Respiratory Medicine (2011) 105, 386e391

ava i lab le at www.sc iencedi rect .com

journal homepage : www.e lsev ie r . com/ loca te / rmed

The prevalence of asthma and COPD in Italy:A practice-based study

Mario Cazzola a,b,*, Ermanno Puxeddu a, Germano Bettoncelli c,Lucia Novelli a, Andrea Segreti a, Claudio Cricelli c, Luigino Calzetta a

aDepartment of Internal Medicine, University of Rome Tor Vergata, Rome, Italyb Pulmonary Rehabilitation Group, IRCCS San Raffaele Pisana, Rome, ItalycHealth Search Institute, Italian College of General Practitioners, Florence, Italy

Received 16 September 2010; accepted 30 September 2010Available online 15 October 2010

KEYWORDSAsthma;COPD;Epidemiology;Administrative healthdatabase

* Corresponding author. Cattedra dipellier 1, IT-00133 Rome, Italy.

E-mail address: mario.cazzola@un

0954-6111/$ - see front matter ª 201doi:10.1016/j.rmed.2010.09.022

Summary

We conducted a population-based cross-sectional epidemiologic survey of asthma and COPD inan adult representative national sample using information obtained from the Health SearchDatabase owned by the Italian College of General Practitioners. General Practitioners whohad a list of patient population of 909,638 individuals (429,962 men and 479,676 women;man/woman ratio [M/WR]: 0.89) �14 years old at the end of December 2009 were selectedto be representatives of the whole Italian population. Cases of asthma and COPD were identi-fied on the basis of the ICD-9 codes. The total sample included 55,500 (6.10% of the entire po-pulation; 5.49% of men and 6.64% of women; M/WR: 0.74) subjects suffering from asthma and25,762 (2.83% of the entire population; 3.51% of men and 2.23% of women; M/WR: 1.41)subjects suffering from COPD. The asthma/COPD ratio in general population was 2.16. Theodds ratio (OR) was chosen because asthma and COPD had a prevalence less than 10%. TheOR of developing asthma decreased with age both in men and women, but in the first groupof age (15e34 years) it was higher in men vs. women (1.69 vs. 1.00) although it became lowerthan 1 from 35 years old and up in men and from 75 years old and up in women. On thecontrary, the OR of developing COPD became higher than 1 from 55 years old and up both inmen and in women and progressively increased with age (in the group 75e84 years, it was6.16 in men and 4.07 in women, respectively).ª 2010 Elsevier Ltd. All rights reserved.

Malattie Respiratorie, Dipartimento di Medicina Interna, Universita di Roma Tor Vergata, Via Mont-

iroma2.it (M. Cazzola).

0 Elsevier Ltd. All rights reserved.

The prevalence of asthma and COPD in Italy 387

Introduction

Asthma and COPD are common chronic diseases of theairways that are mainly diagnosed and treated in generalpractice. The general practitioner (GP) is the provider ofbasic medical care and is the gateway to specialized care.Therefore, it is important that the GPs know the prevalenceof different diseases in their practice.

In Italy, national statistics on morbidity are based on self-reported information, and only few researchers have inves-tigated whether some differences might arise when usingdata from Health Interview Surveys and from medicalrecords.1 Unfortunately, the self-reported prevalence ofasthma or COPD, when compared with GP registration, isassociated with an overestimation in people who suffer fromother respiratory conditions and with an underestimation inelderly persons living in a smoky environmentwho have fewerchronic diseases.2

The aim of this study was to epidemiologically investigatethe prevalence and risk of asthma and COPD in Italy usingmedical records obtained from the Health Search Database(HSD) owned by the Italian College of General Practitioners(SIMG), which stores information on about 1.5% of the totalItalian population served by general practitioners (GPs).

