很多听力检测结果中都可以看到：孩子的听力达到“XX dB”，家长们会自我加以比较。实际上在声学中，“dB”所代表的声音强度是不一样的，故特别提出，并加以简要说明：

　　1、大家知道，分贝是用来表示声音强度大小的单位，在物理声学上，它是以测量点的声压P除以基准声压Pr，然后通过对数计算得出的，即：  dB SPL＝20*lg（P/Pr）

　　其中“SPL”的含义是声压级（Sound Pressure Level），这一参数也是国家计量部门用来校准各种听力仪器的基准值。如用常规的测听耳机测试时，纯音听力计必须要输出7.5dB SPL，才能使1000Hz的纯音信号达到0dB HL（IEC318标准），因篇幅关系，在此不一一列举其他频率的基准值。

       2、通常，如果看到的检测报告中的“dB”后面如果没有标注其他参数，那一般代表是用纯音刺激测出的结果，因为目前有国家标准的听力检测用声信号只有纯音（pure tone），它具有频率特性，如1000Hz的纯音就是每秒振荡1000次的正弦波，常用的频率有125，250，500，1000，2000，3000，4000，6000，8000Hz等，纯音被广泛应用于纯音（电）测听、声导抗、耳声发射、多频稳态等听力检测仪器，纯音的声强用“dB HL”表示，“HL”的意思是听力级（Hearing Level）,可以省略不写，直接表示为“XXdB”；配戴助听器的孩子经常要利用纯音测听仪测试香蕉图，家长们看到检测报告中的“dB”，就代表了孩子配戴助听器后能够听到声音的最小刺激量。

　　3、我们多数聋儿家长比较难理解的，就是大家常直接将听觉脑干诱发电位（ABR）的测试结果所标注的“dB”当成是孩子的听阈，其实不然。因为用纯音信号刺激是检测不出清晰的ABR反应的，于是我们采用了另一种声学信号——短声（Click）进行刺激。短声是一种频谱较宽的短时程信号，它的频谱能量较多集中在4000Hz左右，因此ABR的测试结果仅仅能够代表患者高频的听力损失情况，对于低频部分的听阈，ABR无法评估。所以，有些家长会认为医生测得不准，明明被告知在最大强度刺激下，孩子双耳都没有反应，为什么在家时孩子对关门等声音有反应呢？有可能是孩子的低频听力尚有保留，这是ABR自身存在的不足，也是需要进一步检查多频稳态（ASSR）的原因所在。

　　因为各个医院测试环境的不同，国际上也没有针对短时程信号制定相应的校准标准。为了和纯音测试的结果区别，ABR测出的反应阈值用“dB nHL”或“dB SL”表示，“nHL”的意思是正常听力级（normal Hearing Level），“SL”的意思是感觉级（Sense Level），各家医院采用的参数不完全一样，“dB nHL”和“dB SL”所代表的声强也是不同的，所以必须要在dB后面注明参数是“nHL”还是“SL”！

　　4、所以，我们常所说的“dB”，各种根据刺激声的不同，其代表的声音强度级别是不一样的。说了这么多分贝的参数，有没有相互之间换算的方法呢？答案是肯定的，除了前面提到的：0dB HL＝7.5dB SPL，（1kHz纯音）

　　还有最重要的是，各家医院选购的ABR设备不一，短声的最大输出强度也不同，最大的可以达到140dB SPL（相当于110dB nHL），那么SPL与nHL之间的换算为：
 
　　0dB nHL＝28.7dB SPL，（短声，近似于30dB SPL）

　　这样，大家就不难理解，如果ABR的测试结果仅仅是写了个“90dB”，我们就不清楚孩子的高频听阈到底是达到90dB SPL还是nHL？因为如果是nHL，孩子的听阈应该是120dB SPL。

James Hall博士《Handbook of Auditory Evoked Responses》书中换算的表格：

"Subject age is another factor that may interact with stimulus intensity level, but it is not usually noted in discussions of ABR. The determination of 0 dB nHL for various stimulus combinations and test environments is almost always based on ABR data for young adult normal subjects (ages 20-30 years). Average hearing threshold levels in the 1000-4000Hz region for this group are usually in the range of 10-15 dB. Presbycusis (peripheral hearing loss associated with aging) begins insidiously by age 20. Normal-hearing infants and young children, therefore, if they could be uated as accurately as adults, might very well have hearing threshold levels of 0 dB nHL or even better (-5 to-20 dB HL) in the 1000-4000-Hz region. Interpretation of infant ABRs by adult normative intensity standards probably underestimates actual hearing threshold level in some cases. That is, at 0 dB nHL by adult standards, the intensity level of the stimulation is actually about 20 dB nHL for the infant. It is possible to record reliable ABR waveforms for stimulus-intensity levels as low as -5 dB (RE: adult nHL) from apparently normal hearing infants. After these normative data on the equipment intensity level reading corresponding to 0 dB nHL are determined under different stimulus conditions, then a chart can be prepared, indicating the correction values needed to convert equipment intensity level readings to dB nHL. For example, if, for the normal subjed group, the average behavioral threshold level for a click presented at a rate of 21.1/sec with insert earphones in a standard hospital or clinic room is 10 dB, as indicated on the equipment, then this value becomes 0 dB nHL. Therefore, an equipment intensity level reading of 85 dB would, in fact, be 75 dB nHL, an equipment reading of 45 dB would be 35 dB nHL, and so forth.

The correction factor for bone-conduction stimulation is always much larger, usually on the order of 40-45dB. That is, an equipment intensity level reading of 85 dB might correspond to an effective bone-conduction stimulus level of45 dB nHL. With older evoked response systems, an intensity correction chart should be clearly posted on or near the equipment for all users. Newer software-based evoked response systems permit internal correction for stimulus intensity. That is, after determining the relationship between equipment reading and dB nHL, the clinician can alter the intensity reading on the equipment so that it reflects dB nHL for the equipment in the usual test environment."

Hall, J.W. (1992). Calibration and Normative Data. In Handbook of Auditory Evoked Response (pp. 274). Boston : Allyn and Bacon.

引用其他的材料来解释一下：
Is my clinical ABR equipment calibrated properly for tone-evoked ABR testing?

Probably not. An often overlooked yet important area for obtaining accurate ABR threshold measures is the calibration of tonal and/or click stimuli. In conventional pure-tone behavioural audiometry, behavioural thresholds are expressed in dB HL units, while ABR thresholds are expressed in dB nHL units. Normal hearing level (nHL) refers to the normal threshold for click or brief-tone stimuli. Zero dB nHL will differ for tones of different frequency and duration. Behavioural thresholds for long-duration pure tones, used in conventional audiometry, are significantly lower (i.e., more sensitive) than those for the same frequency brief tones, due to the temporal integration properties of the auditory system. [For example, 0 dB HL for a long-duration 1000-Hz pure tone transduced through a TDH-49 earphone = 7 dB SPL, whereas 0 dB nHL for a brief (2-1-2 cycle) 1000-Hz tone = 23.2 dB ppe SPL.] What is important to realize is that the majority of clinical ABR systems calibrate their tonal stimuli in dB HL units. This is not appropriate for ABR testing, and will lead to an overestimation of ABR thresholds (i.e., conclude ABR thresholds are higher than they actually are). Clinicians must see to it that their equipment is recalibrated, or they must convert the dB HL values into dB nHL values.

dB ppe SPL = dB peak (peak hold) minus 3 dB (for tones)

PS：

TIPS FOR CLINICIANS

THRESHOLD TONE-EVOKED
AUDITORY BRAINSTEM RESPONSE