, and two children between four and five years of age with asymmetrical sensorineural hearing loss were recommended a hearing aid trial. Only two parents followed-up for a hearing aid trial. Both children were fit with appropriate amplification and referred to the nearest rehabilitation centre.

Follow-up rate for 2nd screening

Hearing screening programs have ensured higher participation in initial screenings but a major challenge remains in ensuring subsequent follow-up [33–36]. The follow-up rate for 2nd screening in this program is better than those reported in hospital-based hearing screening programs in India [12]. Even in community-based hearing screening programs, the loss to follow-up for 2nd screening was reported to be 52% despite free transportation and no fees [37]. One community clinic-based program in South Africa reports a follow-up rate of 85%, ranging between 50 and 100% across eight community clinics [31]. The results of this study are similar. Better follow-up for 2nd screening in this community-based program can be attributed to the door-to-door screening protocol, where the onus was on the VHW to complete 2nd screenings. It can be surmised from the above studies that when the onus of follow-up is on the parents, follow-up is poorer.

Refer rate

It is noteworthy that, in this community-based screening program, the 2nd screening refer rates were very low, except in four to five-year-old age group. The refer rate was lower than the reported refer rate (3 to 19.4%) in other community-based programs from the African region [31, 37]. In these programs, two-step screenings using TEOAE/AABR and DPOAE/DPOAE were conducted in immunization clinics with higher noise levels.

In this program, refer rates increased with age. High refer rates of 6.3% were noted only among four to five-year-old children. Acquired permanent conductive hearing loss was ruled out as children identified with hearing loss in this age group had asymmetrical sensorineural hearing loss. Therefore, the higher refer rate can be attributed to older children’s resistance to being tested and transient middle ear conditions that are more common in this age group. A similar trend was reported with TEOAE screening conducted by auxiliary nurses in a community-based program in Nigeria [37].

VHWs were trained to recognize the “noisy” message that appears in the screener when the environment is not conducive to adequate screening and knew to pause screening. In addition, having dedicated personnel for screenings provided sufficient time to make multiple attempts during 2nd screenings to ensure that the “refer” was not due to ambient noise.

As per JCIH 2007, the “refer” percentage of all infants who fail an initial screening and fail any subsequent rescreening before a comprehensive audiological evaluation should be less than 4%.This suggests that the community-based screening program was successful in meeting the standards set by JCIH (2007) for hospital-based programs. The validity of the screening conducted by VHWs was previously evaluated and the negative and positive predictive values were 98.8 and 27.2% respectively [38]. These findings supplement the success of the program.

Follow-up for in-person versus tele-ABR diagnostic testing

The advantage, if any, of a tele- ABR diagnostic testing was studied by comparing it to the traditional in-person ABR follow-up rates. As per JCIH standards for hospital-based programs, 90% of infants requiring diagnostic evaluation should be assessed. The rate of follow-up for tele-ABR nearly achieved this benchmark.

All over the world, achieving 100% follow-up for diagnostics is a challenge. Some hospital-based programs in the US, France, and Malaysia showed higher follow-up rates, between 81 to 91% [39, 40]. Other programs in the US reported follow-up rates as low as 11% [35]. In one hearing screening program in the US conducted on four-year-old children, the follow-up rate was only 10% [41]. All hospital-based programs in India have reported 12 and 21% follow-up for 2nd screening and diagnostic assessment respectively [9, 12]. Though there are very few rural community-based studies, it is noteworthy that community clinic-based studies have shown high (91%) follow-up rates [31, 42], due to shorter travel.

The follow-up rate obtained in this program, irrespective of in-person or tele-ABR follow-up, is better than previous reports of follow-up rates in India, and is comparable to the high follow-up rates obtained in community clinic-based programs around the world. This suggests that in general, community-based programs have had greater success with follow-up. In this program, the improved follow-up compliance maybe attributed to the VHWs efforts in mobilizing and monitoring follow-up and therefore strongly supports a community-based model of hearing screening. Improved follow-up for tele-ABR in this program is comparable to the community clinic-based tele-diagnostic testing conducted in the Californian tele-audiology program [43].

The median number of days taken between 2nd screening and tele-ABR follow-ups were 30 days (10–189 days), and for in-person ABR follow-ups were 31 (30–36 days). The range for tele-ABR follow-up was wider. Tele-ABR was conducted once a month in the community, therefore, it was possible to achieve follow-ups as early as 10 days from the time of 2nd screening. Only one child was brought for tele-ABR after 6 months of 2nd screening, when the mother returned from her maternal home. Despite the minimal difference in the median time between the two follow-up methods, it was possible to achieve much earlier follow-ups for tele-ABR. In a mobile ear-screening service, the time between screening and tele-ENT evaluation consistently diminished over three years of the program. This suggests that with time, the tele- follow-up may show significant time gain [44].

According to Thompson et al., 2001, in the US, quality studies demonstrate that if 2041–2794 low-risk and 86–208 high-risk newborns were screened, then one case of moderate-to-profound permanent hearing loss was found. Though this program included children up to five years of age, four out of the 2815 screened were identified with moderate to profound hearing loss between three to five years of age and were recommended to partake in a hearing aid trial. Even though return rate for diagnostics were good, only two parents followed up for hearing aid trials and fitting. Poor follow-up for intervention can be attributed to a lack of awareness about the consequences of hearing loss, financial constraints in undertaking travel to the hospital to access rehabilitation services, and potential wage loss. Spivak et al., 2009, reported similar non-compliance rates for hearing aid fittings, particularly in infants with unilateral hearing loss.