Material and methods

We conducted an adult population-based cross-sectionalepidemiologic survey of asthma and COPD in a representativenational sample using information obtained from the HSD.The software system used codes all the diagnostic recordsutilizing the International Classification of Diseases (9threvision: ICD-9).3 The participating GPs use the same softwareto record data during their daily practice, and they agree tosend periodically complete, but anonymous, records of theirpatients to the HSD. A unique patient code links demographicand prescription information, clinical events and diagnoses,hospital admission, and cause of death. It is well known thatbefore using any data resource, particularly one that is basedon computer records, it is necessary to determine the qualityand completeness of the available information. Therefore,data are subject to a range of quality checks. Any variationswithin agreed ranges are investigated and submitted to eachparticipating GP. Physicians who fail to meet standard qualitycriteria are not considered for epidemiological studies.4

700 GPs, who had a list of patient population of 909,638individuals (429,962 men and 479,676 women; man/womanratio [M/WR]: 0.89) older than 14 years at the end ofDecember 2009, were selected to be representatives of thewhole Italian population and also because they ensured therequired data quality.

Since the study was a retrospective analysis of dataalready available in the HSD, and confidentiality of patientswas in noway violated, local rules do not request approval byan ethics committee.

Ascertainment of COPD and asthma

Cases of COPD were identified on the basis of the ICD-9 codes491, 492, and 496, and those of asthmaon the basis of the ICD-9 code 493.

Analysis

The study was carried out by a cross-sectional and observa-tional field method. It permitted to describe the situation atthe timeof data collection, at theendofDecember 2009, andit allowed determining the prevalence of investigatedoutcomes. Furthermore, we evaluated the risk factor (RF)calculating the OR or RR. In order to investigate the associ-ation between variables, it was chosen to use OR instead ofRRbecause the diseases of the study (asthmaandCOPD) havea relatively low prevalence that is less than 10%. Positive andnegative OR values mean a co-relative positive and negativeassociation between evaluated variables.5,6

This epidemiological study had a confidence level of99.99% (setting the hypothesized % frequency of outcomefactor in the population at 50% and confidence limit of 0.2%).

The studied population was divided by several clusters:disease, age (15e34, 35e44, 45e54, 55e64, 65e74, 75e84and older than 85 years old), and gender (male and female).

In order to calculate statistical and epidemiologicalvalues, computer software GraphPad Prism (CA, USA) andOpenEpi (Dean AG, Sullivan KM, Soe MM. Open SourceEpidemiologic Statistics for Public Health) were used.

Results

The total sample included 55,500 (6.10% of the entirepopulation; 5.49% of men and 6.64% of women; M/WR:0.74)subjects suffering from asthma and 25,762 (2.83% of theentire population; 3.51% of men and 2.23% of women; M/WR:1.41) subjects suffering from COPD. The asthma/COPDratio in general population was 2.16 (Table 1).

The prevalence of these diseases by age cluster is shownin Fig. 1. In asthma, it was higher in women in all ageclusters except in the 15e34 group. In COPD, the preva-lence was higher in men in all age clusters, with a markeddifference between sexes after 64 years of age.

The OR of developing asthma decreased with age both inmen and women, but in the first group of age (15e34 years)it was higher in men vs. women (1.69 vs.1.00) although itbecame lower than 1 from 35 years old and up in men andfrom 75 years onwards in women. On the contrary, the ORof developing COPD became higher than 1 from 55 years oldand up both in men and in women and progressivelyincreased with age (in the group 75e84 years, it was 6.16 inmen and 4.07 in women, respectively) (Fig. 2).

In Fig. 3 we report the prevalence of smoking habit inthe general Italian population using data recorded by theItalian Central Institute of Statistics (ISTAT). These datashow that in asthmatic patients smoking does not playa role, whereas, as expected, in COPD there is an importantcorrelation between smoking and disease.

Discussion

The first important finding of our study is the documenta-tion that the asthma/COPD ratio in our population was2.16, with a prevalence of asthma and COPD of 6.10% and2.83%, respectively. We must highlight that the prevalenceof asthma in our study was very well in line with results

Table 1 Prevalence (%) and rates of asthma and COPD in population study by gender cluster.

