Tuberculosis screening by anti-TNF prescribers 1169study was designed to address this gap in the knowl- of safety issues associated with anti-TNF agents, ex-edge base by assessing the awareness of TB infection perience prescribing them, types of testing conductedrisk and frequency of TB screening among anti-TNF to identify appropriate candidates for anti-TNF treat-prescribers in European Union (EU) states. It also ment, whether anti-TNF agent risk-related educa-sought to determine whether and to what extent peri- tional materials had been received from pharmaceuti-odic TB re-testing was occurring and to identify fac- cal companies, and if so, what types. The survey wastors predicting adherence to best-practice recommen- developed and pre-tested in English, and then trans-dations for TB screening. lated into the main language of each participating Our study focused on three anti-TNF agents on the country. The one exception was Belgium where French,EU market: infliximab, etanercept and adalimumab. Dutch and German versions of the survey were avail-The Summary of Product Characteristics (SmPC) for able and respondents were able to select their preferredeach of these three products states that prescribers language version. Respondents received financial com-should screen patients for TB prior to prescribing and pensation (between US$50 and US$60 per interviewmonitor them for TB throughout the duration of depending on the country) for survey participation.anti-TNF treatment.23–25 We hypothesized that thelikelihood of conducting TB screening would be asso- Data analysisciated with the receipt of TB screening information Univariate and bivariate statistics were used to ana-and related educational materials on anti-TNF agents. lyze study results. Responses to survey questions were cross-tabulated by physician characteristics. For uni-METHODS variate analyses, the χ2 test was used to compare the frequency distribution of responses across physicianStudy type and population groups. For between-group differences on values ofThe study was a cross-sectional survey of rheuma- continuous variables, the Student’s t-test was used.tologists, gastroenterologists and dermatologists in Significance levels were set at P < 0.05.24 EU countries who prescribed anti-TNF-α agents. Multivariate logistic regression was used to ascer-Countries were classified into two groups: the five tain the relative impact of each variable as a way toforemost industrialized economies, or G5 countries differentiate physicians who performed TB testing(Germany, France, Italy, Spain and the United King- from non-test performers. To assess TB screening per-dom) versus the non-G5 countries.* The survey was formance, the dependent variable was defined as 1 =conducted between 1 March and 31 May 2010. performs a TB history, uses either a purified protein derivative (PPD) skin test or TB blood test, and con-Sampling size and procedure ducts a chest X-ray for all patients to be prescribedA random sample of anti-TNF-α prescribers by spe- any of the three anti-TNF agents (i.e., adalimumab,cialty was generated within each of the participating etanercept or infliximab) vs. 2 = does not perform.EU countries. The sampling frame was based on the Based on a literature review, 20 variables were iden-top 10% of prescribers of adalimumab in the prior tified as potential predictors of TB testing.22 These12 months (n = ~12 000, including 4338 rheumatol- variables included patient caseload characteristics,ogists, 4511 gastroenterologists and 3151 dermatolo- physician training and practice setting characteristics,gists). Sampling was stratified by specialty. physician awareness regarding TB risk, G5 status of Both telephone and Internet-based methods were country where physician practiced, number of anti-used for recruitment. Eligible physicians were 1) spe- TNF agents for which physician had received educa-cialists or sub-specialists in rheumatology, gastro- tional materials and physician receipt of specific typesenterology or