Quelques mythes et légendes de l’exploration fonctionnelle respiratoire

Prof Pierre-Olivier BridevauxService de Pneumologie

Hôpital du Valaispierre-olivier.bridevaux@hopitalvs.ch

Some myths and legend of pulmonaryfunction testing

§ My report fulfills the demands of my colleagues§ The references I used for years are adequate. Change

brings nothing§ Whatever value below 80 percent predicted is

abnormal§ During the last 2 weeks, Mr X lost 400 ml FVC

between the hospital and my practice. His condition is deteriorating quickly

§ Z scores are difficult to understand and useless

«My report fulfills the demands of my colleagues»

Simplicity

Clarity Precision

AccuracyReliability

«My report fulfills the demands of my colleagues»

Record the mandatory anthropometric data with the required precision: • Height in cm (measured not self-reported)• age (day)• Sex• Ethnicity• Weight

Recommendations for a Standardized Pulmonary Function Report An Official American Thoracic Society Technical StatementBruce H. Culver, et al on behalf of the ATS Committee on Proficiency Standards for Pulmonary Function Laboratories

AMERICAN THORACIC SOCIETY statement OCTOBER 2017

«My report fulfills the demands of my colleagues»

Self-reported and measured height comparison

Comparison of self-reported and measured height and weight: implications for obesity research among young adults. Danubio et al in Econ Hum Biol. 2008 Mar;6(1):181-90.

Both sexes overestimate height+ 2.1 cm for males + 2.8 cm for females

…& underestimate weight- 1.5 kg for males- 1.9kg for females

Avoid misunderstanding and overinterpretation– Report only measured parameters: FEV1, FVC, FEV1/FVC – Avoid truncated or dependent or disputable parameters (MIF50, FEV6, FIVS,

Mid flow)

Specify the references values Avoid confusion

– Do not show FEV1/FVC as percent predicted

Recommendations for a Standardized Pulmonary Function Report An Official American Thoracic Society Technical StatementBruce H. Culver, et al on behalf of the ATS Committee on Proficiency Standards for Pulmonary Function Laboratories

AMERICAN THORACIC SOCIETY statement OCTOBER 2017

«My report fulfills the demands of my colleagues»

Exercice 1: Report improvement - parameters

The references I used for years are adequate. Change brings nothing

Building of reference equations

1) Representative sample of the population of interest (age, sex, ethnicity). Selection of healthy never smokers.

2) Data collection (spirometry, plethysmography, TLCO, others)

3) Identification of predictors (most of the time: age, sex, height, ethnicity, others)

4) Building of the equationa. Delineating the lower (or upper) limit of normal value (LLN

/ULN)b. Implementation in PF laboratories

The Swiss SAPALDIA reference equationsRepresentative sample of the adult healthy never smokers in SAPALDIAn=3157/ 9651 from 18 to 60 year old →32.7%

Lung function in healthy never smoking adults: reference values and lower limits of normal of a Swiss population. Brandli O et al Thorax 1996 Mar;51(3):277-83.The average height of 18- and 19-year-old conscripts (N=458,322) in Switzerland from 1992 to 2009, and the secular height trend since 1878, Staub et al SMW 2011

Building of the equation (eg FVC)

SD: standard deviation of the residuals

R2: « R squared » = fraction of explained variance

Predicted values:

Lower limit of normal (LLN)

Example: 40 year old man, 180 cm (LLN =4.56)

H: Height

A : Age

Lung function in healthy never smoking adults: reference values and lower limits of normal of a Swiss population. Brandli O et al Thorax 1996 Mar;51(3):277-83.

Limitations of old reference equations

>60 extrapolated values

<18 no predicted values

Caucasian only

Lung function in healthy never smoking adults: reference values and lower limits of normal of a Swiss population. Brandli O et al Thorax 1996 Mar;51(3):277-83.