Prevalence (%) Rates

Female Male Total Male/female Asthma/COPD female Asthma/COPD male Asthma/COPD total

Asthma 6.64 5.49 6.10 0.83 2.98 1.56 2.16COPD 2.23 3.51 2.83 1.57

388 M. Cazzola et al.

from population studies of representative samples of thegeneral population. This is not very common in registerstudies, as they normally results in underestimations.Therefore, we believe that this is a strength of our studyand of the usefulness of the HSD.

We cannot deny that there may have been misclassifica-tion when our GPs have used specific diagnostic categories.Despite the availability of consensus guideline diagnosticrecommendations, diagnostic confusion between COPD andasthma appears common in primary care patients.7 TheFinEsS-study assessed whether differences existed in preva-lence of asthma, chronic bronchitis, and respiratory symp-toms between three Baltic capitals, and examined the riskfactor profiles for respiratory conditions.8 Prevalence ofrespiratory symptomswas higher in Tallinn than in Stockholmand Helsinki, while physician-diagnosed asthma was morecommon in Stockholm and Helsinki and the prevalence ofphysician-diagnosed chronic bronchitis was three times ashigh inTallinnas inHelsinki or Stockholm. It is likely that theseconsiderable differences in prevalence rates of physician-diagnosed asthma and COPD were due to substantial differ-ence in diagnostic practices between the three countries,while thedifferences between thecapitals in trueprevalenceof disease was small.

In general, COPD is underestimated by GPs. In a pop-ulation-based study in which COPD was diagnosed based onthe assessment of lung function by means of spirometry, theso called Burden Of Obstructive Lung Disease (BOLD) study,prevalences between 8.2 and 19.1% in adults 40 years of ageor older were found.9 In primary care, much lower estimatesof the prevalence of physician-diagnosed COPD have been

Fig. 1 Cases of asthma and COPD in the studied Italian populatCOPD by age and gender clusters.

reported (between 1.5 and 3%).10e12 However, we muststress that the BOLD study included also mild COPD (whichcomprise approximately a half of the COPD-cases) with thediagnosis based only on the FEV1/FVC ratio< 0.79 and theycould have a normal FEV1, so it is not surprising that all thesemild COPD cases not are identified by the health care system.

On the other hand, some studies found an overdiagnosis ofasthma in primary health care.13,14 In particular, Aaron andcolleagues recently found that about one-third of obese andnonobese individuals with physician-diagnosed asthma hadno evidence of asthmawhen their medications were taperedand when they were evaluated with serial assessments ofsymptoms, lung function and bronchial challenge tests.15 Werecognize that GPs are faced with an array of alternativecodes for the same or similar conditions. In effect, practicingphysicians have little faith in the accuracy of the ICD-9diagnoses entered for prescription purposes and this canquestion the real value of diagnoses in the HSD, includingthat of COPD and asthma. It is plausible that misdiagnosis,including possible gender bias, could be reduced by theappropriate use of objective laboratory studies. Unfortu-nately, previously, it was shown that Italian GPs make littleuse of spirometry in their investigation of chronic respiratorysymptoms.16

In the majority of our patients, the diagnosis of asthma orCOPD was mainly clinical and this can be interpreted asa severe limitation. In 2004, Viegi and colleagues17 analyzeddata from two prospective studies (4353 subjects who hadacceptable FVC test results) carried out in Italy, in the ruralarea of Po River delta from1988 to 1991 and in the urban areaof Pisa from 1991 to 1993. Prevalence rates of asthma,

ion. Points shown represent the prevalence (%) of asthma and

Fig. 2 Risk factors of asthma and COPD. Points shown represent odds ratios and 95% confidence limits of associations betweenrisk of asthma and COPD and age and gender. Values >1 indicate a positive association and values <1 indicate negative association,DP< 0.05 vs. non-exposed population.

