Version October 2024
INTRODUCTION — Correct interpretation of pulmonary function tests (PFTs) requires the use of appropriate reference values to which the patient's results are compared [1-4].
Unlike many physiologic parameters for which normal values do not vary with the individual characteristics of the patient, predicted values of pulmonary function depend upon age, sex, and height. Normal values appear to also vary with the patient's race or ethnic background, but the role played by environmental and social determinants of health in this observation is unclear. Therefore, interpretation of PFTs performed for the first time must take multiple factors into consideration. In practice, spirometers and pulmonary function test equipment have software that uses reference equations for calculation of "predicted values," as determined by published studies of large numbers of healthy individuals [3,4].
The American Thoracic Society (ATS)/European Respiratory Society (ERS) statement on the standardization of spirometry, as well as other ATS guidelines, can be accessed through the ATS web site at www.thoracic.org/statements.
The effects of age, sex, and height on the determination of normal predicted values for interpretation of PFTs, and the use of different predicted reference sets, are reviewed here. We also comment on the effects of weight and race/ethnicity. Finally, we discuss using statistical approaches to define normality. The technique and interpretation of PFTs are discussed separately. (See "Office spirometry" and "Overview of pulmonary function testing in adults" and "Overview of pulmonary function testing in children" and "Diffusing capacity for carbon monoxide".)
EFFECT OF AGE, SEX, HEIGHT, AND WEIGHT — PFT results are dependent upon age, sex, and height. Although weight is not a determining factor of lung size or function, body weight may influence lung function results.
Age and sex
Childhood — During childhood up until puberty, lung function increases linearly in proportion to overall growth, which is determined at least in part by age and sex [5,6]. The growth spurt associated with early adolescence is associated with an increase in the rate of overall growth and the rate of increase in lung volumes (and maximal flows) and occurs at slightly different ages in girls and boys (figure 1A and figure 1B).
On average, girls reach both their maximal height and their maximal lung volume earlier than boys, but boys achieve larger lung volumes.
In children, the normal growth of lung function is fast when compared to the rate of changes observed with disease or therapy. As a result, the use of growth charts (looking for trends across percentile lines) most accurately reflects changes that cannot be attributed to growth (figure 1A-B).
Adulthood — Cross-sectional spirometric testing of large healthy populations shows a plateau of lung function between the ages of 20 and 30 (figure 2A-B). However, longitudinal observation of individuals shows that some have a lung function peak in their early 20s, while others, particularly males, may have the peak in their mid-30s [7].
Healthy nonsmokers experience a gradual decline in lung function throughout adulthood and old age:
●The forced expiratory volume in one second (FEV1) decreases approximately 20 to 30 mL per year [8,9]
●The vital capacity (VC) decreases while the residual volume (RV) increases, leaving the total lung capacity intact
●The diffusing capacity for carbon monoxide (DLCO, also known as transfer factor) declines linearly with age
Among adults, sex during development is an important predictor of lung size. While gender identify must be respected, current recommendations include sex at birth (through puberty) as the basis for predicted values of lung function [10]. The effect of gender-affirming hormonal therapy on lung function is poorly understood. The timing of gender reassignment may impact lung development and should be considered when interpreting results [10-12].
Height — Taller persons have a larger frame size and a larger thoracic cage than do shorter persons. Consequently, taller persons have:
●Larger lung volumes
●Higher maximal flow rates
●A greater ability to take up oxygen and carbon monoxide per minute (reflected by a higher diffusing capacity for carbon monoxide [DLCO])
As an example, the predicted (mean) vital capacity of a 40-year-old man who is 6 feet 4 inches (193 cm) is six liters, while that of a man who is a foot shorter at 5 feet 4 inches (163 cm) is four liters. Therefore, the accurate measurement, preferably with a stadiometer, of standing height (to the nearest half inch or tenth of a cm, without shoes) is very important in predicting lung function for a particular patient [3]. Technicians should not rely on the stated height. Sitting height is thought to be potentially a more relevant anthropomorphic determinant of lung size, although the contribution of sitting height is highly variable across studies and only accounts for a small proportion of observed differences in lung size among people [13,14].
