8.1 Prescriptive Fitting Rationales: NAL-NL2 vs. DSL v5

8.1 Prescriptive Fitting Rationales: NAL-NL2 vs. DSL v5

[!NOTE] In modern hearing instrument dispensing, selecting an appropriate prescriptive fitting formula is the foundational step of programmatic verification. Hearing aid programming is not a matter of subjective manufacturer "first-fit" algorithms—which are proprietary, unstandardized, and frequently under-prescribe gain by 5 to 15 dB in the high frequencies—but rather an adherence to validated, peer-reviewed prescriptive targets verified via probe-microphone real-ear measurements (REM). On the NBC-HIS National Competency Examination, specialists must demonstrate command over the distinct psychoacoustic philosophies, mathematical modeling, and clinical indications separating NAL-NL2 from DSL v5.

---

Historical Evolution: From Linear Gain to Non-Linear Modeling

Prior to the establishment of scientific prescriptive formulas, hearing aid selection relied on comparative evaluation methods (e.g., the Carhart 1946 protocol), wherein patients listened to multiple physical instruments in a sound booth while the clinician measured comparative word recognition scores. This approach was time-consuming, prone to high test-retest variability, and devoid of individualized electroacoustic precision.

The Linear Prescriptive Era

The advent of standardized pure-tone audiometry catalyzed the development of linear prescriptive formulas designed to calculate a fixed acoustic gain response regardless of input level:

1. The Half-Gain Rule (Lybarger, 1944): Recognized that individuals with sensorineural hearing loss prefer acoustic gain equal to approximately half of their pure-tone hearing threshold level (Gain ≈ 0.5 × HTL). The remaining 50% of the loss is offset by the natural dynamic range compression of the human cochlea and recruitment.
2. POGO & POGO II (Prescription of Gain and Output; McCandless & Lyregaard, 1983, 1988): Formalized the half-gain rule with a specific low-frequency roll-off (-10 dB at 250 Hz, -5 dB at 500 Hz) to counteract the upward spread of masking. POGO II introduced an additional high-gain correction factor for hearing threshold levels exceeding 65 dB HL.
3. Berger Method (Berger, Hagberg, & Rane, 1977): Applied frequency-dependent fractional multipliers to pure-tone thresholds (e.g., 0.45 × HTL at 500 Hz; 0.55 × HTL at 2000 Hz).
4. NAL-R and NAL-RP (National Acoustic Laboratories - Revised / Revised Profound; Byrne & Dillon, 1986; Byrne, Parkinson, & Ching, 1990): Calculated frequency-specific gain using the formula:

Gain(f) = X + 0.31 × HTL(f) + H

Where X is an overall gain factor based on the three-frequency pure-tone average, and H is a frequency-specific slope factor designed to equalize speech loudness across frequencies for an average 65 dB SPL speech input. NAL-RP extended this formula to severe-to-profound losses by adding a non-linear threshold modifier for thresholds exceeding 60 to 70 dB HL.

The Non-Linear Digital Paradigm Shift

Linear formulas operated under the assumption that the hearing aid provided a constant amount of gain across all acoustic input levels. However, sensorineural hearing loss involves outer hair cell damage, causing loss of cochlear non-linear amplification (recruitment) and a severely constricted dynamic range.

Modern digital instruments employ Wide Dynamic Range Compression (WDRC), necessitating multi-level prescriptive algorithms that simultaneously prescribe distinct target curves for soft speech (50 to 55 dB SPL), moderate conversational speech (65 dB SPL), and loud sounds (75 to 80 dB SPL), alongside maximum output limits (MPO / RESR).