dermatology; 2) prescribers of adalim- of risk-related educational materials. After running theumab for patients with rheumatic diseases, Crohn’s initial analysis with these 20 variables, seven proveddisease or psoriasis at some point within the past to be significant (P < 0.05 level) and were retained in12 months; and 3) in Italy and Spain only, those who the final model: rheumatology specialty (yes/no), gas-had been visited by a sponsor sales representative in troenterology specialty (yes, no), practices in a G5the past year. Country-level response rates ranged from country (yes/no), is aware of TB as a risk associated5.4% (Romania) to 35.73% (Spain).* with use of anti-TNF agents (yes/no), follows inter- national, national or local guidelines for TB screeningSurvey instrument (yes/no), percentage of patients with severe diseaseThe interview consisted of 41 questions and took ap- (i.e., rheumatoid arthritis, Crohn’s disease, psoriasis),proximately 40 min to complete. Topics addressed and number of anti-TNF products for which safetyincluded demographic and clinical training charac- publications had been received.teristics, patient caseload characteristics, awareness The logistic regression was performed using Proc Logistic Statistical Analysis Software, version 9.2 (SAS* A list of the participating countries and responses by G5 status Institute Inc, Cary, NC, USA) on a Windows XP 64 bitcan be obtained from the corresponding author. OS PC (Microsoft, Redwoods, WA, USA). Bootstrap
1170 The International Journal of Tuberculosis and Lung DiseaseTable 1 Physician characteristics by specialty and by G5 vs. non-G5 country status for 2010 survey Rheumatologists Gastroenterologists Dermatologists G5 Non-G5 G5 Non-G5 G5 Non-G5 (n = 229) (n = 212) (n = 173) (n = 93) (n = 138) (n = 70)Characteristic % % % % % %Practice setting Hospital 81 88 92 90 78 76 Office 19 12 8 10 22 24Mean number of years in practice 16 19 16 18 14 16% time in direct patient care 84 75 82 75 82 75Age, years ⩽39 29 19 32 16 49 29 40–49 39 33 38 44 29 28 50–59 26 36 27 32 21 33 ⩾60 5 11 3 8 1 10Sex Male 64 51 80 85 58 54 Female 36 49 20 15 42 46Time prescribing biologics, years <1 3 13 22 17 30 33 1–3 11 25 57 60 57 56 >3 86 62 21 23 13 11Patients of total caseload with indication for anti-TNF 25 26 21 19 14 15TNF = tumor necrosis factor.re-sampling was performed on the final model using and non-G5 countries alike, 95% of rheumatologistsan SAS macro available on the SAS web site called recognized TB reactivation as a risk. Similarly, a high%JackBoot. Iterations (n = 500) of the classic boot- percentage of gastroenterologists (95% in non-G5strap were run using the Proc Logistic module in SAS and 88% in G5 countries) and dermatologists (94%(results available upon request to corresponding au- in non-G5 and 90% in G5 countries) reported recog-thor). Model parameters were stable over repeated re- nizing TB reactivation as a risk.sampling of our data. Estimates were bias-corrected.All estimates were considered statistically significant Percentage reporting following guidelinesat the P < 0.05 level. for TB screening Prescribers were also asked whether they followedRESULTS any TB screening guidelines (e.g., international, na- tional and/or local guidelines) prior to prescribing anPhysician characteristics anti-TNF agent: 92% of rheumatologists in non-G5A total of 915 physicians participated in the sur-vey (441 rheumatologists, 266 gastroenterologists and208 dermatologists). The majority of the respondentshad practices located in a hospital setting as opposedto an office (Table 1). Respondents had practiced fora mean of 14 years (range 2–40); the majority weremale and aged >40 years. Most rheumatologists hadprescribed anti-TNF agents for >3 years (G5 coun-tries, 86%; non-G5 countries, 62%). The majorityof the G5 and non-G5 gastroenterologists had pre-scribed anti-TNF agents for 1–3 years (respectively57% and 60%). Compared to the other specialtygroups, a higher proportion of dermatologists hadbeen prescribing