Until the middle of the last century, the clinical examination was practically the only basis for the diagnosis and treatment of patients and subsequent studies showed that the combination of anamnesis and physical examination could lead to diagnosis in up to 88% of cases, according to the understanding at that time . Despite the widespread acceptance that anamnesis provides the greatest contribution to diagnosis, some evidence has suggested that physical examination can add major elements to diagnosis in 8.7% to 17% of cases   , which resulted in establishing detailed routines to offer a comprehensive examination. However, the growing appearance of complementary technologies, new understandings on the valuation of clinical findings in the management of patients and new models of doctor-patient relationship have led to a simplification of the physical examination and a consequent reduction in the time of care for patients.
In Brazil, in 2019 DATASUS registered 201.7 million outpatient medical consultations in primary and specialized care by the public health system  and in 2018 the National Agency for Supplementary Health  recorded 274.3 million consultations among beneficiaries of private health plans, which totals almost 500 million consultations every year. In the USA, it was estimated that in 2016 there were 883.7 million outpatient consultations .
Regarding the respiratory system, the examination routine is still taught in universities in a very similar way to what was done more than a century ago. This scenario considers the four domains of physical examination as multiple diagnostic tests and designed for interpretation as tests in parallel, considering that an abnormality anywhere is an indication of disease. Within the context of the past, this potentially allowed to increase the sensitivity and predictive value of disease assessment above the result of each separate test, which was appropriate in the absence of additional tests.
There is currently a greater emphasis on auscultation of the chest and some studies show that its performance is in fact more sensitive     . For other findings not related to auscultation, the results seem inferior in relation to sensitivity    . Additionally, there is a significant variability in agreement between examiners, with kappa generally below 0.5, which raises doubts regarding the reliability of the information for clinical application .
If this is true, a thorough examination of the chest may no longer be justified and we should really start to work with the serial testing model, in which the subsequent physical examination tests are only implemented in the event of an indication of illness after auscultation. Thus, we think it is necessary to examine these possibilities in the real field of work and we decided to evaluate the performance of this strategy, which uses only auscultation as a routine initial test on the diagnosis and prediction of the need to perform the other domains of the physical examination.
2. Materials and Methods
Patients admitted to a general hospital were prospectively selected by a researcher, initially by identifying the diagnosis of hospitalization in the medical record, but then by confirming the diagnostic criteria established by reviewing the clinical data and available complementary exams, provided they had a diagnosis of COPD, atelectasis, pneumonia, pleural effusion, or absence of lung disease. The physical examination was carried out by 2 of 4 other examiners, two medical resident doctors or two medical students, unaware of any clinical information. Initially, auscultation of the chest was performed and recorded, and later the rest of the examination, performed in accordance with current recommendations .
The following definition criteria were used: 1) COPD: history of smoking > 10 pack years, presence of chronic respiratory symptoms, diagnosis of COPD by the attending physician, spirometry in the last 12 months showing an FEV1/FVC post-bronchodilator < 0.7 ; 2) Pleural effusion: of any cause, identified by chest radiogram with frontal and lateral view in the last 3 days, with an estimated volume of at least 500 ml, defined by the obscuration of the diaphragm by the meniscus image in the frontal view ; 3) Atelectasis: of any cause, determined by chest radiogram with frontal and lateral view performed in the last 3 days, defined by the collapse of at least one lobe; 4) Pneumonia: diagnosis of pneumonia by the attending physician, presence of acute respiratory symptoms, leukocytosis and new pulmonary radiological infiltrate; 5) Control group: absence of respiratory disease in no smokers and normal chest X-ray.
The data were analyzed considering the number of medical exams performed and defining as a dependent variable a pulmonary auscultation (normal or abnormal) or presence of disease (absent or present), comparing according to each disease and with the findings in the other domains of physical examination. We also evaluated the effect of abnormal auscultation defined by the presence of adventitious breath sounds, regardless of the intensity and symmetry of the sounds. The data were selected according to their sensitivity, specificity, positive predictive value, negative predictive value and likelihood ratio for the meaning of the dependent variables. Agreement between examiners was assessed using the kappa coefficient. Differences with an alpha error probability less than 0.05 were significant. The data were analyzed using the IBM SPSS 2019 program. The likelihood ratio was assessed with the WINPEPI computer programs for epidemiologists and their teaching potential in 2011.
