IMPORTANCE OF EVIDENCE-BASED PRACTICE
Evidence-based practice (EBP) in health care integrates clinical expertise and patient values and preferences with the best available clinical research evidence (Wong & Hickson. Plural Publishing, 2012). A hierarchy of evidence, core to the principal of EBP, ranks studies based on how rigorous the research methods are. Typically, expert opinion is the lowest form of evidence, leading to case studies, cohort studies, randomized controlled trials, and systematic reviews with meta-analyses being the highest level of evidence.
Notably, systematic reviews are not exploratory literature reviews. Rather, systematic reviews collate all the empirical evidence that fits pre-specified criteria to answer a specific research question. As with any study, key elements include clearly stated objectives, pre-defined eligibility criteria, an explicit reproducible methodology, a systematic search to attempt to identify all relevant studies, and an assessment of the validity and systematic presentation of results. The crème-de-la crème of systematic reviews are Cochrane reviews, which are internationally recognized as the highest standard in assessing health care resources. Explicit methods are used to minimize the risk of bias, such as selective reporting of the data, to provide reliable findings for use in clinical decision-making. To maximize transparency, a protocol is peer-reviewed and published before the review starts.
WHY DO A COCHRANE REVIEW ON HEARING AIDS?
The review was prompted for two reasons. First, the previous systematic review on hearing aids included studies that were published only up to 2004 (J Am Acad Aud. 2007;18:151). So it was time to update the evidence from the published literature. Second, in 2014, several health care commissioning groups in the United Kingdom considered withdrawing the provision of hearing aids for adults who had MMHL. Up until then, hearing aids had been provided free at the point of delivery to those who needed them since the inception of the U.K. National Health Service in 1948. There was a clear need to have high-quality and up-to-date evidence on the effectiveness of hearing aids in adults with MMHL. Although the review process had started before the publication of the 2016 National Academies of Science Engineering Medicine report on the accessibility and affordability of hearing aids, the issue of the effectiveness of hearing aids is as relevant to countries such as the United States as it is to the U.K. (NASEM. 2016).
THE REVIEW PROTOCOL
Following the strict guidelines laid down by the Cochrane Collaboration, the team developed a review protocol that underwent rigorous peer review by a Cochrane methodologist (Coch Data Sys Rev. 2015). The review objective was “to evaluate the effects of hearing aids for mild to moderate hearing loss in adults.” The inclusion criteria followed the PICO format to frame the research question and methods, which were:
- Participants: Adults (≥18 years) with mild or moderate hearing loss, based on the WHO definition (http://bit.ly/2xtt1JQ)
- Interventions: Acoustic hearing aids
- Comparison: Passive control (e.g., waiting list control) or an active control (e.g., placebo hearing aids)
- Outcomes: Primary: Hearing-specific health-related quality of life (QoL), with participation as the key domain. Adverse effect: pain. Secondary: health-specific QoL, listening ability. Adverse effect: noise-induced hearing loss
Only randomized controlled trials (RCTs) and the first treatment period of cross-over trials were included. The unit of analysis was individual participants; group settings were not considered.
A series of other predefined criteria were described, such as how studies were selected, data were extracted and managed, studies were considered for risk of bias, the treatment effect was measured, alongside which subgroups (e.g., mild vs. moderate hearing loss) would be analyzed and how heterogeneity (i.e., the differences observed between studies, I2) would be assessed. Nothing went undefined.
THE REVIEW FINDINGS
The final search was done on March 23, 2017. We whittled down 2,840 records (e.g., papers, conference abstracts) to those that met the inclusion criteria (Coch Data Sys Rev. 9:2017). Five RCTs published between 1987 and 2017 were included, of which three were included in the meta-analyses. A total of 825 participants were included, with average age by study between 69 and 83 years, of which 70 percent were from two Veterans Association studies, with a mix of BTE and ITE hearing aids. The studies up to 2005 included controls from a waitlist group, and the two 2017 studies used placebo hearing aids (i.e., no gain, therefore acoustically transparent). One study included patients with Alzheimer's disease only and was not included in the meta-analysis. Forest plots of the treatment effects for each outcome are shown in Figure 1.
For hearing-specific HRQoL, all three studies used the Hearing Handicap Inventory for the Elderly (HHIE) so the actual results (mean difference, MD) and 95 percent CI were shown. For the HHIE, there was a large beneficial effect of hearing aids compared with the controls (MD -26.47, 95% CI -42.16 to -10.77; n=722; moderate quality evidence). There was high heterogeneity (I2=97%), as the Humes study showed a smaller effect than the two VA studies, which showed similar results.
For health-related QoL, the two studies included in the meta-analysis used different questionnaires (SELF, WHO-DAS II), so the standardized mean difference (SMD) is shown. Although only one study showed a significant effect, there was a small overall beneficial effect of hearing aids compared with the controls (SMD -.038; 95% CI -.55 to -.21; n=568; moderate quality evidence). Heterogeneity was low (I2=6%) due to a large degree of overlap in results.
For listening ability, the two studies used different outcomes (PHAP, APHAB) and showed a large overall beneficial effect of hearing aids (SMD -2.57 95% CI -2.84 to -2.30; n=380 participants). Heterogeneity was high (I2=97%) indicating no overlap in the results, although both studies showed large effects.
Only one study measured adverse effects, and none were reported (very low quality of evidence).
The risk of bias (RoB) of included studies (i.e., due to study design flaw) is shown in Figure 2. Unsurprisingly, there was a high RoB for blinding participants and outcome assessment for those who received hearing aids in studies published up to 2005. The use of placebo hearing aids reduced this risk. Further discussion of RoB is found in the full publication.
Quality of evidence, which reflects the extent to which we were confident that our estimate of the treatment effect was close to the true effect, was rated using GRADE (high, moderate, low, or very low). Evidence is downgraded due to RoB, inconsistency, indirectness, imprecision, and publication bias. For our review, the quality of evidence for each outcome domain was rated as moderate. This meant that we were moderately confident in the estimated effect, and that the true effect is likely to be close but possibly substantially different. Reasons for downgrading or not are fully discussed in the review. To put this moderate quality of evidence into perspective, a review of systematic reviews in adult rehabilitation (n=9) showed that all reported low or very low quality of evidence, and of the six ENT Cochrane reviews in adults, only one reported moderate-quality evidence in all domains; none reported high quality of evidence.
Hearing aids are an appropriate intervention, and moderate-quality evidence shows that these devices improve hearing- and health-related QoL and listening abilities of adults with mild-to-moderate hearing loss. However, there needs to be greater consistency in the use of outcome measures and measurement of long-term outcomes (>1 year) in future studies. Self-management of hearing aids should consider other supplementary management to improve hearing aid use, such as multimedia videos for hearing aid users (Coch Data Sys Rev. 2016; Ear Hear. 2016;37:123). We were unable to conduct a subgroup analysis of mild vs. moderate hearing loss. Current research on this is showing some interesting results, but that is another story for another day.
Acknowledgements: Thanks to the review authors Padraig Kitterick, Lee Yee Chong, Mark Edmondson-Jones, Fiona Barker, Derek Hoare. This systematic review presents independent research funded by the National Institute for Health Research (NIHR). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the U.K. Department of Health.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.