biologics for <1 year (30%, G5;33%, non-G5).Awareness of TB reactivation as an adverse event Figure 1 Percentage of rheumatologists, gastroenterologistsassociated with the use of anti-TNF agents and dermatologists who reported identifying TB as an adverse event associated with anti-TNF agents: G5 vs. non-G5 EU mem-Across all three specialties, physicians reported rec- ber countries, 2010. TB = tuberculosis; G5 = the five foremostognizing TB reactivation as an adverse event associ- industrialized economies; TNF = tumor necrosis factor; EU =ated with anti-TNF agents as a class (Figure 1). In G5 European Union.
Tuberculosis screening by anti-TNF prescribers 1171 Figure 3 Percentage of rheumatologists, gastroenterologists and dermatologists who reported ever re-testing their anti-TNF agent patients for latent TB in G5 vs. non-G5 EU member coun- tries, 2010. TB = tuberculosis; G5 = the five foremost industri- alized economies; TNF = tumor necrosis factor; EU = European Union.Figure 2 Percentage of rheumatologists, gastroenterologistsand dermatologists who reported following guidelines for TBtesting of their patients prior to prescribing anti-TNF agents: countries. Physicians who reported being aware ofG5 vs. non-G5 EU member countries, 2010. TB = tuberculosis; risks for TB associated with the use of anti-TNFG5 = the five foremost industrialized economies; TNF = tumor agents were approximately twice as likely to screennecrosis factor; EU = European Union. as physicians who were not; similarly, those who re- ported adherence to TB screening guidelines (eithercountries and 85% in G5 countries reported that local or international) were 2.3 times more likely tothey did follow TB screening guidelines; 80% of gas- screen for TB than those who did not. Furthermore,troenterologists in non-G5 countries reported that other significant predictors of TB screening includedthey followed TB screening guidelines vs. 73% in G5 higher proportion of patient caseload with severe dis-countries, while a similar percentage of dermatolo- ease (OR 1.02), and greater number of anti-TNFgists in both G5 and non-G5 countries (76%) stated agents for which the physician received risk-relatedthat they followed TB screening guidelines when pre- educational materials (OR 1.32).scribing anti-TNF agents (Figure 2). TB re-testingPredictors of TB screening Physicians were asked whether they ever re-tested pa-Results of the logistic regression model showed that tients for TB following initiation of anti-TNF treat-rheumatology and gastroenterology specialty status ment. Differences emerged by specialty and by G5 vs.had odds ratios (ORs) of respectively 1.7 and 2.4, in- non-G5 country status (Figure 3). The lowest level ofdicating that individuals in these specialties were sig- re-testing was reported by dermatologists in non-G5nificantly more likely than dermatologists to report countries (30%). Rheumatologists in G5 countriesconducting TB screening (Table 2). The G5/non-G5 were less likely to re-test for TB than their counter-country status variable had an OR of 2.2, indicating parts in non-G5 countries (41% vs. 50%, P < 0.05),that G5 physicians were significantly more likely to while the percentage of gastroenterologists in G5screen for TB as compared to physicians in non-G5 countries who reported conducting TB re-testing (33%) did not differ appreciably from that reported by their counterparts in non-G5 countries (31%).Table 2 Results of logistic regression for predictors of TBscreening by rheumatologists, gastroenterologists and Among those who reported conducting TB re-dermatologists (n = 915) in select EU member countries, 2010 testing, the average time to re-testing ranged from 12 months (dermatologists in non-G5 countries) to Point estimatePredictors of OR (95%CI) 27.3 months (gastroenterologists in non-G5 coun- tries). Specifically, the average wait time for G5 andRheumatologist (yes /no) 1.73 (1.16–2.60)Gastroenterologist (yes /no) 2.41 (1.52–3.82) non-G5 rheumatologists before TB re-testing was