Consumers today can purchase virtually anything on the Internet, including many medical products and services. Certainly we see that in hearing healthcare, where numerous websites advertise hearing aids at lower prices than consumers typically pay when they are fitted by an audiologist or hearing instrument specialist. With some of these online companies, the purchaser has no face-to-face contact with a licensed and trained hearing healthcare professional at any point in the process.
As reported last month in my first article, “Inquiry into online hearing test raises doubts about its validity,”1 a consumer who decides to purchase a hearing aid by means of an entirely online approach first takes an online hearing test. Based on the results of that test, the company makes a hearing aid recommendation to the customer.
My previous article reported the findings of a study in which 81 subjects had their hearing tested online by the company and also by means of a conventional test conducted in a hearing test suite. That article concluded, “The wide variations between hearing thresholds obtained via the Internet and those obtained in the hearing test suite using traditional techniques suggest that hearing aid fittings based on results obtained online may be inappropriate and at times potentially dangerous to a consumer's hearing health.”1
If, after taking the hearing test, the consumer decides to purchase the recommended hearing aid(s), the company mails him or her an earmold impression kit with instructions on how to make an earmold impression at home, either alone or with the help of a friend or family member. The consumer then mails the impression back to the company so it can manufacture an earmold (for a behind-the-ear instrument) or a custom in-the-ear device.
This raises the question as to how well an average consumer, without professional assistance, can perform this procedure and produce the full and accurate ear impression that a manufacturer needs to make a well-fitting earmold or ITE hearing aid.
ASSESSING AT-HOME IMPRESSION TAKING
The purpose of the investigation reported here was to determine the ability of untrained consumers to make adequate earmold impressions, as they are required to do to purchase hearing aids through some Internet companies. (There are also companies that sell hearing aids via the Internet but have the purchaser receive professional services directly from an audiologist or hearing instrument specialist.)
For our study, the research team obtained one sample impression kit from an online company that sends such kits to hearing aid customers. We then made replicas of the kit.
Eighty-four subjects took part in the study and they were divided into three age groups: 18–30 years, n = 35; 31–59 years, n = 26; and 60 years and over, n = 23. (Data from one subject were disqualified because the earmold impressions were labeled incorrectly; that left impressions from 83 subjects for analysis.)
The subjects were recruited from university classes and from a local church. At the university, pairs of subjects were placed in diagnostic and therapy rooms at the speech and hearing clinic. In the church, round tables were set up approximately 5–7 feet apart and the participating pairs faced away from other pairs.
Each subject was paired with another participant (either a spouse or friend) to best simulate how a customer of the Internet company in question would make the impression at home using another person for assistance. This method was employed because the Internet site suggests that making an impression is easier with assistance from another person.
Each subject was given one of the replica impression kits, which contained a polyethylene syringe, a pack of silicone earmold impression material, a medium cotton otoblock, a cotton swab, and instructions.
In each pair, the subjects took turns making earmold impressions of one of their partner's ears, following the instructions. Members of the research team observed, but did not take part in the process.
After an impression had been made of one ear of every subject, participants' ear canals were checked to make sure no debris re-mained. Then the research team made earmold impressions of all 84 of the subjects' ears from which impress-ions had been made by the untrained subjects. The research team, which consisted of one licensed audiologist and two third-sem-ester doctor of audiology (AuD) students at Illinois State University, used the normal tools that a hearing healthcare professional in a clinical practice would use to make impressions. These included an otoscope, a polyethylene syringe, foam otoblocks of various sizes, silicone impression material, and an Earlite for placing the otoblock in the ear canal. Figure 1 shows a random example of the two impressions made from one subject's ear.
All subjects were then asked to complete an anonymous survey about their overall experience, both with making their own earmold impression and having one made by the research team.
All the completed impressions were sent to a leading hearing aid manufacturer, where two audiologists employed by the company evaluated each impression on the following criteria:
1. Is the impression smooth and complete?
2. Is the canal long enough to show the beginning of the second bend?
3. Is the helix portion of the impression clearly defined?
4. Is the tragus portion of the ear clearly defined?
5. Is the concha complete?
These five criteria were based on information from two earmold laboratories (All-American Mold Labs and Westone Laboratories) about what they have learned are the most important factors in determining a good earmold impression. On each criterion, 1 point was given for a yes response and 0 for a negative response. Each impression was then given a total score ranging from 0 (meaning it met none of the five criteria) to 5 (meaning it met all five).
Every impression was marked either A (for the “amateur” impressions made by the study subjects) or B (for those made by members of the research team). However, the audiologists evaluating the impressions were unaware of who had made the impressions, how and why they were marked, and from where they had come. Only the company's director of audiology knew the details of the study.