                         LINEAR vs. NON-LINEAR GAIN BEHAVIOR

          Linear Amplification (POGO / NAL-R)      Non-Linear WDRC (NAL-NL2 / DSL v5)
       ┌──────────────────────────────────────┐  ┌──────────────────────────────────────┐
       │ Gain is CONSTANT across all inputs.  │  │ Gain is LEVEL-DEPENDENT.             │
       │                                      │  │                                      │
  Gain │        ┌───────────────────┐         │  │ 50 dB In │ ═══════════════ (High Gain)│
  (dB) │ 50 dB  │    Fixed Gain     │         │  │ 65 dB In │ ─────────── (Moderate Gain)│
       │ 65 dB  │    Regardless     │         │  │ 80 dB In │ ────── (Low Gain)          │
       │ 80 dB  │    of Input Level │         │  │          └─────────────────────────── │
       │        └───────────────────┘         │  │ Soft sounds amplified; loud sounds    │
       │ Soft sounds under-amplified;         │  │ compressed to protect residual hearing│
       │ loud sounds exceed discomfort level. │  │ and prevent recruitment distortion.   │
       └──────────────────────────────────────┘  └──────────────────────────────────────┘

---

The Dual Core Audiological Philosophies: Equalization vs. Normalization

Modern non-linear prescriptive science is polarized between two distinct psychoacoustic philosophies: Loudness Equalization and Loudness Normalization.

                       AUDIOLOGICAL PHILOSOPHY COMPARISON

           LOUDNESS EQUALIZATION                     LOUDNESS NORMALIZATION
             (NAL-NL1 / NAL-NL2)                          (DSL 3.0 / DSL v5)
   ┌─────────────────────────────────────┐   ┌─────────────────────────────────────┐
   │ Core Premise:                       │   │ Core Premise:                       │
   │ Equalize loudness contribution      │   │ Restore normal loudness perception  │
   │ across all speech frequency bands.  │   │ across all frequencies and levels.  │
   ├─────────────────────────────────────┤   ├─────────────────────────────────────┤
   │ Speech Intelligibility:             │   │ Speech Intelligibility:             │
   │ Maximizes Speech Intelligibility    │   │ Maximizes audibility across the full│
   │ Index (SII) within comfortable      │   │ speech spectrum (broadband speech   │
   │ overall acoustic loudness.          │   │ cues preserved).                    │
   ├─────────────────────────────────────┤   ├─────────────────────────────────────┤
   │ Perceptual Outcome:                 │   │ Perceptual Outcome:                 │
   │ All frequency bands sound equally   │   │ Soft sounds sound soft; average     │
   │ loud; total loudness ≤ normal ear.  │   │ sounds average; loud sounds loud.   │
   ├─────────────────────────────────────┤   ├─────────────────────────────────────┤
   │ Primary Target Demographic:         │   │ Primary Target Demographic:         │
   │ Post-lingually impaired adults;     │   │ Pediatric fittings (infants/children│
   │ patients seeking comfort in noise.  │   │ and severe-to-profound adult losses.│
   └─────────────────────────────────────┘   └─────────────────────────────────────┘

Loudness Equalization (The NAL Framework)

Developed by the Australian National Acoustic Laboratories, Loudness Equalization posits that speech intelligibility is maximized when all frequency sub-bands of speech contribute equally to the total perceived loudness.

• In unamplified speech, low-frequency vowel formants carry substantially greater acoustic energy (approximately 60 dB SPL) than high-frequency unvoiced fricatives and stops (e.g., /s/, /θ/, /k/, which peak around 30 to 45 dB SPL).
• If all frequencies were amplified to match normal loudness, low frequencies would dominate total perceived loudness, leading to the upward spread of masking and listening fatigue.
• NAL algorithms deliberately adjust frequency gain so that each critical band of speech generates equal specific loudness in sones, subject to the mathematical constraint that the overall loudness of the amplified signal does not exceed the overall loudness experienced by a normal-hearing listener hearing the same sound.

Loudness Normalization (The DSL Framework)

Developed at the University of Western Ontario, the Desired Sensation Level (DSL) philosophy is grounded in Loudness Normalization. Its objective is to restore normal loudness growth at each frequency:

• A sound presented at 50 dB SPL should sound "soft" to the impaired listener, exactly as it sounds to a normal listener.
• A sound at 65 dB SPL should sound "comfortably loud".
• A sound at 85 dB SPL should sound "loud, but OK".
• DSL does not equalize loudness across frequencies; instead, it preserves the natural spectral tilt of acoustic stimuli while ensuring that the full dynamic range of speech remains audible within the patient's residual auditory area (between hearing threshold and uncomfortable loudness level).