ap-Practicing in a G5 country (yes /no) 2.19 (1.56–3.08) proximately the same (respectively 25.6 months, 95%Aware of risks of TB associated with use of confidence interval [CI] 1–12 vs. 25.3 months, 95%CI anti-TNF agent 1.97 (1.10–3.53)Follow any guidelines for TB testing (yes /no) 2.25 (1.51–3.33) 1–60). The mean wait time for gastroenterologists dif-% of patients with severe disease (rheumatoid fered between G5 and non-G5 countries (18.8 months, arthritis, Crohn’s disease, psoriasis) 1.02 (1.01–1.03) 95%CI 1–24 vs. 27.3 months, 95%CI 1–60). In con-Number of anti-TNF agents for which they had received safety publications 1.32 (1.15–1.54) trast, dermatologists reported the shortest mean in- terval between TB re-testing in both G5 and non-G5TB = tuberculosis; EU = European Union; G5 = the five foremost industrial-ized economies; OR = odds ratio; CI = confidence interval; TNF = tumor countries (respectively 13.6 months, 95%CI 2–60 vs.necrosis factor. 12.4 months, 95%CI 3–24).
1172 The International Journal of Tuberculosis and Lung DiseaseDISCUSSION These wide differences in monitoring practices may reflect the fact that there is no EU or internationalOur survey of TB risk awareness and screening is consensus recommendation on whether or how oftenthe first to be reported for EU prescribers of anti- to conduct TB re-testing.19TNF-agents. Our study results provide a snapshot of This study had several limitations. First, TB screen-physicians’ awareness of latent TB reactivation risk ing performance was based on physician self-report.and reported TB screening practices for patients A number of factors, including social desirability andwith chronic inflammatory diseases being considered deficits in recall, may have affected the accuracy offor anti-TNF treatment. The vast majority of physi- these responses. Second, we limited recruitment tocian respondents reported being aware of the risk of high-volume prescribers of anti-TNF agents. This ap-reactivation of LTBI in patients taking anti-TNF proach increased sampling efficiency but may haveagents. Most physicians reported utilizing guidelines introduced some selection bias. Third, we did not de-—typically national guidelines—for conducting TB termine whether prescribers who reported not per-screening tests prior to prescribing anti-TNF agents. forming TB re-testing had referred their anti-TNF pa-Consistent with this, reported rates of TB screening tients to pulmonary specialists for such testing.ranged from 73% (gastroenterologists in G5 coun-tries) to 92% (rheumatologists in non-G5 countries). CONCLUSION Study results revealed room for improvement inclinical practice across EU countries. About one in Receipt of anti-TNF safety information was associ-every 10 anti-TNF prescribers reported not follow- ated with higher prescriber awareness of the risk ofing any guideline for pre-treatment TB screening, and LTBI with anti-TNF agents as well as an increasedbetween 8% and 27% of physicians reported not likelihood of performing TB screening. These findingsscreening their patients for TB.11,26 suggest that the provision of pertinent educational ma- We found that rheumatologists and gastroenterol- terials can be an effective component of a risk mini-ogists were at least twice as likely as dermatologists mization strategy to promote TB screening amongto screen potential anti-TNF users for LTBI; likewise, anti-TNF prescribers. Physicians who prescribe anti-physicians in G5 countries were twice as likely to TNFs, particularly those in non-G5 countries, mayscreen for TB as those in non-G5 countries. Physi- benefit from additional interventions that prompt andcians who reported being aware of the risk of TB re- support them to conduct TB screening and monitor-activation in patients on anti-TNF agents were ap- ing.30–34 To improve the effectiveness of risk minimi-proximately twice as likely to screen patients as those zation efforts in this area, further studies are neededwho were unaware of or did not adhere to any guide- to understand