Weight — Body weight is much less important than height when predicting most pulmonary function values; as a result, weight is not included in spirometry prediction equations. However, extremes in weight are associated with changes in lung volumes [15-20]. The body mass index (BMI), which is obtained by dividing the weight (in kilograms) by the square of the height (in meters) (calculator 1), can be used to quantify obesity.
The most consistent effect of obesity is reduction in FRC and expiratory reserve volume (ERV) (figure 3) [19,20]. Mildly reduced lung volumes in patients with a BMI above 30 kg/m2 may be entirely due to obesity [18]. Further increases in BMI above 30 kg/m2 result in additional decreases in FRC and ERV [20]. The effect on spirometry tends to be small. FEV1 and FVC may be slightly reduced and the ratio is usually preserved, particularly in adults. (See "Overview of pulmonary function testing in adults", section on 'Lung volumes'.)
EFFECT OF RACE/ETHNICITY — The role of race and ethnicity in reference values for healthy individuals is a controversial topic in pulmonary medicine. The traditional approach used by the standard reference equations (ie, the Global Lung Function Initiative [GLI] and the National Health and Examination Survey [NHANES]) assigns different normal ranges for spirometry according to race and ethnicity. However, this approach assumes that there are true biologic differences in these values among different populations and that these differences are not medically relevant.
A growing body of evidence has called these assumptions into question:
●In the Atherosclerosis Risk in Communities cohort, the association between survival and lung function was similar between Black and White populations when lung function was interpreted using NHANES III White equations for everyone, but mortality was higher for a given lung function in Black patients when assessed using NHANES III race-specific equations [21]. The authors suggested that because mortality correlated with absolute lung size (forced vital capacity [FVC]) in both populations similarly, use of race-specific equations normalizes the lower lung function of Black patients and leads to an apparent excess mortality when comparing the same percent predicted values. This finding was confirmed in the larger NHANES III population using NHANES III White versus race-specific reference equations [22] and separately by using the GLI-Other versus race-specific GLI reference equations [23].
●For the broader population, one multiethnic cohort study of 3600 individuals in the United States demonstrated that use of race/ethnicity-based spirometric equations failed to improve the prediction of chronic lower respiratory disease events and mortality compared with race-neutral equations [24]. Similarly, use of race-specific equations based on GLI failed to improve prediction of breathlessness or prognosis in the NHANES III population compared with race-neutral equations (GLI-Other) [25]. Recent data confirm that race-specific reference equations do not improve models of dyspnea nor relate better to quantitative CT measures of lung disease than race-free or race-neutral reference equations [26].
●A separate line of research has more specifically examined underdiagnosis of COPD and COPD severity. Two large cohort studies of patients with or at-risk for COPD (SPIROMICS and COPDGene) have shown that race-neutral equations greatly improve the correlation between spirometry measures and other markers of COPD severity such as the COPD assessment test, modified Medical Research Council dyspnea score, six-minute walk distance, St. George's respiratory questionnaire, and CT measures of emphysema or airway wall thickness [27,28]. These findings were driven by significant underestimation of COPD severity when using race-specific spirometric equations for Black participants. A retrospective clinical cohort of over 10,000 individuals with COPD demonstrated that reclassification of disease severity using race-neutral reference equations corresponded well with risk of COPD exacerbation or hospitalization. Individuals who shifted to a more severe disease category had an increased risk of these outcomes, and those who shifted to a less severe disease category had a reduced risk of COPD exacerbation and hospitalization [29].
In a population-based cohort study (CARDIA), 6.5 percent of Black males and 6.6 percent of Black females with low-normal lung function (forced expiratory volume in one second [FEV1] between 80 percent and 99 percent predicted) based on race-including equations demonstrated emphysema on chest CT, a rate 3.9- and 1.9-fold higher than that seen in White males and females, respectively [30]. The use of race-neutral equations approximately halved these racial differences in undiagnosed emphysema prevalence in males and eliminated the racial differences in females.
Together, these data suggest that an unintended clinical consequence of race/ethnicity-specific equations can be to underestimate the severity of pulmonary disease in Black populations. Based on these findings, there have been calls to use one set of normative values rather than race-based equations for standardization of lung function interpretation [27,31].