One hundred and four patients were examined whose sociodemographic characteristics are described in Table 1. Of the total, 88 were seen by 2 examiners and another 16 by only one, making a total of 192 physical examinations. COPD diagnosis was observed in 27 cases, atelectasis in 20 (at least lobar), pleural effusion in 29 (at least 1/4 of the hemithorax), pneumonia in 7 and the remaining 21 did not have lung disease.
Some abnormality on physical examination was identified in 97.3% of patients with diseases of the respiratory system, but the findings showed very low specificity for detecting the disease. An abnormal auscultation had a sensitivity of 85.2% and a PPV of 84.1% to identify the presence of any disease, showing that in fact it is a tool that is useful in the clinical scenario, whereas when we consider, alternatively, as criteria for defining abnormal auscultation only the presence of ABS, we observed a lower sensitivity (Figure 1). Considering an abnormal pulmonary auscultation as a positive diagnostic test, the positive likelihood ratio to indicate disease was 1.53 (95% CI; 1.16 to 2.01) and negative likelihood ratio 0.33 (95% CI; 0.2 to 0, 56). The meaning of abnormal auscultation according to each diagnosis is detailed in Table 2.
On the other hand, the initial finding of normal pulmonary auscultation showed low accuracy to identify healthy individuals, with sensitivity, specificity, NPV and PPV, respectively, of 44%, 43%, 41% and 46% (Figure 2). In case we consider only the absence of ABS as the definition of the diagnostic pattern of the test, we observed some improvement in performance, with sensitivity, specificity, VPN and PPV, respectively, of 59%, 57%, 86% and 23%. However, the accuracy of finding normal auscultation with either of the two definition approaches remains low to be used to conclude that there is no disease in the clinical setting.
Another aspect assessed was the probability of finding abnormalities in the other domains of the physical examination (inspection, palpation and percussion) when identifying normal or abnormal auscultation, in order to assess whether it is really necessary to complete the entire physical examination of the chest. In patients whose examination identified some abnormality on auscultation, 87% showed some additional findings on inspection, 67% on vocal resonance, 78% on chest expansion, 67% on tactile fremitus and 66% on chest percussion. Abnormal auscultation had sensitivity, specificity, NPV and PPV, respectively, of 97%, 17%, 63% and 81% to indicate the presence of additional findings in the other domains of the examination (Figure 3).
Table 1. Demographic and clinical characteristics of the participants.
Table 2. Identification of respiratory disease by auscultation.
*Defined by any type of abnormality; **defined exclusively by the identification of any abnormal sound. Data described in percentage.
Figure 1. Identification of respiratory disease by physical examination.
Figure 2. Meaning of normal auscultation to exclude disease. *Patients with normal lung auscultation. Data showed in percentage.
Figure 3. Meaning of abnormal auscultation to indicate the presence of additional abnormal findings in other domains of physical examination.
The positive LR of an abnormal pulmonary auscultation was 2.23 (95% CI; 1.02 to 4.9) to indicate the presence of additional abnormalities in the remaining physical examination and the negative LR was 0.3 (95% CI; 0.17 to 0.51). In the case of considering only the presence of ABS as abnormality criteria, sensitivity, specificity, NPV and PPV were, respectively, 96%, 10%, 70% and 56% to indicate the presence of other abnormalities.
Analyzing the other possibility, the finding of a normal pulmonary auscultation showed sensitivity, specificity, NPV and PPV, respectively, of 64%, 81%, 97% and 17% to indicate that the other domains of the physical examination are normal, that is, a very low PPV to choose to skip a full physical exam and consequently lose potentially useful information.
Another analysis was made considering the effect of the combined assessment of inspection and auscultation, due to the impossibility of dissociating these two assessments in a real scenario, and the presence of any abnormality in one or both domains showed similar performance than the exclusive use of auscultation, with sensitivity, specificity, NPV and PPV to indicate the presence of some disease, respectively, 97%, 11%, 50% and 79%. On the other hand, when both auscultation and inspection were normal, the performance to indicate that the individual is healthy was lower, with sensitivity, specificity, NPV and PPV, respectively, 3%, 89%, 21% and 50%.
Agreement between examiners considering normal versus abnormal findings was assessed in 88 patients and showed the following results: Kappa = 0.53 for auscultation (p < 0.0001); kappa = 0.44 for identification of adventitious noises (p < 0.0001); kappa = 0.39 for vocal resonance (p < 0.0001); kappa = 0.52 for inspection (p < 0.0001); kappa = 0.37 for thoracovocal fremitus (p = 0.001); kappa = 0.35 for percussion (p = 0.001); kappa = 0.76 for any change in the physical examination present (p < 0.0001).