The prevalence of asthma and COPD in Italy 389

chronic bronchitis, and emphysema were 5.3%, 1.5%, and1.2% in the Po delta, and 6.5%, 2.5%, and 3.6% in Pisa. Forboth genders, the frequency of isolated asthma decreasedwith age, while that of isolated airflow obstruction, chronicbronchitis-emphysema, and the combination of asthma andchronic bronchitis-emphysema increased. It is difficult tocompare the results of that study and our results, since thereare considerable differences in the methods used and thestudy populations included. In particular, our analysis wasrestricted to those patients who consulted their GPs.Therefore, the conclusions of our study are probably notvalid for the healthiest people: those who did not consulttheir GPs at all. Nonetheless, our research, using a largerpopulation with a different approach, confirmed that theprevalence of asthma was higher than that of COPD and,moreover, age was a strong and positive risk factor of COPD,whereas in asthma the correlation between disease and agewas weakly negative.

Similarities in the conclusions of our study and those ofViegi and colleagues17 are important considering that it hasbeen highlighted that distinction between COPD and asthmamay be relatively easy in patients with severe disease or in

Fig. 3 Prevalence (%, A) and correlation (rate, B and C) of smokin>0 indicate a positive correlation and values <0 indicate negativeobtained from the Italian Statistical Yearbook 2009 of Italian Centra

older patients, but in many cases it is not so clear as clinicalsigns and symptoms show overlap. This is particularly thecase in family practice, where the diagnostic process isprimarily based on symptoms and signs presented by thepatient.18 In any case, it must be mentioned that only25e50% of asthma or COPD patients are known to theirdoctor.19 In fact, it seems that the majority of patients witha decreased FEV1 do not complain of any bronchial symp-toms.20 Poor perception of dyspnoea is a probable cause ofunderpresentation by patients, so neither the physician northe patient is to blame for the underdiagnosis of asthma orCOPD in the general population.10

Our data confirm that young adults are at highest risk ofhaving a diagnosis of asthma. In the current study, theprevalence was highest in the age group of 15e34 years old(physician-diagnosed asthma 7.19%) and declined progres-sively to become 4.12 in the cluster of age >85 years old.However, the decrease in prevalence was progressive inmen, but women showed an increase in prevalence up to thecluster of age 45e54 years old and soon after a progressivedecrease, although the prevalence remained always higherin women than in men. This last finding contrasts with

g habit and asthma or COPD by gender clusters. B and C: valuescorrelation. Smoking habit data of general population (A) werel Institute of Statistic (ISTAT), available at http://www.istat.it.

390 M. Cazzola et al.

a British documentation that in people older than 70 years,current asthma was more prevalent in men than women(5.1% compared to 1.8%).21 Also a cross-sectional Frenchstudy documented that in men, the prevalence rate washigher that in women (7.3% for cumulative asthma and 2.8%for current asthma vs. 5.2% and 2.2%, respectively). In anycase, as in our study, the rates were lower at advanced ages(>85 years) for both men and women.22

Consistent with the international epidemiology of COPD,9

a progressive rise of prevalence with age in both sexes untilthe age of 84 years was observed. The prevalence was higherinmen than inwomen. For those aged 75 years and older, theprevalencewas 14.15% formen and 6.61% for women. Reportfrom the Obstructive Lung Disease in Northern Sweden(OLIN) Studies confirms that increasing age, together withsmoking, is the dominating risk factor for COPD.23

