Tuberculosis (TB) is a global pandemic, and multi-drug resistant tuberculosis (MDR-TB) presents additional, unique challenges for those who suffer from this debilitating disease. Existing drug treatments are multifactorial and complex with lengthy treatment regimens, complicated administration, and toxic secondary effects. The costly treatment of drug- resistant TB can take up to two years and presents significant challenges to health systems, payers, and patients. Recognizing that significant hearing loss can be a side effect of the current treatment regimen for MDR-TB, there is a need to closely monitor patients undergoing treatment with existing and new drug therapies.

The regimen used for MDR-TB patients is longer and more complex than for those with non MDR-TB. It involves a combination of both oral and injectable drugs and although treatment times vary, these regimens must be administered in clinic. Given the length and complex nature of treatment, it is not uncommon for patients – who are often very sick and perhaps unable to travel– to miss appointments, therefore interrupting these strict courses of treatment.

In May of 2016 the World Health Organization (WHO) recommended a shorter treatment regimen reducing patient management time by almost half. Even still, therapy occurs over a period of 9 to 12 months, and is comprised of a combination of up to seven different drugs. Although some drugs need to be administered daily, treatment and duration completely depends on the individual’s stage of MDR-TB as well as the presence of other health conditions or limitations.

Aminoglycosides are used in the treatment of serious bacterial infections. Certain tuberculosis strains have become resistant to several drugs in this important group. Kanamycin is one of the last choices of effective agents in this classification group to treat drug-resistant tuberculosis. It is generally administered in doses based on bodyweight meaning some patients could receive very high dosages. Due to its reported side effects – and specifically the risk of ototoxicity – it is not a drug of choice, but often prescribed as a last resort.

Many factors come into play when considering the option to use this drug and its possible side effects on hearing. It has been suggested that co-morbidity with other health conditions, as well as a genetic predisposition to the effects of ototoxicity, may lead to a higher incidence and/or higher level of hearing loss.

And so, it is highly recommended that anyone who is on a regimen that includes kanamycin have a baseline audiogram established, and then regular follow-up hearing tests over the period of their treatment as well as for several months afterwards. It is important to note that ototoxic effects can emerge as much as six months after treatment has ended.

When performing the baseline audiogram, keep in mind that there can have been other factors in this person’s medical history, unrelated to tuberculosis, that could affect hearing. It is possible, and even likely, that the baseline hearing test will not be within normal limits. And that is fine, because that’s not the purpose of ototoxic monitoring. The purpose of the monitoring is to detect changes in hearing. The baseline audiogram is a starting point from which all other subsequent test results will be compared.

Mobile, tablet-based audiometry is often best suited for performing hearing testing in general clinic environments. Firstly, they are highly transportable to remote locations and are shown to work well in low-resource settings. Once the tablet is charged, testing can be performed even if the power is out at the clinic. The best systems offer automated testing, since the highly-skilled and specialized healthcare workers who are treating MDR-TB patients rarely have any background in audiology. It is also highly recommended that you utilize a system that has been clinically validated to produce accurate results in any reasonably quiet location as it is unlikely the TB clinic will offer the luxury of a sound booth.

In many of the MDR-TB studies that we support, testing is performed either monthly, every second month, or when a change in hearing is reported by the patient. The purpose of these tests is to look for any significant change that has occurred since that first baseline test was performed.

To determine whether a shift has occurred, it is important to identify which classification scale your program is opting to subscribe to. These classification scales are being used globally for ototoxicity monitoring, as well as hearing monitoring after other medical conditions. Some of these peer-reviewed and published classification scales have slightly different requirements for adult and pediatric patients. This is in large part because young children are in a critical period of auditory, cognitive, and language development so to give them their best opportunity to achieve their potential we choose to be a bit more aggressive with this group. For example, the CTCAE (Common Terminology Criteria for Adverse Events, version 4.03) is a classification scale often used in U.S. clinical trials that offers different definitions for what it considers a significant shift between children and adults.

Others include the BROCK criteria, SIOP scale, ASHA (American Speech and Hearing Association) classification protocols, and the CHANG scale. Again, some of these have slightly different classification definitions for adult and pediatric patients.

Once the classification scale has been decided upon, testing protocols follow the frequencies taken into consideration by that specific scale, and at each test frequency thresholds are compared against the initial baseline test. For example, if you subscribe to a scale that looks at five frequencies, perhaps starting at 500 Hz, then you would have 10 comparisons to make – five frequencies for each of the patient’s ears.

The classification scales discussed above are used to decide how much of a shift is considered a significant change. 15 years ago, the common consensus amongst clinical audiologists was that the test/retest reliability limit within patients was +/- 10dB. In other words, only a change of more than 10dB was considered a significant shift. However, with advancements made in technology and standardized test protocols, an emerging and debated topic of conversation challenges this number and suggests that the acceptable test/retest limit be lowered to +/- 5dB.

What many these classification scales state is that for a significant shift to be identified, a change in threshold level at more than one frequency per ear would be observed. For example, often in the first level of grading, a minimum of two frequencies in one ear would need to be affected for that change in grading to be met and hence, a shift in hearing recorded.

It is easy to see how analyzing shifts in hundreds – or possibly even thousands – of patients over a 6- to 9-month period could become a data management nightmare. Not to mention the fact that a shift as low as only 5dB becomes somewhat difficult to identify by simply “eyeballing” two audiograms. It is recommended that you look for a solution that will automate the collection and analysis of test results, providing flags and alerts when a shift has been recorded, based on the specific classification scale you subscribe to, as compared to the initial baseline exam. This will help ensure consistency, and provides a certain level of standardization across patients. Two very important factors when you consider that a recorded shift may result in decisions around changing drug regimens and treatment options.

An important consideration in ototoxic monitoring is the test configuration: upper and lower limits of testing, which frequencies to include, what constitutes a significant shift etc. One of the important limiting factors is related to the patient’s overall state of health. Often patients with active TB are not at their best, and energy is limited. The time allotted for testing should be mindful of that. Compounding this, due to the high ambient background noise common in clinical environments, the lower frequencies of 250 and 500Hz can be de-prioritized as these are typically the last frequencies to be affected by ototoxicity. The suggested standard frequency hearing monitoring protocol, taking these factors into consideration, would be to start testing at 1000Hz and include 2000, 4000, 6000 and 8000Hz. Measuring hearing at 6000Hz is typically considered ‘advanced testing’ however as the high frequencies are where ototoxic effects are typically first observed, as well as where tinnitus (ringing in the ears) can affect threshold testing, measuring this additional frequency can serve as not only an added data point, but also as a ‘control’ for 8000 Hz. Inclusion of this frequency could enable the identification of ototoxic effects sooner.

We also recommend adding 1, 2, or 3 of the extended high frequencies depending on the time that you have available to you. The key consideration here would be the health of the patient so if your patient can handle a test of 9 frequencies then we recommend adding these in. These higher frequencies can provide the earliest detection of a shift that may need to be addressed.