10/14/20091Todd A. Ricketts, PhDAssociate Professor Vanderbilt Medical Center; and, Director, Dan Maddox Memorial Hearing Aid Research LaboratoryVanderbilt Bill Wilkerson Center For Otolaryngology and Communication Sciences Nashville, TN, USA2 How do you select a hearing aid? What decisions do you need to make? What information do you need?41. Choice of treatment for a patient should be based on the best available, specifically relevant, up-to-date data about treatment effectiveness. Furthermore….2. Practitioners can and should learn to evaluate evidence from original research to support their recommendations. Finally..3. Unless there is valid evidence for real-world effectiveness, we must acknowledge uncertainty about the value of any treatment, no matter how “logical” it seems.1. Impact on patient counseling? American Academy of Audiology Task Force: Guidelines for the audiological management of adult hearing impairment. AudiolToday 2006; 18(5); 32-36.  http://www.audiology.org/resources/documentlibrary/Documents/haguidelines.pdf Were the study subjects similar to my patients on variables such as age, financial situation, health, living arrangement, lifestyle, etc. Comparisons to “past” or “previous” hearing aids are nearly always flawed – Gnewikow dissertation results. May be poorly fit or not matched for features Even if matched, subject AND investigator bias may be present. How similar were study conditions to those your patients experience? Was the study an efficacy study or an effectiveness study?610/14/20092Efficacy: Can the treatment work under optimal conditions? Best possible scenario. Subjects: highly motivated, uncomplicated hearing problems, etc. Practitioners: highly trained, plenty of time, etc. Measures: might only be made in laboratory settings.Effectiveness: Does the treatment work under typical conditions? Patients: may be motivated, health problems, daily pressures, financial concerns, etc. Practitioners: minimal training, minimal time, distractions & pressures, etc. Measures: should always include real life trial.1.1.Systematic reviews and metaSystematic reviews and meta--analyses of analyses of randomized controlled trials.randomized controlled trials.--Possible problems with digital and directional?Possible problems with digital and directional?2.2. Randomized controlled trials.Randomized controlled trials.3.3. NonNon--randomized intervention studies.randomized intervention studies.4.4.NonNon--intervention studies: crossintervention studies: cross--sectional sectional surveys, cohort studies, casesurveys, cohort studies, case--control studies.control studies.5.5. Case reportsCase reports6.6. Expert opinionExpert opinion Most effects of individual features or treatment differences are rather small, too small to show up in many general outcome measures. Not many situation specific outcome measures that are realistic AND balanced exists. One outcome of these hurdles? The impact of nearly all hearing aid features looks weak! 9 10 Absolutely Not! Be skeptical - But don’t assume weak or nonexistent effectiveness in the presence of strong efficacy+ means there is no effect Likely that we just need a better (more directed) outcome measure However, acknowledge the possibility that there may be no average effect in the real world.  Example - A feature that has positive or negative effects depending on the specific environment. Ecological framework is formed by applying concepts of biological systems to the field of auditory communication (Borg, 2003) Auditory Ecology is the auditory lifestyles of listeners in terms of the types of listening circumstances that they experience, the frequency of their occurrence, and their importance to that individual in everyday existence (Gatehouse, 2003) As applied to hearing aid selection: Are there tests that we can conduct that will allow us to predict how the hearing aid user will interact with their environment and/or what technologies may improve this interaction? The features in modern hearing aids are Not all created equal and some features have multiple settings. Though for some features there will be no real differences.  Importantly, how the feature operates in environments that our patients are actually in, and that are important to them much be considered - Potentially improved expectations and use counseling. What is the most important factor leading to satisfaction? There will be some trading relationships  For example, the open comfortable fit of the OC style will come at the cost of less directional benefit.10/14/2009313 Bentler “Hype” Study In one condition listeners fit with a digital hearing aid. Half told it was “analog”, half it was “state-of-the-art” digital After a trial the hearing aid was removed, the investigator left and came back with the SAME hearing aid. Listeners were then told they were wearing a different hearing aid in the opposite condition and sent out for another trial. 14 Strong and significant preference for the “digital” condition.  “Sounds Clearer” “Easier to listen to” “Helps get rid of the background noise” …and other digital advertisements of the time were nearly directly quoted.  Ability to understand speech in quiet? Ability to understand speech in noise? Patient’s perceptions of ability to understand speech in noise? Patient’s “acceptance” of background noise Patient’s Threshold of Discomfort? Problems and Expectations? High Frequency Amplification? Chasing the highs Extending the highs? Bilateral versus unilateral amplification? Cognition? That these tests will help determine hearing aid candidacy? That the scores will help predict how a person is doing in the real world? That the scores will help predict hearing aid benefit and satisfaction? That these tests will help diagnose auditory pathology? That they will help determine differences between ears? PB-Max DOES NOT consistently occur at 30 or 40 dB above the patient’s SRT or pure-tone average PB-Max DOES NOT consistently occur at the patient’s MCL IT DOES consistently occur at 75-80 dB HL (for the average patient).10/14/20094 Important NOT to use differences that are not really different ! Check out chart from Thornton-Raffin article (in 1978 JSHR, also in