and overcome barriers that constrainlines. This finding underscores the value of providing knowledge transfer and behavior change in prescrib-TB risk-related educational materials to prescribers. ers of anti-TNF agents.The positive linear relationship between the numberof anti-TNF agents for which physicians had received Acknowledgementssafety-related educational materials and the likeli- The authors thank R Hoffman, B Paperiello and S Williamson forhood of TB screening supports the learning concept input in the survey design and study implementation, and C Hof-that the repeated receipt of reinforcing information mann for editorial assistance. Funding for this study was providedfrom multiple sources can promote desired behav- by Abbott Laboratories. MS and LM are full-time employees of Ab- bott Laboratories. BA of Psyma International Inc (King of Prussia,ioral change among physicians. PA, USA) and TE were paid as consultants by Abbott Laboratories. A significant predictor of physician TB screeningwas G5 country status. LTBI is much more prevalentin non-G5 areas. TB incidence has increased in Europe Referencesover the past two decades, from 37 cases per 100 000 1 Lin J, Ziring D, Desai S, et al. TNF-alpha blockade in humanpopulation in 1990 to 49/100 000 in 2007.27 The TB diseases: an overview of efficacy and safety. Clin Immunolburden is relatively low in G5 countries (a low of 2008; 126: 13–30.5 cases/100 000 in Germany) compared to non-G5 2 Gardam M A, Keystone E C, Menzies R, et al. Anti-tumournations (e.g., 128/100 000 in Romania). Prior studies necrosis factor agents and tuberculosis risk: mechanisms of ac- tion and clinical management. Lancet Infect Dis 2003; 3: 148–of TB screening practices in the general population 155.have also found that screening rates are lower among 3 Wallis R S, Broder M S, Wong J Y, et al. Granulomatous infec-physicians in less industrialized nations.28,29 tious diseases associated with tumor necrosis factor antago- TB re-testing rates were approximately one in ev- nists. Clin Infect Dis 2004; 38: 1261–1265.ery three patients for gastroenterologists and non-G5 4 Keane J, Gershon S, Wise R P, et al. Tuberculosis associated with infliximab, a tumor necrosis factor alpha-neutralizingdermatologists, one in every two patients for all rheu- agent. N Engl J Med 2001; 345: 1098–1104.matologists, and two in every three for G5 dermatol- 5 Bongartz T, Sutton A J, Sweeting M J, et al. Anti-TNF antibodyogists. When re-testing was conducted, the average therapy in rheumatoid arthritis and the risk of serious infec-interval between tests was 20 months (range 14–27). tions and malignancies: systematic review and meta-analysis of
Tuberculosis screening by anti-TNF prescribers 1173 rare harmful effects in randomized controlled trials. JAMA rheumatoid arthritis. Semin Arthritis Rheum 2010; 39: 327– 2006; 295: 2275–2285. 346. 6 Nam J L, Winthrop K L, van Vollenhoven R F, et al. Current 20 Arend S M, Breedveld F C, van Dissel J T. TNF-alpha blockade evidence for the management of rheumatoid arthritis with bio- and tuberculosis: better look before you leap. Neth J Med logical disease-modifying antirheumatic drugs: a systematic lit- 2003; 61: 111–119. erature review informing the EULAR recommendations for the 21 Acevedo-Vasquez E, Ponce de Leon D, Gamboa-Cardenas R. management of RA. Ann Rheum Dis 2010; 69: 976–986. Latent infection and tuberculosis disease in rheumatoid arthri- 7 Flynn J L, Chan J. Immunology of tuberculosis. Annu Rev tis patients. Rheum Dis Clin North Am 2009; 35: 163–181. Immunol 2001; 19: 93–129. 22 Gomez-Reino J J, Carmona L, Angel Descalzo M. Risk of tu- 8 Mohan A K, Cote T R, Block J A, et al. Tuberculosis following berculosis in patients treated with tumor necrosis factor an- the use of etanercept, a tumor necrosis factor inhibitor. Clin tagonists due to incomplete prevention of reactivation of latent Infect Dis 2004; 39: 295–299. infection. Arthritis Rheum 2007; 57: 756–761. 