The average height of 18- and 19-year-old conscripts (N=458,322) in Switzerland from 1992 to 2009, and the secular height trend since 1878, Staub et al SMW 2011

Mean height of adult males in CH:

1950: 169 cm

2010: 178.2 cm, SD 6.5 cm

Global Lung Initiative (GLI) equationsfor spirometry (2012) and TLCO (2017)

• 74’000 healthy never smokers• Multiethnic

• Included subjects aged from 3 to 95 • LLN for each age

MULTI-ETHNIC REFERENCE VALUES FOR SPIROMETRY FOR THE 3–95 YEAR AGE RANGE: THE GLOBAL LUNG FUNCTION 2012 EQUATIONS Quanjer et al Eur Resp J 2012

Official ERS technical standards: Global Lung Function Initiative reference values for the carbon monoxide transfer factor for Caucasians, Stanojevic S, Graham BL, Cooper BG, Thompson BR, Carter KW, Francis RW, Hall GL; Global Lung Function Initiative TLCO working group; Global Lung Function

Initiative (GLI) TLCO. Eur Respir J. 2017 Sep 11;50(3).

«Puberty» factor arbitraly introduced in the Rosenthal equation at age 15.5

GLI 2012

Exercise 2: Serial FEV1 (expressed as %predicted value Rosenthal & GLI 2012) in a patient with CF

Rosenthal

Exercise 3: Serial FEV1 (expressed as %pred & GLI 2012) in a patient with well controlled asthma

Automatic switch from Rosenthal to ECCS93 at age 18

GLI 2012

Exercise 4: Are the reference equations adapted for thissubject?

ECCS included subjects aged 18+. Extrapolated values?

«Whatever value below 80 percent predicted is abnormal»

«Whatever value below 80 percent predicted is abnormal»

Exercice 4: 80-year-old lady, 160 cm, healthy, asymptomatic, never smokerreferred before abdominal surgery

.

Predicted LLN meas % pred Z scoreFVC 2.45 1.72 1.73 71% 1.63

→Traditional interpretation: FVC is reduced. Further diagnostic procedures are required.

→Correct interpretation: FVC in the normal range for age and size. No further test necessary.

Normal or abnormal FVC?

Probability of FEV1 <0.8 or FEV1/FVC <0.7 as a function of age

Stanojevic et al, Breathe, December 2013

«During the last 2 weeks, Mr X lost 400 ml FVC between the hospital and my practice. His condition isdeteriorating quickly!»

Spirometer types

Pneumo-tachograph

Ultrasonic

Bell

Turbine

Pulmonary waveform generator

Adjustable temperature, humidity, flow volume

Sensormedics « hot wire » spirometer

EasyOne Ultrasonicspirometer

Spirometers replacement in SAPALDIA

Sensormedics

FVC decline over 19 years

SAPALDIA 1991- 2010

Spirometers-specific equations

EasyOne

Spirometer Replacement and Serial Lung Function Measurements in Population StudiesPierre-Olivier Bridevaux*, Elise Dupuis-Lozeron et al American Journal of Epidemiology 2015

Spirometers-specific equations

Δ=~400 ml

Sensormedics

EasyOne

FVC 1991 - 2002 - 2010

Spirometer Replacement and Serial Lung Function Measurements in Population StudiesPierre-Olivier Bridevaux*, Elise Dupuis-Lozeron et al American Journal of Epidemiology 2015

American Journal of Epidemiology, Volume 181, Issue 10, 15 May 2015, Pages 752–761, https://doi.org/10.1093/aje/kwu352

Figure 2. Predicted lung function indices with different reference equations for women (left column) and men (right column).

Exercice 5: Any additional problem which may lead to measurement bias?

Calibration not performed

«Z scores are difficult to understand and useless»

Z scoresFVC distribution over age

Z scores are easy!

§.

Advantages of Z scores over percent predicted

§ Statistically sound for all ages

§ Independent of sex, height, ethnicity

§ For one individual, Z score can accurately predictchanges over time despite growth or physiological decline

§ Z scores allow comparisons between indivduals of different age, sex, ethnicity.

Z scores facilitate interpretation of spirometry

§.

Recommendations for a Standardized Pulmonary Function Report An Official American Thoracic Society Technical StatementBruce H. Culver, et al on behalf of the ATS Committee on Proficiency Standards for Pulmonary Function Laboratories

AMERICAN THORACIC SOCIETY statement OCTOBER 2017

§.

Z scores facilitate PFT interpretation of plethysmography

Recommendations for a Standardized Pulmonary Function Report An Official American Thoracic Society Technical StatementBruce H. Culver, et al on behalf of the ATS Committee on Proficiency Standards for Pulmonary Function Laboratories

AMERICAN THORACIC SOCIETY statement OCTOBER 2017

§.