                           LOUDNESS GROWTH FUNCTIONS

          Perceived
          Loudness
          (Category)

          Uncomfortably ┼                                    / Normal Hearing
          Loud          │                                   /  Ear
                        │                                  /  
          Loud          ┼                         ╭───────/ ◄── DSL v5 Target
                        │                        ╭╯      /      (Normalizes slope)
          Comfortable   ┼                       ╭╯      /   
                        │                      ╭╯      /  ◄──── NAL-NL2 Target
          Soft          ┼                     ╭╯      /         (Balances intelligibility)
                        │                    ╭╯      /    
          Barely Audible┼                   ╭╯      /   ◄────── Impaired Ear
                        │                  ╭╯      /            (Abnormal recruitment,
                        │                 ╭╯      /              steep loudness growth)
          Inaudible     ┼─────────────────┴──────/──────────────────────────────
                        0   10  20  30  40  50  60  70  80  90  100  Input (dB SPL)

---

National Acoustic Laboratories (NAL) Rationales: NAL-NL1 to NAL-NL2

The Limitations of NAL-NL1

Introduced in 1998, NAL-NL1 was the first non-linear fitting formula derived via mathematical optimization of the Speech Intelligibility Index (SII) (ANSI S3.5) combined with the Moore & Glasberg loudness model. Although revolutionary, real-world clinical deployment uncovered significant limitations:

1. Excessive High-Frequency Gain: For listeners with severe-to-profound high-frequency hearing loss, NAL-NL1 prescribed substantial high-frequency amplification. When cochlear dead regions were present, this amplification provided no usable intelligibility benefit and instead generated auditory distortion, screeching, and acoustic feedback.
2. Excessive Compression Ratios: NAL-NL1 frequently prescribed high compression ratios (CR > 3.0:1) across the speech frequencies, which smeared the temporal envelope of speech, flattened phonemic intensity contrasts, and degraded speech-in-noise perception.
3. Poor Patient Acceptance: Mild-to-moderate adult wearers frequently rejected NAL-NL1 initial fittings due to perceived unnatural sharpness and excessive high-frequency background noise.

The NAL-NL2 Empirical Model

Released in 2011, NAL-NL2 (Keidser, Dillon, Flax, Ching, & Brewer) departed from purely theoretical mathematical modeling. The designers trained a neural network on an empirical database of over 1,500 hearing aid users across multiple international clinical trials to determine the exact acoustic gain and compression settings that real-world patients actually prefer and utilize effectively.

Core Acoustic Modifications in NAL-NL2 Relative to NAL-NL1:

• Gain Reductions for Mild-to-Moderate Losses: NAL-NL2 prescribes approximately 2 to 4 dB less gain across conversational input levels (65 dB SPL) for mild-to-moderate hearing losses, substantially improving initial sound comfort.
• Low-Frequency and High-Frequency Redistribution: Prescribes slightly more low-frequency gain and significantly less gain in severe high-frequency regions, recognizing the physiological limits of damaged basal cochlear hair cells.
• Lower Compression Ratios: Decreases compression ratios across channels (averaging 1.5:1 to 2.2:1) to preserve speech envelope temporal cues.