Pulmonary hyperinflation, consolidation, atelectasis and pleural effusion are the main pleuropulmonary syndromes whose diagnosis is possible through physical examination and represent a set of diseases of relevant prevalence in the respiratory system. In this study, we demonstrate the importance of physical examination in the diagnosis of these syndromes and also that it is not prudent to restrict it only to auscultation, even when combined with inspection, when we want to enjoy the best performance.
The first point to highlight is that our sample of patients does not represent diseases at an early stage, for which the accuracy of the physical examination would certainly be lower. However, the physical examination showed some abnormality in 97.3% of the patients and also revealed a homogeneous distribution with very high sensitivity for diagnosis in all the conditions studied. And it demonstrated a positive and negative likelihood ratio with statistical significance to include or exclude the diagnosis of some disease by physical examination alone, disregarding the pre-test probability arising from the anamnesis, confirming that the physical examination effectively adds power to the clinical diagnosis.
The objective of the study was mainly to understand if it is possible to use it as a routine physical examination of the chest auscultation alone or in combination with inspection to identify if there is a disease. If these were true, it would provide support to carry out this practice, which is becoming more frequent every day, and to reduce the examination time with the patient. However, our results showed that it has low accuracy to be used as a single element of examination. The finding of normal auscultation had a low yield for the identification of healthy patients, occurring in only 46% of them, suggesting that it cannot be used as a marker of normality and the physical examination or further investigation must continue. We understand that one of the possible reasons for this unsatisfactory result is the subjectivity in assessing the symmetry and intensity of the sounds, especially considering that the examiners did not receive any clinical information before the examination. Also, about 15% of individuals with the disease had normal auscultation, perhaps representing the proportion of patients with pulmonary hyperinflation syndrome, which can occur without any abnormality in pulmonary auscultation.
The other evaluated aspect revealed that normal auscultation showed a sensitivity of 64% and PPV of only 17% to indicate that the other domains of the physical examination would result in normal findings, which indicates that the addition of new information obtained in the inspection, palpation and percussion is still probable, that is, if we do the complete physical examination we will identify new abnormalities, which could be the basis for the diagnosis of each disease. The capture of other information is especially important because the possibility of syndromic diagnosis results from the joint analysis of abnormalities in different areas of the exam, allowing the identification of a sufficient composition for the diagnosis of a pattern. The yield was slightly better when both auscultation and inspection were normal, especially in specificity, but this was also shown to be insufficient for the decision process.
Considering the finding of an abnormal pulmonary auscultation as a positive test, we found a relatively good sensitivity and PPV, 85% and 84% respectively, suggesting that auscultation when abnormal is efficient to detect disease, but the low specificity and NPV indicate that the reverse (normal auscultation) also occurs in a significant number of sick individuals, that is, as already mentioned, we were unable to exclude the presence of disease. On the other hand, an abnormal auscultation revealed a PPV of 81% to indicate that there would be some other abnormality in the other physical examination domains and, in the analysis of subgroups classified by physical examination domains, the PPV surpassed 90% for chest expansion, tactile fremitus or percussion abnormalities.
This logic favors the idea that the physical examination needs to continue, particularly when auscultation is normal. Analyzing subgroups, we observed different trends, as in the case of pulmonary consolidation, where the NPV was relatively good, suggesting that auscultation is of greater importance in this situation. This makes sense, as we know that the different pleuropulmonary syndromes have their diagnostic basis in different domains of physical examination.
Thus, this study demonstrates that auscultation alone is an insufficient strategy for tracking diseases or establishing whether the continuity of the physical examination or complementary tests is necessary or not. If the goal is to enjoy the best possible performance through this tool, we still need to do a thorough physical examination of the chest, as we did 100 years ago. We must not forget that the rational use of complementary tests comes from our ability to generate the best hypotheses during patient care. It is possible that the reduction of time spent in consultations and the overuse of complementary tests are reducing the medical ability to perform and interpret the clinical examination and, as a vicious circle, further increasing the request for complementary tests and the cost of the health system.
Auscultation of the chest alone, may not be a sufficient strategy to track diseases or establish whether continuity of the examination is necessary or not.
PHS, Pulmonary hyperinflation syndrome; ATE, atelectasis; CON, consolidation; PE, pleural effusion; Sn, sensitivity; Sp, specificity; NPV, negative predictive value; PPV, positive predictive value; PA, pulmonary auscultation; Abn, abnormal; ABS, Adventitious breath sounds; (+), present; (−), absent; VR, vocal resonance; CE, chest expansion; TF, tactile fremitus; SD, standard deviation.