Z scores facilitate PFT interpretation of TLCO

Recommendations for a Standardized Pulmonary Function Report An Official American Thoracic Society Technical StatementBruce H. Culver, et al on behalf of the ATS Committee on Proficiency Standards for Pulmonary Function Laboratories

AMERICAN THORACIC SOCIETY statement OCTOBER 2017

Conclusions

• Report quality can/should be standardized in the pulmonary community

• Z scores to be implemented in each modern laboratory

• Manufacturers have a key role to promote good practice and provide adequate reporting

Thanks for your attention

Pierre-Olivier Bridevaux

Interpretation of lung function testsin daily life

Current practice of lung function laboratoriesØ Reference equations not always reported (ERS? Sapaldia? NHANES? Crapo? Knudson?

Zapletal?) Ø Varying definition for obstruction (LLN, fixed cut off, <88%, etc)Ø Multiple references equations for spirometry, bodyplethsymography and TLCO

Limitations of older reference equationsØ Data 1960-1990 → Change in nutrition, smoking and ETS prevalence, SESØ Developed for adult OR pediatric populationØ Few non caucasian subjects included

ConsequencesØ Inconsistent interpretation between and within countries, lab, physiciansØ Limited understanding of epidemiology of lung disease

Daily life Case 1…Woman, caucasian, aged 80, 160 cm, smoker, chronic dyspnea and productive cough, admitted for 2 acute exacerbations in the last 12 months. AECOPD rate: 2 exacerbation/yr, CAT score: 22 points

ERS 1993 GLI 2012

meas pred LLN % pred pred LLN % pred

FEV1 , L 0.92 1.72 1.1 53% 1.87 1.31 49%

FVC, L 2.25 2.12 1.41 106% 2.45 1.72 92%

FEV1/FVC 0.41 0.74 0.63 <0.7 0.77 0.62 <LLN

Conclusions : Physician 1: COPD stage 2 using the FEV1/FVC fixed ratio and ERS/ECSC-93Physician 2: COPD stage 3 using the LLN for FEV1/FVC and GLI-2012

Physicians (1-2): COPD group D

Δ

150 ml

330 ml

Daily life Case 2 …Man, caucasian, aged 65, 178 cm, non smoker, no symptom, referred before surgery

ERS 1993 GLI 2012

meas pred LLN % pred pred LLN % pred

FEV1 , L 2.85 3.28 2.44 87% 3.41 2.52 84%

FVC, L 4.38 4.22 3.22 104% 4.46 3.35 98%

FEV1/FVC 0.65 0.76 0.64 <0.7 0.77 0.64 >LLN

Conclusions :

Physician 1: COPD stage 1 using the fixed ratio for FEV1/FVCPhysician 2: Normal spirometry

:

Δ

130 ml

240 ml

|

Les résultats des manœuvres sont interprétables si… (6 critères ATS/ERS):

Quanjer PhH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC. Lung volumes and forced ventilatory flows. Official Statement of the European Respiratory Society. Eur Respir J 1993; 6 suppl. 16: 5-40. Erratum Eur Respir J 1995; 8: 1629. American Thoracic Society. Standardization of spirometry. 1994 update. Am J Respir Crit Care Med 1995; 152: 1107-1136.

La CVF est reproductible < 200 ml Δ entre les 2 meilleures CVF

Le VEMS est reproductible < 200 ml Δ entre les 2 meilleurs VEMS

Le Peak Flow est reproductible <10% Δ entre les meilleurs PF

Le volume d’extrapolation est petit

volume d’extrapolation < 5% de la meilleure CVF ou <150 ml (=absence d’hésitation lors de la mesure du VEMS)

Le plateau de fin de test est atteint Flux est nul en fin de manoeuvre

L’expiration dure plus de 6 sec

Temps expiratoire lors de la meilleure manœuvre >6 secondes (>3 secondes pour les enfants)