9 Gomez-Reino J J, Carmona L, Valverde V R, et al. Treatment 23 European Medicines Agency. REMICADE summary of prod- of rheumatoid arthritis with tumor necrosis factor inhibitors uct characteristics. London, UK: EMEA, 2011. http://www. may predispose to significant increase in tuberculosis risk: a ema.europa.eu Accessed May 2012. multicenter active-surveillance report. Arthritis Rheum 2003; 24 European Medicines Agency. ENBREL summary of product 48: 2122–2127. characteristics. London, UK: EMEA, 2011. http://www.ema.10 American Thoracic Society/Centers for Disease Control and europa.eu Accessed May 2012. Prevention. Targeted tuberculin testing and treatment of la- 25 European Medicines Agency. HUMIRA summary of product tent tuberculosis infection. Am J Respir Crit Care Med 2000; characteristics. London, UK: EMEA, 2011. http://www.ema. 161 (4 Pt 2): S221–S247. europa.eu Accessed May 2012.11 Bijlsma J W. Optimal treatment of rheumatoid arthritis: 26 Smolen J S, Landewe R, Breedveld F C, et al. EULAR recom- EULAR recommendations for clinical practice. Pol Arch Med mendations for the management of rheumatoid arthritis with Wewn 2010; 120: 347–353. synthetic and biological disease-modifying antirheumatic12 Centers for Disease Control. Tuberculosis associated with block- drugs. Ann Rheum Dis 2010; 69: 964–975. ing agents against tumor necrosis factor-alpha—California, 27 World Health Organization. Global tuberculosis control: epi- 2002–2003. MMWR Weekly 2004; 53: 683–686. demiology, strategy, financing. WHO report 2009. WHO/HTM/13 Doherty S D, Van Voorhees A, Lebwohl M G, et al. National TB/2009.411. Geneva, Switzerland: WHO, 2009. Psoriasis Foundation consensus statement on screening for la- 28 Shimeles E, Aseffa A, Yamuah L, et al. Knowledge and practice tent tuberculosis infection in patients with psoriasis treated of private practitioners in TB control in Addis Ababa. Int J with systemic and biologic agents. J Am Acad Dermatol 2008; Tuberc Lung Dis 2006; 10: 1172–1177. 59: 209–217. 29 Richardson N L. Evaluating provider prescribing practices for14 Furst D E, Keystone E C, Fleischmann R, et al. Updated con- the treatment of tuberculosis in Virginia, 1995 to 1998: an as- sensus statement on biological agents for the treatment of sessment of educational need. J Contin Educ Health Prof 2000; rheumatic diseases, 2009. Ann Rheum Dis 2010; 69 (Suppl 1): 20: 146–155. i2–29. 30 Brandt B, Shanedling J. Is the CME system obsolete? Minn15 Hasan U. Tumour necrosis factor inhibitors⎯what we need to Med 2010; 93: 35–37. know. N Z Med J 2006; 119: U2336. 31 Davis D, Galbraith R. Continuing medical education effect on16 Hochberg M C, Lebwohl M G, Plevy S E, et al. The benefit/risk practice performance: effectiveness of continuing medical edu- profile of TNF-blocking agents: findings of a consensus panel. cation: American College of Chest Physicians Evidence-Based Semin Arthritis Rheum 2005; 34: 819–836. Educational Guidelines. Chest 2009; 135 (Suppl): S42–S48.17 Ledingham J, Wilkinson C, Deighton C. British Thoracic Soci- 32 Gray J. Changing physician prescribing behaviour. Can J Clin ety (BTS) recommendations for assessing risk and managing Pharmacol 2006; 13: e81–84. tuberculosis in patients due to start anti-TNF-α treatments. 33 Patow C A, Bryan D J. Engaging physicians in CME: the power Rheumatology (Oxford) 2005; 44: 1205–1206. of theater. Minn Med 2010; 93: 38–40.18 Solovic I, Sester M, Gomez-Reino J J, et al. The risk of tuber- 34 Rodriguez E, Marquett R, Hinton L, et al. The impact of edu- culosis related to tumour necrosis factor antagonist therapies: cation on care practices: an exploratory study of the influence a TBNET consensus statement. Eur Respir J 2010; 36: 1185– of ‘action plans’ on the behavior of health professionals. Int 1206. Psychogeriatr 2010; 22: 897–908.19 Furst D E. The risk of infections with biologic therapies for