┌─────────────────────────────────────────────────────────────────────────────┐
│                     NAL-NL2 EMPIRICAL ADJUSTMENT FACTORS                     │
├─────────────────────┬───────────────────────────────────────────────────────┤
│ Factor              │ Specific Modeling Adjustment                          │
├─────────────────────┼───────────────────────────────────────────────────────┤
│ User Experience     │ Experienced users receive +2 to 3 dB higher overall   │
│                     │ gain than new users across all input levels.          │
├─────────────────────┼───────────────────────────────────────────────────────┤
│ Gender              │ Female listeners receive ~1 to 2 dB less gain than    │
│                     │ male listeners (accounting for dynamic range norms).  │
├─────────────────────┼───────────────────────────────────────────────────────┤
│ Language Tonality   │ Tonal languages (e.g., Mandarin, Thai) receive +2 to  │
│                     │ 4 dB low-frequency boost (<500 Hz) for pitch contours.│
├─────────────────────┼───────────────────────────────────────────────────────┤
│ Pediatric vs. Adult │ Children receive +5 dB higher gain for low-level      │
│                     │ speech (50 dB SPL) to support incidental learning.    │
├─────────────────────┼───────────────────────────────────────────────────────┤
│ Cognitive Status    │ Lower compression ratios prescribed for older adults  │
│                     │ with reduced cognitive / temporal processing speed.   │
├─────────────────────┼───────────────────────────────────────────────────────┤
│ Bilateral Fitting   │ Bilateral gain reduction of 2 to 6 dB applied across  │
│ (Summation)         │ channels to account for central binaural summation.   │
└─────────────────────┴───────────────────────────────────────────────────────┘

Specific Modifiers in NAL-NL2

1. Bilateral Summation Correction

When sound is presented to both ears simultaneously, the central auditory nervous system integrates neural firing from both cochleae at the level of the brainstem (superior olivary complex), resulting in binaural loudness summation:

• If a hearing aid is fitted unilaterally, the instrument must provide full target gain to achieve comfortable loudness.
• When fitted bilaterally, binaural summation increases perceived loudness by approximately 3 to 6 dB.
• To prevent over-amplification, NAL-NL2 automatically applies a bilateral gain reduction of 2 to 6 dB across channels. The reduction scales dynamically with input level: approximately 2 dB reduction for soft inputs (50 dB SPL), scaling up to 5 to 6 dB reduction for high inputs (80 dB SPL) where loudness summation is most pronounced.

2. User Experience Level

Auditory acclimation is a documented neurological phenomenon. New hearing aid users are unaccustomed to environmental high-frequency sounds and possess lower loudness tolerance ceilings:

• New Users: NAL-NL2 prescribes 2 to 3 dB less gain across all frequencies to reduce sensory overload and enhance adaptation.
• Experienced Users: Prescribes full prescriptive gain (+2 to 3 dB higher than new users), maximizing acoustic contrast and speech intelligibility.

3. Language Type: Tonal vs. Non-Tonal Languages

Speech acoustics vary fundamentally across linguistic typologies:

• Non-Tonal Languages (e.g., English, Spanish, German): Semantic meaning is conveyed primarily via phonemic consonant identification (formants, bursts, frication noise between 1500 Hz and 6000 Hz). NAL-NL2 emphasizes mid-to-high frequency amplification.
• Tonal Languages (e.g., Mandarin, Cantonese, Vietnamese, Thai): The lexical meaning of a monosyllabic word changes completely depending on its fundamental frequency (F0) pitch contour (e.g., in Mandarin, the syllable "ma" can mean mother, hemp, horse, or scold depending on pitch contour). Because F0 and the first harmonic reside in the low frequencies (100 Hz to 400 Hz), NAL-NL2 provides an empirical boost of 2 to 4 dB below 500 Hz for tonal language speakers while slightly reducing high-frequency gain.

4. Gender Differences

Empirical data collected during NAL-NL2 development revealed that female listeners systematically selected lower gain settings than male listeners with identical pure-tone audiometric configurations. NAL-NL2 applies a 1 to 2 dB reduction in overall target gain for female listeners, reflecting differences in dynamic range and loudness discomfort levels.

---

Desired Sensation Level (DSL) Rationales: DSL 3.0 to DSL v5

The Pediatric Roots of DSL

Developed by Richard Seewald and colleagues at the University of Western Ontario's Child Amplification Laboratory, the Desired Sensation Level (DSL) rationale was conceived specifically to address the unique audiological requirements of infants, toddlers, and young children.

Unlike adults who acquire hearing loss post-lingually, infants and young children:

1. Cannot provide subjective behavioral feedback during programming.
2. Must acquire language, speech phonology, and grammar through incidental acoustic exposure (overhearing conversations at distances greater than 3 feet).
3. Require higher Signal-to-Noise Ratios (SNRs) (+10 to +15 dB) and higher sensation levels across the entire speech frequency spectrum (250 Hz to 8000 Hz) to develop stable mental phonemic representations.