Δ= différence

|

Essai 1 Essai 2 Essai 3 Essai 4 Valeursretenues

VEMS (L) 2.41 2.75 2.78 2.80 2.80

CVF (L) 3.60 2.91 3.61 3.21 3.61

Temps exp(Sec) 6.3 4.2 6.4 5.1

42juin 19

Reproductibilité du VEMS et de la CVF

Δ entre les 2 meilleurs VEMS et CVF = 2 et 1 ml (< 200 ml)

|

Essai 1 Essai 2 Essai 3 Essai 4 Valeursretenues

VEMS (L) 2.41 2.75 2.78 2.80 2.80

CVF (L) 3.60 2.91 3.61 3.21 3.61

Temps exp(Sec) 6.3 4.2 6.4 5.1

PEF (L/sec) 3.8 4.2 5.8 6.0 6.0

43juin 19

Critères ATS/ERS (exemples)Reproductibilité du Peak Flow

Δ entre les 2 meilleurs PEF = 0.2 (< 10% [soit 0.6 dans ce cas])

| 44http://www.spirxpert.com/indices7.htm

Volume d’extrapolation <150 ml ou 5% de la CVF

Volume d’extrapolation: « back extrapolated volume » doit être <5% de la CVF ou <150 ml

Ex: CVF: 4L, Volume d’extrapolation doit être <150 ml

Ex: CVF: 2.8L, Volume d’extrapolation doit être <140 ml (5%)

Le plus souvent signe une hésitation du sujet

45juin 19

Courbes débit volume (exemples artéfacts)

Fischberg et al, RevMed 2009

Doit-on accepter les manœuvres imparfaites?

OUI ! Si après 3 manoeuvres, les volumes restent non reproductibles avec unevariation >200 ml dans le VEMS ou la CVF, alors…

– Répéter les manoeuvres (8 au maximum) – Accepter les valeurs de VEMS et CVF les plus

hautes, indépendamment de la “beauté” de la courbe D/V

46juin 19

Les paramètres mesurés sont acceptables même si les manœuvressont imparfaites et non reproductibles!

Interprétation des résultats

juin 19 http://www.spirxpert.com/gli_intro.html48

Valeurs de références

Les valeurs de références (ECCS 93, NHANES, Sapaldia, etc) ne sont valables que pour des sujets adultes (20 à 75 ans) et caucasiens. L’extrapolation des prédictions pour les autres groupes ethniques et les personnes plus agées n’est pas recommandable.

Global Lung Initiative reference value (GLI 2012)

-Basée sur 74’000 sujets non fumeurs et en bonne santé-Multi-ethniques-Tous les âges (3-95 ans)-Limites inférieures de la norme (LIN) et Z score pour chaque âge

Sheet 1 of 1 Ref: 2012S0045 © GOSH NHS Foundation Trust March 2013

Lung Function UpdateExciting News! New recommendations for the interpretation of spirometric lung function measurements

Great Ormond Street Hospital for Children NHS Foundation Trust: Information for Staff

Following a recent publication (Quanjer, Eur Resp Journal, 2012) new recommendations for spirometry growth charts, from the Global Lung function Initiative (GLI), have now been endorsed by all international respiratory societies and have been implemented at GOSH.

As we know, normal ranges for lung function are dependent on body size, age, sex and ethnicity (see Figure 1).

Previously each hospital selected their own preferred reference equations with which to interpret spirometry results. However, since use of different equations has been shown to lead to different interpretation, this meant that results from any given individual could differ depending on where they were assessed. This was particularly important during periods of rapid growth and transition of adolescents into adult care.

What’s New?International collaboration of more than 40 countries has resulted in standardised GLI spirometry reference equations that can be used globally for people aged from 3 to 95 years.

How does this impact clinical practice?

� Absolute (raw) values of lung function will not be affected.

� The predicted value for a given height and age may differ but will be a better reflection of your patient’s results.

� You will need to refer to updated trend reports when comparing new measurements with those taken prior to this change to avoid unnecessary misinterpretation and anxiety of patients

� In addition to standardising the results for each patient within each department, results will be standardised across different institutions.

It is important to remember that for every person there is a range of results that is considered normal for their age and height. In future, lung function reports may include a diagram like Figure 2.

Figure 1: lung function increases rapidly during childhood to a peak in early adulthood, followed by a gradual decline.