 Peterson, M.C., Holbrook, J.H., Hales, D., Smith, N.L. and Staker, L.V. (1992) Contributions of the History, Physical Examination, and Laboratory Investigation in Making Medical Diagnoses. The Western Journal of Medicine, 156, 163-165.
 Hampton, J.R., Harrison, M.J.G., Mitchell, J.R.A., Prichard, J.S. and Seymour, C. (1975) Relative Contributions of History-Taking, Physical Examination, and Laboratory Investigation to Diagnosis and Management of Medical Outpatients. British Medical Journal, 2, 486-489.
 Datasus.gov.br [Homepage on the Internet]. Brasilia: Ministério da Saúde—Sistema de Informações Ambulatoriais do SUS (SIA/SUS).
 Ans.gov.br [Homepage on Internet]. Rio de Janeiro: Mapa assistencial da saúde suplementar.
 Rui, P. and Okeyode, T. (2019) National Ambulatory Medical Care Survey: 2016 National Summary Tables.
 Wipf, J.E., Lipsky, B.A., Hirschmann, J.V., Boyko, E.J., Takasugi, J., Peugeot, R.L. and Davis, C.L. (1999) Diagnosing Pneumonia by Physical Examination: Relevant or Relic? Archives of Internal Medicine, 159, 1082-1087.
 Heckerling, P.S., Tape, T.G., Wigton, R.S., Hissong, K.K., Leikin, J.B., Ornato, J.P., Cameron, J.L. and Racht, E.M. (1990) Clinical Prediction Rule for Pulmonary Infiltrates. Annals of Internal Medicine, 113, 664-670.
 Kalantri, S., Joshi, R., Lokhande, T., Singh, A., Morgan, M., Colford, J.M. and Pai, M. (2007) Accuracy and Reliability of Physical Signs in the Diagnosis of Pleural Effusion. Respiratory Medicine, 101, 431-438.
 Badgett, R.G., Tanaka, D.J., Hunt, D.K., Jelley, M.J., Feinberg, L.E., Steiner, J.F. and Petty, T.L. (1994) The Clinical Evaluation for Diagnosing Obstructive Airways Disease in High-Risk Patients. Chest, 106, 1427-1431.
 Straus, S.E., McAlister, F.A., Sackett, D.L., Deeks, J.J. (2000) The Accuracy of Patient History, Wheezing, and Laryngeal Measurements in Diagnosing Obstructive Airway Disease. CARE-COAD1 Group. Clinical Assessment of the Reliability of the Examination-Chronic Obstructive Airways Disease. JAMA, 283, 1853-1857.
 Garcia-Pachon, E. and Padilla-Navas, I. (2006) Frequency of Hoover’s Sign in Stable Patients with Chronic Obstructive Pulmonary Disease. International Journal of Clinical Practice, 60, 514-517.
 Diehr, P., Wood, R.W., Bushyhead, J., Krueger, L., Wolcott, B. and Tompkins, R.K. (1984) Prediction of Pneumonia in Outpatients with Acute Cough—A Statistical Approach. Journal of Chronic Diseases, 37, 215-225.
 Bourke, S., Nunes, D., Stafford, F., Hurley, G. and Graham, I. (1989) Percussion of the Chest Revisited: A Comparison of the Diagnostic Value of Auscultatory and Conventional Chest Percussion. Irish Journal of Medical Science, 158, 82-84.
 Benbassat, J. and Baumal, R. (2010) Narrative Review: Should Teaching of the Respiratory Physical Examination Be Restricted Only to Signs with Proven Reliability and Validity? Journal of General Internal Medicine, 25, 865-872.
 Mattos, W. and Junior, S.S. (2017) Roteiro de Exame—Inspeção, Palpação, percussão e Ausculta. In: Mattos, W., Hilbig, A., Tovo, C.V., Meyer, E.L.S., Lima, M.R.A.A. and Silva, N.B., Eds., Semiologia do Adulto-Diagnóstico Baseado em Evidências, Medbook, Rio de Janeiro, 236-246.
 Blackmore, C.C., Black, W.C., Dallas, R.V. and Crow, H.C. (1996) Pleural Fluid Volume Estimation: A Chest Radiograph Prediction Rule. Academic Radiology, 3, 103-109.