Tuberculosis screening by anti-TNF prescribers i RÉSUMÉC O N T E X T E : Des bureaux des médecins et des contextes les rhumatologues dans les pays non-G5. Des facteurshospitaliers dans 24 pays de l’Union Européenne. prédictifs du dépistage de la TB comportent une spécialitéO B J E C T I F S : Evaluer le degré de prise de conscience du de rhumatologie ou de gastroentérologie, une meilleurerisque de tuberculose (TB), les performances du dépi- conscience du risque que représente la réactivation de lastage de la TB et les facteurs qui permettent le dépistage TB, une meilleure adhésion aux directives des tests pourde la TB chez les prescripteurs des agents facteur de né- la TB, une charge plus importante de patients atteintscrose tumorale-alpha (TNF-α). d’une maladie grave, le fait de pratiquer dans un paysM É T H O D E S : En total, 915 prescripteurs (441 rhumato- avec une large industrialisation et un nombre plus grandlogues, 266 gastroentérologues et 208 dermatologues) d’agents anti-TNF pour lesquels des informations liéesd’agents anti-TNF ont participé à une enquête compor- au risque avaient été reçues.tant 41 données entre mars et mai 2010. On a mené des C O N C L U S I O N S : La plupart des médecins ont signaléanalyses multivariées pour identifier les facteurs prédic- être conscients du risque qui accompagne les traitementstifs du dépistage de la tuberculose (TB). anti-TNF en matière de réactivation d’une infection TBR É S U LTAT S : La réactivation de la TB a été identifiée latente. Les résultats suggèrent que la distribution depar ⩾88% des médecins comme effet indésirable associé documents éducatifs pertinents est une composante effi-à l’utilisation des anti-TNF. Le dépistage auto-rapporté ciente de la stratégie de minimisation du risque visant àde la TB a été de 73% chez les gastroentérologues dans promouvoir le dépistage de la TB par les prescripteursles cinq pays les plus industrialisés (G5) et de 92% chez de médicaments anti-TNF. RESUMENMARCO DE REFERENCIA: Se realizó un estudio en me- rólogos de los cinco países más industrializados (G5) ydios hospitalarios y en consultorios médicos de 24 países 92% por los reumatólogos de los demás países. Los fac-de la Unión Europea. tores que predijeron la práctica de la detección siste-O B J E T I V O S : Evaluar el conocimiento del riesgo de apa- mática de la TB fueron la especialidad en reumatologíarición de tuberculosis (TB), la práctica de la detección de o gastroenterología, un mejor conocimiento del riesgoesta enfermedad y los factores asociados con la realiza- de reactivación de la TB, un mayor cumplimiento de lasción de la investigación sistemática, por parte de quienes directrices sobre las pruebas diagnósticas de la TB, larecetan medicamentos antagonistas del factor de necro- atención a un mayor número de pacientes con enferme-sis tumoral alfa (TNF α). dad grave, la práctica en un país industrializado y la uti-M É T O D O S : Participaron en el estudio 915 médicos que lización de una mayor cantidad de medicamentos conrecetan medicamentos anti-TNF (441 reumatólogos, efecto anti-TNF, sobre los cuales se había recibido in-266 gastroenterólogos y 208 dermatólogos), los cuales formación relacionada con los riesgos.respondieron a un cuestionario con 41 elementos entre C O N C L U S I Ó N : La mayoría de los médicos afirmó po-marzo y mayo del 2010. Mediante análisis multifacto- seer conocimientos sobre el riesgo probable de reacti-riales se definieron los factores que permiten predecir la vación de una infección tuberculosa latente con la ad-práctica de la detección sistemática de la TB por parte ministración de tratamientos antagonistas del TNF.de estos profesionales. Estos resultados indican que la distribución de materialR E S U LTA D O S : En las tres especialidades, 88% y más de didáctico pertinente, constituye un componente eficazlos médicos reconocieron la reactivación de la TB como de una estrategia de disminución de riesgos, tendente auna de las reacciones adversas asociadas con el uso de fomentar la detección sistemática de la TB por parte delos medicamentos anti-TNF. La autonotificación de la los médicos que recetan medicamentos anti-TNF.detección osciló entre 73% por parte de los gastroente-