The Individualized RECD Requirement

Infant ear canals are substantially smaller in physical volume (0.3 to 0.6 cm³) compared to the standardized 2.0 cm³ coupler or adult ear canal (1.0 to 1.5 cm³). Under acoustic physics:

SPL = Acoustic Power / (Acoustic Impedance / Volume)

A hearing aid delivering 100 dB SPL into a 2cc coupler will generate 115 to 125 dB SPL in an infant's ear canal. DSL pioneered the mandatory clinical integration of the Real-Ear-to-Coupler Difference (RECD), converting all 2cc coupler electroacoustic specifications into individualized dB SPL at the tympanic membrane to ensure acoustic safety and audibility precision.

Evolution to DSL v5 (DSL m[i/o])

In 2005, Scollie et al. introduced DSL v5 (DSL multi-stage input/output), expanding the rationale from a purely pediatric algorithm into a comprehensive lifespan fitting rationale.

                               DSL v5 ARCHITECTURE

                             ┌───────────────────────┐
                             │     DSL v5 ENGINE     │
                             └───────────┬───────────┘
                                         │
                 ┌───────────────────────┴───────────────────────┐
                 ▼                                               ▼
     ┌───────────────────────┐                       ┌───────────────────────┐
     │   DSL v5 PEDIATRIC    │                       │     DSL v5 ADULT      │
     ├───────────────────────┤                       ├───────────────────────┤
     │ • Maximum Audibility  │                       │ • Reduced Gain Targets│
     │   (Broadband 250-8kHz)│                       │   (3-5 dB lower than  │
     │ • Higher Sensation    │                       │    pediatric targets) │
     │   Levels for Soft In  │                       │ • Prioritizes Comfort │
     │ • Separate Quiet vs.  │                       │ • Prevents Loudness   │
     │   Noise Prescriptions │                       │   Discomfort in Noise │
     │ • Strict RESR Limiting│                       │ • Real-Ear Alignment  │
     └───────────────────────┘                       └───────────────────────┘

Key Features of DSL v5:

1. DSL v5 Pediatric vs. DSL v5 Adult:
   • Pediatric targets maximize sensation level and audibility across all frequencies.
   • Adult targets prescribe 3 to 5 dB less gain than pediatric targets for identical audiograms. DSL v5 recognizes that post-lingually deafened adults experience sensory overload, annoyance, and speech degradation in noise when subjected to full pediatric audibility targets.
2. Quiet vs. Noise Prescriptions: DSL v5 calculates separate target curves depending on the acoustic listening environment:
   • DSL Quiet: Maximizes audibility for low-level and conversational speech.
   • DSL Noise: Reduces low-frequency gain and slightly lowers mid-frequency gain to prevent the upward spread of masking and reduce listening effort in noisy backgrounds.
3. Real-Ear Saturation Response (RESR) Targets: DSL v5 generates explicit frequency-specific maximum output limits based on measured or age-predicted Uncomfortable Loudness Levels (UCL), ensuring the hearing aid amplifier compresses loud transient sounds prior to reaching patient discomfort.

---

Head-to-Head Comparative Analysis: NAL-NL2 vs. DSL v5

                        PRESCRIPTIVE TARGET DISPARITY
             Target Real-Ear Aided Response (REAR) for 65 dB SPL Input

      Output (dB SPL)

       100 ┼
        90 ┼                           ╭───────╮ ◄── DSL v5 Pediatric Target
        80 ┼                  ╭────────╯       ╰───────╮ (Higher overall gain;
        70 ┼         ╭────────╯   ┌───────────────────┐ ╰─── broad bandwidth)
        60 ┼─────────╯            │  3 to 6 dB Delta  │
        50 ┼                      └───────────────────┘ ╭─── NAL-NL2 Target
        40 ┼                  ╭────────────────────────╯     (Loudness equalized;
        30 ┼         ╭────────╯                               comfort in noise)
        20 ┼─────────╯
           └─────────┴────────┬────────┬───────┬───────┬────────
            250     500      1000     2000    4000    8000  Frequency (Hz)