The 2022 European Respiratory Society (ERS)/American Thoracic Society (ATS) technical standards acknowledge that interpretation in reference to ancestral grouping may only be useful in certain contexts and that there are trade-offs involved in a population-based rather than a universal standard [32]. The GLI 2022 "race neutral" equation (termed "GLI-Global") is an alternative equation derived as a weighted average of available GLI data from European, African American, North East Asian, and South East Asian racial/ethnic groups; it can be applied universally, but has wider limits of normal than race- and ethnicity-specific equations [14]. Following the publication of the ERS/ATS technical standards, a working group convened by the American Thoracic Society recommended replacing race and ethnicity-specific equations with race-neutral averaged reference equations such as GLI-Global while also acknowledging that additional research is merited to evaluate the consequences of these changes [33]. Although the shift from race-specific to race-neutral approaches does lead to tradeoffs, discussed in more detail below, our authors favor this transition away from use of race and ethnicity in spirometric interpretation.
Several controversies remain regarding the use of the GLI-Global equation. Some have questioned whether the "GLI-Global" equation is truly race-neutral since it is still derived from data from people of different self-reported racial or ethnic backgrounds. It has been suggested that "GLI-Global" more appropriately be considered a "race composite" equation [34]. While this is more accurate, the term "race-neutral" was meant to convey that selection of race is not necessary to employ this equation. The "GLI-Other" equation (2012) can be used similarly to allow a race-neutral approach when using software in which the "GLI-Global" equation is not available. Clearly listing the reference equations used to generated predicted values on pulmonary function test reports, as recommended by the ATS/ERS, may help to facilitate mindful interpretation of results.
A multisociety statement has highlighted the importance of the need for research on the tradeoffs involved in the transition to race-neutral approaches to spirometric interpretation [35]. Subsequent publications assessing the impact of using race-neutral GLI-Global reference equations compared with using GLI race-specific equations have demonstrated the consequence of diagnosing more Black individuals with respiratory impairment, particularly restrictive pattern by spirometry [36-42]. This effect is expected for the following reasons:
●While differences in absolute values of FEV1, forced vital capacity (FVC), lung volumes, diffusing capacity for carbon monoxide (DLCO), and maximal inspiratory and expiratory pressures (MIP and MEP) have been described for different race/ethnic groups, ratios such as FEV1/FVC and residual volume (RV)/total lung capacity (TLC) appear to be relatively independent of a patient's background.
●For a given absolute value in FEV1 or FVC, the predicted values for Black individuals are generally lower using a race-specific approach compared with a race-neutral approach. Therefore, the percent of predicted values are generally higher for Black individuals under a race-specific compared with race-neutral approach.
●Because obstruction is defined based on FEV1/FVC and concern for restriction is based on FEV1 and FVC (algorithm 1), using race-neutral equations has the greatest impact on the detection threshold for restrictive defects and the classification of disease severity.
One large epidemiologic study comprehensively assessed downstream effects of using GLI-Global rather than GLI-2012 race-specific equations among more than 360,000 participants in the United States and United Kingdom lung-health surveys and the Organ Procurement and Transplantation Network [41]. Although the race-specific GLI-2012 equations and the race-neutral GLI-Global equations performed similarly in predicting symptoms, health care utilization, and death (including death on the transplant waiting list), changes in classification of ventilatory deficits impacted nearly 5 percent of the adult population. In particular, there were large shifts in the prevalence of nonobstructive ventilatory impairments, with prevalence among self-identified White and Hispanic individuals decreasing from approximately 4.5 to 1 to 1.5 percent and increasing among self-identified black individuals from 5.7 to 14 percent. The investigators further demonstrated how these changes can have a large practical effect due to thresholds currently in use for occupational disqualification, Medicare pulmonary disability assessment, priority for lung transplantation, and Veterans Health Administration (VA) disability compensation. In general, the largest trade-offs were seen among Black persons, who were more often classified as having restrictive disease and more severe COPD, resulting in decreased eligibility for firefighting jobs but increased eligibility for disability payments and lung transplantation. The projected redistribution of USD $1.94 billion dollars of disability funds represents thousands of dollars in annual income increases or decreases for many of the 400,000 individual veterans affected. This work highlights inherent problems in cutoff-based approaches to assessment of functional capacity, disability, and diagnosis of pulmonary disease.
The effects seen above should be taken into account when using race-neutral equations for clinical diagnosis and decision making, particularly when comparisons are made to prior studies.