RATINGS OF IMPRESSIONS
We averaged the two audiologists' ratings of each earmold impression to come up with a combined score for each ranging from 0 (meaning both audiologists felt the impression met none of the criteria) to 5.0 (meaning that both felt the impression met all five criteria). Figure 2 shows how many of the impressions made by the subjects (A) and those made by the research team (B) received each rating.
To determine if the difference between the ratings of the A and B impressions was statistically significant, we used a simple two-sample T-test. Results at the .05 confidence level indicated a high level of significance (P = 0.000) for the difference between the scores of the two groups of impressions (Figure 3).
We also performed a regression analysis to see if any significance was noted as a function of age. The results indicated that age had no effect on the outcome and cannot be used to explain the data (Figure 4).
SUBJECT SURVEY RESULTS
Eighty-three subjects completed and returned the survey, which first asked them on a scale of 1 (not comfortable) to 5 (very comfortable) how comfortable they felt at the beginning of the earmold impression process. The mean score was 3.3. The next two questions asked subjects their comfort level with making the earmold impression themselves (mean of 2.8) and having the impression made by a member of the research team (mean 4.5).
When asked if there was a noticeable difference between the feel of the impressions they made and those made by the research team, 66 of 83 participants described the impression taken by the research team as being either firmer, deeper in the canal, tighter, fuller, more uncomfortable or exerting more pressure.
Finally, we asked subjects how willing they would be to make their own ear impression to purchase a hearing aid. Fifty-eight subjects said they were “not willing” to do so, 17 were “possibly willing,” and 8 said that they were “very likely” to do so.
The results of this study suggest that it is a difficult task for untrained consumers to make their own earmold impressions, even with a partner. Approximately half of the earmold impressions produced by the subjects received a score of 2 or lower from the two audiologists who judged them and almost 80% scored 3 or lower. Only 20.5 % of the “amateur” impressions received an average score of 3.5 or higher.
In contrast, the judges awarded more than 97% of the impressions produced by the research team a score of 3.5 or higher, and only .02% of the “professional” impressions scored as low as 3. About 93% of the impressions made by the research team rated 4.0 or higher, and 25% earned a perfect 5 from both judges.
It is uncertain how high a score an earmold impression needs for a hearing aid manufacturer to be able to provide a well-fitting hearing aid from it. According to the manufacturer that assisted with this study, the minimum acceptable impression quality varies from case to case and depends also on the style of the hearing aid being ordered.
The subjects in our study were clearly more comfortable when a member of the research team made the impression than when they (with their partner) made their own. Some of the written comments expressed apprehension about placing the cotton block too deeply in their partner's ear canal. Some subjects said that as soon as a research team member placed an otoblock into their ear canal they realized they “hadn't gone deep enough” in placing an otoblock in their partner's ear. That the research team members “knew what they were doing” had a clear impact on the subjects' feelings.
Concerns about home-made impressions
Several concerns arise when consumers make their own earmold impressions at home. One is safety, the risk that individuals may injure themselves or may unknowingly leave debris in their ears. While none of the participants in this study experienced any discomfort or left debris inside the ear canal, one subject appeared to have a scratch in his ear canal that reddened after his partner removed the original amateur-made impression. Instead of putting more impression material into the same ear, the patient's other ear was chosen for the professional impression.
This illustrates one of the potential risks of at-home, consumer-made earmold impressions. However, there are no data about the rate of injuries during impression making at home.
Another major concern about having consumers make their own impressions is that it will result in a poorly fitting hearing aid. Pirzanski has outlined critical issues for making earmold impressions.2 Regarding how far the otoblock must be placed into the ear canal, he states, “With a shallow otoblock position, the impression material will not stretch the cartilage within the seal area, and the resulting earmold may fit loosely, have retention problems, and be susceptible to acoustic feedback.”
Pirzanski also recommends using an Earlite to place the otoblock. In our study, subjects clearly felt uncomfortable placing the otoblock very deep in their partner's ear. Additionally, the use of a cotton swab rather than a pen light seems insufficient for proper insertion of the otoblock.
The results of this study suggest that having hearing aid purchasers make their own earmold impressions, even with the help of a partner, is likely to result in, at best, ill-fitting hearing aids and, at worst, injury to the ear canal or tympanic membrane. While it is no surprise that trained professionals make better earmold impressions than amateurs, the findings of how poor many of the impressions made by amateurs in our study were raise serious doubt as to whether hearing aids, which are FDA-regulated medical devices, should be permitted to be manufactured for hearing-impaired consumers based on impressions made in this manner.
The final article in this series will look at the performance of and user satisfaction with hearing aids purchased online in accordance with the procedures reported on in my first two articles.
The author would like to acknowledge Dr. Thomas Powers, Nina Atchley, and Linda DiCamillo of Siemens Hearing Instruments; Michael Sharp and Shanna Davis, Illinois State University (ISU) graduate assistants; and Dr. Jinadasa Gamage of the ISU Statistical Center for their gracious help with this project.