In our study, male gender was a risk factor of asthmaonly in young people, being the ORs in men always lowerthan 1 from 35 years old and up. This finding contrasts withthe documentation in the “Pollution Atmospherique etAffections Respiratoires” (PAARC) study that, from aged 50yrs upwards rates in men increased.24 On the contrary, inour study female gender was a significant risk factor only inthe age group 45e54 years, with an OR that was 1.11. From75 years old and up, the ORs were lower than 1 also inwomen. Our results differ from those reported by Osmanand colleagues25 documenting a higher prevalence ofasthma in women from 15 years old and up. The timing ofthe gender reversal in asthma prevalence has generallybeen reported to occur between 10 and 20 years of age inmost studies, with exceptions locating the switch beyond20 years.26 However they fit with other Italian reports. Theprevalence of an “episode of asthma” (defined as a self-reported attack of asthma or treatment for asthma) and of“current asthma” (defined as a self-reported attack ofasthma, treatment for asthma or wheezing other than dueto a cold with dyspnoea in the last 12 months) was 3.47%(3.74% in men and 3.14% in women) and 5.01% (5.07% inmen and 4.90% in women), respectively, in randomlyselected subjects, aged 20e44 years, living in three areasof northern Italy: Turin, Pavia, and Verona.27 Interestingly,no differences were found on the basis of age and sex. Ina French survey, asthma was inversely correlated to age(higher prevalence in the youngest) but was not related tosex.28 Nonetheless, it is important to highlight that ananalysis based on a total of 9,486,173 hospital records inCanada for a 3-year period, including 204,304 asthmapatients and 288,977 asthma-related records, showed thatthe incidence ratio for women vs. men for asthma hospi-talization reached 2.8 for individuals 25e34 years of age,decreased gradually with increasing age, and thenapproached unity for those aged 80 years or more.29

Contrary to that observed by others who reported thatmen would have a higher risk of COPD than women in theyounger age groups (45e49 years, 50e54 years) and womenwould show a higher risk in the older age groups up to the75e79 years,30 in our population the risk of COPD wasalways higher in men than in women regardless of age.

In conclusion, our findings show that the prevalence ofasthma is higher than that of COPD in an adult Italian generalpopulation with higher prevalence of asthma in women andhigher prevalence of COPD in men. It is important to stress

that this information comes from an administrative healthdatabase. We believe databases provide a valuable methodfor identifying sex differences in the epidemiology of asthmaand COPD, because they provide large sample sizes, havestrong statistical power, are relatively inexpensive studies toconduct, and can link information on diagnosis, and treat-ment. An administrative health database may represent the‘real world’ more accurately and characterize sex differ-ences more fully than do other data sources.

Conflict of interest statement

The authors declare that they have no competing interests.

References

1. Cricelli C, Mazzaglia G, Samani F, Marchi M, Sabatini A, Nardi R,et al. Prevalence estimates for chronic diseases in Italy:exploring the differences between self-report and primarycare databases. J Public Health Med 2003;25:254e7.

2. Mohangoo AD, van der Linden MW, Schellevis FG, Raat H.Prevalence estimates of asthma or COPD from a health inter-view survey and from general practitioner registration: what’sthe difference? Eur J Public Health 2006;16:101e5.

3. World Health Organisation. International classification ofdiseases, Manual of the international statistical classificationof diseases: injuries and causes of deaths. Geneva: WHO; 1977[9th revision].

4. Lawrenson R, Williams T, Farmer R. Clinical information forresearch: the use of general practice databases. J PublicHealth Med 1999;21:299e304.

5. Schmidt CO, Kohlmann T. When to use the Odds Ratio or therelative risk? Int J Public Health 2008;53:165e7.

6. Motulsky H. Intuitive biostatistics. New York, Oxford: OxfordUniversity Press; 1995.

7. Breyer MK, Spruit MA, Celis AP, Rutten EP, Janssen PP,Wouters EF. Highly elevated C-reactive protein levels in obesepatients with COPD: a fat chance? Clin Nutr 2009;28:642e7.

8. Pallasaho P, Lundback B, Meren M, Kiviloog J, Loit HM,Larsson K, et al. Prevalence and risk factors for asthma andchronic bronchitis in the capitals Helsinki, Stockholm, andTallinn. Respir Med 2002;96:759e69.

9. Buist AS, McBurnie MA, Vollmer WM, Gillespie S, Burney P,Mannino DM, et al. International variation in the prevalence ofCOPD (the BOLD Study): a population-based prevalence study.Lancet 2007;370:741e50.

10. White P. Prevalence of COPD in primary care: no room forcomplacency. Fam Pract 2009;26:1e2.

11. Chapman KR. Chronic obstructive pulmonary disease: arewomen more susceptible than men? Clin Chest Med 2004;25:331e41.