PULMONARY FUNCTION REFERENCE EQUATIONS — The patients or population being tested should be considered when applying reference equations. Reference equations should not be used (extrapolated) for patients whose age or height is outside the range of participants included in the reference study. The reference equations that are being used to create predicted values should be included in the pulmonary function report, as outlined in the “Recommendations for a Standardized Pulmonary Function Report” [43].
Spirometry — Large reference studies for spirometry have been performed in healthy persons in Europe and North America [2-4]. These studies, National Health and Nutrition Examination Survey (NHANES) III and Global Lung Function Initiative (GLI), have been used to generate reference equations for use in analyzing spirometry results, although additional data are needed for individuals under age 3 and over age 95.
●Global Lung Function Initiative – In 2012, the Global Lung Function Initiative (GLI) published spirometric prediction equations for ages 3 to 95 years for ethnic and geographic groups in 26 countries comprising White American/European, African American, Northeast Asian (Japanese and Korean), and Southeast Asian (Thai and Chinese) individuals. These reference equations are endorsed by the European Respiratory Society, American Thoracic Society, American College of Chest Physicians, the Australian and New Zealand Society of Respiratory Science, Thoracic Society of Australia and New Zealand, and Asian Pacific Society for Respirology. In 2022 the GLI also published a "race-neutral" weighted average of their previously published equations (called "GLI-Global") that can be used universally. The GLI reference equations are recommended for use in North America, as well as in Europe and Australia-New Zealand [43].
Improved data are needed for other population groups, particularly Arab, Indian, Polynesian, African, and Latin American, and other underrepresented groups [4]. Although the latest European Respiratory Society/American Thoracic Society (ERS/ATS) Interpretation Standard [32] recommends using GLI race-specific or "GLI-Other" reference equations, a special workshop convened by the ATS has endorsed using the "GLI-Global" equations for all people, at least until better options are available [33]. As with all recommendations, it will take time for industry to respond to these recommendations and provide GLI-Global as an option in their PFT equipment.
●NHANES III – In 2005, the American Thoracic Society (ATS)-European Respiratory Society (ERS) task force published a listing of pulmonary function reference values and recommendations for interpretative strategies [3]. For the United States, race/ethnically appropriate NHANES III reference equations were recommended for spirometry for those 8 to 80 years old [2]. Specific recommendations were not made for use in Europe at that time. A separate set of equations were recommended for those under eight years [5] but, subsequently, reference equations that can be applied across the spectrum of ages spanning childhood and adulthood were published [6]. For laboratories that are using NHANES III reference equations and choose to maintain continuity, simulation study has demonstrated that GLI and NHANES III are generally concordant with the most notable difference among those at the extremes of height and age [44].
Criteria for selecting pulmonary function reference equations include the following [3]:
●A population sample (with a wide range of age and height) is preferred to a convenience sample (eg, using volunteers or patients referred to a clinic).
●All participants should be never-smokers.
●Participants should be free of heart disease, lung disease, and chronic respiratory symptoms.
●The spirometers and test methods should meet current ATS/ERS recommendations. Almost all studies performed prior to 1981 did not conform to ATS standards.
Refinement of reference equations may be needed for patients of certain backgrounds. As an example, use of background-specific reference equations may be more accurate for patients in the United States who are of Dominican or Puerto Rican background [45]. In the Hispanic Community Health Study/Study of Latinos (16,415 participants, aged 18 to 74 years), those of Dominican or Puerto Rican background had lower predicted and lower limit of normal values for forced vital capacity (FVC) and forced expiratory volume in one second (FEV1), than those of other Hispanic or Mexican American backgrounds [45].
Diffusing capacity for carbon monoxide (DLCO) and lung volumes — In 2017, the GLI published prediction equations for DLCO, also known as carbon monoxide transfer factor (TLCO) in Europe and other areas of the world [46]. These reference equations apply to White populations who are age 5 to 85 years. The equations are based on data from over 12,000 individuals in 14 countries and 85 percent of the data were from White populations.
In 2021, the GLI published reference values for lung volumes for individuals of European ancestry between the ages of 5 and 80 years [47]. The equations are derived from 7190 observations in 11 countries and were from White populations only.