6

4

Pred

icte

d F

EV1 (

L)

2

00 20 40

Age (years)60 80

FEV1

-1.64 +1.64

FEV1/FVC

FVC

Normal Range

Figure 2: Pictogram showing subject’s results for FEV

1 (upper bar), FVC (middle

bar) and FEV1/FVC ratio (as large black

arrows) in relation to the normal range (shown in white and determined by ± 1.64 SD scores), mild reduction (light grey area) and lower than normal range (dark grey). Thus the patient illustrated has a reduced FEV

1, a normal FVC, and a reduced FEV

1/FVC.

For further information please see the GLI website at www.lungfunction.org

Compiled by the Lung Function Lab in collaboration with the Child and Family Information Group Great Ormond Street Hospital for Children NHS Foundation Trust, Great Ormond Street, London WC1N 3JH www.gosh.nhs.uk

juin 19 http://www.spirxpert.com/gli_intro.html49

Pourcentage de la valeur prédite ou Z scores

Sheet 1 of 1 Ref: 2012S0045 © GOSH NHS Foundation Trust March 2013

Lung Function UpdateExciting News! New recommendations for the interpretation of spirometric lung function measurements

Great Ormond Street Hospital for Children NHS Foundation Trust: Information for Staff

Following a recent publication (Quanjer, Eur Resp Journal, 2012) new recommendations for spirometry growth charts, from the Global Lung function Initiative (GLI), have now been endorsed by all international respiratory societies and have been implemented at GOSH.

As we know, normal ranges for lung function are dependent on body size, age, sex and ethnicity (see Figure 1).

Previously each hospital selected their own preferred reference equations with which to interpret spirometry results. However, since use of different equations has been shown to lead to different interpretation, this meant that results from any given individual could differ depending on where they were assessed. This was particularly important during periods of rapid growth and transition of adolescents into adult care.

What’s New?International collaboration of more than 40 countries has resulted in standardised GLI spirometry reference equations that can be used globally for people aged from 3 to 95 years.

How does this impact clinical practice?

� Absolute (raw) values of lung function will not be affected.

� The predicted value for a given height and age may differ but will be a better reflection of your patient’s results.

� You will need to refer to updated trend reports when comparing new measurements with those taken prior to this change to avoid unnecessary misinterpretation and anxiety of patients

� In addition to standardising the results for each patient within each department, results will be standardised across different institutions.

It is important to remember that for every person there is a range of results that is considered normal for their age and height. In future, lung function reports may include a diagram like Figure 2.

Figure 1: lung function increases rapidly during childhood to a peak in early adulthood, followed by a gradual decline.

6

4

Pred

icte

d F

EV1 (

L)

2

00 20 40

Age (years)60 80

FEV1

-1.64 +1.64

FEV1/FVC

FVC

Normal Range

Figure 2: Pictogram showing subject’s results for FEV

1 (upper bar), FVC (middle

bar) and FEV1/FVC ratio (as large black

arrows) in relation to the normal range (shown in white and determined by ± 1.64 SD scores), mild reduction (light grey area) and lower than normal range (dark grey). Thus the patient illustrated has a reduced FEV

1, a normal FVC, and a reduced FEV

1/FVC.

For further information please see the GLI website at www.lungfunction.org

Compiled by the Lung Function Lab in collaboration with the Child and Family Information Group Great Ormond Street Hospital for Children NHS Foundation Trust, Great Ormond Street, London WC1N 3JH www.gosh.nhs.uk

Interprétation historique (%age de la valeur prédite): >80% vp = normal

Interprétation correcte (Z scores): Les Z-scores contiennent par définition95% de la population (1,96 Z-score) ou 90% de la population (1,64 Z-score).

Le Z-score indique la distance, en nombre de déviation standard, entre une valeur mesurée et une valeur médiane, normale. Résultat «normal» si entre -1.64 et +1.64 Z scores

Ex: femme de 80 ans, 160 cm, asymptomatique et non fumeuse a une valeur prédite pour la CVF de 2,45L et une LIN a 1,72L (correspondant au 5ème percentile de la valeur médiane prédite, soit 1,64 Z-score). CVF mesurée 1,73 L (71%vp) soit 1,63 Z-score de la valeur médiane prédite→Interprétation historique: CVF abaissée à 71% de la valeur prédite. A investiguer!→Interprétation correcte: CVF dans la limite de la norme. Pas d’investigation supplémentaire nécessaire.

Nota bene: Les Z-scores restent encore trop peu utilisés en pratique courante.