In 1980, Guarino published the results of a controlled blind study in which 30 patients with suspected lung disease were examined by one examiner using both ConP and AusP. Using the chest radiograph as the gold standard, the examiners found AusP to be more sensitive than ConP detecting 27 of 28 lung abnormalities.
Recent work has not confirmed initial observations of the usefulness of AusP. In 1989, Bourse et al attempted to determine the diagnostic value of ConA and ConP again using the chest radiograph as the gold standard. Fifty random patients who had a chest radiograph at the time of hospital admission were examined by two physicians alternating between AusP and ConP. Twenty-six of 100 lungs had radiographic abnormalities. No time frame was given as to when radiographs were taken in relation to when the examination was performed. These examiners found ConP to be 15.4 percent sensitive and 97.3 percent specific and AusP to be 19.2 percent sensitive and 85.1 percent specific. mycanadianpharmacy
Bohadana and Kraman tried to determine the mechanism of sound transmission during AusP. In five healthy subjects, they recorded the output audio signal in four regions of the lung while the sternum was being percussed. They concluded that the resonance of the chest cage was the primary factor determining the transmission of sound during AusP with the parenchyma of the lung having only minimal influence.
Each technique we employed in our study relies on sound transmission through all or part of the thorax. The performance characteristics may be influenced by examiner and the type and location of the disease process (spectrum bias). The ConA examination used was limited primarily to breath sounds. Voice sounds such as whispered pectoriloquy and egophony were not elicited. The population we chose to study has a high prevalence of radiographic pulmonary abnormalities heavily weighted toward interstitial lung disease as demonstrated in Table 3. Our study differed from others not only in patient population and in the multiple examiners but in localization of the lesion. We chose to classify each lung as normal, abnormal, or equivocal.
Physical characteristics of the study population may also play a role in the effectiveness of each technique. Our study population has tended to be thin by nature of their disease. Personal observations lead us to believe that AusP is not as effective in obese individuals. Guarino also noted that the sound may be shunted through the bony skeleton in emphysematous chests with increased AP diameters, when the thoracic cage is bony and fixed or when percussion is too intense.
The likelihood ratios of AusP, ConA, and ConP are consistent with a variety of physical diagnostic techniques as reported in the literature. Diagnostic accuracy was good and quite similar across examiners as shown by the areas under the ROC curve. The apparent discrepancy between the diagnostic accuracy reflected in the ROC analysis and the relative pool level of agreement among examiners as reflected in the kappa coefficient may be explained by the following considerations.