The GLI reference equations provide normative values based on a large, internationally representative dataset and expand on the previously existing reference equations that were based on smaller populations [3,48-59]. The GLI plans to validate these reference values in broader populations in future studies. The GLI reference equations for DLCO and lung volumes are endorsed by the ATS, ERS, American College of Chest Physicians, and the Asian Pacific Society of Respirology.
Other pulmonary function tests — Reference values for maximal inspiratory pressure (MIP) and maximal expiratory pressure (MEP) in older adults have also been published [60]. These studies have been limited to a narrower age range and therefore may have limitations in applicability [56].
Reference values for the six-minute walk distance for adults were published in 1998; this was followed by an ATS guideline on the subject in 2002 [61,62], which was updated by ATS/ERS in 2014 [63]. Sex, age, height, and weight must all be considered when interpreting this test [63]. (See "Overview of pulmonary function testing in adults", section on 'Six-minute walk test'.)
UPPER AND LOWER LIMITS OF NORMALITY — The American Thoracic Society (ATS) and European Respiratory Society (ERS) recommended that the fifth percentile be used as the lower limit of the normal range (LLN) [32]. By definition, only 5 percent of healthy individuals will have a result below this LLN. Note that the 5th percentile occurs at 1.65 standard deviations below the mean, defining a "z-score" of -1.65. Thus, any low value can be remembered easily by having a z-score of less than -1.65.
Prior to this, a traditional rule of thumb was to use 80 percent of predicted as the LLN for most pulmonary function test results. This standard works fairly well for forced vital capacity (FVC), forced expiratory volume in one second (FEV1), diffusing capacity for carbon monoxide (DLCO), and total lung capacity (TLC) in middle-aged adults, but gives a high rate of false positive or false negatives for FEV1/FVC, forced expiratory flow (FEF, also called maximum midexpiratory flow) 25 to 75 percent, and maximal respiratory pressures, especially when used for adolescents and adults over the age of 60 [64].
The use of the LLN rather than a fixed ratio cut off (eg, FEV1/FVC 0.7) to define airflow obstruction reduces the misclassification that occurs with a fixed ratio approach. The fixed ratio tends to under-diagnose younger individuals and over-diagnose older individuals with airflow obstruction and the LLN incorporates the changes in FEV1/FVC that occur with age and therefore, reduces this misclassification. Categorizing results as "borderline" abnormal if they fall near the LLN is also a useful concept. (See "Office spirometry", section on 'Ratio of FEV1/FVC'.)
Of note, use of the LLN to define normal is controversial [65], as some argue that normality should be defined by clinical and not statistical criteria [66], and either the fixed ratio or the LLN may variably correlate better with clinical outcomes [67].
Static lung volumes and DLCO values are also abnormal when values are above or below a reference standard, so that the normal range for these studies should include both an upper limit of normal (ULN) and an LLN.
Even when age, sex, height, and race/ethnicity are taken into consideration, the "normal" range for measurements of pulmonary function remains wide. This means that large changes with disease progression or therapy can easily occur while the patient's values remain within the normal range. Therefore, follow-up pulmonary function tests in adults should be compared with the patient's previous (baseline) absolute values, not the predicted values, and interpreted within the clinical context of the patient’s condition [33].
ENVIRONMENTAL INFLUENCES — Exposure to indoor air pollution, including passive and active exposure to cigarette smoke, and outdoor air pollution can affect lung growth and function [68]. As an example, early childhood exposure to parental smoking is associated with reduced lung function in school age children [69]. All cigarette smoking (starting during the early teens) is associated with an earlier peak in lung function and therefore an earlier onset of decline [7]. In addition, all of the changes due to aging are accelerated in susceptible cigarette smokers.
Childhood exposure to air pollution has been associated with reduced pulmonary function in a number of studies [70-75]. It is thought that children are particularly susceptible to the effects of poor air quality as lung development continues during childhood [76] and because the higher ventilatory rate and greater likelihood of outdoor exercise result in increased exposure [70]. Both restrictive and obstructive defects have been identified. One cohort study did not find a correlation between air pollution exposure and lung function, but the exposure models used may not have included adjustments for changes in exposure over time [77].
Other factors related to the social environment, such as socioeconomic status [13,78] and pre- and perinatal nutrition [79], may influence lung development and health.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pulmonary function testing".)