---

Clinical Decision Framework: Rationale Selection

                             ┌─────────────────────────────┐
                             │     PATIENT EVALUATION      │
                             └──────────────┬──────────────┘
                                            │
                 ┌──────────────────────────┴──────────────────────────┐
                 ▼                                                     ▼
     ┌───────────────────────┐                             ┌───────────────────────┐
     │   PEDIATRIC PATIENT   │                             │     ADULT PATIENT     │
     │   (Infant to Teen)    │                             │  (Post-Lingual Onset) │
     └───────────┬───────────┘                             └───────────┬───────────┘
                 │                                                     │
                 ▼                                                     ▼
     ┌───────────────────────┐                             ┌───────────────────────┐
     │    MANDATE: DSL v5    │                             │ Evaluate Audiometric  │
     │      (Pediatric)      │                             │ & Lifestyle Profile   │
     │ • Maximize audibility │                             └───────────┬───────────┘
     │ • Full speech cues    │                                         │
     │ • Individual RECD     │                 ┌───────────────────────┴───────────────────────┐
     └───────────────────────┘                 ▼                                               ▼
                                   ┌───────────────────────┐                       ┌───────────────────────┐
                                   │   MILD-TO-MODERATE    │                       │  SEVERE-TO-PROFOUND   │
                                   │    SLOPING LOSS       │                       │    OR PRE-LINGUAL     │
                                   ├───────────────────────┤                       ├───────────────────────┤
                                   │   SELECT: NAL-NL2     │                       │   SELECT: DSL v5      │
                                   │ • Conversational ease │                       │       (Adult)         │
                                   │ • Comfort in noise    │                       │ • High audibility need│
                                   │ • Natural loudness    │                       │ • Established acoustic│
                                   │ • Acclimation options │                       │   sensation dependence│
                                   └───────────────────────┘                       └───────────────────────┘

Clinical Case Applications

Case 1: The New Adult User with Presbycusis

• Patient: A 71-year-old retired engineer with bilateral moderate sensorineural hearing loss sloping from 25 dB HL at 500 Hz to 65 dB HL at 4000 Hz. Reports severe difficulty hearing grandchildren and understanding speech in restaurants.
• Rationale Selection: NAL-NL2 (New User Profile).
• Clinical Justification: Post-lingually deafened adults with presbycusis possess normal low-frequency hearing and are highly vulnerable to loudness intolerance and background noise complaints. NAL-NL2's loudness equalization maximizes speech intelligibility in the consonant range without over-amplifying low-frequency ambient sounds. Selecting the "New User" profile reduces gain by 2 to 3 dB initially, allowing comfortable sensory adaptation before stepping up to full targets.

Case 2: The Infant Identified via Newborn Hearing Screening

• Patient: An 8-month-old infant diagnosed with bilateral moderate sensorineural hearing loss (45-50 dB HL across all frequencies) via diagnostic Auditory Brainstem Response (ABR).
• Rationale Selection: DSL v5 Pediatric (with measured RECD).
• Clinical Justification: The infant must acquire syntax, phoneme categorization, and self-monitoring vocal control. DSL v5 provides the critical sensation levels required across the entire frequency range. Coupling must be verified via individualized RECD to avoid acoustic overexposure in the small infant ear canal.

Case 3: The Severe-to-Profound Long-Time Adult Wearer

• Patient: A 52-year-old teacher with congenital, severe bilateral sensorineural loss (75 to 85 dB HL flat), fitted since childhood with power BTEs.
• Rationale Selection: DSL v5 Adult.
• Clinical Justification: Having developed speech and language using early linear and DSL amplification, this patient relies on maximum acoustic sensation levels. Switching this patient to NAL-NL2 would result in an immediate complaint that the instruments sound "dead, weak, muffled, and lacking clarity" due to NAL-NL2's lower overall gain targets.