12. Soriano JB, Maier WC, Egger P, Visick G, Thakrar B, Sykes J,et al. Recent trends in physician diagnosed COPD in women andmen in the UK. Thorax 2000;55:789e94.

13. Marklund B, Tunsater A, Bengtsson C. How often is the diag-nosis bronchial asthma correct? Fam Pract 1999;16:112e6.

14. Tinkelman DG, Price DB, Nordyke RJ, Halbert RJ. Misdiagnosisof COPD and asthma in primary care patients 40 years of ageand over. J Asthma 2006;43:75e80.

15. Aaron SD, Vandemheen KL, Boulet LP, McIvor RA, Fitzgerald JM,Hernandez P, et al. Overdiagnosis of asthma in obese andnonobese adults. CMAJ 2008;179:1121e31.

16. Lusuardi M, De Benedetto F, Paggiaro P, Sanguinetti CM,Brazzola G, Ferri P, et al. A randomized controlled trial onoffice spirometry in asthma and COPD in standard general

The prevalence of asthma and COPD in Italy 391

practice: data from spirometry in asthma and COPD:a comparative evaluation Italian study. Chest 2006;129:844e52.

17. Viegi G, Matteelli G, Angino A, Scognamiglio A, Baldacci S,Soriano JB, et al. The proportional Venn diagram of obstructivelung disease in the Italian general population. Chest 2004;126:1093e101.

18. Albers M, Schermer T, Molema J, Kloek C, Akkermans R,Heijdra Y, et al. Do family physicians’ records fit guidelinediagnosed COPD? Fam Pract 2009;26:81e7.

19. van Schayck CP, Chavannes NH. Detection of asthma andchronic obstructive pulmonary disease in primary care. EurRespir J 2003;39(Suppl.):16se22s.

20. Tirimanna PRS, van Schayck CP, den Otter JJ, et al. Prevalenceof asthma and COPD in general practice in 1992: has it changedsince 1977? Br J Gen Pract 1996;46:277e81.

21. Burr ML, Charles TJ, Roy K, Seaton A. Asthma in the elderly: anepidemiological survey. Br Med J 1979;1:1041e4.

22. Nejjari C, Tessier JF, Letenneur L, Dartigues JF, Barberger-Gateau P, Salamon R. Prevalence of self-reported asthma symp-toms in a French elderly sample. Respir Med 1996;90:401e8.

23. Lindberg A, Bjerg-Backlund A, Ronmark E, Larsson LG,Lundback B. Prevalence and underdiagnosis of COPD by diseaseseverity and the attributable fraction of smoking. Report from

the obstructive lung disease in Northern Sweden studies.Respir Med 2006;100:264e72.

24. Kauffmann F, Becklake MR. Maladies obstructives pulmonaires:un paradigme de la complexite des differences de sante entrehommes et femmes. In: La sante des femmes. Paris: Medecineet Sciences; 1996. p. 209e33.

25. Osman M, Hansell AL, Simpson CR, Hollowell J, Helms PJ.Gender-specific presentations for asthma, allergic rhinitis andeczema in primary care. Prim Care Respir J 2007;16:28e35.

26. Zannolli R, Morgese G. Does puberty interfere with asthma?Med Hypotheses 1997;48:27e32.

27. Marinoni A. Prevalence of asthma and asthma symptoms ina general population sample from northern Italy. Allergy 1995;50:755e9.

28. Neukirch F, Pin I, Knani J, Henry C, Pison C, Liard R, et al.Prevalence of asthma and asthma-like symptoms in threeFrench cities. Respir Med 1995;89:685e92.

29. Chen Y, Stewart P, Johansen H, McRae L, Taylor G. Sexdifference in hospitalization due to asthma in relation to age.J Clin Epidemiol 2003;56:180e7.

30. Camp PG, Chaudhry M, Platt H, Roch M, Road J, Sin D, et al.The sex factor: epidemiology and management of chronicobstructive pulmonary disease in British Columbia. Can RespirJ 2008;15:417e22.