SUMMARY AND RECOMMENDATIONS
●Utility of reference values
•Correct interpretation of PFTs requires comparison of the patient's values with appropriate reference values. The majority of spirometers and pulmonary function equipment use software with reference equations to generate predicted values for each patient. (See 'Introduction' above and 'Pulmonary function reference equations' above.)
•Reference equations should be listed on the pulmonary function testing report as recommended by the American Thoracic Society (ATS) statement on Recommendations for a Standardized Pulmonary Function Report. (See 'Pulmonary function reference equations' above.)
•Reference equations should not be used (extrapolated) for patients outside the range of participants included in the reference study. (See 'Pulmonary function reference equations' above.)
●Reference equation parameters
•Age and sex – During childhood up until puberty, lung function increases linearly in proportion to overall growth, which is determined at least in part by age and sex. Healthy never-smoking adults without exposure to air pollution experience a gradual decline in lung function. The forced expiratory volume in one second (FEV1) decreases approximately 30 mL per year. The vital capacity decreases while the residual volume (RV) increases, leaving the total lung capacity intact. The diffusing capacity for carbon monoxide (DLCO) declines linearly with age. (See 'Age and sex' above.)
•Height – Taller persons have larger lung volumes, higher maximal flow rates, and a greater ability to take up oxygen and carbon monoxide per minute. Accurate height measurement is important in lung function prediction; technicians should not rely on stated height. (See 'Height' above.)
•Weight – Body weight is much less important than height when predicting most pulmonary function values and is not included in reference equations. Obesity may result in mildly reduced lung volumes, but this reflects physiologic restriction and should not be corrected for. (See 'Weight' above.)
•Race and ethnicity – The traditional approach assigns different normal ranges for spirometry, lung volumes, and other parameters according to race and ethnicity. Accumulating data suggest that an unintended clinical consequence of race/ethnicity-specific equations can be to underestimate the severity of pulmonary disease in some populations. Although there are some trade-offs to the shift in interpretative approach, our authors support the use of the "Global Lung Function Initiative (GLI)-Global" (or "GLI-Other") spirometry reference equations for all races/ethnicities. (See 'Effect of race/ethnicity' above.)
●Reference standards
•Spirometry – The GLI reference standards for spirometry are based on data from large, internationally representative populations and are recommended for use in North America, Europe, and Australia-New Zealand. The NHANES III reference standards for spirometry have been used for continuity over time in many labs in North America. Differences between GLI and NHANES III are most notable at the extremes of height and age. (See 'Spirometry' above.)
•Diffusing capacity for carbon monoxide – The GLI reference equations for DLCO are based on large internationally representative data and provide reference values for White populations from ages 5 to 85 years. (See 'Diffusing capacity for carbon monoxide (DLCO) and lung volumes' above.)
•Lung volumes – The GLI has published lung volume reference equations for individuals of European ancestry between the ages of 5 and 80 years. These are reference standards for static lung volumes and capacities (eg, total lung capacity, RV, and functional residual capacity (figure 3)) in White populations. (See 'Diffusing capacity for carbon monoxide (DLCO) and lung volumes' above.)
●Interpretation of normal values
•Spirometry – Airway obstruction should be determined by a FEV1/forced vital capacity (FVC) ratio below the age and sex-specific lower limit of normal (LLN), determined from the fifth percentile or z-score of 1.65. The use of other indices (eg, 0.7) increases the misclassification rates above the accepted 5 percent level. The LLN should also be used for the interpretation of FEV1/FVC separately based on current guidelines.(See 'Upper and lower limits of normality' above.)
•Lung volumes and DLCO – Static lung volumes and DLCO values are abnormal when values are above or below a reference standard, so that the normal range for these studies should include both an upper limit of normal (ULN) and LLN, determined by the upper and lower fifth percentiles (z-score of +1.65 and 1.65). This approach should replace the traditional rule of thumb of using 80 and 120 percent of predicted as the LLN and ULN for most pulmonary function test values. (See 'Upper and lower limits of normality' above.)
•Assessing change over time – Even when age, height, sex, and race/ethnicity are taken into consideration, the "normal" range for measurements of pulmonary function remains wide. It is, therefore, important to compare absolute values in lung function parameters when assessing change over time in a given individual while taking into consideration the natural loss of lung function with age. (See 'Upper and lower limits of normality' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Paul Enright, MD, who contributed to earlier versions of this topic review.
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟
